POWER DESIGNS PRECISION POWER SUPPLIES

[jpc6204]

Dave has listed some supplies that he recommends under his Amazon tab:

http://www.eevblog.com/amazon/

There you will see:

1. BK Precision Power Supply, 1670A, Triple Output, Digital
2. Extech 382260 80 Watt Switching Mode DC Power Supply
3. Mastech TRIPLE LINEAR DC POWER SUPPLY 30V 5A HY3005F-3

I've been drooling over the Mastech supply for a while now.

Echo the fact that building your own supply is always a good learning experience if you haven't don so yet.


Jeff...

[robrenz]

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.

I tried that on my 2010 and 2020 and got similar results to you.  Then I thought let me see what my Mastech 3030 triple supply does.  It only required 6.5mV to drive the full 3A.  A Chinese unit performing better than the PD supplies  how can that be? Then I realized I shorted the output on the Mastech with my 3" heavy 8awg banana jumpers I made for paralleling the outputs.  When I tested the PD units I used a 7" long piece of 20awg solid wire that was laying around. I retested the PD units with the heavy jumper and They both drive full amperage at approx. 6mV also.  It makes sense to me now, the power supply has to put out at least the voltage drop across the short. So the 6.5mV at 3A means the total "short" resistance seen at the output was 2.16mOhm.  My lesson learned,  use very short shorts.

[saturation]

Nice!  I did the test shorting the inputs with 1252a DMM and read the current of it, so it was also working against the voltage drop across the DMM.  I did this because the built in ammeters need to be recalibrated.  I presume you were reading the current off the PSU built in ammeters?  I'll try that later on and report back.  Nevertheless, your results are even more impressive that what I reported earlier.  Its a more ideal response; maybe my Mastech isn't as bad as it looks too [ in which case, kudos to Mastech!]

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.

I tried that on my 2010 and 2020 and got similar results to you.  Then I thought let me see what my Mastech 3030 triple supply does.  It only required 6.5mV to drive the full 3A.  A Chinese unit performing better than the PD supplies  how can that be? Then I realized I shorted the output on the Mastech with my 3" heavy 8awg banana jumpers I made for paralleling the outputs.  When I tested the PD units I used a 7" long piece of 20awg solid wire that was laying around. I retested the PD units with the heavy jumper and They both drive full amperage at approx. 6mV also.  It makes sense to me now, the power supply has to put out at least the voltage drop across the short. So the 6.5mV at 3A means the total "short" resistance seen at the output was 2.16mOhm.  My lesson learned,  use very short shorts.

[robrenz]

So just to prove to myself that the short resistance is a big deal in CC testing I made a super short pictured here.  .625 diameter tellurium copper with gold plated beryllium copper banana jacks.



This is the 3" heavy 8awg banana jumpers I used in my first test described before.



I checked the calibration of the amp meter on the power supply and it is spot on. I set the current limit to 1 Amp using my fluke 87 but had to use a higher voltage because of the voltage drop across the shunt in the 87.  Left the current setting alone, plugged in the super short, dialed down the voltage until it just bumped the 1 Amp reading and measured the output voltage with the 8846A.  Sub 1mV,  that's 5mV less than the 3" heavy 8awg banana jumper. 

EDIT#1: I checked the 2020B at 2 amps and got 1.39mV but noticed you need to wiggle the shunt to get the lowest contact resistance/voltage. So I went back to the 2010 at 1 amp and wiggled for lowest resistance and got 0.7mV instead of the .8749mV. So it looks like .7mv/Amp.  Becoming obvious that Milli and micro Ohms matter down here.

EDIT#2: So the effective resistance of the "3" heavy 8awg banana jumper" is 0.00216 Ohm and the "super short effective resistance is 0.000695 Ohm. Only 0.00147 Ohm difference.

Sorry for the crap picture, hand held long exposure to get the whole reading of the VFD




So I wondered why the PS current measuring shunt (.2 Ohm) voltage drop isn't showing up in the output voltage? I think it is because the voltage sense is after the shunt voltage drop in the power supply. 

[SoftwareSamurai]

3. Mastech TRIPLE LINEAR DC POWER SUPPLY 30V 5A HY3005F-3

I've been drooling over the Mastech supply for a while now.

What about the Mastech HY1803D?

[SeanB]

Robenz, I think you are measuring the resistance of the posts themselves, rather than anything else.........

And there I though using a thick copper foil as a zero ohm resistor was good enough........

[robrenz]

Robenz, I think you are measuring the resistance of the posts themselves, rather than anything else.........

And there I though using a thick copper foil as a zero ohm resistor was good enough........

