EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: joeqsmith on June 19, 2018, 05:07:11 pm
-
We lost our power a couple of times last week and my UPS is on the blink, again. Most power dropouts in my area only last a second or so and I am thinking to setup a UPS using super capacitors rather than lead acid. UPS requires 4 cells. Cost is about $60/ea and they can last up to a couple of years.
I ran some numbers on what I need for capacitance and started looking on Digikey. I am really tempted to go with some of these ultra cheap caps and suspect they are like cheap CAT III rated handhelds and 10,000 mAH 9v transistor batteries.
What's been your experience with cheep supercaps?
-
Are there such things as cheap high voltage supercaps? Voltage scaling, even to say 5V over 3.3V, seems to double the price per Farad or more... I can't imagine trying to make a high voltage array with them.
I think the usual approach for high voltage large cap arrays is to have lots of much smaller high voltage caps in parallel with some sort of load balancing mechanism. Each cap is probably only going to be 1-10mF, but with a large array you can get a fair bit of capacity. I guess first things first, though, how much charge do these things need to store? How much power is used up in the couple seconds it would take to switch on, regain input power, and switch off gracefully? Would you still need a UPS for longer outages?
Also worth noting that "conversion" is not realistic. This would be a totally new build. Some bits of the UPS may be the same as the new one, but the response and charging mechanism for a lead acid cell is totally different from a capacitor array and when dealing with this sort of energy, they will not be compatible to just swap in caps.
-
Voltage scaling, even to say 5V over 3.3V, seems to double the price per Farad or more...
51 percent more voltage, E=1/2CV^2 so 128 percent more energy at 5V than 3.3 for the same capacitance, so the cost per joule stored remains about the same ;)
-
I ran some numbers on what I need for capacitance and started looking on Digikey. I am really tempted to go with some of these ultra cheap caps and suspect they are like cheap CAT III rated handhelds and 10,000 mAH 9v transistor batteries.
What's been your experience with cheep supercaps?
Dunno what numbers you calculated but it's not possible to replace lead-acid batteries with reasonably sized supercapacitors and retain reasonable capacity at the same time. Not to say that charge circuit in UPS is completely unsuited for this. Maybe it will somewhat work out if you attach pre-charged supercapacitors. As of discharge, after consuming relatively small part of already inferior capacity, UPS will detect that battery is empty and will cut off as voltage will drop with discharging unlike on battery which will hold around the same voltage until almost depleted.
-
Are there such things as cheap high voltage supercaps? Voltage scaling, even to say 5V over 3.3V, seems to double the price per Farad or more... I can't imagine trying to make a high voltage array with them.
I think the usual approach for high voltage large cap arrays is to have lots of much smaller high voltage caps in parallel with some sort of load balancing mechanism. Each cap is probably only going to be 1-10mF, but with a large array you can get a fair bit of capacity. I guess first things first, though, how much charge do these things need to store? How much power is used up in the couple seconds it would take to switch on, regain input power, and switch off gracefully? Would you still need a UPS for longer outages?
Also worth noting that "conversion" is not realistic. This would be a totally new build. Some bits of the UPS may be the same as the new one, but the response and charging mechanism for a lead acid cell is totally different from a capacitor array and when dealing with this sort of energy, they will not be compatible to just swap in caps.
Keep in mind, you are the one who mentions high voltage. If it could hold my loads for 5 seconds, that would cover the vast majority of drop outs I see. If it is out for days, I am running a genset. Looks like someone has tried it.
https://www.youtube.com/watch?v=W7CtqmfMrWI (https://www.youtube.com/watch?v=W7CtqmfMrWI)
I ran some numbers on what I need for capacitance and started looking on Digikey. I am really tempted to go with some of these ultra cheap caps and suspect they are like cheap CAT III rated handhelds and 10,000 mAH 9v transistor batteries.
What's been your experience with cheep supercaps?
Dunno what numbers you calculated but it's not possible to replace lead-acid batteries with reasonably sized supercapacitors and retain reasonable capacity at the same time. Not to say that charge circuit in UPS is completely unsuited for this. Maybe it will somewhat work out if you attach pre-charged supercapacitors. As of discharge, after consuming relatively small part of already inferior capacity, UPS will detect that battery is empty and will cut off as voltage will drop with discharging unlike on battery which will hold around the same voltage until almost depleted.
It seems you missed the second sentence: "Most power dropouts in my area only last a second or so and I am thinking to setup a UPS using super capacitors rather than lead acid. "
Obviously I am not looking for it to run for several minutes. It's very rare our power drops out for more than seconds. The calculations I have done are just to get an idea on the size I need to pull this off with my loads.
Again, what I am asking is peoples experiences with these cheap caps. If I buy a cap from a name brand supplier, I expect it will perform to their datasheets. I have my doubts about these cheap caps I see. I can see buying a 1000F cap and receiving a 10F part.
-
Well I know the supercaps used in some of the Fluke multimeters and Davis weather stations are a known weak point. Failure with a leaky mess seems to be fairly common.
-
Keep in mind, you are the one who mentions high voltage. If it could hold my loads for 5 seconds, that would cover the vast majority of drop outs I see. If it is out for days, I am running a genset. Looks like someone has tried it.
Fair, but what voltage do you actually need? I'm not used to seeing UPSs lower than 12, and 12V is a fair bit for a supercap... and when I hear UPS, I generally think of mains output, in which case your DC storage voltage would not be too low. If you mentioned what the output or the load was, it would be easier to tell what sort of application you're looking at ;)
Regardless, if you're using an array with integrated charge controllers and your UPS doesn't need to last for any duration, then you're probably still losing out on recharge speed using the UPS's charging circuit, but maybe that isn't much of a problem. I'd look for caps with a brand name on the cheap end, rather than just unnamed or unspecified, then just derate a bit for voltage. Supercaps are always going to be less tolerant of ripple currents and heat, have a much higher ESR, and don't have the same long lifespan ratings as their electrolytic counterparts, so if your output can draw spikes of current it may be worth adding some standard electrolytics to effectively bypass the array to reduce ripple. You can measure the capacitance yourself with a lab power supply with a current limit (or a voltage source and a resistor), a stopwatch, and a multimeter to verify the rated claim. Maybe doesn't help pick a brand if you're buying them for a one-off all at once, but you can at least be sure you get what you paid for.
-
Ok, I've just never considered 12V or even 120V high voltage. Yes, the caps are rated for a lower voltage but it's not a problem gang them up. That's not really my concern, nor is how to evaluate their performance. I am really interested in hearing about the experiences people have had buying cheap brands. Some of these are a third the price. Hard to believe they would perform very well. Are there cheap brands people have had good results with, ones that should be avoided.
Some of the new Eaton parts look interesting.
-
This is precisely the range where supercaps shine: time constants of seconds. :-+
Batteries are probably cheaper, smaller even, despite not being suited to such short durations -- probably just because they're a bigger commodity. So, you never(?) see supercaps here.
Batteries are fine: it works just like electrolytic bypass of a much faster example, a switching supply. A 10mJ electrolytic won't handle a 10us switching cycle, it'll have far too much ESR to do the job, or to do it without overheating. But 1J together will.
Likewise, a 1kJ lead-acid won't handle a 1s dropout, but 100kJ together will. As a bonus, it'll keep you propped up for minutes, not just seconds, on the off chance that you need such.
To put it another way: you're fundamentally worsening one performance metric of your system: support time. What are you getting in return for that? You aren't gaining efficiency, because this isn't one of those places where you only need what you use -- while that would be soothing to the OCD, :P the real need is availability, and if you aren't saving money or space on supercaps, why bother? :)
Now, if you're getting these drops so frequently that they're doing substantial damage to a battery that therefore needs frequent replacement, then you've got such a motivation: reliability. You might still want to ask if it's worthwhile to sacrifice the support time by going to supercaps, versus just getting a bigger battery (that therefore can handle the surge demand better -- again, short time demands get distributed better and better over larger reservoirs). Or getting a different chemistry, like a LiPo or LiFeP that can handle the demand with better life.
So, yeah, fundamentally, you're right in the sweet spot for supercaps. :) I just wonder if that's a sufficient cause, though, or if there may be other considerations that make you think, "oh, well, I suppose... oh well". :-//
Tim
-
This is precisely the range where supercaps shine: time constants of seconds. :-+
Batteries are probably cheaper, smaller even, despite not being suited to such short durations -- probably just because they're a bigger commodity. So, you never(?) see supercaps here.
Batteries are fine: it works just like electrolytic bypass of a much faster example, a switching supply. A 10mJ electrolytic won't handle a 10us switching cycle, it'll have far too much ESR to do the job, or to do it without overheating. But 1J together will.
Likewise, a 1kJ lead-acid won't handle a 1s dropout, but 100kJ together will. As a bonus, it'll keep you propped up for minutes, not just seconds, on the off chance that you need such.
To put it another way: you're fundamentally worsening one performance metric of your system: support time. What are you getting in return for that? You aren't gaining efficiency, because this isn't one of those places where you only need what you use -- while that would be soothing to the OCD, :P the real need is availability, and if you aren't saving money or space on supercaps, why bother? :)
Now, if you're getting these drops so frequently that they're doing substantial damage to a battery that therefore needs frequent replacement, then you've got such a motivation: reliability. You might still want to ask if it's worthwhile to sacrifice the support time by going to supercaps, versus just getting a bigger battery (that therefore can handle the surge demand better -- again, short time demands get distributed better and better over larger reservoirs). Or getting a different chemistry, like a LiPo or LiFeP that can handle the demand with better life.
So, yeah, fundamentally, you're right in the sweet spot for supercaps. :) I just wonder if that's a sufficient cause, though, or if there may be other considerations that make you think, "oh, well, I suppose... oh well". :-//
Tim
I could care less about weight, size or efficiency and I've been clear about the lack of need for long uptimes. Since I have been using a UPS, the number of times I have needed more than 5 seconds of up time is maybe a handful. This is over the course of many years.
As I mentioned in my first post, every two or three years I seem to be replacing my lead acids. I do have one now that has been in service for about five but it's rare. The power where I have lived has always been generally stable. It may drop out 4 or 5 times a year where it would have effected something I was doing.
I mentioned what it costs per battery. I just checked pricing and it seems on par with the last set I bought. There could be a break even assuming the super caps don't prematurely fail but it was not really my goal to save money.
Based on my personal experience with the mains, I am looking for a solution that gets me out of the battery swapping mode. It's a perfect time to look into it as I mentioned in my first post, I currently need to replace them.
