POWER DESIGNS PRECISION POWER SUPPLIES

[Vgkid]

Do you have a model number of the diode, I have a reference book from 1965.

[timb]

2005A Circa 1977

Silicon Zener:
TS823H2
TS823G
TS823

Silicon Diode:
GI44

2005 Circa 1966

Silicon Zener:
AC359C
AC359D
AC359BT

Edit: You may have better luck adding 1N instead of the first two letters to some of these. It seems either PD added a two letter manufacturers code to part numbers or that's how it was back in the day? No idea who TS or AC is, but 1N823 turns up hits on Google. Perhaps the suffix was a grade or binning code added by PD?

Maybe I should sacrifice this 2005A reference and desolder the parts so I can analyze them in detail. 

Sent from my Smartphone

[timb]

One other thing I've noticed about PD, is they used mainly plastic film and tantalum/tantalytics in these precision supplies. Very few AE or ceramic caps.

I replaced all the AE's in my 2020B, even though they were still spot on after 34 years, just in case. (There were two small 1uF ones, then three larger ones.)

Should I be worried about the wet tantalum caps leaking? They're in fully sealed little metal axial can packages and the values are still perfect.


Sent from my Smartphone

[David Hess]

Should I be worried about the wet tantalum caps leaking? They're in fully sealed little metal axial can packages and the values are still perfect.

They may eventually leak sulfuric acid but inspection of the ends will reveal any corrosion if this starts to happen.

I do not have a good idea of what an appropriate replacement is for wet tantalum capacitors.

[SeanB]

If the tantalum caps have a glass end don't bother, they will outlive you. If the end is rubber they will need to be replaced. Having the caps made for them was probably to get a low ESR ( for the time) or a higher ripple current capacitor, so internally they might have thicker foil and multiple tabs for each electrode. As well they would likely have had a spec for both tolerance and voltage rating with surge rating.

TS is TRW Semiconductors and AD is AC Delco. Both no longer exist. The parts were most likely house coded parts specially made for PD, either a selected grade of regular production, a house coded part to control spares availability ( lock in for OEM spares) or a specially made part. My bets are they were regular parts with the house code so as to get the units returned to PD for service.

[David Hess]

If the tantalum caps have a glass end don't bother, they will outlive you. If the end is rubber they will need to be replaced.

I have seen photos of hermetically sealed ones which failed as I described but it is apparently so rare that it is not something I would worry about.

Having the caps made for them was probably to get a low ESR ( for the time) or a higher ripple current capacitor, so internally they might have thicker foil and multiple tabs for each electrode.

Their claim to fame now seems to be long life, high reliability, and high temperature operation.  Once I have one fail, I will do a serious search to find suitable replacements.

[timb]

Interesting info, thanks. I'll check the ends and see how they're sealed.

Here's six in a small area of the 2020B board:



(They're the silver (metal) axial lead cans.)

I had a grey beard who restores old radios once tell me tantalytics were the best type of cap you could buy (in the 1960's at least). He said that he has to replace about 1 in 1000 he sees (this was at least 14 years ago though so things have perhaps changed).


Sent from my Smartphone

[MarkPalmer]

There were several tantalytics in a Dymec 2401A I just finished refurbishing, made in 1962.  Not a single one of those was bad, this in a machine with a LOT of operating hours on it, so it's safe to say they are durable have a very long lifespan if in circuits where they aren't stressed too much.  If in doubt you can lift a lead and do the standard capacitance/ESR/current leakage tests, but if they are within tolerance IMHO it's probably best to leave them be in the not-broken-don't-fix-it tradition. 

Like I mentioned earlier I only found 4 bad electrolytics in the 5020.  All others tested to be in excellent condition. 

-Mark-
 

[timb]

Yeah, I randomly spot checked about 5 of the highest stressed ones I could and they all had beautiful ESR and spot on capacitance values, so I concur. Same with the low value orange plastic film caps (also seen in the photo above). I was curious about their use as well, as most modern (and even older equipment) would simply use ceramics.

In fact, the only ceramic on this thing is C10, which is a 1kvdc 100pf going from the output of the voltage regulation OpAmp to the inverting input. (So I guess they're just AC coupling the feedback?)


Sent from my Smartphone

[robrenz]

As you can see from this thread I have purchased quite a few PD supplies and I was fortunate that none of them had any serious problems other than needing a thorough switch cleaning. I have also sold many of them since jumping ship to the HP 611X supplies. All that I sold were burnt in for at least 24 hours at max load before calibration without any problems.