SeanB, I agree, I am making a set of kelvin probes for my IET LOM-510A micro Ohm meter
http://www.ietlabs.com/decaderes-1/resistance-meter/lom-510.html
and as soon as their done I will be able to measure the resistance directly to 1 micro ohm resolution +/- (0.02% rdg. + 4 micro Ohm) accuracy. I then will be able to actually measure the  short resistance and separate out the jack resistances from the bar itself.

[saturation]

I did prelim tests with a just a shorting bar and my results are as you demonstrated with yours: the PD were able to source its rated current, 2ADC, for my 3 units measured at at 6mV, 2mV, 3mV, as compared to when I measured current with the DMM.  I then double checked the output voltage on the terminals with the DMM.  These units are a 2020B x 2, and a 5020.

My 3030D Mastech, alas, still outputed 3A at ~800mV.  Clearly, this was not designed as well as your version.

Nevertheless, its clear from the more meticulous checks you've made that the PD can output close to an ideal current source, that is its output is ~ independent of the voltage across the outputs, as whether its approaching zero volts up to its rated maximum output voltage, the rated current can be pumped out as you desire.  Awesome!

As resistances is getting fairly small, the jacks or any other minor series resistance are starting to play bigger roles in the voltage drop measured.  I think you can still push it to see what the theoretical minimum can be, but driving at 1-6mV output voltage at 1-3A is far from what any circuit I know can be powered!  So, I think the point is very well made at this level.

In my units, the still uncleaned adjustment pots are likely playing large roles in not getting adjustment down further or improving the output stability.  So, it should be interesting to revisit the tests when the units are spruced up.

So just to prove to myself that the short resistance is a big deal in CC testing I made a super short pictured here.  .625 diameter tellurium copper with gold plated beryllium copper banana jacks.
...

[krivx]

Can you the Power Designs supplies sold in the US run off 230V 60Hz? Or is a step-down transformer required? Getting a supply from the states to Europe might be worth it for me if I can use it without much modification or having to buy extra equipment.

edit: After looking at a few manuals, it seems some models have a jumper on the power transformer for 240V operation and some may not. If anyone knows for sure please let me know.

[saturation]

Hello krivx

I have not tested it personally.  But the manual for the 2020B, 5020 and C500 state it will work from 47 Hz to 440 Hz; a jumper has to be set internally for 240V, so spec wise its ready for EU.

You may have to check manuals for any specific model of this level of supply, as the older ones can be different.  Except voltage, many PD supplies before 1980 labeled the key specs on the rear of the units.

I have an ongoing discussion on another forum about obtaining more manuals and schematics.  The fellows are doing it on their own time, so patience, please.

As the users post them, they will be here.  Some are already loaded:

http://www.ko4bb.com/manuals/

Can you the Power Designs supplies sold in the US run off 230V 60Hz? Or is a step-down transformer required? Getting a supply from the states to Europe might be worth it for me if I can use it without much modification or having to buy extra equipment.

edit: After looking at a few manuals, it seems some models have a jumper on the power transformer for 240V operation and some may not. If anyone knows for sure please let me know.

[krivx]

As the users post them, they will be here.  Some are already loaded:

http://www.ko4bb.com/manuals/

Thanks, there are quite a few manuals here awaiting approval. http://www.ko4bb.com/manuals/index.php?dir=10%29_Recent_Uploads

These could be very helpful! I'll take a look through them when they're available and start asking for shipping quotes for suitable models. Even with shipping costs some of these supplies are still bargains.

[saturation]

There is mirror server, here files are available, its not secret its indexed on google:

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

[krivx]

It looks like the newer models have a transformer jumper, the older ones do not. Thanks for that link, it could save a lot of hassle.

[robrenz]

I bought 2 later model 2005 units off ebay in "not functioning" condition.  They lit up but did not output any voltage. Both units had the 1 amp DC fuse (internal) blown but seem ok after fuse replacement.  (good thing since that is about the extent of my troubleshooting  skills)

[robrenz]

Here are some pics of one of the two 2005's.  I tried doing a voltage cal "as is" and couldn't get to zero with the zero adjust pot.  Did an electrical cleaner spray and air blast on all the gang switch contacts and then used de-oxit on all the many many wipers on the gang switches.  Now the zero cal pot can get it to zero. Adjusted the 20V setting to 20.00000V

These are some random voltage checks using a 8846A 6.5 digit 24ppm meter. % error values do not include meter error. 
The stated accuracy for this supply is 0.1% +/- 1mV
SETTING    READING   % ERROR
20             20.00000         0%
15             15.00020        .0013
10 on 20    10.00046        .0046
10 on 10     9.99980         .002
5                5.00000         0
1                1.000116       .011
1.555555    1.555370       .011