-
As I mentioned in my first post, every two or three years I seem to be replacing my lead acids.
Either you buy crap batteries or charging circuit in your UPS sucks. Good long life batteries easily last 5+ years in a decent UPS. BTW if by chance you buy original batteries from APC, usually they are a cheap Chinese crap with very expensive APC sticker slapped on top.
-
As I mentioned in my first post, every two or three years I seem to be replacing my lead acids.
Either you buy crap batteries or charging circuit in your UPS sucks. Good long life batteries easily last 5+ years in a decent UPS. BTW if by chance you buy original batteries from APC, usually they are a cheap Chinese crap with very expensive APC label slapped on top.
I think there's a lot of truth in that.
We see lots of ppl replacing UPS batteries on what seems like too regular basis of 1-3yrs and we all know a properly cared for SLA can last much longer than this.
Personally I think UPS designers universally cock up by setting charging voltages for cyclic use instead of standby usage and by doing so just the additional few hundred mV shortens SLA life dramatically. Do they do this to artificially inflate their UPS specs, who knows ?
Add to this that SLA do not like being deep cycled and it's no wonder their life is often shortened in UPS usage.
In the only basic knowledge I have of UPS it seems one solution might be to use a modern Cadmium lead automotive battery that can withstand slightly higher charge voltages.
-
UPS have been APC and Tripp lite. Panasonic batteries for the most part. Although I tried some from battery plus or something that lasted about as long. I have an early APC that was the first one I bought. I later bought a larger APC. Battery life has been about the same between them all.
I use Odyssey AGMs in the race bikes. These are a total loss system (no charging system) and the batteries get charged between runs. These take a lot of abuse but about 2 years is what I expect for a service life. My street bike typically does much better but no where near what I see with my car or truck batteries. These seem to last about six. The longest use LA battery I have in use looks to be 8 years old now. It charges from the solar panel and runs the lights in my trailer. This is a deep cycle marine battery.
What brand/model UPSs and batteries do you feel are high quality?
-
What brand/model UPSs and batteries do you feel are high quality?
UPS's, I have no idea and I base some of my previous comments on being given a APC 1000 that liked eating SLA's too. After some investigation, measurements and tests I came to the conclusion the charging circuitry was entirely responsible for battery life and all this was some years back. Of what I've seen online about UPS's since hasn't changed my view one iota.
Batteries, well SLA's are the most convenient for small portable UPS's and you can't get around that unless you go down the path of your Supercap proposal Joe or look at deep cycle LA's.
If we look at proper 'off grid' solutions, large flooded cell LA's are the traditional power storage solutions and are usually matched to charge capability and load conditions in order not to deep cycle them.
But Li Ion is starting to gain a foothold in solar installations as technology improves to manage them properly and obtain good life.
However campers and marine use also needs deep cycle capabilities and marine further mechanical strengthening to withstand the banging about they can get.
My buddy uses those Odyssey's too but not as total loss.
My overall preference of high capacity battery brands is Optima bar none and I've had 10yrs+ on occasion for agricultural/commercial use. (red top)
However for UPS it's still the charge voltage and deep cycle abuse that need be accounted for and properly managed. I'd be ripping the charging system down and making some adjustments as it may be the cheapest and best investment of your time.
Thinking out loud.... if the charge voltage sense divider resistors are some shit 10% part could that throw the sense circuit all out of whack compared to a prototype ?
Of course it could and the plonkers that build these things don't even insert a trimmer so they can be corrected within the proper range so to not f**k SLA's, one set after the other. ::)
If these twits also have shares in a battery supply company, well that's pure genius ! >:D
-
I hate sealed lead acid batteries. It seems you never know they have died until you go to use them and they are absolutely dead, even when they have been on float charge all the time. Wet type NiCad batteries seem to last forever and I love them but are very expensive and aren’t suited for all applications.
I recently bought the same cheap 6 supercap bank (6 x 500F @16.2V/82F) shown in the previous video just to see if it would be usable for long term backup (2 days) on a piece of equipment I’m building that only draws about 40 ma at around 15 volts. After doing some testing like charging, discharging, and just letting them self discharge I found that of the six 500F/2.7V caps in the bank, 2 would not hold a charge that well but 4 had fairly low leakage. The 2 with high leakage were not useable for more than several hours before dropping to around .8 volts but the other 4 would stay at 2.6 volts for days. I removed the 2 bad supercaps to make sure it was the caps leaking and not the protective circuitry causing the discharge and it was the caps.
Even the 2 caps with high leakage were still able to produce high current output and when I tried draining the remaining .5 volt from one by shorting it with a resistor lead the lead got extremely hot and I had to drop it. Judging from how long it takes each cap to charge the capacity of all caps is probably within reason but it is leakage causing the self-discharge.
I also am not too sure of the protective circuit across each cap as it should put a load (2.5 ohms) across the cap when its voltage gets too high to try to keep the voltage across the cap under about 2.7 volts. It seems that when charging the bank some caps could charge above 3 volts and the resistors were not getting hot because the circuit had not turned on. A couple of circuits did work and the resistors got quite hot to the touch indicating the load had been put on the cap to discharge it to a safe voltage. As to charging the supercapacitor bank I have charged it at a rate as high as 5 amps with no problem.
If you are basically going to use these supercaps in a UPS which is always keeping them on charge and only using them for seconds or minutes every so often, I think they would work great for that application. I would do a few charge/discharge cycles and measure the voltages on each cap to make sure you don’t have any defective caps like I found but other than that I think your plan should work. You would still need 4 banks even though 3 of the 16.2 volt banks would be close.
-
What about lipo kept charged at 4-4.1V/cell?
-
As I mentioned in my first post, every two or three years I seem to be replacing my lead acids.
Either you buy crap batteries or charging circuit in your UPS sucks. Good long life batteries easily last 5+ years in a decent UPS. BTW if by chance you buy original batteries from APC, usually they are a cheap Chinese crap with very expensive APC sticker slapped on top.
Something nobody has mentioned yet is the Achilles heel of lead acid batteries - TEMPERATURE. The ups next to my work computer has the battery at easily 30deg C. I’ll bring in the temp probe tomorrow and see. The 13.65V charging voltage is only relevant at 20 deg C and should decline with increasing temperature.
The optimum operating temperature for the lead-acid battery is 25°C (77°F). As a guideline, every 8°C (15°F) rise in temperature will cut the battery life in half. VRLA, which would last for 10 years at 25°C (77°F), will only be good for 5 years if operated at 33°C (95°F). Theoretically the same battery would endure a little more than one year at a desert temperature of 42°C (107°F).
http://prod.sandia.gov/techlib/access-control.cgi/2004/043149.pdf (http://prod.sandia.gov/techlib/access-control.cgi/2004/043149.pdf)
-
UPS have been APC and Tripp lite. Panasonic batteries for the most part. Although I tried some from battery plus or something that lasted about as long. I have an early APC that was the first one I bought. I later bought a larger APC. Battery life has been about the same between them all.
I use Odyssey AGMs in the race bikes. These are a total loss system (no charging system) and the batteries get charged between runs. These take a lot of abuse but about 2 years is what I expect for a service life. My street bike typically does much better but no where near what I see with my car or truck batteries. These seem to last about six. The longest use LA battery I have in use looks to be 8 years old now. It charges from the solar panel and runs the lights in my trailer. This is a deep cycle marine battery.
What brand/model UPSs and batteries do you feel are high quality?
Panasonic and others have different battery series. The same company may have series which are rated for 2 or 3 times longer life than others. Say compare first one with the last one (the last one is for extreme temperatures).
http://www.csb-battery.com/upfiles/dow01524110112.pdf (http://www.csb-battery.com/upfiles/dow01524110112.pdf)
http://www.csb-battery.com/upfiles/dow01524110240.pdf (http://www.csb-battery.com/upfiles/dow01524110240.pdf)
http://www.csb-battery.com/upfiles/dow01520317548.pdf (http://www.csb-battery.com/upfiles/dow01520317548.pdf)
http://www.csb-battery.com/upfiles/dow01520318304.pdf (http://www.csb-battery.com/upfiles/dow01520318304.pdf)
In my experience Riello UPS are very good. Very reliable, silent, batteries inside or in separate enclosure stay cool and last a long time. APC IME is a turd, not reliable at all. Supplied batteries likely will be some crap like Kung Long. Eaton is reliable, but fans were extremely loud the last time we bought those (9130 series).
I buy long life CSB batteries for UPS used at my job. They are affordable and quiet good. They also often come in decent UPS like Riello and Eaton. BTW Hitachi bought them 2 years ago, brand remains the same though.
-
You could always try some Enersys Cyclon cells (cylindrical wound SLA). They tend to have very long life (I've got a few 20yr old ones), high energy density and low self-discharge, also very low internal resistance. They're about as reliable as it gets in SLA technology.
https://www.enersys.com/EMEA/CYCLON_Batteries.aspx?langType=1033 (https://www.enersys.com/EMEA/CYCLON_Batteries.aspx?langType=1033)
-
What brand/model UPSs and batteries do you feel are high quality?
Panasonic and others have different battery series. The same company may have series which are rated for 2 or 3 times longer life than others. Say compare first one with the last one (the last one is for extreme temperatures).
In my experience Riello UPS are very good. Very reliable, silent, batteries inside or in separate enclosure stay cool and last a long time. APC IME is a turd, not reliable at all. Supplied batteries likely will be some crap like Kung Long. Eaton is reliable, but fans were extremely loud the last time we bought those (9130 series).
I buy long life CSB batteries for UPS used at my job. They are affordable and quiet good. They also often come in decent UPS like Riello and Eaton. BTW Hitachi bought them 2 years ago, brand remains the same though.
[/quote]
I see the models for the batteries you recommend but not for the UPSs. Both companies make a broad range of products.
-
As I mentioned in my first post, every two or three years I seem to be replacing my lead acids.
Either you buy crap batteries or charging circuit in your UPS sucks. Good long life batteries easily last 5+ years in a decent UPS. BTW if by chance you buy original batteries from APC, usually they are a cheap Chinese crap with very expensive APC sticker slapped on top.
Something nobody has mentioned yet is the Achilles heel of lead acid batteries - TEMPERATURE. The ups next to my work computer has the battery at easily 30deg C. I’ll bring in the temp probe tomorrow and see. The 13.65V charging voltage is only relevant at 20 deg C and should decline with increasing temperature.
