Power Supply Shopping, Round 2

[Electro Fan]

Ok, I’m loving the whole EE/EEVblog thing.  I’ve been learning a lot and it just keeps getting more fun and more addicting.  Thanks (sincerely) to you guys and your posts for fueling the education and addiction. 

Unfortunately (or maybe fortunately) with Marc’s Korad post I think I might have just hit a power supply speed bump.  I have a Korad 3005P and it’s been great (knock on wood); however after seeing his experience and reading Bored@Work’s diagnosis, maybe I’ve been lucky (knock on wood some more).

I find myself wiring up lots of stuff with the power supply driving the circuitry and I’d like to avoid doing something that takes out my DUT or maybe my other test gear.

So - I’m thinking about moving up the power supply food chain.  It’s not like it takes a lot to come up with a reason to buy another piece of test equipment but this one seems like a prudent move.  I have a bunch of projects and some more gear in mind and I don’t want to vaporize more than an occasional small resistor.

Here are some ideas/questions – any advice is welcome, especially if it helps justify a new or used Agilent power supply:

1.   I’ve found myself cranking up the Korad to over 10 volts, and occasionally over 20 volts all the way to 30.  Often it’s just to help me clearly see some Ohm’s Law tradeoffs (and in the process I’ve learned to be cognizant of watts as well as volts, amps, and ohms).  It’s not like I couldn’t live with less but it would be nice to stay in the 30V 5 amp range (or at least 15-20 volts and 2-3 amps). It seems like Agilent is very financially proud of every volt and amp supported, so maybe for now I could get by with less but I don’t know what I’m going to wish I could do in the future.  Based on your experiences, any suggestions on the sweet spot in terms of volts and amps?
2.   Mostly the single channel has been sufficient but lately I find myself powering stuff with the Korad and then looking for 9V batteries to try other stuff at the same time.  Maybe it’s time to go 2 or 3 channels?  Or maybe stick with 1 channel for the next PS, and save 2-3 channels for unit #3 when I discover still more features I "need".
3.   I’m starting to get to the point where “almost” measurements aren’t real satisfying.  I’m thinking that remote sense might be a good feature.  How often do you guys use remote sense?
4.   I like having the 4 memories on the Korad and it would be nice to have programmability to turn voltage up and down at various intervals, probably with a PC interface.
5.   What's the typical remaining life expectancy of a used Agilent from a recent generation?  Seems like the E3640A, E2642A, E3645A, etc. generation might be a good deal if you could find one used in nice condition.  If you have a favorite Agilent I’m all ears.
6.   What’s up with the new tall Agilents (E3632A, etc.)  I thought technology was making stuff smaller, not larger?  What are they putting inside these units that they left out of the older units?  Any compelling reasons to avoid a used Agilent and go for a new one?
7.   Price: If it turns out that one of the new Agilents is cooler than cool maybe I could get my head around $1k - $1500, but I probably should have my head examined.  Maybe it would be possible and smarter to find a good used Agilent in the $350 - $700 range?  (And save the $ for a Rigol or Agilent scope  )

Any other ideas/suggestions (including manufacturers other than Agilent) are all welcome.

Thanks, EF 

[robrenz]

Please fill in your country in your personal profile.  No point in suggesting stuff you cant afford the shipping on.

[ovnr]

Well, I was in your situation a couple years ago (had a single 30V 3A powersupply, but with two fixed auxiliary outputs of 5V and 12V).

I ended up purchasing a TTi CPX400A (2x420W 60V/20A (powerflex, so 60V @ 7A / 20V @ 20A, etc)) PSU, and I'm not going to recommend it to you. 
Don't get me wrong, it works splendidly, but I practically never load it enough for it to be worth it - the only reasonably large load I subject it to regularly is around 150W. It also has rather slow constant-current limiting, which has bitten me in the ass one time too many. On the other hand, it'll weld things pretty darn well! 

It also has remote sense, which I've never ever used (on account of being lazy, and not trusting the readouts anyway - even though they've turned out to be pretty darn accurate), answering that question too.