IMO on these PD supplies if it aint broke don't fix it

[SharpEars]

I just got myself a Power Designs 2020B (with a pretty high serial number, woohoo) after reading this thread in its entirety, so please welcome me to the PD club. After properly calibrating the 0 V value and the highest voltage value per the calibration instructions using a 6 1/2 digit voltmeter as a reference, I have found some non-linearity in the voltage selection.

When switching between 5.000 V and 6.000 V (maybe it was between 4 V and 5 V), I noticed that the voltage coming out of the power supply goes from almost dead-on correct to about 1.4 mV higher than it should be. I realize that this is a small error, but it continues for all of the other unit voltage numbers above the threshold number (say 5 V, so e.g., 6 V, 7 V, 8V and so on). It is as if there is a resistor in place for the range 5-9 (whole) volts that kicks in after you cross the threshold and it is slightly off.

If someone can point me to the exact resistor this is among the many resistors connected to the rotary switch, I would really appreciate it, because it's really hard to tell what's what.

[motocoder]

I just got myself a Power Designs 2020B (with a pretty high serial number, woohoo) after reading this thread in its entirety, so please welcome me to the PD club. After properly calibrating the 0 V value and the highest voltage value per the calibration instructions using a 6 1/2 digit voltmeter as a reference, I have found some non-linearity in the voltage selection.

When switching between 5.000 V and 6.000 V (maybe it was between 4 V and 5 V), I noticed that the voltage coming out of the power supply goes from almost dead-on correct to about 1.4 mV higher than it should be. I realize that this is a small error, but it continues for all of the other unit voltage numbers above the threshold number (say 5 V, so e.g., 6 V, 7 V, 8V and so on). It is as if there is a resistor in place for the range 5-9 (whole) volts that kicks in after you cross the threshold and it is slightly off.

If someone can point me to the exact resistor this is among the many resistors connected to the rotary switch, I would really appreciate it, because it's really hard to tell what's what.

Have you cleaned the switch contacts? This almost always needs to be done, and leads to voltages being off as you move between different values. There's instructions earlier in this thread, but in general the process is:

Spray out any dust and dirt with compressed air
Clean everything with a mild, general purpose electronic cleaner
Clean the contacts with some DeOxit Red
Use the electronics cleaner to remove any deoxit red
Apply some deoxit gold to the contacts and leave this in place
Re-lube the switch bearings and points where the metal rubs with some lightweight grease (just don't get this on the contacts).

Then recalibrate, and I bet your unit will work fine.

[SharpEars]

I just got myself a Power Designs 2020B (with a pretty high serial number, woohoo) after reading this thread in its entirety, so please welcome me to the PD club. After properly calibrating the 0 V value and the highest voltage value per the calibration instructions using a 6 1/2 digit voltmeter as a reference, I have found some non-linearity in the voltage selection.

When switching between 5.000 V and 6.000 V (maybe it was between 4 V and 5 V), I noticed that the voltage coming out of the power supply goes from almost dead-on correct to about 1.4 mV higher than it should be. I realize that this is a small error, but it continues for all of the other unit voltage numbers above the threshold number (say 5 V, so e.g., 6 V, 7 V, 8V and so on). It is as if there is a resistor in place for the range 5-9 (whole) volts that kicks in after you cross the threshold and it is slightly off.

If someone can point me to the exact resistor this is among the many resistors connected to the rotary switch, I would really appreciate it, because it's really hard to tell what's what.

Have you cleaned the switch contacts? This almost always needs to be done, and leads to voltages being off as you move between different values. There's instructions earlier in this thread, but in general the process is:

Spray out any dust and dirt with compressed air
Clean everything with a mild, general purpose electronic cleaner
Clean the contacts with some DeOxit Red
Use the electronics cleaner to remove any deoxit red
Apply some deoxit gold to the contacts and leave this in place
Re-lube the switch bearings and points where the metal rubs with some lightweight grease (just don't get this on the contacts).

Then recalibrate, and I bet your unit will work fine.

Why use deoxit gold if there are no gold contacts in there?

[robrenz]

It is an excellent contact lube in general. The only exception is cermet or carbon pots where Caig Fader lube is a better choice

[motocoder]

Yes, +1 to what robrenz said. The DeOxit Gold is a preservative / protectant and lube for the contacts.

[timb]

I personally prefer the DeOXIT Shield myself on anything fully open to the free air, like the switch banks on these units are.

While putting my 2020B back together I did some experiments with lubricants and found WD-40's spray-on White Lithium Grease and Silicone Lubricants to both be excellent choices, but I feel like the Silicone Spray edges the other out in the least mess and protection factor areas, so that's what I'm going with. Though, if you want a good long lasting lube that's easy to apply and find, you can't go wrong with Hopp's #9 Gun Lube (orange bottle), available at Wal-Mart or anywhere firearms are sold.

