POWER DESIGNS PRECISION POWER SUPPLIES

[manticore00]

The 6050A arrived today and outwardly everything appears to be good and aside from being a bit dusty the inside appears good too, all fuses are intact and no obviously failed or damaged components.

That said, I've found that the 6V range won't output anything above 4.1 volts and when I switch to the 60V range I can dial it up to 9 volts but anything beyond that and the voltage drops suddenly and I hear a faint clicking noise coming from inside the unit. The overvoltage crowbar is rotated to the maximum clockwise position and I've dialed up the current limited pot and switch but that made no difference.

I plan on trying a more thorough examination over the weekend but I thought I'd see if anyone here has any immediate thoughts or suggestions?

[onlooker]

For those interested, 3 PD6050A are on ebay BIN now (not by me) for ~$28 each with FS. I guess they are not going to be there for long.

http://www.ebay.com/itm/Power-Designs-6050A-Power-Supply-/181156023030?pt=LH_DefaultDomain_0&hash=item2a2dbd8af6

[c4757p]

That was fast.

[don.r]

Real bargains there but the time shipping goes in... its no longer a bargain for me. There where a couple of working 6050D's a few weeks ago for $49 a piece but shipping to me was $125 for the two. These supplies often get overlooked and undervalued because they are not HP/Agilent or Lambda.

[onlooker]

They are gone! Who had them?

[SLJ]

Just an FYI and I don't know if this has been posted yet but here is a nice redo of a Power Designs model 2005 manual:
http://www.rako.com/Articles/29.html

[manticore00]

Ordered a Power Designs 2010 about a week ago and it showed up today. Right off the bat I was kind of worried because the front panel was bowed out from the case in a way that looked like it had been dropped...

Went into my lab to power it on and immediately noticed that the V/A meter wasn't moving as I moved up and down the voltage values, but after hooking it up to my bench DMM the output itself was absolutely perfect to all four digits so that made me much happier

I cracked open the case and after probing around a bit I couldn't see anything wrong with the wiring or parts(other than the meter) so I ended up taking the meter out to examine it and in the course of that effort the needle migrated from the bottom of the scale to the middle and wouldn't budge at all... Long story short, I ended up taking the case of the meter apart and discovered that some ancient blue insulation tape on the inside was fouling the free flow of the needle and after replacing it with some brand new black electrical tape everything now works perfectly.

Thought I'd share incase anyone else ever sees something like this in the future and could do with a troubleshooting idea.

[SLJ]

Thought I'd share incase anyone else ever sees something like this in the future and could do with a troubleshooting idea.

Thanks for sharing.  I have a 2005 on the way.  Looks like the output is OK but the meter/dials are a little off so it will need some calibration and there is no oven light or main power indicator lit so it will be an interesting project.

[robrenz]

It looks like the epidemic is spreading

[SLJ]

It looks like the epidemic is spreading

This will be number two for me (already have a 5005R) and I'm keeping an eye out for a good deal on a 6050A.

[manticore00]

I've also got a TP340 that works perfectly(but had a huge dent in the case that needed some percussive maintenance to correct) and a 6050A that appears to have a faulty current pot. I stripped the 6050A down a couple days ago and near as I can tell when the voltage go over about 8v there's an arcing noise and the current goes through the roof. I looked at the pot itself and it doesn't look like there's an easy way to get inside of it 

[KarlMonster]

Many thanks for this thread.

I was Power(Supply)less and needed something to drive Op Amps and help me to release the majik smoke from my dormant ICs.
[I used a looped AND-NOT logic gate to verify that my new 'Scope (with wonky CAL function) was really working!]

I literally had a Sinometer/Mastech HY3005F-3 triple linear PS in my shopping cart when I found this thread. Instead of that, I got one of many TPI 340s that are on eBay right now. I know I'm going to be pleased with it - 'cause it was half the cost.

[SLJ]

The 2005 arrived yesterday.  After a new power cord and a slight tweak it's dead on at 5.000 volts.  When you get it up to 20 volts it's off by about .02 volts.  Close enough for my use and until I find a service manual.  It took me longer to get three cal and inventory stickers and their goop off the front panel.
They do look nice stacked.