As for recommendations - I'm going to buy a Tek (née Keithley) PWS4205 20V 5A programmable PSU when I get around to it, for these reasons:

* I'm tired of having to hook up two multimeters to get accurate voltage/current source readings.
* I'm tired of potentiometers drifting overnight and having to twiddle them. Or basically having to twiddle pots at all.
* I'd like to set up an automated test bench to characterize some projects, and doing it manually is such a pain. This way I can have a PC deal with adjusting and measuring the parameters.
(Oh, and there are other ranges too, up to 72V @ 1.2A and 30A @ 5A)

It's $900 or so (well, more like $1200 with VAT here in Norway). The CPX400 was $1900 when I bought it, and will fill my role of "excessively big beefy PSU" - hence why I can get away with a single-output 20V 5A PSU for everyday stuff.


I'm sure you'd be happier with a simpler multi-output powersupply, but here are some things to consider:
* Some times it's better to have several powersupplies as opposed to a single giant 4-output box.
* There are other brands with Agilent-like performance - Tek/Keithley (same thing these days), TTi, and the like.
   TTi has these models you may like: EX-R series mixed-mode supplies, EL-R series fully linear supplies.


Good luck!


-

[Electro Fan]

Please fill in your country in your personal profile.  No point in suggesting stuff you cant afford the shipping on.

Thanks.  I'm in the U.S.  Once I understand the desirable criteria for a PS I can track down a source that is cost-effective including shipping.

[Electro Fan]

Well, I was in your situation a couple years ago (had a single 30V 3A powersupply, but with two fixed auxiliary outputs of 5V and 12V).

I ended up purchasing a TTi CPX400A (2x420W 60V/20A (powerflex, so 60V @ 7A / 20V @ 20A, etc)) PSU, and I'm not going to recommend it to you. 
Don't get me wrong, it works splendidly, but I practically never load it enough for it to be worth it - the only reasonably large load I subject it to regularly is around 150W. It also has rather slow constant-current limiting, which has bitten me in the ass one time too many. On the other hand, it'll weld things pretty darn well! 

It also has remote sense, which I've never ever used (on account of being lazy, and not trusting the readouts anyway - even though they've turned out to be pretty darn accurate), answering that question too.


As for recommendations - I'm going to buy a Tek (née Keithley) PWS4205 20V 5A programmable PSU when I get around to it, for these reasons:

* I'm tired of having to hook up two multimeters to get accurate voltage/current source readings.
* I'm tired of potentiometers drifting overnight and having to twiddle them. Or basically having to twiddle pots at all.
* I'd like to set up an automated test bench to characterize some projects, and doing it manually is such a pain. This way I can have a PC deal with adjusting and measuring the parameters.
(Oh, and there are other ranges too, up to 72V @ 1.2A and 30A @ 5A)

It's $900 or so (well, more like $1200 with VAT here in Norway). The CPX400 was $1900 when I bought it, and will fill my role of "excessively big beefy PSU" - hence why I can get away with a single-output 20V 5A PSU for everyday stuff.


I'm sure you'd be happier with a simpler multi-output powersupply, but here are some things to consider:
* Some times it's better to have several powersupplies as opposed to a single giant 4-output box.
* There are other brands with Agilent-like performance - Tek/Keithley (same thing these days), TTi, and the like.
   TTi has these models you may like: EX-R series mixed-mode supplies, EL-R series fully linear supplies.


Good luck!


-

Thanks, the TTi has the remote sense terminals on the front; makes sense to me (if you want remote sense).  I'm sort of surprised you don't use them much given that you found them to be accurate - but maybe it's a hassle to wire them up except for when you really need them.

I looked at your Tek/Keithley.  The PWS4305 (30V/50A) looks a lot like what I think I'd like - seems like a respectable Agilent contender.  Thanks for the tip.  EF

[robrenz]

Check out the Power designs power supplies on ebay.  They were good enough for Jim Williams.

Here is an example

[ovnr]

Thanks, the TTi has the remote sense terminals on the front; makes sense to me (if you want remote sense).  I'm sort of surprised you don't use them much given that you found them to be accurate - but maybe it's a hassle to wire them up except for when you really need them.

Well, in my case I'm either powering something where I don't care about getting the exactly correct voltage, or it's low power enough for any cable losses to be negligible. And I just can't bring myself to actually trust these big bright LED displays; it just seems so... iffy.

Plus my desk is cluttered enough without having sense wires running across it too!