I use DeOXIT Gold inside the  wirewound pots, plus Silicone Lubricant for the shaft and Lithium Grease for the bearings.

By the way, for really bad problems with pots or switch banks, don't be afraid to desolder the connections, remove them, use the original spray on DeOXIT and then scrub them in a sink full of warm water and dishwashing soap with a toothbrush. Followed by a hot water rinse to get the soap out, then a rinse with 90+% IPA. Followed by the prescribed lube and protect procedure above.

You can remove pots and take the cover off them (a couple of metal bend tabs) and do the same thing.


Sent from my Smartphone

[SharpEars]

OK, I may have found what is causing the 1.4 mV drift going from 5 V to 6 V. There is a 1 k 0.05% resistor in there that I measured at 930 ohms. I don't know if this is the circuit contribution or if it has fallen out of spec over time.

If it has fallen out of spec, I am wondering if that is enough out of tolerance on such a precise part to cause the problem. It seems that this (black) resistor goes in-circuit on the switch from 5 V to 6 V. Because it measures less, I am wondering if it skews the divider circuit enough to cause the 1.4 mV drift for voltages where the 1's digit is greater or equal to 6.

I'll see if I can get a picture of the particular resistor I am talking about over the weekend and perhaps some of you with the 2020B can help out by measuring the resistance on your unit of that resistor in circuit and let me know if it's supposed to measure low (e.g., due to the contribution of other circuit elements).

I am very hesitant to clip or desolder it in order to measure it in isolation, unless I find no other way. It's easy to decrease resistance in circuit with a parallel resistor, but increasing it is kind of tough without doing some "damage," so that's my path of last resort.

[timb]

Set all switches to 0 and put it on the 0-10V range, after that you you should be able to measure the resistors individually.

How are you doing the measurements? With a 4-wire kelvin clip I assume?


Sent from my Smartphone

[Vgkid]

That 1k resistor at .05% should have a tolerance of +/- 0.50 ohms. So it is out of spec.

[timb]

Yeah, all the switch resistors in my 2020B are dead on. Even the 5% and 1% tolerance ones are +/- 0.001Ohms. With all the switches in their full positions, I get exactly 10,000.01Ohms on my 6.5 Digit HP'57. So yours seems out for sure.

I might have a replacement here if you want it.


Sent from my Smartphone

[timb]

Boom goes the dynamite! My 2020B is alive, again!



Set for 10V and my 3456A is showing 10.00008V there in the background.

So, what was it? It appears to be the OP05 was the culprit (one of my initial guesses).



Popped in the new LT1097 and she seems to work fine with one issue: I'm about -26mV at a switch setting of zero, with no more turns on the zero pot left. I guess I could always add some inline resistance to the offset input, right? Or a larger pot.

Either way I'll read up on the LT1097 today and figure out how to modify the 2020B accordingly.

I also pretty much got everything done on the front panel. All I have left is to hook up the 10-20V switch, plus the indicator lamps and finally, cover the hole where the meter goes temporarily.

This time I used all crimp connections for the banana jacks; well worth it IMHO if I ever need to remove the panel again!






Sent from my Smartphone

[robrenz]

Nicely done. 

[SharpEars]

Yeah, all the switch resistors in my 2020B are dead on. Even the 5% and 1% tolerance ones are +/- 0.001Ohms. With all the switches in their full positions, I get exactly 10,000.01Ohms on my 6.5 Digit HP'57. So yours seems out for sure.

I might have a replacement here if you want it.


Sent from my Smartphone

Say where are you measuring from/to to get that 10,000.01. I have a 3456A as well, so I can perform the measurement and see what I come out to...

[David Hess]

Popped in the new LT1097 and she seems to work fine with one issue: I'm about -26mV at a switch setting of zero, with no more turns on the zero pot left. I guess I could always add some inline resistance to the offset input, right? Or a larger pot.

Either way I'll read up on the LT1097 today and figure out how to modify the 2020B accordingly.

The LT1097 has a much smaller offset null range (+/- 600uV) than the OP-05 (+/- 4mV).  If they were relying on this to correct for other errors which is generally a bad practice, then something else may have to be done.

[SharpEars]

Here are the measurements I've got. If someone can compare them to their unit I would appreciate it:

(Click on any image to view it full screen)

2020b 4-wire Kelvin resistance measurement connection location:



2020b settings (all maxed out, including vernier):



4-wire measurement results on HP 3456A: