EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: josecoelho96 on June 16, 2016, 05:46:57 pm
-
Hi everybody,
I got a 750W (12VDC, 62.5A) power supply (I don't have it with me right now so I can't remember the model) from a old server and I was thinking on using it to create a lab bench power supply. My question is if it's worth it. I don't have any power supply and I can't afford spending much money on one. I'm a beginner in electronics so it would also be a project to get some experience from. I was thinking on using some components to have some fixed voltage levels and also a adjustable one. I would like if someone could give me some advices on this.
The components I was thinking of are:
3.3V - LM317 (1.5A max) and other regulator for more max current.
5V - LM7805 or LM317 (both 1.5A max)
12V - Directly from the power supply
Adjustable: LTC3780 (1-30V) (10A), using a display to show voltage and current and a few potentiometers.
Is this worth the money and time? Both for usability and experience.
Also, is it possible to have both positive and negative voltages?
Thanks in advance!
-
Hi everybody,
I got a 750W (12VDC, 62.5A) power supply (I don't have it with me right now so I can't remember the model) from a old server and I was thinking on using it to create a lab bench power supply. My question is if it's worth it. I don't have any power supply and I can't afford spending much money on one. I'm a beginner in electronics so it would also be a project to get some experience from. I was thinking on using some components to have some fixed voltage levels and also a adjustable one. I would like if someone could give me some advices on this.
The components I was thinking of are:
3.3V - LM317 (1.5A max) and other regulator for more max current.
5V - LM7805 or LM317 (both 1.5A max)
12V - Directly from the power supply
Adjustable: LTC3780 (1-30V) (10A), using a display to show voltage and current and a few potentiometers.
Is this worth the money and time? Both for usability and experience.
Also, is it possible to have both positive and negative voltages?
Thanks in advance!
Do you want current limiting? if no, A simple LM317 with external pass transistor on rather large heat sink will be good.
-
Do you want current limiting? if no, A simple LM317 with external pass transistor on rather large heat sink will be good.
That would be nice to have, but I can always start on a simple design and then upgrade it. My main focus right now is on converting the 12V to, at least, 5V and 3.3V.
-
Do you want current limiting? if no, A simple LM317 with external pass transistor on rather large heat sink will be good.
That would be nice to have, but I can always start on a simple design and then upgrade it. My main focus right now is on converting the 12V to, at least, 5V and 3.3V.
Nice to have but bitch to design.
Like I said before, I think LM317 with external pass transistor on a heat sink will work great.
-
Hi
Ok, the first part of this is always:
How much are you saving?
How useful is the result?
Is this a design exercise or a test gear project?
For a very modest amount of money, you can buy brand new "wall wart" power adapters. For a lot less than that, you can find them at the local flea market. Are they perfect / ideal / exciting ... nope. They likely are as good as a simple add on to the supply you have.
Please don't take this the wrong way. I am not saying you should just buy the wall warts. I am only pointing out that they form a low cost alternative. Spending a lot more money that they would cost does not make sense if this is a test gear project. Even as a learning / design project, there may be more interesting ways to invest your time and money.
=====
Just walked past the rack full of Dell servers (it's on the way to the beer ...). A couple of things I noticed:
1) The fans in the server also put air through the power supply. There is a fan in the supply, but it does not do the whole job.
2) The firmware that controls the supply is on the motherboard. You can get the supply to power up (otherwise no boot). All the smarts are elsewhere.
3) That's one really weird connector on those supplies ...
Bob
-
I have a Dell PSU that I modified. In my case the PSU was from one of their R200 servers so has a standard motherboard power connector. The PSU does not have -5v or -12v outputs so the connector has 12v wired to both of those instead!
I only use it when I want to take a large current for a long amount of time. Otherwise I just use a 12V battery.
From either of the 12V sources I feed it into a DC-DC converter ( http://www.ebay.co.uk/itm/Mini-DC-DC-Converter-Digital-Step-Up-Step-Down-Boost-Buck-Module-38V-6A-UK-W1U0-/191859761767 (http://www.ebay.co.uk/itm/Mini-DC-DC-Converter-Digital-Step-Up-Step-Down-Boost-Buck-Module-38V-6A-UK-W1U0-/191859761767) ).
-
Nice to have but bitch to design.
Like I said before, I think LM317 with external pass transistor on a heat sink will work great.
Thanks for your advice, I will search about external pass transistor circuits in order to get a idea of what that is.
Hi
Ok, the first part of this is always:
How much are you saving?
How useful is the result?
Is this a design exercise or a test gear project?
For a very modest amount of money, you can buy brand new "wall wart" power adapters. For a lot less than that, you can find them at the local flea market. Are they perfect / ideal / exciting ... nope. They likely are as good as a simple add on to the supply you have.
Please don't take this the wrong way. I am not saying you should just buy the wall warts. I am only pointing out that they form a low cost alternative. Spending a lot more money that they would cost does not make sense if this is a test gear project. Even as a learning / design project, there may be more interesting ways to invest your time and money.
=====
Just walked past the rack full of Dell servers (it's on the way to the beer ...). A couple of things I noticed:
1) The fans in the server also put air through the power supply. There is a fan in the supply, but it does not do the whole job.
2) The firmware that controls the supply is on the motherboard. You can get the supply to power up (otherwise no boot). All the smarts are elsewhere.
3) That's one really weird connector on those supplies ...
Bob
I don't have a defined amount of money for this,but arround 25€ would be the max I'm affording on this right now.
The result is to have a power supply to all my electronics projects, both college and personal.
It was just a idea I had since I got the power supply.
I am just trying to get some experience doing some projects and I got this idea since I don't have any power supply and I thought it would be a good idea to have one. I have one of those "wall wart" already.
I already made it work and yes, I repaired that with some load the fan on the power supply might not be enough.
I have a Dell PSU that I modified. In my case the PSU was from one of their R200 servers so has a standard motherboard power connector. The PSU does not have -5v or -12v outputs so the connector has 12v wired to both of those instead!
I only use it when I want to take a large current for a long amount of time. Otherwise I just use a 12V battery.
From either of the 12V sources I feed it into a DC-DC converter ( http://www.ebay.co.uk/itm/Mini-DC-DC-Converter-Digital-Step-Up-Step-Down-Boost-Buck-Module-38V-6A-UK-W1U0-/191859761767 (http://www.ebay.co.uk/itm/Mini-DC-DC-Converter-Digital-Step-Up-Step-Down-Boost-Buck-Module-38V-6A-UK-W1U0-/191859761767) ).
I also got some DC-DC converter as a circuit I would add to output of the power supply.
-
Personally, I think it's not worth the bother, especially if you can lay your hands on a desktop PC power supply which already have those rails avaialble.
I would put some effort into adding something like a Cuk converter to a ~50V blade server power supply though to make it variable at high current and the topology of a Cuk means it's able to boost as well so that may be worth investigation for the 12V supplies.
-
Personally, I think it's not worth the bother, especially if you can lay your hands on a desktop PC power supply which already have those rails avaialble.
I would put some effort into adding something like a Cuk converter to a ~50V blade server power supply though to make it variable at high current and the topology of a Cuk means it's able to boost as well so that may be worth investigation for the 12V supplies.
Hi
Once I have scrapped out the server, there are all these supplies lying around. I keep them because there is *always* one more to pull out of the rack and I'm to lazy to port all the stuff on it over to the new one. Net result is that in a few years, the supplies are indeed free. Actually they have negative value since I may have to pay to recycle them.
So yes, I can understand why people do this sort of thing. The beast is sitting there doing absolutely nothing of value. Even if you didn't pull it from the server yourself, you picked it up at the recycle for nothing (or some guy named Bob gave it to you).
Free is free (sort of).
Bob
-
Personally, I think it's not worth the bother, especially if you can lay your hands on a desktop PC power supply which already have those rails avaialble.