-

[WBB]

Check out the Power designs power supplies on ebay.  They were good enough for Jim Williams.

Here is an example

+1

Their line if power supplies are excellent. They lack some of the more modern features, but they generally work when you press the button. That seems to be becoming an optional extra these days. For the money they're hard to beat. Their precision models are hard to beat for any price.

[Corporate666]

In my opinion, one would be *VERY* hard pressed to find something better than the Rigol DP832 in terms of bang-for-the-buck. 

Under $400 shipped from TE (with EEVBlog discount code), three channel, digital display, excellent precision and resolution, LAN/USB ports, 30V/3A on two channels 5V/3A on the third.  0.02% voltage accuracy on the 30V channels, 0.05% current accuracy. 

For under $400 new.

It's the bargain of the century, IMO, especially considering the prices of competing Agilent supplies.

[Electro Fan]

they generally work when you press the button. That seems to be becoming an optional extra these days.

LOL, that does seems to be the main feature motivating this thread

[Electro Fan]

In my opinion, one would be *VERY* hard pressed to find something better than the Rigol DP832 in terms of bang-for-the-buck. 

Under $400 shipped from TE (with EEVBlog discount code), three channel, digital display, excellent precision and resolution, LAN/USB ports, 30V/3A on two channels 5V/3A on the third.  0.02% voltage accuracy on the 30V channels, 0.05% current accuracy. 

For under $400 new.

It's the bargain of the century, IMO, especially considering the prices of competing Agilent supplies.

Corporate - good suggestion; does seem to be very feature rich plus a good price.  Thx


- Funny what shapes thinking on a piece of equipment; my 2nd reaction after looking at the specs was "but do I want a Halloween (orange and black theme) on the dispay?"; turns out Rigol gives 4 monochrome choices.

[ovnr]

Re: the DP832: Ah, but I'll gladly pay some extra to get something I wouldn't be embarrassed to have sit on my bench. Because the DP832 is so horribly ugly, it's not even funny.

[Zbig]

My Siglent SPD3303S has just arrived a minute ago. I'm going to perform some basic tests and will report on my findings.

[bingo600]

Re: the DP832: Ah, but I'll gladly pay some extra to get something I wouldn't be embarrassed to have sit on my bench. Because the DP832 is so horribly ugly, it's not even funny.

Like a HP/Agilent 6632B 0..20v/5A and afaik ... Same or even better specs than your mentioned Tek/k
http://www.ebay.co.uk/itm/AGILENT-HP-6632B-Bench-DC-Power-Supply-0-20VOLT-5-AMP-Programmable-/350838184127

Datasheet
http://s3-eu-west-1.amazonaws.com/content.microlease.com/static/Catalogue/Datasheets/Agilent_6632B.pdf

I have 2 of those , and i absolutely love them ... (They have some fan noise)  , but are more accurate than most of your DMM's.

KJDS is a EEVBLOG user , and might offer a discount (for non *Bay sales) .. And pls. do a ask for a "realistic" shipping price.

Ohh and as a bonus you can use it as a "load also" , same specs as power

/Bingo

[Electro Fan]

Re: the DP832: Ah, but I'll gladly pay some extra to get something I wouldn't be embarrassed to have sit on my bench. Because the DP832 is so horribly ugly, it's not even funny.

No doubt, test equipment should not only have a good UI but ideally the UI and the overall device should be attractive as well as effective.  I'm not sure how effective the UI is on the DP832 but the unit could probably be more attractive.

If you wanted a more "attractive" PS, what would be the next unit up in price that would not only be "attractive" but would also have most of the same functionality as the DP832.   (ie, the DP832 looks like it has reasonable specs and great functionality, so what is the next unit up the food chain that can compete strongly on functionality regardless of price)?

[Corporate666]

I just got my DP832 today - I love it!

I don't see what is unattractive about it?  The numbers around the dial?  That's the only thing.  All the rest of it looks like any other PSU, except maybe nicer.  It's a hell of a lot nicer looking than my Fluke PM2811, and much easier to use with more functionality to boot.