The optimum operating temperature for the lead-acid battery is 25°C (77°F). As a guideline, every 8°C (15°F) rise in temperature will cut the battery life in half. VRLA, which would last for 10 years at 25°C (77°F), will only be good for 5 years if operated at 33°C (95°F). Theoretically the same battery would endure a little more than one year at a desert temperature of 42°C (107°F).
Good point. My home lab will get well into the 90s F (30s C). The UPSs internals will raise that.
My one APC UPS must be well over 25 years old now. UL and CSA listed. The only thing I ever do with it is replace the battery.
-
I see the models for the batteries you recommend but not for the UPSs. Both companies make a broad range of products.
I think all Riello models we have are 8kVA or more. They have like 20y difference in age. Probably new models came out anyway since we bought the last batch. I don't recall a single one to fail. While APC were pain in the ass, even expensive 10 kVA.
EDIT: It's not that I recommend all of those batteries. It was more like for comparison. I would avoid GP series which are cheapest but have relatively short life and would go for at least GPL series. And would prefer XTV if they will be placed in hot environment.
-
GP series are what often comes in UPS from a factory (cheapest). As you can see in the datasheets all of them are rated for standby service. Do you really think that batteries used in UPS have written "for UPS" in the datasheet. Some of them may have but it's basically the same as "for audio" electrolytic capacitors.
BTW, check again:
Quote
Design for Standby Power Applications
Quote
Design for Standby Power Applications at Extreme Temperatures
« Last Edit: Today at 07:57:04 pm by wraper »
Which Riello models are you recommending?
-
Which Riello models are you recommending?
Already deleted before you posted, understood your post wrong. To recommend something, requirements are needed first.
-
I have messed around with this idea a bit in the past and found that some UPS charge circuits become very unhappy with the low impedance load presented by a discharged capacitor bank. So besides charge balancing, some type of charge control should be added.
-
Which Riello models are you recommending?
Already deleted before you posted, understood your post wrong. To recommend something, requirements are needed first.
You claim long battery life. That's the requirement.
-
I have messed around with this idea a bit in the past and found that some UPS charge circuits become very unhappy with the low impedance load presented by a discharged capacitor bank. So besides charge balancing, some type of charge control should be added.
What capacitors did you use and did you collect any data from them? This is really what I am after.
I have not dug into the details for the UPS I plan to use beyond basic stuff to get a ballpark for the capacitors required. It's interesting that some of the SuperCaps will list UPSs as their targeted application.
-
Which Riello models are you recommending?
Already deleted before you posted, understood your post wrong. To recommend something, requirements are needed first.
You claim long battery life. That's the requirement.
How one can recommend a model if you don't want to post at least a power required. I could advise 10kVA model but would you want to pay for that (you could just buy Tesla Powerwall)?
-
Which Riello models are you recommending?
Already deleted before you posted, understood your post wrong. To recommend something, requirements are needed first.
You claim long battery life. That's the requirement.
How one can recommend a model if you don't want to post at least a power required. I could advise 10kVA model but would you want to pay for that (you could just buy Tesla Powerwall)?
If you take the time to read my first post, you will find I never asked about battery brands or UPS models. I never stated I needed extended my UPS to to have extended periods of uptime. This is all something you have brought up.
I am not asking you for a recommendation. You claim to have a combination/s that give you what you feel is an acceptable battery life. I am only asking that you document the specifics of that combination/s. If you keep records on battery life where you work, that also may be of interest for people looking for general information on UPS systems. I am not suggesting I will do anything with this data but others may find it helpful.
It would also be interesting to see some pictures of your UPS systems internals just to get a general idea about them. Maybe the next time you pull one down for service, you could post a few.
-
I have messed around with this idea a bit in the past and found that some UPS charge circuits become very unhappy with the low impedance load presented by a discharged capacitor bank. So besides charge balancing, some type of charge control should be added.
I'd not be too worried about balance; leakage is exponential with voltage so they self balance.
Tim
-
Which Riello models are you recommending?
Already deleted before you posted, understood your post wrong. To recommend something, requirements are needed first.
You claim long battery life. That's the requirement.
How one can recommend a model if you don't want to post at least a power required. I could advise 10kVA model but would you want to pay for that (you could just buy Tesla Powerwall)?
If you take the time to read my first post, you will find I never asked about battery brands or UPS models. I never stated I needed extended my UPS to to have extended periods of uptime. This is all something you have brought up.
I am not asking you for a recommendation.
:-// :palm:
-
:-// :palm:
If you cherry pick, you will continue to be confused.
You claim to have a combination/s that give you what you feel is an acceptable battery life. I am only asking that you document the specifics of that combination/s.
-
:-// :palm:
If you cherry pick, you will continue to be confused.
You claim to have a combination/s that give you what you feel is an acceptable battery life. I am only asking that you document the specifics of that combination/s.
It's what you say now. Back then you asked for models I recommend. Why asking at all if not interested to begin with. What I can say that all our high power Riello UPS over various locations easily lasted 5+ years on the original set of batteries.
-
"UPS requires 4 cells."
OK, not trying to be pedantic or a smart arse, but I suspect you mean four (4) batteries, to give a nominal 48 volt system. The standard lead acid 12 volt battery consists of six (6) cells. So what I suspect you have is 4 x 12V 7AH or such batteries. Therefore 24 cells with a charging float voltage of around 2.27 volts per cell so that the voltage across the battery bank is 54.5 volts. When a load is applied to the batteries, as when the input voltage fails and the UPS is called upon, the voltage across the battery bank, and therefore inverter input, will fall to approximately 50 volts and then stabilise. It will then only fall slowly until the battery end voltage is reached which is normally around 1.7 volts/cell or around 41 volts for the bank. At this end voltage the battery has basically given up all its energy and if it is allowed to keep running the voltage will 'fall off a cliff' very rapidly. Take it too far down and the battery can be irreparably damaged.
The inverter of the UPS will therefore be designed to operate within the extremes of the battery operational points, so 40 volts at the low end and 55 volts at the high end. So doing a quick BoE calc you'll find that about half of the energy stored in the capacitor is never going to get used, so you will probably need to take that into consideration when selecting the caps.
The other thing that you may need to take into account is the reaction time of the UPS. In this I mean how long does the UPS stay on battery power even with a short break. Many UPSs will have an extended time on battery even for the shortest of power outages. They may stay on battery power for ten (10) or Twenty (20) seconds even for a votage fail of only a few cycles. This hysteresis values varies between UPS manufacturers, so this would need to be verified before selection and sizing of any capacitor could take place.
-
It would also be interesting to see some pictures of your UPS systems internals just to get a general idea about them.
We only have old Powerware (now Eaton) 9120 series at the location where I am. In think one with the lowest power is 3kVa. I won't be taking them apart as they are in work, I cannot just randomly walk and take them apart. They use quality components.
What I can say about APC is that they use cheap Chinese JH electrolytic capacitors, cheap relays and cheap batteries even in their expensive models.
-
What's been your experience with cheep supercaps?
I think this has not been answered yet for almost two pages of replies.
I guess no one ever experienced these Joe.
-
Changing SLAs to supercaps would require a different charging circuit and a modified battery monitoring. It might be better to design a completely new supercap UPS. Another possibility to simply things is to go for a centralized 12V power supply since most SOHO network/telephone stuff is powered by 12V DC anyway (use DC/DC converters for other voltages). This would be also more efficient than using a cheap SMPSU wall warts for each device. BTW, be aware that some UPS' perform regular self-tests (mine does a 10s test every two weeks).
From my experience with running UPS' I can say that Panasonic SLAs run fine for about 5 years. They also have special SLAs for UPS usage which last longer but are also more expensive. If possible check and adjust the float voltage every few years because a too high voltage reduces the SLA's life span. APC's SmartUPS' have a hidden service menu for that and APC are also known to set the float voltage quite high. Sadly I have to agree that the quality of APC's UPS' went down and that their stickers for battery packs are quite expensive. Since they use Kung Long and other inexpensive brands I wouldn't buy any genuine replacement packs. The change of the control protocol to a top-secret one adds to the impression that APC's main goal became profit optimization. You have to buy a special translation module to be able to use a proper UPS management software.
-
I have messed around with this idea a bit in the past and found that some UPS charge circuits become very unhappy with the low impedance load presented by a discharged capacitor bank. So besides charge balancing, some type of charge control should be added.
I'd not be too worried about balance; leakage is exponential with voltage so they self balance.
Tim
I've seen some posts (other forums) where people have posted about leakage being all over the place with some of the cheap caps. I never see any specifics.
At some point I will need to have a look at the charger in the UPS I plan to use. I only looked at the open circuit voltage. With the UPS unplugged and unloaded, I measured the current draw and various thresholds where the UPS detects low battery, disables the output and eventually turns off. I based the capacitor size on the target up time max rated load for the UPS and the point where the UPS detects a low battery, plus some pad.
It seems very doable but like the video I linked, the person spent most of their time reverse engineering and changing the charging. I suspect this is going to be the problem in my setup as well.
-
What's been your experience with cheep supercaps?
I think this has not been answered yet for almost two pages of replies.
I guess no one ever experienced these Joe.
AFAIK, good quality supercaps have a live span of about 10 years. My guess for the cheapies is 5 years.
-
"I think this has not been answered yet for almost two pages of replies.
I guess no one ever experienced these Joe."
Read reply #14 where I went into quite some detail about my initial experience with the exact same cheap supercaps used in the previous video the OP linked to. Perhaps no one noticed. :-[
-
I have messed around with this idea a bit in the past and found that some UPS charge circuits become very unhappy with the low impedance load presented by a discharged capacitor bank. So besides charge balancing, some type of charge control should be added.
What capacitors did you use and did you collect any data from them? This is really what I am after.
Before buying 35 supercapacitors for a 70 volt system, I did tests using a 100,000uF bank of aluminum electrolytics just to find out what modifications of the UPS charge circuits would be necessary. My test UPS was a Powerware Prestige EXT which uses an inconvenient 5x12V battery configuration.
I have not dug into the details for the UPS I plan to use beyond basic stuff to get a ballpark for the capacitors required. It's interesting that some of the SuperCaps will list UPSs as their targeted application.
They should be great for high reliability high power density line conditioners which is no surprise. The problem is that UPSes intended to use batteries expect a narrow operating voltage range.
The inverter of the UPS will therefore be designed to operate within the extremes of the battery operational points, so 40 volts at the low end and 55 volts at the high end. So doing a quick BoE calc you'll find that about half of the energy stored in the capacitor is never going to get used, so you will probably need to take that into consideration when selecting the caps.