I would put some effort into adding something like a Cuk converter to a ~50V blade server power supply though to make it variable at high current and the topology of a Cuk means it's able to boost as well so that may be worth investigation for the 12V supplies.
Hi
Once I have scrapped out the server, there are all these supplies lying around. I keep them because there is *always* one more to pull out of the rack and I'm to lazy to port all the stuff on it over to the new one. Net result is that in a few years, the supplies are indeed free. Actually they have negative value since I may have to pay to recycle them.
So yes, I can understand why people do this sort of thing. The beast is sitting there doing absolutely nothing of value. Even if you didn't pull it from the server yourself, you picked it up at the recycle for nothing (or some guy named Bob gave it to you).
Free is free (sort of).
Bob
Completely understand that urge, I have a shed full of 'free' and tinker often, the urge to use it is only tempered by the lack of spare time I have.
Fascinates me how people will find ways to re-use defunct equipment and the innovation some display is incredible.
My supply of those supplies dried up a few years ago when I moved away from server work, I probably have some hiding somewhere that I would re-purpose if I needed high quality high current power.
Just seems an ATX would be a much simpler way to do it and probably be better as the rails will supply more current.
But, a good linear reg will clean them up nicely if done right.
-
I don't have a defined amount of money for this,but arround 25€ would be the max I'm affording on this right now.
The result is to have a power supply to all my electronics projects, both college and personal.
It was just a idea I had since I got the power supply.
I am just trying to get some experience doing some projects and I got this idea since I don't have any power supply and I thought it would be a good idea to have one. I have one of those "wall wart" already.
I already made it work and yes, I repaired that with some load the fan on the power supply might not be enough.
The problem with constructing a power supply for testing purposes is that it's something that will become the foundation for other tests. There could be a deviation in accuracy or that only occurs under certain circumstances and is difficult to detect, which would mess up tests you'll perform in the future (unless you keep a multimeter permanently hooked up to the output).
For digital projects, a stable linearly-regulated 5V and 3.3V is good enough, and cheap. The ability to adjust the amperage limit is useful, but even a known, fixed limit (say 1A) is fine.
But when you start getting into analog and power circuits the ability to make accurate, fine voltage and current adjustments becomes important, especially if you're relying on the power supply to report the current output accurately on a readout.
There's a relatively cheap Gopher power supply that was reviewed on Voltlog a year ago, and its updated version was reviewed a few weeks ago:
https://www.youtube.com/watch?v=t49P2nh9_t0 (https://www.youtube.com/watch?v=t49P2nh9_t0)
https://www.youtube.com/watch?v=yIx_VWWgKUI (https://www.youtube.com/watch?v=yIx_VWWgKUI)
I have the older one, and it's fine as long as you're aware of the potential problems which he covers in the review.
There are several versions of them. I have a 32V/5A version, but there's also a 60V/3A version and a 16V/20A version.
The new one's around 90 euros:
http://www.ebay.com/sch/items/?_nkw=cps-3205c&_sacat=&_ex_kw=&_mPrRngCbx=1&_udlo=&_udhi=&_sop=12&_fpos=&_fspt=1&_sadis=&LH_CAds=&rmvSB=true (http://www.ebay.com/sch/items/?_nkw=cps-3205c&_sacat=&_ex_kw=&_mPrRngCbx=1&_udlo=&_udhi=&_sop=12&_fpos=&_fspt=1&_sadis=&LH_CAds=&rmvSB=true)
You can find one of the old ones for around 55 euros:
http://www.ebay.com/itm/CPS-3205-0-32V-0-5A-Portable-Adjustable-DC-Power-Supply-110V-220V-Test-Equipment-/131575925163?hash=item1ea28949ab:g:DukAAOSwu4BVyXAa (http://www.ebay.com/itm/CPS-3205-0-32V-0-5A-Portable-Adjustable-DC-Power-Supply-110V-220V-Test-Equipment-/131575925163?hash=item1ea28949ab:g:DukAAOSwu4BVyXAa)
The internal construction is very good for the price. It's not one of the chinese bench power supplies which you open up and witness a disaster site.
