POWER DESIGNS PRECISION POWER SUPPLIES

[resistcircuitresist]

Hey guys, new to the forum, so please don't castrate me if I've posted in the wrong section.


Im looking for a good bench power supply or kit with lcds that has constant current and voltage modes.

I have been looking at something like this http://www.uni-trend.com/UTP3701.html

Any feedback or recommendations....... Also are there power supplies that also measure current draw as well.

[electronwaster]

Hi there,

I am a beginner so take everything I say with a shovelful of salt.

It looks like a very neat power supply, but I just searched around can't find anywhere that publishes a price, so it's hard to say whether it is good value.

You asked about monitoring current draw, well any single channel power supply with 2 screens will be showing you voltage and current. If it's constant voltage, constant current, as you mentioned, it should show you what current is being drawn. Only the very basic and cheap kits (e.g. LM317 based) will not show you the current being drawn.

The standard advice from the experts is to go for an old or new HP/Agilent (same products, company buyout, I believe). e.g. http://www.ebay.com.au/itm/190663291908?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649#ht_500wt_949 I would go for a big Agilent supply if I had plenty of money.

It seems that you can also get trusty old power supplies from other smaller manufacturers for pretty good money, and for a lot of them, the schematics are available and are usually simpler to follow than the complicated new stuff.

I just (yesterday) bought a Farnell dual output 0-30V, 0-2A power supply with digital readout and constant current and constant voltage mode for $155. I do have to pay $60 for shipping because I don't live on the mainland of Australia, but that's true for me with any power supply. This is much cheaper than the $200-300 for the HP supply, and it has much more output (120W output vs 30W), and with luck it will all work and be of good quality.

If you are not absolutely wedded to the idea of digital output (I am too), there are lots of analogue output supplies going for much less on ebay. The circuitry could be identical, but getting a precision reading would require 2 multimeters in parallel and series, during use.

electronwaster

Edit: I forgot to mention, there are two power supply projects going on at the moment: Dave (the guy in the videos on the eevblog youtube channel) has a video series of about 10+ videos describing the design of a power supply he is making for production. I believe these will be a kit. They won't be available for at least a couple of months though (That is a guess, I'd love to know for sure)

Also Richard (forum username amspire) is designing a kit (check the forum) that will be slightly simpler, and available for/with through-hole components (easier for me to solder), and it won't require the use of a microcontroller. This is great for me as I spent several years as a programmer, and am sick of it :-) I want analog!

[saturation]

This has been discussed at eevblog for quite a bit, but here's a summary of key points so you can search the archives.

Beware of Chinese branded supplies, don't judge quality by appearance.  PSU must provide clean DC power before anything else.  Even within brand names, quality control can vary. 

Choose between linear: capable of cleanest output but less efficient and thus big, heavy and bulky, versus switching mode, high power but noisier, but more efficient and compact size. 

See Agilent's papers on how to choose and evaluate PSU, to give you a benchmark to compare the others.

http://www.home.agilent.com/agilent/product.jspx?cc=US&lc=eng&ckey=856757&nid=-35673.0.00&id=856757

Digital programmability: while convenient it can make the output noisier than a manual one, particularly on lesser branded supplies.

FWIW there are many lab grade very high quality PSU sold on eBay for under what they are really worth, its just a question of waiting for the deal and knowing how to spot a dud from a gem.

[robrenz]

Just got a Power designs 2010 in very nice condition for $60.00 on ebay today.  Anxious to see how it performs.

[saturation]

Nice!  Happy for you.  I posted elsewhere how good it is, and unusual.  They don't make PSU of this caliber since the early 1980s, and the only other one who competed in the category of "precision power supplies was an short running HP series.

If you finish your review before I do, please post your findings on a new thread, and I'll add to it.  I have 2 units currently being tested, and one more enroute. 

Some tidbits I'll string together later on, and add to a separate thread about this very nice PSU.

One unit is holding to 2ppm accuracy going on 100+ hours now of testing.  Its incredible.  However, while the dials are calibrated the linear adjust pot, a 100 ohm 10 turn Bournes, is erratic and can vary 300-500uV the knob setting!  Fairly easy to clean, adjust or change.  But, even if left as is its impressive setting output voltage to uV level and then pumping out 1A or so, and watch it hold to this value!  The unit came from a Boeing factory lab and looks like its been kept in better environmental conditions than the next supply.

The other unit is older and while it appears to be in pristine shape, there's rust on the outer body.  The dials are off 5mV, and bobbles in the uV, but the bourns pot is good.  Its 48hrs and its wandering +/-10uV, still very good for any PSU.   Its very likely oxidation or contaminants on the contacts of the rotary switch, from the environment it either worke din or was stored in.