Utilizing half of the energy of the capacitor is actually pretty good and what I would aim for minimum. Switching regulators are commonly designed to operate with a 2:1 input range (75%) but wide range 4:1 units (94%) also exist. A UPS intended to use supercapacitors would be designed to handle a wider input voltage range but most of the available energy is returned without going wild.
In a retrofit application, a wide input range high current boost converter could be used to extent the operating time but if you are already getting 50%, I would declare victory and move on.
AFAIK, good quality supercaps have a life span of about 10 years. My guess for the cheapies is 5 years.
This can be extended by voltage derating. Electrolysis of the electrolyte increases 10 times for every 0.4 volt increase in voltage and 30C increase in temperature. If you expect the capacitors to last more than 1 year, then do not operate them above 2.6 volts at 35C.
-
Incidentally, this will prove useful for determining capacitance:
https://www.seventransistorlabs.com/Calc/PSHoldUp.html (https://www.seventransistorlabs.com/Calc/PSHoldUp.html)
Tim
-
:-// :palm:
If you cherry pick, you will continue to be confused.
You claim to have a combination/s that give you what you feel is an acceptable battery life. I am only asking that you document the specifics of that combination/s.
It's what you say now. Back then you asked for models I recommend. Why asking at all if not interested to begin with. What I can say that all our high power Riello UPS over various locations easily lasted 5+ years on the original set of batteries.
Again, take the time to read. From my previous post:
I am not suggesting I will do anything with this data but others may find it helpful.
You may have felt I was looking for general help on UPSs and was offering your experience and advice but it was never my intent of the thread.
-
"I think this has not been answered yet for almost two pages of replies.
I guess no one ever experienced these Joe."
Read reply #14 where I went into quite some detail about my initial experience with the exact same cheap supercaps used in the previous video the OP linked to. Perhaps no one noticed. :-[
I had read it and it was why I posted about the difference in the leakage currents. Did you measure their capacitance as well? ESR? I wonder how close some of these parts are to their specs.
-
I have messed around with this idea a bit in the past and found that some UPS charge circuits become very unhappy with the low impedance load presented by a discharged capacitor bank. So besides charge balancing, some type of charge control should be added.
What capacitors did you use and did you collect any data from them? This is really what I am after.
Before buying 35 supercapacitors for a 70 volt system, I did tests using a 100,000uF bank of aluminum electrolytics just to find out what modifications of the UPS charge circuits would be necessary. My test UPS was a Powerware Prestige EXT which uses an inconvenient 5x12V battery configuration.
I have not dug into the details for the UPS I plan to use beyond basic stuff to get a ballpark for the capacitors required. It's interesting that some of the SuperCaps will list UPSs as their targeted application.
They should be great for high reliability high power density line conditioners which is no surprise. The problem is that UPSes intended to use batteries expect a narrow operating voltage range.
The inverter of the UPS will therefore be designed to operate within the extremes of the battery operational points, so 40 volts at the low end and 55 volts at the high end. So doing a quick BoE calc you'll find that about half of the energy stored in the capacitor is never going to get used, so you will probably need to take that into consideration when selecting the caps.
Utilizing half of the energy of the capacitor is actually pretty good and what I would aim for minimum. Switching regulators are commonly designed to operate with a 2:1 input range (75%) but wide range 4:1 units (94%) also exist. A UPS intended to use supercapacitors would be designed to handle a wider input voltage range but most of the available energy is returned without going wild.
In a retrofit application, a wide input range high current boost converter could be used to extent the operating time but if you are already getting 50%, I would declare victory and move on.
AFAIK, good quality supercaps have a life span of about 10 years. My guess for the cheapies is 5 years.
This can be extended by voltage derating. Electrolysis of the electrolyte increases 10 times for every 0.4 volt increase in voltage and 30C increase in temperature. If you expect the capacitors to last more than 1 year, then do not operate them above 2.6 volts at 35C.
I had stacked a few caps together when I first started to look at the UPS. I just don't have anything near large enough on hand. I do plan to derate them a fair amount.
I found another person that had made an attempt.
http://www.instructables.com/id/Converting-a-UPS-to-run-on-Super-Capacitors/ (http://www.instructables.com/id/Converting-a-UPS-to-run-on-Super-Capacitors/)
It appears several companies offer them now including the two companies that wrapper mentions:
https://powerquality.eaton.com/emea/supercapacitor/default.asp (https://powerquality.eaton.com/emea/supercapacitor/default.asp)
http://www.riello-ups.com/products/6-solutions/76-supercaps-ups (http://www.riello-ups.com/products/6-solutions/76-supercaps-ups)
https://www.marathon-power.com/supercapacitor-ups.html (https://www.marathon-power.com/supercapacitor-ups.html)
https://www.ecopowersupplies.com/supercapacitor-ups (https://www.ecopowersupplies.com/supercapacitor-ups)
-
I had read it and it was why I posted about the difference in the leakage currents. Did you measure their capacitance as well? ESR? I wonder how close some of these parts are to their specs.
By the way, ESR varies with time, much as it does with batteries. Though at least, I don't think this will be much of a problem at modest discharge rates. (Here, it should basically look like whatever the nominal ESR is.)
It is disgustingly apparent over long rates, however -- what would ordinarily be called dielectric absorption, is really ionic diffusion as charges equalize over the deeply porous (activated charcoal?) electrodes. So the leakage has a very long (days) tail, and the ratio of immediate capacity to full capacity (i.e., as measured at a discharge rate of, say, minutes versus weeks -- discounting leakage, naturally!) is surprisingly large, like 20 or 30% (i.e., the absorption recovery fraction).
The same effect, in batteries, limits how fast you can discharge, and especially charge, the chemistry.
On that note, an aside: I think it's neat to consider what's going on in batteries. At modest charge or discharge rates, behavior is linear, ionic diffusion isn't dominant, and efficiency is high. At high discharge rates, there is a large voltage drop through the electrolyte (and any semiconductors or insulators involved -- for example, lead dioxide is a semiconductor, and lead sulfate is an insulator), which saturates the reaction at the facing electrode surfaces, and drives more reaction deeper into the pores. The effect is, as you increase discharge rate, you get more and more current, albeit at less voltage -- less efficiency. This is great news for cold cranking applications, like cars and UPSs.
The opposite isn't true, though: when charging, the intended reaction has a lower overpotential, so it dominates at light charge rates. But a rapid charge quickly saturates the facing electrode surfaces, which undergo higher voltage reactions -- namely, the production of oxygen and hydrogen -- while the deeper pores continue to charge at a modest rate. This directly loses electrolyte solvent as gas (or recycles it, via catalyst, generating heat, in a sealed type), while the bubbles increase resistance further. I tested this recently, on a whim: a motorcycle-sized 12V lead acid battery can sustain over 60A discharge, keeping terminal voltage above 8V; but saturates quickly on charge, climbing to, say, 18V at only 20A. That was only a few seconds test, by the way.
I wouldn't recommend testing this in any lithium technology, for obvious reasons. The higher-voltage reactions (electrolysis of the solvent), and their consequences (gas buildup inside a sealed bag of pyrophoric chemicals, anyone?), are well known. ;) The same physics ought to be relevant, though. This is good news for RC enthusiasts, but not so much for Tesla Motors. :P
Tim
-
I am still scratching my head over WHY this idea seems like it solves anything useful. :scared:
Big, expensive, and complicated over the option of just using the typical SLA batteries? Is it possible? Sure.
-
Changing SLAs to supercaps would require a different charging circuit and a modified battery monitoring. It might be better to design a completely new supercap UPS. Another possibility to simply things is to go for a centralized 12V power supply since most SOHO network/telephone stuff is powered by 12V DC anyway (use DC/DC converters for other voltages). This would be also more efficient than using a cheap SMPSU wall warts for each device. BTW, be aware that some UPS' perform regular self-tests (mine does a 10s test every two weeks).
From my experience with running UPS' I can say that Panasonic SLAs run fine for about 5 years. They also have special SLAs for UPS usage which last longer but are also more expensive. If possible check and adjust the float voltage every few years because a too high voltage reduces the SLA's life span. APC's SmartUPS' have a hidden service menu for that and APC are also known to set the float voltage quite high. Sadly I have to agree that the quality of APC's UPS' went down and that their stickers for battery packs are quite expensive. Since they use Kung Long and other inexpensive brands I wouldn't buy any genuine replacement packs. The change of the control protocol to a top-secret one adds to the impression that APC's main goal became profit optimization. You have to buy a special translation module to be able to use a proper UPS management software.
Do you have any further info about this ?
Some obscure service manual link maybe ?
This menu might even allow some tweaking of charging voltages to better suit various configurations of super caps.
-
I am still scratching my head over WHY this idea seems like it solves anything useful. :scared:
Big, expensive, and complicated over the option of just using the typical SLA batteries? Is it possible? Sure.
Isn't that just what we do, explore alternative solutions to problems ?
Member Psi started using super caps instead of LA's in his car years back:
https://www.eevblog.com/forum/beginners/supercapacitor-circuit-design-help-please!/msg655164/#msg655164 (https://www.eevblog.com/forum/beginners/supercapacitor-circuit-design-help-please!/msg655164/#msg655164)
Link fixed.
-
I am still scratching my head over WHY this idea seems like it solves anything useful. :scared:
Big, expensive, and complicated over the option of just using the typical SLA batteries? Is it possible? Sure.
I thought I made my goal clear. Did you take the time to read any of the literature I linked from name brand companies that are offering it?
-
How about simply mounting a much bigger capacitor in the PSU of the device? That will probably give a much better return on investment. The problem with capacitors is that their voltage has a relation to their charge so you can only use a very small part of the energy stored before the under voltage protection kicks in. A PSU OTOH has been designed to deal with that.
-
I am still scratching my head over WHY this idea seems like it solves anything useful. :scared:
Big, expensive, and complicated over the option of just using the typical SLA batteries? Is it possible? Sure.
I thought I made my goal clear. Did you take the time to read any of the literature I linked from name brand companies that are offering it?
I hope your goal won't end up with house in flames.
-
I'm curious if some of the posters here have seen any of the videos where a car battery has been replaced with supercaps with success.
https://www.youtube.com/watch?v=gzaLF5tFf88 (https://www.youtube.com/watch?v=gzaLF5tFf88)
Ebay
https://www.ebay.com/p/Gravity-600-Amp-Car-Battery-Capacitor-Gr-600bc/1985231668?iid=392043817345&chn=ps (https://www.ebay.com/p/Gravity-600-Amp-Car-Battery-Capacitor-Gr-600bc/1985231668?iid=392043817345&chn=ps)
-
I'm curious if some of the posters here have seen any of the videos where a car battery has been replaced with supercaps with success.