[Zbig]

I've done some basic tests of my new Siglent SPD3303S and I'm happy with it so far. The first thing I checked after reading the horror story of Marc's Korad is the behavior during "hard" power on. While I haven't noticed anything like that on two its main outputs, there is some voltage spike on the third, fixed 2.5/3.3/5V output during powering the unit on, regardless of the fact that the "soft" switch for that output is off. It's nothing nearly as brutal as in the Marc's Korad's few second-long 30V surge, though. It's more transient-like, a short ~8-10V spike and then the voltage ramps down back to zero in some ~300 ms. Not a show stopper, but something to keep in mind, I guess. One important difference comparing to Rigol would be that the third channel is not as fully featured as the main two: apart from it having only three fixed voltage settings, there's no current reading, only the hard limit which is supposed to switch it into CC mode upon reaching 3A. Other than that, I'm pretty happy just about everything else. Here's just a few random things I found nice about it:

• Clean, functional, uncluttered design (I'm looking at you, Rigol ;-))
• Nice LCD, with C/V graph plotting feature
• Parallel/Series tracking mode does the connections internally with relays (no need for a shorting bar)

I haven't opened it (I'd void the warranty this way) but I took a few good peeks through the venting holes and the build quality seems good to me. Nice, clean front controller PCB with ST ARM chip, lots of AD chips, some Rubycon caps. Big toroidal transformer and "fin-to-fin" double heatsink setup on the output stage. The fan is temperature-controlled and in my personal opinion the noise level ranges from barely noticeable at idle to quite loud under serious load but then again, they had do dissipate all that heat somehow. If anyone is curious about some other aspect of this fella, I'll try my best to provide some more details.

[LaurenceW]

Hi, "EF"!

I think we need to put the horse back in front of the cart, before we go any further. What power requirements do the projects that you have in mind (and those that might cross your mind in the next few years!) have?

Probably 30V and maybe 3A will cover most requirements. A power supply with TWO such channels will allow you combinations of +/- 30V at 3A, 30V at 6A (parallel the two supplies), or 60V at 3A (two supplies in series). Unless you have any high power aspirations, that's going to cover most hobby electronics, I'd suggest.

You've talked of precision - cheaper supplies will display volts to only 1dp (i.e 100mV resolution) and 2 with current (i.e 10mA resolution). Better ones will offer 2 and 3 decimal places respectively, so you are down to 10 mV and 1 mA display resolution (which, admittedly, is NOT the same as accuracy!). For a POWER supply, I'd suggest three decimals is very nice and handy to have.

Remote sensing? You'll only ever need to use such a facility if
1. you are powering something at a significant (>2A?) current
2. your interconnect leads are either long, or thin, or worse, both!
3. your load is really fussy about a "sagging" power supply (which suggests a design issue with your load, straight away)

I have remote sensing on my main bench supply, but have yet found the need to use it! Appropriately dimensioned interconnects answers most issues, here, IMHO.

Some cheaper power supplies - especially fancy digitally controlled  ones - do VERY RUDE THINGS when you turn them on. Enough voltage spikes to fry your sensitve load, in some cases. These power supplies are clearly a FAIL , and are to be avoided at all costs.

Well, You've seen my own analogue TTI tripple channel power supply at work, first hand! I've owned it for two years, and do not regret the decision to buy. They are built here in Blighty, so not too sure how many of them will have found their way across the pond. eBay is always a good source...

I admit that some modern power supplies now give the added function of being able to set voltage and current limits digitally. Nice, but not essential (to me). The Rigol unit seems very well spec'd for the money (if you can get your head around the fact that the 12-year-old who designed the front panel didn't attend school the day they taught man-machine interfaces and ergonomic keyboard layout - EURGH!), and meets the above requirements (of mine!). There are some more professional looking, even more capable (more than necessary??) supplies from Rigol, but they cost a little more, of course.

You will know what people (IBM people??  ) used to say about buying IBM!. "Nobody ever got fired for buying IBM..." Well (probably) nobody (with the budget) ever got fired for buying Aglient, either. Of course, their gear will be beyond reproach, but boy, do you pay for that.  Some EVBLOGGER will now tell us that they knew a bloke whose friend heard of somebody who bought an HP/Agilent power supply from a mate once, and it blew up. Well, we can put such stories some way further down the list than exploding Korads...