This is more than you intended to spend, but having an vulnerable/inaccurate power supply could lead to hair-pulling frustration.
If you don't have the money to spend and you won't need much power, use linear regulation (the old LM78xx and LM317). It's cheap, accurate, low-noise, and requires fewer components.
-
I had plans for the longest time to make my own power supply, but I just never got around to it. I did have a velleman 3 in 1 lab unit with a (crappy) power supply built in, however, it was not fully adjustable, and was only able to output 3, 4.5, 6, 7.5, 9, and 12V. It had an overload light but no current limiting. I was able to get 24V out if you carefully adjust the knob between the 8V and 12V setting. However I had to repair it 3 times because the regulator kept blowing up I did reverse engineer the entire circuit and learn how it works. The overload light was made by simply having a lowish value shunt resistor in series with the the regulator input and the unregulated input, and tacking in an LED/resistor in series with it. When the shunt resistor had over 2V across it the overload LED would light. The switch had lots of 1% metal film resistors for setting the voltage.
So really I am not a fan of the LM317T, they will tolerate some abuse but not much, and not for very long. You cannot easily adjust the voltage and current limit easily, (if using a potentiometer, I beleive the voltage setting would be nonlinear) so you have to calculate the values, and you cannot set it to 0V. (1.25V minimum)
There are actually many of those cheap switch mode regulators on amazon that have a display and let you set voltage and current limit. Those are pretty nice, although because of the output capacitors they do not have a true constant current output (capacitors need to dump their charge before the power supply can limit current by dropping voltage)
-
I had plans for the longest time to make my own power supply, but I just never got around to it. I did have a velleman 3 in 1 lab unit with a (crappy) power supply built in, however, it was not fully adjustable, and was only able to output 3, 4.5, 6, 7.5, 9, and 12V. It had an overload light but no current limiting. I was able to get 24V out if you carefully adjust the knob between the 8V and 12V setting. However I had to repair it 3 times because the regulator kept blowing up I did reverse engineer the entire circuit and learn how it works. The overload light was made by simply having a lowish value shunt resistor in series with the the regulator input and the unregulated input, and tacking in an LED/resistor in series with it. When the shunt resistor had over 2V across it the overload LED would light. The switch had lots of 1% metal film resistors for setting the voltage.
So really I am not a fan of the LM317T, they will tolerate some abuse but not much, and not for very long. You cannot easily adjust the voltage and current limit easily, (if using a potentiometer, I beleive the voltage setting would be nonlinear) so you have to calculate the values, and you cannot set it to 0V. (1.25V minimum)
There are actually many of those cheap switch mode regulators on amazon that have a display and let you set voltage and current limit. Those are pretty nice, although because of the output capacitors they do not have a true constant current output (capacitors need to dump their charge before the power supply can limit current by dropping voltage)
Hi
If you spend some "quality time" shopping on eBay, you can find some very good analog meter supplies for cheap prices.
Bob
-
I went with the 0-30V eBay kit.
http://www.ebay.com.au/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.X0-30V+power+supply+kit.TRS0&_nkw=0-30V+power+supply+kit&_sacat=0 (http://www.ebay.com.au/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.X0-30V+power+supply+kit.TRS0&_nkw=0-30V+power+supply+kit&_sacat=0)
Gutted a ATX PSU except for the fan and input filter, found a old 25V transformer at the junk sale for $2, got 10 turn pot, dual V&A LED panel meter, and terminal fittings. Cost about $35AUD and it has current limiting adjustability.
It's not serious, but I only need it for small automotive/Arduino stuff.
-
I went with the 0-30V eBay kit.
http://www.ebay.com.au/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.X0-30V+power+supply+kit.TRS0&_nkw=0-30V+power+supply+kit&_sacat=0 (http://www.ebay.com.au/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.X0-30V+power+supply+kit.TRS0&_nkw=0-30V+power+supply+kit&_sacat=0)
That design is infamous on Eevblog and other EE forums, it turns up like a bad penny on a regular basis. It can be made to work reasonably well with but has many design flaws.