Did you follow the link I posted (now fixed) in reply#48 ?
-
I am still scratching my head over WHY this idea seems like it solves anything useful. :scared:
Big, expensive, and complicated over the option of just using the typical SLA batteries? Is it possible? Sure.
I thought I made my goal clear. Did you take the time to read any of the literature I linked from name brand companies that are offering it?
I hope your goal won't end up with house in flames.
How about simply mounting a much bigger capacitor in the PSU of the device? That will probably give a much better return on investment. The problem with capacitors is that their voltage has a relation to their charge so you can only use a very small part of the energy stored before the under voltage protection kicks in. A PSU OTOH has been designed to deal with that.
That would be a major undertaking. But again, to be clear I am looking for peoples experiences with cheap super capacitors.
-
I am still scratching my head over WHY this idea seems like it solves anything useful. :scared:
Big, expensive, and complicated over the option of just using the typical SLA batteries? Is it possible? Sure.
I thought I made my goal clear. Did you take the time to read any of the literature I linked from name brand companies that are offering it?
I hope your goal won't end up with house in flames.
None of them have so far but I appreciate your concern.
-
I'll add my limited but very real experience with commercial UPS's.
I once designed and installed redundant server farms that we placed in rental space at colocation facilities all over the western USA. We consumed a LOT of UPS's, mostly APC's expensive (and supposedly high end) SmartUPS rackmount units. This gave me a lot of exposure to UPS's in general and APC UPS's in particular.
The earlier comments were correct about APC charging circuits: They suck. They REALLY suck. They suck so terribly I've run out of superlatives to describe how bad they are. And this is across the entire product line, not just the supposedly high end SmartUPS.
Within 12 months the batteries would have swelled so much that they could not be physically slid out of the rackmount frame. We initially presumed it was low quality batteries so we started researching and buying the very best we could, any chemistry, any price. Didn't matter. No battery could stand up to the electrical abuse that the APC charging circuit dished out. I've lost count of how many times I had to remove a rackmount APC UPS from the rack and take the entire enclosure apart, including multiple internal walls, just so the batteries could be PRIED out and replaced.
Meanwhile, I have many lead-acid and AGM batteries for other purposes that we maintain with Battery Tenders and similar aftermarket intelligent chargers and those batteries last years and years.
I will never again willingly purchase an APC UPS. I've since moved on in my career and don't have need for racks of UPS's so I don't have experience with other brands to share. But APC? Never again. Too many last-minute plane tickets and chasing around remote cities looking for a Batteries Plus and all-nighters disassembling and reassembling entire rackmount enclosures.
I agree with the other respondents who have openly wondered if the battery life is controlled by the UPS's charging circuit. Makes me wonder if one could run a quick experiment by taking an existing UPS, disconnecting its internal charging circuit, and substituting a respected external aftermarket battery charger/conditioner. Use the UPS to generate power, but rely on someone respected to recharge the batteries afterward. Just thinking out loud.
-
I'm curious if some of the posters here have seen any of the videos where a car battery has been replaced with supercaps with success.
Did you follow the link I posted (now fixed) in reply#48 ?
Sorry, my remarks were not directed at you but SOME think its black magic.
-
I'm curious if some of the posters here have seen any of the videos where a car battery has been replaced with supercaps with success.
I've seen a few of these. Actually, the cranking time on that video you link is longer than what I normally experience for a line drop out and I would not be surprised if the car did not present a much larger load then I pull off the UPS. The interesting thing about this video is the caps are mounted under the hood of a car. That's a much worse environment than I plan to run them in.
There are a few boxes of used Maxwells on eBay. I have not looked to see what they are. May be an option just to run some experiments with.
-
I'll add my limited but very real experience with commercial UPS's.
I once designed and installed redundant server farms that we placed in rental space at colocation facilities all over the western USA. We consumed a LOT of UPS's, mostly APC's expensive (and supposedly high end) SmartUPS rackmount units. This gave me a lot of exposure to UPS's in general and APC UPS's in particular.
The earlier comments were correct about APC charging circuits: They suck. They REALLY suck. They suck so terribly I've run out of superlatives to describe how bad they are. And this is across the entire product line, not just the supposedly high end SmartUPS.
:clap:
Some further hunting on this topic today reveals yours and my thoughts about APC charging are widespread.
A interesting link I just stumbled upon from chasing a hint madires gave us earlier: hidden service menu
http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage (http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage)
I might have to dig out the APC 1000 I parked up years ago and have another look at it ! :scared:
-
Yeah, those apc smart ups's they have set the floating voltage ridiculously high (29V if I remember correctly?) by default,
it can be lowered in some menu.
Had to replace the batteries on a couple of 19" units after 5 years of mostly standby use and maybe a handful of discharges.
The batteries where already completely deformed and swollen and took brute force to get them out.
I think apc's idea on this is that costumers should replace battery's every two or three years, $.
-
i have been dumping price and data on a spreadsheet for sometime, so i could meddle with some kWh capacity and voltage, and sort it to tell me some ranking data.
so i sized all units into some 14v modules with minimum 1Wh. i could see that supercaps that could nearly catch the tails of lead acids on my spreadsheet are something like 8.4v MAL219690118E3 vishay, this is around $25/Wh/year. the nearest lead acid (7.2AH) is at around $20/Wh/year. a 38AH is at around $9/Wh/year.
a 2.7v 400F cap (XV3560-2R7407) to compare, stands at around $45/kWh/year. i think this similar to what the youtube tripplite mod used 6x 2.7v? 67 farads?
if to compare with this tripplite mod in terms of similar amount of energy (1 Wh ?), then MAL219690117E3 or MAL219690118E3 might be useful for long lifespan.
but how much power outage can it bridge out of a "1 Wh battery"? if the usable UPS voltage is 14v to 10v, the UPS can only use about 28% of that 1 Wh from a supercapacitor. which means, bridge for about 3seconds for a 350w load? did i math this correctly?
or maybe, there is a statistics of brownouts etc to help size this?
on the other hand, the sustained current output of a "small" supercap stack (this vishay or string of 400F), i dont think could match that of a bulk leadacid. if i am seeing correctly, this (either the vishay or the 400F version) would need maybe alot more in parallel like 4 or 5?
by itself, the caps overall lifespan is around 23x more than the SLA, for about at least 2x the cost (just for 1 Wh, or comparatively 1/15 the capacity)
just some rough estimates, if the capacitor is around 50C/W in thermal resistance. a load dump for 3s could mean the end of life for the caps. 1 single use hmmm?
in any case, very interesting to dive inside some specs to find some numbers
-
As I mentioned in my first post, every two or three years I seem to be replacing my lead acids.
Either you buy crap batteries or charging circuit in your UPS sucks. Good long life batteries easily last 5+ years in a decent UPS. BTW if by chance you buy original batteries from APC, usually they are a cheap Chinese crap with very expensive APC sticker slapped on top.
Something nobody has mentioned yet is the Achilles heel of lead acid batteries - TEMPERATURE. The ups next to my work computer has the battery at easily 30deg C. I’ll bring in the temp probe tomorrow and see. The 13.65V charging voltage is only relevant at 20 deg C and should decline with increasing temperature.
The optimum operating temperature for the lead-acid battery is 25°C (77°F). As a guideline, every 8°C (15°F) rise in temperature will cut the battery life in half. VRLA, which would last for 10 years at 25°C (77°F), will only be good for 5 years if operated at 33°C (95°F). Theoretically the same battery would endure a little more than one year at a desert temperature of 42°C (107°F).
http://prod.sandia.gov/techlib/access-control.cgi/2004/043149.pdf (http://prod.sandia.gov/techlib/access-control.cgi/2004/043149.pdf)
Okay... got into my little tin shed office this morning and the ambient temp was 8.2 deg C. UPS has a single 12V 7AH SLA. Underside of the battery was 18 deg, side was 25 deg. Top was 33 deg. Battery housing surface next to the top of the battery was 37 deg C. Remember, this is in 8 deg C ambient so add another 15-25 deg to those figures. The battery was 2-3/4 years old and with a 2R2 ohm load dropped to 10.5 volts immediately. Also had some corrosion around one of the terminals, so basically it was way dead.
Edit -> After replacing the battery I've now stood the unit up vertically so that the battery is at the bottom and hopefully the internal heat from the transformer will now exit the top.
-
A interesting link I just stumbled upon from chasing a hint madires gave us earlier: hidden service menu
http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage (http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage)
It's not a "hidden service menu". It's a set of commands for factory calibration of the units that isn't documented unless you have the factory calibration notes. That page (and most of the net) is quite unclear about the float voltage on these units, so here are some real take home facts.
The APC UPS always target the upper end of the float voltage range for the batteries to get maximum runtime at the expense of ultimate battery life. The plethora of reports about APC SmartUPS cooking batteries isn't because APC are deliberately doing anything to cause this, it's because a combination of tolerances drift significantly as the units age and the divider that measures the battery voltage tends to drift low, forcing the charging ASIC to increase the voltage. It seems to happen to most of them, but then those 3G units are all >15 years old now. An occasional service isn't out of the question for a unit that has > 100,000 hours on the clock.
The SmartUPS family comes in 5 basic generations at this point. The most commonly found are G3 (beige SU series) and G4 (black SUA series). Some of the later G3 variants were black also, but they're kind of a half-breed (mostly G3 with some G4ish improvements but not the G4 charging smarts).
Generation 1&2 were configured with trimpots and had RS232 status line reporting. I've never seen one in the flesh, only the service manuals.
Generation 3 (SU) has the ability to report and alter the reported battery voltage in firmware. They have no ability to alter the *real* battery voltage in firmware other than to apply a boost during recharge (and that boost is an on/off thing only). To alter the charge voltage you must alter the resistive divider at the input to the charger ASIC. There is no way to alter the float voltage in firmware. All you are doing by "re-calibrating" the UPS is to change the voltage the UPS *thinks* the battery is at.