AC RMS (no load) is 10.1uV so P-P would be 28.56uV
Spec is 100uV P-P

Stability is not as good as the 2010 I have.  The 2010 after 14 hours had a total min. max. span of 3.75uV and a SD of 763nV .  This 2005 is out of spec with a 1 hour min. max. span of 196uV.
EDIT: After a couple of hours on time the min.- max. span has dropped to 9.1uV with a SD of 2.5uV  for a half hour test. maybe electrolytics are reforming some?
EDIT2: No wonder they burn them in for 100 hours before calibration.  Recalibrated after about 4 hours of on time and a 2 hour stability test gave a min.- max. span of 5.95uV and a SD of 1.48uV

[T4P]

MY GOD the parts looks orgasmic

[robrenz]

The crazy part is that supply cost me $60.99 + shipping on ebay plus a $3.45 $0.37 fuse.  But the perfomance cannot be touched even by $400 to $600 modern supplies.

[T4P]

The crazy part is that supply cost me $60.99 + shipping on ebay plus a $3.45 fuse.  But the perfomance cannot be touched even by $400 to $600 modern supplies.

Sure. Pity it's not like widely available. What's the weight like? Colossal or Lighter then expected?

[robrenz]

2005 weighs 9 pounds.  2020B weighs 13 pounds.  I only see 4 of the precision units on US Ebay currently.  Dont worry I am not buying any more of these.

[T4P]

2005 weighs 9 pounds.  2020B weighs 13 pounds.  I only see 4 of the precision units on US Ebay currently.  Dont worry I am not buying any more of these.

I don't live in the states you see. Shipping can be a killer but i remembered they are precision power supplies and not general high power power supplies it makes sense.

[robrenz]

Tested one of the 2005 supplies for CC mode with the super short at its full 500mA.  Needed to set 1.3mV to get the full 500mA. Measured the voltage drop right where the banana plugs screw into the copper bar (probe tip touching the plug and the bar simultaneously). 3uV drop across the short itself.  The remaining voltage is dropped in the banana jacks and sockets getting to the copper bar. 

Did the same thing on the 2010 supply at 1A and needed 2mV to get there. Voltage drop across the copper bar was 6uV as expected.  So the actual copper bar is approximately 6uOhm when you don't count the banana plug resistances.

EDIT: measured the short bar resistance at the same locations with my microOhm meter and got 6uOhm also. Even though I calculate the copper resistance for the effective bar length at 1.85uOhm. This is the effect of the actual effective path of the current through the bar being a smaller cross sectional area than the bar itself.  I proved this out with a 4" square piece of copper .0625" thick.  Measured the resistance on one edge with the sense leads of the uOhm meter .5" apart.  The theoretical resistance for a cross sectional area of .0625" x 4"  .5" long is 1.32uOhm but the meter reads 13uOhm. The electric field does not uniformly spread out and use the whole cross sectional area, It takes the shortest path.  (sounds good but who knows if it is technically correct )
 
 It was interesting to see the voltage rise as I moved the probes on the banana jacks away from the copper bar and toward the power supply jacks. When I measured the voltage right at the face of the binding posts on the supply it is 1mV so the other 1mV (minus 6uV) is dropping on the  internal wiring up to the binding posts.

OK, I know I am obsessed with this but: It would be interesting to hook up the remote sense leads right at the junction of the banana jacks and the copper bar. I suspect then you could dial roughly 6uV on the 0-1 mV dial and drive the full 1A. All the  other voltage drops would be compensated for.

[terabyte]

You guys are a bad influence.  Look what I just got;
http://www.ebay.com/itm/200731066755?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649


I will report once it arrives. 

[terabyte]

Question.  Does anyone know what the difference is fromt he TW5005T model that I got compared to the TW5005W on Jim William's bench in the first page?



The only difference I can spot is the pot for voltage selection.  Does the "W" model have a 10 turn pot and the "T" a dual pot setup with the fine adjustment on the inner dial?

[robrenz]

Nice buy! It looks like you are probably correct about the pot difference.  The manual for the "W" is on eBay for $25.00 and the manual links in this post only go to the "D" on the TW5005.  Just keep in mind the specs and performance that have been shown in this thread so far are for the precision versions with the decade voltage setting.  But the specs on the analog setting versions are extremely good also. It was good enough for Jim Williams, that says a lot.

I am definitely NOT implying don't post here if not a "precision" supply. I just want people to be aware of the difference.

[terabyte]

Thanks for the reply robrenz.

I'm a beginner and I don't have anything more accurate than a Fluke 87V so yeah I'll start a new thread once it arrives.  I looked around a lot and even for a beginner this bench power supply seemed like a bargain.  And yea my thinking was exactly that; "If it was good enough for Jim Williams..."