Ripple check via a floating scope at it max 5mV/div show flat line up to 1A draw.  Superb for an old, or any age, PSU.

Power Designs have linear PSU that are not labeled "precision"; they have the similar specs as most any good quality stand alone supplies available today like the Agilent 36xx family.

Finally, look at this photo of Jim Williams' desk, note his PSU at lower left, with the "pd" symbol, and the signature aluminum face plate style with orange knobs, that's a Power Designs dual tracker.

Just got a Power designs 2010 in very nice condition for $60.00 on ebay today.  Anxious to see how it performs.

[robrenz]

Just got a Power designs 2020 (20V 2A other specs same as 2005) in excellent condition for $85.00 on ebay today.  They are hiding in many places on ebay.  A simple "Power Designs" search does not show them all.

[saturation]

You bought another?  Good for you!  I'm watching many of them, so you must be the one buying some as I see only single buyers on 2 units in the past week  ... but if our review means anything later on, it could be hard to find cheaply.  Enjoy!

There are several generations of the precision model line.  The most modern are fairly ugly, with plastic bodies, and pricey.  Not sure why.  This design looks fairly 1980s, probably about the time PD started to fall into hard times.



Service, user and schematics for the 2005, 2010, 5020 and C500 are available as pdfs.  The schematics show some of the designs are so simple, its amazing such stability can be had so simply.
Likewise,  many of the popular general PD PSU have manuals, such as that pictured in the Jim Williams post.

Their name, "Power Designs" can make it hard to search, since its common search terms.  Add the word 'precision' helps narrow it down.

Just got a Power designs 2020 (20V 2A other specs same as 2005) in excellent condition for $85.00 on ebay today.  They are hiding in many places on ebay.  A simple "Power Designs" search does not show them all.

[olsenn]

I own a BK Precision GPS-4303 PSU which I absolutely love. It is a 4 channel power supply, but you can get a 2 or 3 channel one for cheaper (around $200 - $400).

One thing I would like to mention as you say you are a beginner, is don't expect to find a PSU that offers quality constant current operation. When we refer to setting the current on a PSU it means the maximum current. It is always the voltage that is kept constant, and Ohm's Law can be used to determine how much current that will yeild (I = V/R). If you set the voltage to a greater value than can be achieved while maintaining the max current dialed in, then effectively it will be a constant current source, but don't rely on this; it will over/under shoot and may fail altogether.

[chimera_786]

my two cents: make on your self. It will teach you a lot. If you want help, pm me and we can start a new thread! 

[robrenz]

You bought another?  Good for you!  I'm watching many of them, so you must be the one buying some as I see only single buyers on 2 units in the past week  ... but if our review means anything later on, it could be hard to find cheaply.  Enjoy!

This was the first. I was the only bidder  http://www.ebay.com/itm/251031626964?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

This was the second. Was asking $119.99 I offered $85.00 and they took it.  http://www.ebay.com/itm/120758873482?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

There are many in much worse condition and more money I was very fortunate thanks to your heads up on these supplies

Service, user and schematics for the 2005, 2010, 5020 and C500 are available as pdfs.  The schematics show some of the designs are so simple, its amazing such stability can be had so simply.
Likewise,  many of the popular general PD PSU have manuals, such as that pictured in the Jim Williams post.

saturation,  do you have links to these pdfs?  I have looked some and only found them available for sale not free.
Thanks, robrenz

[saturation]

Yes, to all.

I know those 2, just know the second one you won has a broken switch to move between 10V and 20V range, that switch is vintage aluminum handled locking pull to set type.  But its easy to substitute until you can find a matching replacement.

There are other units that have not been put back into auction that are in my tracking list.  They should be back shortly.

2005: Paul Rako's site.  He wrote for EDN, and now works for analog.

http://www.rako.com/Articles/29.html

2010, 5020, and C500

https://docs.google.com/viewer?a=v&q=cache:EwpcBMGQ5TAJ:128.238.9.201/~kurt/manuals/manuals/Other/POWER%2520DESIGNS%25202020B,%25205020,%2520C500%2520Instruction.pdf+&hl=en&gl=us&pid=bl&srcid=ADGEESgOKHKiERW-JexT4B1lKdgXdxFrlR3FKBaAISArSRNrlmm-L65VRQh2WZMvUBY3QvNaKRhZXi3nvdJuMMzLNhgGF1Zjzdz0wmT83-bRw-oCoH3jyU819g4ik3Z5_13PKKNwF4RS&sig=AHIEtbT9w8oh_e9PK_kT7JD8rqPC1kPWRA