G4 -> (SUA) has the ability to alter the reported battery voltage in firmware. The processor uses the reported battery voltage to alter a PWM value to provide trimming to the charger ASIC which does alter the *actual* battery float voltage. This ability is primarily so the UPS can apply temperature compensation to the float voltage, but it has the side affect of being able to change the float voltage by altering what the UPS *thinks* the battery voltage is. As you change the battery voltage constant, the UPS alters the charge voltage to keep it where it thinks it should be based on the battery temperature (NTC hanging in the back of the battery box). There is no way of changing the target setpoint and changing the divider means the UPS isn't accurately reporting the battery voltage, but at the end of the day you are changing the float voltage.
The G4 also encompasses the new G5 (SMX/SMT/...) as the actual UPS back end is relatively the same. There is a second UI processor on those that handles the front panel and the Microlink interface, but if you dig in you'll find that processor is polling an essentially unmodified G4 back end using the same old UPSLink protocol. So if you get between those two you can tweak the UPS like you could with the older generations.
Each generation has a series of sub-generations. I've seen at least 3 variants of the SU series, the latest of which was entirely SMD and with a dedicated FET driver (the original through hole 3G SU series had a discrete FET driver that was awesome at cooking the board on the 48V models). No matter which variant, they all rely on a fixed resistive divider to set the battery voltage.
The reality with the UPS units is the batteries live inside the box with the electronics. That makes the batteries much warmer than they would be in a separate enclosure and *that* is the major compromising factor in their life. I've got a couple of units modified for external batteries, and that takes the battery life from ~3 years to ~5 years. No change in voltages or battery management strategy.
I also modify the SU series to run the fans 24/7 at half speed. That drops the internal battery temperature by about 5C off the bat. The SUA series and up do that out of the box.
As for quality, things certainly started to dive after the Schneider buyout. I don't think there is any dispute of that.
-
I think the first UPS I used was made by Sola and used ni-cads. Maybe. That was early 80s.
The first APC I bought for home has turned yellow over the years. Its still works fine and is use 24/7. This unit stays in a part of the house where it is cooler and has the battery that I mentioned that's the oldest. It looks like these:
http://www.mysynergyups.com/m7/2%20-%20BK600--two-apc-back-ups-600.html (http://www.mysynergyups.com/m7/2%20-%20BK600--two-apc-back-ups-600.html)
I bought a newer APC that was gray/black and rated for higher power. This unit developed a problem after several years and I junked it. I had some no name brand, space age name. It could not do the one thing it was sold for. Simple transients could cause a disruption in the output. After the batteries failed, I junked it.
-
I've added an SSR to my AC line power breakout box to do line dropout tests. This will give me an easy way to test the UPS.
I put together a small 70000uF capacitor bank w/ alum elec. The UPS has no problem charging them from a short. You can see the holdup time in the attached plots change.
Trace1: Vin to the UPS
Trace2: Current through the UPS
Trace3: SSR control Signal
Trace4: Voltage across load
I'll start but looking at a few low cost caps and see how they perform.
-
New channel in the making Joe ?
Watching with interest. ;)
-
im not sure if i interprete correctly, so the 0.7F cap allows a mitigation of around 0.2ms to 0.25ms? we are subtracting the difference from the "no battery"? but this is without load right?
-
There's not much to interpret. It will be a while before I can do anything meaningful with it.
New channel in the making Joe ?
Watching with interest. ;)
I wasn't planning on making a video about it but may post some data once things start coming together. It may be a week or so.
-
About half way in he talks about the parts being counterfeits.
https://www.youtube.com/watch?v=OglAziNcTxA&t=2s (https://www.youtube.com/watch?v=OglAziNcTxA&t=2s)
-
The results shown in the video in post #69 pretty much agrees with what I found with my limited testing. Of the 6 'Green-Cap' caps in the bank I bought, 2 were almost useless but 4 of them looked quite good. The video may be correct that these caps are rejects for any number of reasons and the sellers aren't too fussy about how good the caps are that they use. This would explain why I found 2 out of 6 to be defective.
I didn't check for capacity because the bank of 6 would give me far more power-up time than I would need but I would need them to hold their charge for a reasonable length of time. After I replace the 2 defective caps with good ones I plan to solder a resistor/LED across the equalizing resistors to give me a visual indication that each protective circuits is working properly.
I knew when I bought these cheap caps that there might be problems but I wanted to give them a try to see if I could use supercaps in this application. If you're looking to buy a bank of supercaps that you can just drop in a circuit and have them work with no problems, I wouldn't recommend these at all.
-
Some of those prices are so low, they just draw you in... :-DD
In the case of the UPS, it's always on so leakage shouldn't cause much of a problem out side the added heat.
-
Soldered up the bank and cycled it several times. Installed into the UPS. Showing the input voltage/current, trigger and output voltage with 300W resistive load.
105 is showing a 100ms dropout with the same load. Better than the 75000uF caps anyway.
-
One second dropout, 300W resistive load.
-
I am guessing with a 500W load, if it could ride out 5 seconds it would cover the vast majority of the dropouts we have. Most of the time, I am using about 200W and I would guess the dropouts are around a second. It's enough to have the lights flash and upset any tests I am running.
Looks like it can hold the 300W for roughly 17 seconds. In the ballpark anyway. After reading the datasheets, I had decided to run the caps well below their max operating voltage. I think the next step is just button the thing up and see what happens over the next 5 years. I will program the UPS to run a test on the bank every week or so.
-
Coincidentally I just had my Powerware 9120 announce battery failure a couple days ago with a "new" set of batteries off of Ebay that are about 2 months old. I am beginning to not trust UPS batteries bought off of Ebay.
-
2 months? And to think I was moaning over every 2-3 years.
-
2 months? And to think I was moaning over every 2-3 years.
There is nothing wrong with the UPS and this is the second set of Ebay batteries which had an unusually short operating life. They should last 5 years at least.
-
I'd recommend to get Panasonic SLAs or whatever you prefer from a reputable distributor.
-
I'd recommend to get Panasonic SLAs or whatever you prefer from a reputable distributor.
That is what I am concluding. I wish there was some way I could leave negative feedback on Ebay over the situation.
-
There is nothing wrong with the UPS and this is the second set of Ebay batteries which had an unusually short operating life.
I has what I felt was a low usage model from batteries purchased from BatteriesPlus locally.
It was so bad that I "gave up" having a ups on units.
Last year; I noticed Amazon had some really good deals vs local prices.
Amazon Prime had a set of two batteries delivered for under $70usd if I recall. BP would have wanted twice that.
-
2 months? And to think I was moaning over every 2-3 years.
There is nothing wrong with the UPS and this is the second set of Ebay batteries which had an unusually short operating life. They should last 5 years at least.
UPS optimized batteries are a special breed, but if you dig into the manufacturers datasheet, the advantage is clear. Compare the runtime charts of a UPS/extra high rate battery and a normal or deep cycle cell of the same dimensions, and at 10-30 minute discharge range under what a UPS operates, the UPS batteries come out on top by a significant margin. 25%+ more runtime on a 8AH ups battery vs a 9AH normal battery for when I was buying some, IIRC.
And that is before you count in the fact the Ebay battery is probably whatever trash battery comes in the right size case so they can call it compatible.
My replacement set of batteries for my APC1200VA UPS (the old beige tower style one) outlasted the UPS.
-
After another day of testing, it's back together and on-line.
-
Surprised that the UPS does not keep metrics on the number of brownouts, dropouts, etc. It keeps an error log of other faults. Shown is my quick and dirty dropout counter using one of those fitness recorders.
UPS has been running for 2-days now. Caps are cold. Looks as boring as it should be.
-
I have another thread going for a supercap board I'm designing for a boxing robot. For shits and giggles i took a 6 x 350F 2.7V array and connected it to an inverter and plugged a large portable AC to it and it powered it for about 3 to 4 seconds.
-
i was reading into the fine print of the EATON XB pdf. in the 500k cycles, if i calculate the cycles, the charge up and discharge @ 15A means that the capacitors are suppose to experience a power dissipation of only 0.06W in order to get to a meaningful lifespan of 500k cycles.
i also saw the same cycle specs in maxwells, these calculate out to about 0.1W.
both seem to imply that temp rise should not be over 1Celcius for max lifespan?
(there is also similar spec in Vishay 85C rated SC, but they limited charging to 3A)
which then seem to also suggest, the large number of base pins is likely for better heat dissipation/cooling?
jus wondering, what ballpark kind of temperature rise is yours in the UPS doing when it is cranking? (if you had a chance to measure?)
-
I am more concerned with the operating lifetime. Maxwell had a paper on calculating the life based on the ambient temperature and operating voltage. It looked like temperature dominates the equation.
I am not sure how we would know the temperature of a capacitor. They would need to publish theta of the core to the case, case to the sleeve or something like that. When I have had to measure transformers for example, I would embed a thermocouple into the windings. If we could sort it out, we may be able to determine the peak temperature and could get some idea of duty cycle based on how often the power drops. I suspect that the majority of the heat will come from the UPS and the actual room temperatures. The capacitors are mounted internal to the UPS. If you look at that first picture I posted in the last set, you know it's a double conversion type and I run the inverter 24/7. The UPS is fairly efficient and with the loads I normally run does not put out a lot of heat. Airflow is also decent. Inlet air temp to outlet is between 2.1 and 2.5 deg C rise.
-
It's been 25 days and so far, everything works fine. We had a few dropouts in the power since then. Once where I intentionally removed the power. Two were short drops like I normally see. During this time, we had power lines get damaged and lost power for a few hours where obviously the USP did not save me. So far, I like the concept. The caps continue to run cold.
-
Peal the sticker off of the battery and order the based on what is printed on the battery.
-
Two months into this little project. You can see we lost power again. This event was fairly long. I was in the lab when it happened. Everything stayed up but by the time the power did come back up the UPS decided the battery was too depleted and turned off. I changed the UPS to economy mode which will switch it back to mains rather than continue to run on the super caps. When I tested this mode, the system was a bit more tolerant of the line.
The caps themselves continue to run cold. It's a boring test.
-
The UPS has been running 24/7 for three months now. With the double conversion disabled, running in economy mode, I do hear it cycle over from time to time. Beyond that, there is nothing new to report. It's still running cool.
-
Had a bit of a storm come through today. Lost power a couple of times for a few seconds with each dropout. Had some stuff running at the time. The supercap mod held up with no trouble. Almost 4 months of run time on it now. Just not seeing the downside.
-
Usually I get a few power outages of a few minutes each year, mostly caused by thunderstorms in the summer. So I have to stay with the SLAs.