Based on what you and I have spoken of, before, an Agilent supply (of appropriate voltage and current specs) is going to more than meet your needs, EF, I would suggest. And that's no bad thing, but you are paying for it.

Siglent gets a mention in this thread - they tend to be towards the budget end of electronics equipment, but possibly (I just don't know) fine for hobbiest use. But you have to ask yourself how well any product that you couldn't fix yourself is going to be supported in your country.  To quote Siglent's own web site:

We believe when SPD3303 on sale, instruments market will meet a power storm, it worths the wait!

A "power storm", eh? I think I can imagine what any manual (which may or may not be supplied) will read like

[Zbig]

[..]
Siglent gets a mention in this thread - they tend to be towards the budget end of electronics equipment, but possibly (I just don't know) fine for hobbiest use. But you have to ask yourself how well any product that you couldn't fix yourself is going to be supported in your country.  To quote Siglent's own web site:

We believe when SPD3303 on sale, instruments market will meet a power storm, it worths the wait!

A "power storm", eh? I think I can imagine what any manual (which may or may not be supplied) will read like

Wait, it gets worse. When you try to adjust the voltage or current outside the limit, the message says "Voltage/Current setting overspray!" But broken English aside, I'm really happy with the unit so far.

[LaurenceW]

That'll be overspray from the power storm, I expect

[echen1024]

[..]
Siglent gets a mention in this thread - they tend to be towards the budget end of electronics equipment, but possibly (I just don't know) fine for hobbiest use. But you have to ask yourself how well any product that you couldn't fix yourself is going to be supported in your country.  To quote Siglent's own web site:

We believe when SPD3303 on sale, instruments market will meet a power storm, it worths the wait!

A "power storm", eh? I think I can imagine what any manual (which may or may not be supplied) will read like

Wait, it gets worse. When you try to adjust the voltage or current outside the limit, the message says "Voltage/Current setting overspray!" But broken English aside, I'm really happy with the unit so far.

It's not even as bad as the Technologytronix oscilloscopes I saw in China. From "Horizon Knob" to things such as "Range is adjustment out," it was a mess.

[Electro Fan]

Hi, "EF"!

I think we need to put the horse back in front of the cart, before we go any further. What power requirements do the projects that you have in mind (and those that might cross your mind in the next few years!) have?

Probably 30V and maybe 3A will cover most requirements. A power supply with TWO such channels will allow you combinations of +/- 30V at 3A, 30V at 6A (parallel the two supplies), or 60V at 3A (two supplies in series). Unless you have any high power aspirations, that's going to cover most hobby electronics, I'd suggest.

- Hi Laurence, as always, thanks very much for taking the time.  Your info and guidance is much appreciated.

Regarding amps, I've been happy with my 30 volt 5 amp PS and I suspect I could get by on 3amps (or 2x3 easily) or for a long time.  Your answer helps confirm what I suspected 3-5 amps or so should be sufficient, Thx.

You've talked of precision - cheaper supplies will display volts to only 1dp (i.e 100mV resolution) and 2 with current (i.e 10mA resolution). Better ones will offer 2 and 3 decimal places respectively, so you are down to 10 mV and 1 mA display resolution (which, admittedly, is NOT the same as accuracy!). For a POWER supply, I'd suggest three decimals is very nice and handy to have.

- Good point about precision vs. accuracy, when I see more digits I feel like I'm getting closer to accuracy but I realize it's just precision - maybe not accuracy.  I should have said accuracy.  I'd go for accuracy first and precision second; it would be nice to have both - to a at least one milliamp and one millivolt if possible.  If that isn't economically practical I'll live with hundredths. 

Remote sensing? You'll only ever need to use such a facility if
1. you are powering something at a significant (>2A?) current
2. your interconnect leads are either long, or thin, or worse, both!
3. your load is really fussy about a "sagging" power supply (which suggests a design issue with your load, straight away)

I have remote sensing on my main bench supply, but have yet found the need to use it! Appropriately dimensioned interconnects answers most issues, here, IMHO.

- Ok, based on your feedback and that of others I'm starting to think Remote Sense might not be that big of a deal, but I'd kind of like to experience it at least once so I can write it off from personal experience

Some cheaper power supplies - especially fancy digitally controlled  ones - do VERY RUDE THINGS when you turn them on. Enough voltage spikes to fry your sensitve load, in some cases. These power supplies are clearly a FAIL , and are to be avoided at all costs.