More, Courtesy of KO4BB [ many thanks! Curt]:

http://www.ko4bb.com/

[ the above is the main link, but he has far more manuals and not indexed on the main site]

http://128.238.9.201/~kurt/manuals/manuals/Other/

of interest:

POWER DESIGNS 2005A Technical Data.pdf
POWER DESIGNS 2015R Instruction.pdf
POWER DESIGNS 2020B, 5020, C500 Instruction.pdf
POWER DESIGNS 5015T Instruction.pdf
POWER DESIGNS TB343A Instruction.pdf
POWER DESIGNS TP325 Technical Data.pdf
POWER DESIGNS TP340A Instruction.pdf
POWER DESIGNS TW347D, TW5005D, TW6050D Instruction.pdf
POWER DESIGNS TW347D,TW5005D,TW6050D Instruction.pdf


Enjoy.

This was the first. I was the only bidder ..
This was the second. Was asking $119.99 ..

There are many in much worse condition and more money I was very fortunate thanks to your heads up on these supplies

Service, user and schematics for the 2005, 2010, 5020 and C500 are available as pdfs.  The schematics show some of the designs are so simple, its amazing such stability can be had so simply.
Likewise,  many of the popular general PD PSU have manuals, such as that pictured in the Jim Williams post.

saturation,  do you have links to these pdfs?  I have looked some and only found them available for sale not free.
Thanks, robrenz

[robrenz]

saturation,  thanks for the help.  I still need to find the 2010 manual.  The googledocs was for 2020 not 2010 (I am not complaining ).
I got the first unit (the 2010) today.  Wow, as Dave would say "pure electronic equipment pornography" As a machinist I especially appreciate the quality of the knobs and dials. Voltage output is well within spec and front and rear panel are mint condition. Only the case paint needs a little help but I cant believe I got this for $60.00
Thanks again for all the info.

[resistcircuitresist]

I own a BK Precision GPS-4303 PSU which I absolutely love. It is a 4 channel power supply, but you can get a 2 or 3 channel one for cheaper (around $200 - $400).

One thing I would like to mention as you say you are a beginner, is don't expect to find a PSU that offers quality constant current operation. When we refer to setting the current on a PSU it means the maximum current. It is always the voltage that is kept constant, and Ohm's Law can be used to determine how much current that will yield (I = V/R). If you set the voltage to a greater value than can be achieved while maintaining the max current dialed in, then effectively it will be a constant current source, but don't rely on this; it will over/under shoot and may fail altogether.

Thanks for the information about the current settings. The reason I'm interested in that setting is to test circuit designs under less than ideal power. Ie failing or low battery.

[saturation]

You're welcome.  The 2010 and 2020 are very similar, but mea culpa, I made that presumption.  I wouldn't be surprised if they are for the most part identical except for specifications for the main power elements: diodes, transformer, and power transistors.   Since the components are heavily derated, its even possible most if not all are identical except for the transformer.  So, until one can be located, the 2020 manual could serve as a close guide.

saturation,  thanks for the help.  I still need to find the 2010 manual.  The googledocs was for 2020 not 2010 (I am not complaining ).
I got the first unit (the 2010) today.  Wow, as Dave would say "pure electronic equipment pornography" As a machinist I especially appreciate the quality of the knobs and dials. Voltage output is well within spec and front and rear panel are mint condition. Only the case paint needs a little help but I cant believe I got this for $60.00
Thanks again for all the info.

[robrenz]

I agree about the manual similarity and that these power suppies are unbelievable. 
100mV setting.   stat mode on the 8846A 6.5 digit 100nplc digital filter
 
1 hour AVG 99.99839 mV    1.75 uV min-max span   SD of .392 uV
3 hour AVG 99.99873 mV    2.93 uV min-max span   SD of .667 uV
6 hour AVG 99.99935 mV    3.33 uV min-max span   SD of .855 uV

My 8846A shorted input DC 152 hr stability min-max span  is 1.08 uV !

Full 1A load output ripple measured with 8846A on AC is 15.8 uV max.
 I know this is way out of the guaranteed specs of the 8846A but I did some crude testing here https://www.eevblog.com/forum/product-reviews-photos-and-discussion/low-level-ac-performance-of-fluke-8846a/msg76514/#msg76514 that shows it is way better than the specs and does not put out bogus numbers down here like some meters aparently do.