-
I suspect I jinxed us as another tree came down and took the lines with it. It seems we loose it for seconds or several hours. This one may take a few days. :-DD
-
A generator might be a good idea. ;) This reminds me to perform some preventive maintenance ... ::)
-
I've had a generator for about 30 years now. Currently I have two inverter type generators. Ones a small KW Honda that I really like. That generator has a lot of hours on it and still starts easy and runs smooth. Well worth the price.
About six years ago I picked up a Chinese knock off Honda. Before even starting it, I tore it all apart and changed a few things. Did a little break-in with it before putting it under load and have been using full synthetic oil since. Its about as loud as my little Honda but can handle about 3KW. I never expected it to survive very long but it's what is running the house now.
-
Where I am, about 30 minutes west of St. Louis, we lose power several times a year and at least twice a year for half a day or more so my 3000 watt inverter style Kipor generator gets a regular workout. Even with an inverter type of generator, my UPSes are required so if I bought another generator, it would be a non-inverter type but with pressurized lubrication.
-
We have had it go down for several days at a time. This one wasn't too bad. Just under three days and roughly 5 gallons of gasoline burned. I don't do much with my hobbies once the power goes down and I just left the modified UPS off the entire time.
-
Coming up on a half year of service. The more times it saves my ass, the more I like it. I've gone back to the double conversion and so far it has only dropped out that one time but I'm not sure if it would have stayed up even with it off. You gain a little but it's not double. The UPS continues to autotest the capacitors weekly but nothing becomes of it. The caps continue to run cold. I'm still not seeing the downside but the test has a long way to go.
-
Oh it'll last for a long time, don't worry. It just doesn't have much up-time! :)
Tim
-
I'm still not seeing the downside but the test has a long way to go.
For your specific use case where it's not about availability or graceful/safe shutdown it's probably the perfect solution. What you have is the electronic equivalent of the old motor/flywheel/generator power conditioner. Smooth power and the ability to ride out minor blips, plus isolating your load from the mains just long enough to bring up the generators.
-
Oh it'll last for a long time, don't worry. It just doesn't have much up-time! :)
Tim
When I started the thread, I was asking for experiences with some of these very cheap capacitors. While most of the discussions surrounded brands of UPSs and what batteries to buy, my concern is still the same, how will these cheap capacitors perform over time. I derated them a fair amount for voltage and capacity to try and give them every chance to have a long and useful life. We will see....
I'm still not seeing the downside but the test has a long way to go.
For your specific use case where it's not about availability or graceful/safe shutdown it's probably the perfect solution. What you have is the electronic equivalent of the old motor/flywheel/generator power conditioner. Smooth power and the ability to ride out minor blips, plus isolating your load from the mains just long enough to bring up the generators.
For the most part, that is correct. My goal was never to ride out more than five seconds or so.
I mentioned early on that the cost of the batteries vs these low cost, bottom end super caps was a wash. Someone mentioned size but as I mentioned, with the short durations I am looking at, it's a tight fit but the bank is small enough to fit into the UPS. Some were concerned over the recovery time (how long to recharge the bank after a dropout). So far this has not been an issue. Someone mentioned their concern of a fire hazard and the possibility of burning down our house. The case of the UPS is steel and the caps are well insulated. I guess a battery could also rupture and leak acid, maybe explode. I would guess that the case is designed to handle fault conditions like that but then again...
It does seem like a good application for them. If the system holds up for a few years, I may invest some time in making something a little higher quality than this unit.
-
It's been a little over 2 months from my last update. We only had one dropout that caused the counter to increment during this time. It was the typical secondish that we see. I had been running some long term test at the time it occurred. The UPS did it's job and rode it out.
I've left the double conversion active and havn't done anything with it since turning that function back on. The capacitors continue to run cold. I have some filter material in front of the unit that is starting to get a bit dirty but no other service beyond the UPU's built in battery test that I have programmed to run.
I've been thinking that after a year, I will pull the UPS and repeat my dropout test. It may provide some insight in how these low cost capacitors are aging.
-
We had another large storm which caused another tree to take down a line. The power company was fairly quick and the generator was only running for a day. Of course the UPS didn't save me with that one. The rest were short dropouts like we normally will see and the UPS held up just fine.
-
A few days ago I was logged onto EEVBLOG, posting some BS and the power dropped for several seconds. I thought for sure it was going to drop but it still rode it out just fine. No other power drops to report but it was a reminder that the project is coming up on a year of testing. The house has not burned down and the caps continue to run cold. It continues to be one of those boring projects that just does it's job.
On 6/28 it will mark one year and I will repeat the test below to see if I detect any changes in the hold up time. No idea what to expect.
I am guessing with a 500W load, if it could ride out 5 seconds it would cover the vast majority of the dropouts we have. Most of the time, I am using about 200W and I would guess the dropouts are around a second. It's enough to have the lights flash and upset any tests I am running.
Looks like it can hold the 300W for roughly 17 seconds. In the ballpark anyway. After reading the datasheets, I had decided to run the caps well below their max operating voltage. I think the next step is just button the thing up and see what happens over the next 5 years. I will program the UPS to run a test on the bank every week or so.
-
Just a couple of minutes ago, we had another dropout. This seems pretty close to the same time as the last event a couple of days ago. Which, come to think of it, the UPS used to frequently beep about an hour earlier (it beeps to indicate an event). We have this stupid daylight saving time that changes the time zones.
I wonder if they switch sections of the grid at certain times of the day and if there is something going on with some of their gear. Maybe next year, make up something better than the pedometer to record events which would log the date, time, type of event and length of dropout. Maybe there is more to this story than just using the supercaps to ride through these dropouts..
-
Has the ups got an interface? Can you monitor it with something line nut?
I monitor all mine with apcupsd. Really useful to get the emails and logs letting me know when, where and what the fault was.
-
NUT = Network UPS Tools -> https://networkupstools.org/
-
It has an interface and came with some software but I have never used it. As soon as I saw it was JAVA based, I discarded it.
So, I just checked and basically it looks like with a PC running 24/7 it could track the times when it switches to the battery bank. It has other metrics like the input/output line voltage, frequency, temperature...... Just basic stuff.
It can't seem to collect the metrics I would be looking for. Ideally, I would like something battery powered, that can record the date/time and length of each dropout, down to a 10th of a second.
-
It can't seem to collect the metrics I would be looking for. Ideally, I would like something battery powered, that can record the date/time and length of each dropout, down to a 10th of a second.
Old smartphone/laptop, Raspberry Pi, or whatever low power computing platform recording the mains with a sound card connected via an attenuator.
-
@joeqsmith
Just noticed thread. In response to first post. In a Davis weather station sending unit acquired 2005, that sits outside with a 3x4 inch solar cell and a supercap to power the rest of the day. Failed in 2009 when after about 5 heavy overcast days no signal. The leads were severely corroded. Replaced with 2 NessCap 2.7V, 10F from digikey. Lasted better than 10 years when some other chip failed from corrosion. The caps are still good looking for a project. They were discontinued at Digikey but are the same physical size as these 50F ones for the same price today: https://www.digikey.com/product-detail/en/nesscap-co-ltd/ESHSR-0050C0-002R7/589-1004-ND/946804 (https://www.digikey.com/product-detail/en/nesscap-co-ltd/ESHSR-0050C0-002R7/589-1004-ND/946804)
I highly recommend them.
Edited: Just read the whole thread. Here's some more data. The transmitter operates correctly when the supercars are as low as .31V. Typically around 2.6V, but never higher than 2.65V. Never measured the currents. At the time I replaced the original cap I didn't suspect electrolyte leakage so didn't check for it. The ~4x5 inch board is not conformally coated and gets lots of spiders and bug droppings and the 2019 failure was on the main transmitter chip which is only rated for 80% humidity! (This is up on a pole with the rain gage, temp and humidity sensor, hot days in the sun the caps must be over 90F 90% humidity) I replaced the original cap with two in parallel with no other circuitry.
To Davis Instruments credit the current boards ship with conformal coating.
-
Old smartphone/laptop, Raspberry Pi, or whatever low power computing platform recording the mains with a sound card connected via an attenuator.
I have yet to own a cell phone. :-DD If I roll something for it, I will just convert the AC to a digital signal and time the pulses. I could just use the relay that I have today that runs from the wallwart drive it.
@bicycleguy
A friend of mine gave me a Fluke 189 that still has the original super cap installed. I've had to replaced a few of them that started to show some oxidation and one that had really high leakage.
After writing this, I thought I've had that meter for a while now and had bought spare caps. Decided to pull it back apart and replace it while I was thinking about it. Noticed the batteries were low and sure enough. See attached. :-DD All four of my old Fluke 18x's now have new caps.
On the flip side, I had a tablet that the PCB was damaged beyond repair after the cap leaked after a a few years. I had two of these tablets and when I looked at the second one, the cap was just starting to go.
So depending on several factors, the life could be a couple years to well over 20.
-
It's been almost a year since I started this thread. Over the past year, the UPS has saved me roughly 20-30 times. We had 4 or so outages that lasted more than an hour. There was really only one time that it dropped out and had the holdup time been even 10 seconds longer, it would have rode it out. All in all, I have been pleased with the results.
As promised, I have repeated the 300W holdup test using a resistive load (incandescent lamps) and the UPS set to 110V 60Hz double conversion. If I reapply power right after the system shuts down, wait for the output to enable and immediately remove the power, the hold uptime is much shorter as expected. It takes a few minutes to recharge the bank. So far, this has not been a problem.
-
Sounds like tacking on a few more caps might be a worthwhile investment, but still quite reasonable as-is. :)
Tim
-
I checked the voltage balance and it doesn't see to have changed much from the original data I took. They are all still within 500mV. The parts all continue to run cold. I really can't tell how they have degraded yet.
As far as trying to increase the size of the bank. I may try and find a physically larger UPS now that I have a feel for the system. Most likely, I would pull the trigger on a name brand for the capacitors. In the meantime, I plan to put this one back online soon.
I had a look today and came across where Eaton and Riello both offer smaller units but they don't seem to offer a 120V single phase system. If you look at the SentinelProSC, they seem to fall right into the area I am interested in. The recharge times are much longer than I would have expected but the uptime looks good. I haven't been able to find any details on them.
https://www.ecopowersupplies.com/super-capacitor-ups (https://www.ecopowersupplies.com/super-capacitor-ups)
https://www.riello-ups.com/uploads/file/822/822/DATCAPX0Y18CREN.pdf (https://www.riello-ups.com/uploads/file/822/822/DATCAPX0Y18CREN.pdf)
-
Is the ProSC a super cap version ?