- This is #1 on my list.  Even more than bells and whistles I want reliability.  Turn-on's that spike and fry stuff are a "turn-off".  So, whether it's a Rigol or an Agilent or anything else, I'd like something solid.

Well, You've seen my own analogue TTI tripple channel power supply at work, first hand! I've owned it for two years, and do not regret the decision to buy. They are built here in Blighty, so not too sure how many of them will have found their way across the pond. eBay is always a good source...

- I have indeed heard from various sources that TTi is good stuff, but I'm not sure what exactly makes a PS "analogue" vs. digital.  It seems we can start with analog meters, then we can go to digital meters - but beyond the LEDS or LCD it seems like to do almost anything with a PS (like set volts and amps) the PS is going to have some control circuitry that allows the user to select volts and amps, and it seems like that is going to be digital circuitry, no?  And what got this whole thing started was Marc's post and Bored@Work's reply that surmised that perhaps there was a race condition which allowed the output to turn on before the input had stabilized.  Now maybe that race condition was on an analog circuit in an all analog PS, but I thought maybe there was some digital circuitry in almost any "modern" PS, and certainly in any programmable PS.  Feel free to straighten me out - I always feel better when you do that   Thx

I admit that some modern power supplies now give the added function of being able to set voltage and current limits digitally. Nice, but not essential (to me). The Rigol unit seems very well spec'd for the money (if you can get your head around the fact that the 12-year-old who designed the front panel didn't attend school the day they taught man-machine interfaces and ergonomic keyboard layout - EURGH!), and meets the above requirements (of mine!). There are some more professional looking, even more capable (more than necessary??) supplies from Rigol, but they cost a little more, of course.

- So the TTi PS's aren't digital?  The new Rigol's must be digital? No?  I'm pretty sure what I'm looking for is a programmable (and therefore "digitally-enabled" PS).  I just need reliable and sufficiently programmable, with decent accuracy; precision would be a plus

You will know what people (IBM people??  ) used to say about buying IBM!. "Nobody ever got fired for buying IBM..." Well (probably) nobody (with the budget) ever got fired for buying Aglient, either. Of course, their gear will be beyond reproach, but boy, do you pay for that.  Some EVBLOGGER will now tell us that they knew a bloke whose friend heard of somebody who bought an HP/Agilent power supply from a mate once, and it blew up. Well, we can put such stories some way further down the list than exploding Korads...

Based on what you and I have spoken of, before, an Agilent supply (of appropriate voltage and current specs) is going to more than meet your needs, EF, I would suggest. And that's no bad thing, but you are paying for it.

- I could be talked into an Agilent or out of an Agilent, my jury is still out on Agilent.  Someday I'm sure I'd like an Agilent something or other.

Siglent gets a mention in this thread - they tend to be towards the budget end of electronics equipment, but possibly (I just don't know) fine for hobbiest use. But you have to ask yourself how well any product that you couldn't fix yourself is going to be supported in your country.  To quote Siglent's own web site:

- I don't know that much about Siglent.  What I can say is that my Korad (from China) is so far a joy to use (and if it keeps working it will be a great value), knock on wood.  And based on your (VERY helpful) input along with input from others here I am the proud owner of a Mayuno Electronic Load (it arrived in record shipping time from China today, more on that later).  So, as long as products from China with not a lot of local support keep working, I'm ready to declare them a very good value.  I realize hope is not the best strategy, but I'm chalking this first round of "lab building" up to "R&D" as I use the gear to help learn the basics.  I'm either going to reach the end of the line in my passion for experimenting in which case I wasn't in for all that much (less than some other hobbies), or I'm going to learn from the first round what would be a better second round.  So far, I'm loving test equipment and the feedback it helps provide in the learning process.

We believe when SPD3303 on sale, instruments market will meet a power storm, it worths the wait!

A "power storm", eh? I think I can imagine what any manual (which may or may not be supplied) will read like

- All I can say is that if I had to translate English into Chinese the people who read Chinese would probably not be highly impressed   Seriously, Laurence - thanks for the deluxe info sharing, I appreciate it.  EF