I don't know about you but I might replace that mV pot with a 10 turn and a turns counter. It is very hard to set 1uv levels with that 1 turn pot.

[saturation]

WOW!  That's fantastic.  Much better values that mine. 

Yes, I would consider changing the pot to a 10 turn, do you have a source for a turns counter?

I also have a PD 5020 which uses a 100 ohm 10 turn Bournes, you can set uV levels for a 50V output!, but the pot needs servicing or replacement.  It has a beautiful machined turns counter, with lock and mV vernier scale for setting 100uV increments.  Its erratic currently, but its stable once its settles.  Its tracked by the 3rd 3456a in the bottom photo.





Some porn for you, 2020B x 2, above and a 5020 at bottom:



Tracking output going on 4+ days.  That's variance, to get SD square root those numbers.  Each 3456a is tracking one PSU.  That's about 13uV, 5uV and 4uV SD for each unit.  For 99% confidence interval multiple each by 6 [3 SD above and 3 SD below the mean], so its roughly spans 78uV, 30uV and 24uV.  This is as close I've seen a PSU get battery like variation.

I agree about the manual similarity and that these power suppies are unbelievable. 
100mV setting.   stat mode on the 8846A 6.5 digit 100nplc digital filter
 
1 hour AVG 99.99839 mV    1.75 uV min-max span   SD of .392 uV
3 hour AVG 99.99873 mV    2.93 uV min-max span   SD of .667 uV
6 hour AVG 99.99935 mV    3.33 uV min-max span   SD of .855 uV

My 8846A shorted input DC 24hr stability min-max span  is 1.15 uV !

Full 1A load output ripple measured with 8846A on AC is 15.8 uV max.
 I know this is way out of the guaranteed specs of the 8846A but I did some crude testing here https://www.eevblog.com/forum/product-reviews-photos-and-discussion/low-level-ac-performance-of-fluke-8846a/msg76514/#msg76514 that shows it is way better than the specs and does not put out bogus numbers down here like some meters aparently do.

I don't know about you but I might replace that mV pot with a 10 turn and a turns counter. It is very hard to set 1uv levels with that 1 turn pot.

[robrenz]

I just switched to 1V instead of .1V and the min-max span is 16uV with a SD of 4.6uV for 2 hours run time. So it looks like it may be a percentage of output voltage just like the spec implies.  What voltage are you testing at?

This is the turns counter I quickly found. This is the model with no brake. It is available with brake also.
http://www.alliedelec.com/search/productdetail.aspx?SKU=70125994

[saturation]

Thanks for the link! that saved a lot of time.  It looks like Kilo is the source for some of PD knobs as its a dead ringer for what's on these, 30+ years later.  Those knobs cost over 50% the cost of the PSU 

For the picture, I used 0.1V too, adjusted to 100.000 00 mV.  I think a good fraction of the instability is coming from the pots as they contribute more in adjustments at lower output voltages.  At higher voltages, they are less contributory so I get values closer to yours, 1-10uV, e.g. when I set it to 1V SD is only 0.5uV.

FYI, cleaning precision switches use Dexoit 5 or equivalent as other sprays may leave residues that alter contact resistance.  Its similar to maintenance on older voltage sources that use rotary switches for resistance decade dividers:

http://www.krohn-hite.com/htm/technotes/technical_note.htm#Technical%20Note%20#012

I just switched to 1V instead of .1V and the min-max span is 16uV with a SD of 4.6uV for 2 hours run time. So it looks like it may be a percentage of output voltage just like the spec implies.  What voltage are you testing at?

This is the turns counter I quickly found. This is the model with no brake. It is available with brake also.
http://www.alliedelec.com/search/productdetail.aspx?SKU=70125994

[robrenz]

@ saturation
I have deoxit 5 for the switches and also gold, and I use the fader lube for pots. but its good to get a thumbs up on using it. I get the bottled liquid instead of the spray.  If I need to spray it I put a drop or two in a fine solder paste dispensing needle and put it on my paste dispenser and give it a 70psi blast. With that you can get into places the spray cans cant touch.

Should we start a new thread for these PD suppies and ask for the PD related stuff here to be moved/copied to it?
10 volts is showing 16uv span and 2.3uV SD for 15 mins

[chrome]

Damn, 10 Mega Volt

[robrenz]

And that is the least signicant digit!!!!!!

[saturation]

Those are great ideas for applying deoxit, you think the spray can with nozzle is still not fine enough?

As for moving the thread, yes, why not?  I still have a way to go with getting mine up to factory shape, and a business trip coming up so I won't be able to post tear downs for at least a month. 