[don.r]

The 2005 arrived yesterday.  After a new power cord and a slight tweak it's dead on at 5.000 volts.  When you get it up to 20 volts it's off by about .02 volts.  Close enough for my use and until I find a service manual.  It took me longer to get three cal and inventory stickers and their goop off the front panel.
They do look nice stacked.

My god that's attractive.... ahem...

Try putting it to 10V on the 20V range (dial reading 0.000) and see if its still 0.02V out. If it is then its probably the 10K range resistor moving out of tolerance (assuming the 2005 is similar in the divider network to my 2020B).

[c4757p]

Remember that each voltage is switched in by individual, discrete resistors. There are a million places where that error could come from. If the error is linear with the set voltage, it's a calibration error, just adjust the trimpot (you're supposed to do it at 20.000V). Let both the PSU and the voltmeter you're checking it with warm up (fully turned on, not just standby) for a long time before you do this. If if's not, you'll have to track down which resistor it's from. I too suspect the +10V switch - if you get the same 20mV error going from 10 to 20 and from 0 to 10, check that. Otherwise, open it up and start probing resistors (for each one, you'll have to find a knob position that gets it out of the circuit). I suggest adjusting them by adding a small series resistor or large parallel resistor. The ones that are in there are already quite aged and should not drift too much more, and it's probably best to take advantage of that. Adjust the resistance according to output voltage, not to what the resistor is marked with.

[robrenz]

The 2005 arrived yesterday.  After a new power cord and a slight tweak it's dead on at 5.000 volts.  When you get it up to 20 volts it's off by about .02 volts.  Close enough for my use and until I find a service manual.

The manual is here from post#10 of this thread . Calibration procedure is there.

[SLJ]

Thanks,  I had forgotten about the link to the manual for the 2005 being posted here.  It's only the 10-20V range thats out (across the entire range) so it looks like I will need to check the trimmer and resistor for the upper range.

[c4757p]

Lucky... Mine had a whole bunch of resistors out. Not sure if it was a bad batch of resistors, or if the unit was just treated with particular disrespect.

Actually, given how the front panel was mangled to fit in replacement neons (a different size), I'd say it's probably the latter...

[robrenz]

There is only a zero adjust and a 20V adjust. You dial in the 20 after the zero. If 10V is off after that check the 10k resistor that is shorted by the 10-20 range switch. If 5V is off after that a Deoxit cleaning of the decade switches should get it in tolerance from my experience.

[don.r]

Thanks,  I had forgotten about the link to the manual for the 2005 being posted here.  It's only the 10-20V range thats out (across the entire range) so it looks like I will need to check the trimmer and resistor for the upper range.

Its almost certainly that 10K 0.1% (R45) range resistor. Pull the RV to S+ strapping out and measure across the resistor. Mine measured high so I paralleled a couple of 10Mohm resistors to it and its bang on the money now.

[robrenz]

We cant make that assumption until SLJ calibrates zero and at 20V and then checks 10V on the 10V scale. The 5V is affected by the decade switches and resistors.  The 20V being off means it needs calibrated before you can do any other testing.

[don.r]

We cant make that assumption until SLJ calibrates zero and at 20V and then checks 10V on the 10V scale. The 5V is affected by the decade switches and resistors.  The 20V being off means it needs calibrated before you can do any other testing.

Thus my "almost". But if its out in the entire 20V range by the same amount and not in the 10V range at all, then that would be my guess. Its easy to check anyhow.

[GK]

The 2005 arrived yesterday.  After a new power cord and a slight tweak it's dead on at 5.000 volts.  When you get it up to 20 volts it's off by about .02 volts.  Close enough for my use and until I find a service manual.

The manual is here from post#10 of this thread . Calibration procedure is there.

Very simple design, electrically. Checked out some of the other stuff on Mr Rako's site. Yet another personality from the semiconductor industry appears to be somewhat "out there":

http://www.rako.com/Apostate/44.html

 

[GK]

So what kind of legitimate applications are these PD supplies being used for, where their accuracy is actually required?

The lack of remote sensing facilities strikes me as odd.

[robrenz]

Discussion of that on this thread starts here Besides being more precise than the meters most many people would use to check them they have incredibly low noise and excellent long term stability.  You have to pay over $5000 dollars from Agilent to get a supply with noise performance this low. Why would you not buy these when they are less expensive than a lot of new junk

They do have remote sensing and programming.

Signed:
Power Designs Cult Co-leader