This all I could find but no 120VAC version:
https://www.riello-ups.com/uploads/file/784/1784/DATSEPA1Y18CREN.pdf (https://www.riello-ups.com/uploads/file/784/1784/DATSEPA1Y18CREN.pdf)
Our ISP put a 1KW version of these in here last week:
https://www.riello-ups.com/products/1-ups/45-sentinel-dual-low-power (https://www.riello-ups.com/products/1-ups/45-sentinel-dual-low-power)
We host some of his customers via point to point 5 GHz links and some are far enough away that they're not on the same powerco network and we might lose power and therefore drop their internet connection.
Anyways, it came in an always ON config with it's fan cooling on full blast so we stuck it in ECO mode where the fan is only running when it's actually doing something. :P
-
Is the ProSC a super cap version ?
It would seem so based on the couple of links I had found.
If the non-SC version you found is indeed the same except for the storage, I would try to find something else. I suspect the larger system you show is more like what I have now. It sounds like yours defaults with the converter enabled. The one I am using for this experiment will run the fan even with the converter off but it is running at the lower speed until the charger circuit becomes active. It works out well as the bank always has some airflow.
The ones with SCs and Lithium don't seem to be offered in the USA. I was surprised with people mentioning Riello that they don't offer 120V systems.
-
We are coming up on a year and a half of full time operation. Recently we lost power and I was reminded of this little project. It's been one of those projects that I seldom think about because it just works. The first drop was less than a second and of course the UPS rode it out just fine. However, shortly after it dropped again for several seconds (15ish). I was only running a light load on the UPS and it was holding up just fine. The power came back on and I was pleased that everything was still working just fine. It takes some time to recharge the bank but after a few seconds, it went down for a 3rd time and stayed down for several hours.
Over the summer, we had lost power several more times and I spoke with one of the crews doing the repairs. Since then, they have upgraded some of the cables which has improved the reliability. Of course, downed tress and people running their cars into power poles are a little more difficult problems to solve.
I'm still very pleased with the system, even with it's cheap capacitor bank.
-
Just a few more days and the UPS will have 2 years of 24/7 operation with the cheap caps. Still set for double conversion. I plan to take some measurements from it again at the end of the month. I no longer keep track of the number of line dropouts but it continues to be a problem and the UPS continues to ride out the vast majority of them with ease. Normally I would be changing batteries by now.
-
Those are some interesting results. I fixed a couple of UPS's recently that just needed new batteries and after looking at datasheets for 9Ah SLA batteries it occurred to me just how hard these things get hammered. In many cases the capacity printed on the batteries is for a 20 *hour* discharge rate and drops by around 10% for a 10 hour discharge with a maximum surge current of somewhere around 8 times the Ah rating. The last UPS I was messing with was a 1500VA/1kW unit with a 24V battery pack so even if we're very optimistic and assume 100% efficiency and no sag below the nominal battery voltage that's still over 40A at full load. At that current those 9Ah batteries are probably going to deliver more like 3Ah and go figure it's not going to take too many cycles to kill them, nevermind overly aggressive charging. Then there is the fact that a lot of times general purpose SLA batteries designed for alarm system and other low power backup applications get sold for UPS duty instead of batteries designed for high discharge rates.
LiFePO4 batteries seem like a perfect fit for UPS duty but even though a retrofit is possible I don't think I'd want to do that, you'd want a UPS designed from the ground up to use lithium cells. Given the developments that have taken place to enable practical electric vehicles I would expect a properly designed LiFe based UPS battery could last 10+ years, easily offsetting the substantially higher cost. By the time you factor in the much lighter weight and resulting lower shipping cost it gets even more competitive.
-
Sorry for the delay. After 2 years 3 months of continuous operation still no house fires to report, the capacitors continue to run cool and the balance still looks good.
Shown again using the 300W resistive load. The holdup time has dropped from 17 seconds to a little over 12. I allowed it to recharge and attempted to run it a second time. The holdup time remained the same. The cheap supercaps are starting to degrade but have plenty of life left for this application.
***
Wording.
-
Given the developments that have taken place to enable practical electric vehicles I would expect a properly designed LiFe based UPS battery could last 10+ years, easily offsetting the substantially higher cost. By the time you factor in the much lighter weight and resulting lower shipping cost it gets even more competitive.
Oddly enough if you properly manage and use SLA batteries with an appropriate design life you can get 10+ years out of them also.
-
It's been almost a year since I started this thread. Over the past year, the UPS has saved me roughly 20-30 times. We had 4 or so outages that lasted more than an hour.
Out of interest what part of the US do you live in? This hugely more events than I am used to (southern UK, suburban), I guess you have a lot more overhead distribution but it still seems like a lot.
-
Longer outages are from overhead lines and trees. As I have mentioned in this thread, these events are fairly rare. I have no idea what causes the short duration drops. These are by far the most common everywhere I have lived. There are some that appear to be timed but are not long enough to cause a problem.
It will be interesting to see how these degrade over the next year. Maybe they degrade more quickly with time. I wonder too if they are all degrading at the same rate.
-
I've just re-read the whole thread. Did you ever post details or pics on what you actually did? I can't seem to find any data on the conversion itself, just load / hold-up tests.
-
Only what you have seen posted in this thread. People seemed more interested in discussing UPSs and battery brands than discussing the actual project. Longer term, I plan to put something together or possibly buy one.
-
Only what you have seen posted in this thread. People seemed more interested in discussing UPSs and battery brands than discussing the actual project.
Unless I'm blind, you've shown some outcomes, but nothing about the unit itself. I can't even seem to find what UPS you used, let alone which caps or how they're configured. I do know it's double conversion but you've put it in conventional mode.
The outcomes look good, but I have no idea how you got them.
-
I do know it's double conversion but you've put it in conventional mode.
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2016721/#msg2016721 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2016721/#msg2016721)
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2207082/#msg2207082 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2207082/#msg2207082)
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2494173/#msg2494173 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2494173/#msg2494173)
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg3101123/#msg3101123 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg3101123/#msg3101123)
-
I do know it's double conversion but you've put it in conventional mode.
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2016721/#msg2016721 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2016721/#msg2016721)
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2207082/#msg2207082 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2207082/#msg2207082)
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2494173/#msg2494173 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg2494173/#msg2494173)
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg3101123/#msg3101123 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg3101123/#msg3101123)
Sorry, mis-read that then.
-
Longer outages are from overhead lines and trees. As I have mentioned in this thread, these events are fairly rare.
Four a year would be considered more than rare here. I have been in my current house 5 years and have had one. I probably had 5-10 in the house I lived in previously during the 25 years I was there. Short dips, enough to reset a PC or similar maybe 2-3 times per year.
-
.... let alone which caps or how they're configured.
I didn't see much of a point of posting which caps I procured as I doubt very much people would be able to buy the same ones. My guess is they are rejects from a larger company that are rebranded. I was going for cheap to get a feel what is the worse performance I could expect.
Looking at datasheets, obviously the rated voltage of an SC is much too low to be of use and they will need to be placed in series. I doubt that is what you are asking.
-
Pretty much what I am seeing, the leakage current continues to decrease along with the capacitance.
https://passive-components.eu/wp-content/uploads/2017/10/3.3.-BUT-Kuparowitz-Supercap.pdf (https://passive-components.eu/wp-content/uploads/2017/10/3.3.-BUT-Kuparowitz-Supercap.pdf)
https://www.osti.gov/etdeweb/servlets/purl/20671677 (https://www.osti.gov/etdeweb/servlets/purl/20671677)
-
We are quickly approaching three years of 24/7 operation. The UPS continues to provide more than enough uptime to ride out the short dropouts we see.
Once I have some time, I will pull it back apart and take some measurements.
-
The modified UPS has now been in operation for close to 4 years. It's been almost 2 years since I looked at the hold up time.
I checked the balance of the caps and nothing unusual appears to be happening.
The cheap supercaps continue to degrade. Using the 300W resistive load, the holdup time has dropped to about 12 seconds. Still plenty of life left and no house fire to report.
-
Thanks for the updates! I've retired my classic UPS after the SLA battery pack went bad because the mains became more stable over the last few years. Meanwhile you can also get a UPS with a Li-Ion battery pack. The claimed lifetime is about 7 years which is a bit more than the typical 5 years of standard SLAs, but still not as long as special SLAs optimized for UPS usage.
-
The amount the caps are degrading does not appear linear. We started at roughly 17 seconds holdup.
https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg1635044/#msg1635044 (https://www.eevblog.com/forum/projects/ups-supercapacitor-conversion/msg1635044/#msg1635044)
It's been over a year since the last time I checked it and we lost about a half second or so was all. It seems to be settling in.
-
Thank you for sharing your project and continued updates! I live in a rural area with plenty of power problems, and I also got sick of replacing lead batteries all the time. I switched a couple of our UPS's to a123 LiFePo batteries and these have held up quite well. With initial testing I used a small 8-cell pack to make coffee, and my espresso machine was pulling around 55A from the pack without issue. I got the cells used (I think they were made in 2009) and I've had them on the UPS for around 5 years now. They still have around 85% of the datasheet capacity. This pack is around 15AH with a 1200w space heater as a test load:
(https://i.postimg.cc/tgx9w2z0/APCa123.jpg)
I have two more smaller UPS's that now need replacement 12v batteries, and I'm looking at the capacitor route. I may have missed it, but what do you use for balancing the pack/array? Also, how much are you de-rating the capacitors? Have you tried higher loads on the UPS?
-
I started this experiment on 6-28-2018. Its been almost five years now and the house has yet to burn down as a result. I had some free time and decided to make some measurements a couple of months early.
...
I may have missed it, but what do you use for balancing the pack/array? Also, how much are you de-rating the capacitors? Have you tried higher loads on the UPS?
Sorry for the year delay but I don't check this thread often.
I did not care about efficiency and balanced them by measuring the leakage when they were new and increased it using a resistor in parallel with each. When I measured the voltage drop today they are all still well balanced.
I had measured the UPS charge and dropout voltage to determine the capacitance required to achieve the hold up time. It looks like I doubled the capacitance required.
I have ran it at higher loads but continue to collect the data a 300W to give some idea how the parts are degrading. I wasn't expecting the hold time to be as good as it is and had to repeat the test. Looks like about 11 seconds. Consider we started out with about 17 second holdup, we have lost a fair amount. Because you asked, I increased the load to a bit over 700W and the holdup time dropped to about 3.7 seconds. I let it recharge the bank and tried it again with roughly the same result.
As we still have these short dropouts, I plan to let it continue to run.