I think these supplies are very unique and need a detailed going over.  With the available schematic, and with interest shown by readers in Dave's recent videos on PSU designs, this supply would make a great DIY build.  Why build any LM317 based supply if its easier to buy one and mod it for better performance?  Not so with these supplies, so far.  DIY effort would be worth it because these designs are no longer produced so building is the only alternative; any patents on the circuit would be expired by now.   

Further, if the eBay supplies are exhausted, is unavailable or priced too high to ship to other countries, or if prices rise due to any interest we generate on the thread, it will remain cost effective and well worth one's time to build.

@ saturation
I have deoxit 5 for the switches and also gold, and I use the fader lube for pots. but its good to get a thumbs up on using it. I get the bottled liquid instead of the spray.  If I need to spray it I put a drop or two in a fine solder paste dispensing needle and put it on my paste dispenser and give it a 70psi blast. With that you can get into places the spray cans cant touch.

Should we start a new thread for these PD suppies and ask for the PD related stuff here to be moved/copied to it?
10 volts is showing 16uv span and 2.3uV SD for 15 mins

Yes, that an error they've propagated through decades on the precision series PSU, methinks its because their stencil is all upcase and in the old days it wasn't as cheap or easy to change on demand

Damn, 10 Mega Volt

[robrenz]

Those are great ideas for applying deoxit, you think the spray can with nozzle is still not fine enough?

I don't have a problem with spray cans. I just like the versatility and cost effectiveness of the bottles or tubes. You never run out of propellant pressure with my method and you can swab every last drop out of the bottle. I keep a micro brush applicator http://www.microbrush.com/automotive/products/microbrush/index.asp with each bottle for non spray application. I have more control with the hypodermic/paste dispenser method as to the amount and where it goes. a carefully located drilled source (and drain if needed) micro holes in a switch or pot can allow cleaning and lube without disassembly. You can put as much IPA or your favorite cleaning solvent into a glue dispenser syringe as you feel appropriate and inject it under pressure. Then take the piston out of the syringe and blow it out with just the air. Then blast the conditioner or lube in.

[saturation]

Thanks for tips, robrenz, and those links!

Here's another test you can run on the PD supplies, current sourcing.  An idea current source is fully independent of output voltage, so in theory it outputs full current near "zero" volts.

In reality most supplies can't.  However, the 2020B can pull rated amps, >2.2A on my test, set at 200mV, 1.3A at 100mV, and still gradually down the line as you approach zero VDC.

By comparison, my Chinese Mastech supply requires 1.0VDC to output its rated 3A, and drops quickly to 30mA at 600mV.

Those are great ideas for applying deoxit, you think the spray can with nozzle is still not fine enough?

I don't have a problem with spray cans. I just like the versatility and cost effectiveness of the bottles or tubes. You never run out of propellant pressure with my method and you can swab every last drop out of the bottle. I keep a micro brush applicator http://www.microbrush.com/automotive/products/microbrush/index.asp with each bottle for non spray application. I have more control with the hypodermic/paste dispenser method as to the amount and where it goes. a carefully located drilled source (and drain if needed) micro holes in a switch or pot can allow cleaning and lube without disassembly. You can put as much IPA or your favorite cleaning solvent into a glue dispenser syringe as you feel appropriate and inject it under pressure. Then take the piston out of the syringe and blow it out with just the air. Then blast the conditioner or lube in.

[saturation]

Some comments for beginners, such a supply is often not a good supply as if it performs this way the voltage regulation will also certainly be poor, or there are other flaws in the design.  Just test of response time will show the over/under shoot or other failures.

A constant voltage source can be converted to a constant current source, its the essence of Norton's theorem.   Thus, the quality of voltage regulation is also proportionate to its capability in current regulation and vice versa.

I own a BK Precision GPS-4303 PSU which I absolutely love. It is a 4 channel power supply, but you can get a 2 or 3 channel one for cheaper (around $200 - $400).

One thing I would like to mention as you say you are a beginner, is don't expect to find a PSU that offers quality constant current operation. When we refer to setting the current on a PSU it means the maximum current. It is always the voltage that is kept constant, and Ohm's Law can be used to determine how much current that will yield (I = V/R). If you set the voltage to a greater value than can be achieved while maintaining the max current dialed in, then effectively it will be a constant current source, but don't rely on this; it will over/under shoot and may fail altogether.

Thanks for the information about the current settings. The reason I'm interested in that setting is to test circuit designs under less than ideal power. Ie failing or low battery.