POWER DESIGNS PRECISION POWER SUPPLIES

[SeanB]

Use a higher wattage one instead, a 1W resistor will fit and will run cooler. Did you measure R34 with one lead unsoldered, you will get a false reading in circuit.

[timb]

Ugh. Fucking site just ate my long reply…

Short version: Can’t find a 1W in that size at 0.1% but it might not be needed as I was testing in-circuit and see now there’s a 20k trim pot in parallel with the resistor, so that’s most likely what’s causing my reading. Since the unit seems to be working and in-spec, I may not even touch it.

[SeanB]

Working means leave well alone.

[timb]

Also I just ordered a 5-pack of those transistors you linked and a fuse holder for the back. Now I just have to fix the panel meter (it sticks at the bottom, you have to give it a tap to get it to working), do some cleaning and she'll be good as new.

By the way, the seller gave me a $25 refund due to the packing and broken fuse holder.

Not a bad deal for $50+$10 in parts, eh?


Sent from my Smartphone

[timb]

I might be addicted to these... I just picked up a 2005A for $50 and a 1025P (0-100VDC Precision Remote Programmable) for $35.

Also, it looks like the edgewise panel meter on the 2020B is shot. Unless I can find someone who has one from a parts unit, I'm going to need to get creative.


Sent from my Smartphone

[timb]

Oh yeah, can anyone with a 2020B measure the voltage on the 10V/20V indicator lights for me? Thanks. =)


Sent from my Smartphone

[robrenz]

Oh yeah, can anyone with a 2020B measure the voltage on the 10V/20V indicator lights for me? Thanks. =)


Sent from my Smartphone

125V according to the schematic

[iampoor]

Do the Power designs power supplys have floating outputs that I could connect togeher to make a +/- supply? Im hinking somehing along the lines of a tw5005

[robrenz]

Do the Power designs power supplys have floating outputs that I could connect togeher to make a +/- supply? Im hinking somehing along the lines of a tw5005

Yes

[timb]

Oh yeah, can anyone with a 2020B measure the voltage on the 10V/20V indicator lights for me? Thanks. =)


Sent from my Smartphone

125V according to the schematic

Yeah, I saw that, just wanted to verify as the schematic I have is a pretty poor quality scan, so it’s hard to read some of the digits. I wanted to make sure it wasn’t 12.5V!

I’m trying to figure out the best way to replace the bulbs with LEDs.

Do the Power designs power supplys have floating outputs that I could connect togeher to make a +/- supply? Im hinking somehing along the lines of a tw5005

Just make sure the DC- terminal isn’t strapped to the GND terminal on the front panel!

[robrenz]

IMO save yourself a lot of trouble and just buy some new neon bulbs. I have replaced several on the PD units.

[timb]

IMO save yourself a lot of trouble and just buy some new neon bulbs. I have replaced several on the PD units.

One of the assemblies was broke, so I'm going to have replace the holder anyway. I guess I could get a neon panel mount assembly.

Thinking about it though, there should be no reason an LED wouldn't work so long as you used a high enough wattage resistor. Right? (I found some Amber colored panel mount LEDs in my junk bin that fit perfect. I'd need nearly a 1W resistor though.)


Sent from my Smartphone

[SeanB]

Would work with the original value resistor, just use a 1W unit. A hyperbright led would give more than enough light at low current. If the original neon lamps are dim then reverse them, they run on DC so only one electrode wears. Eventually the whole envelope gets sputtered and it stops working.

[timb]

So my 2005 came in today. After plugging it in and turning it on, I began to smell something burning, but not in the normal Magic Smoke sense. This smelled like, I dunno, it reminded me of of those toy cap guns that took those rolls of paper caps? Anyway, I don’t *think* there was any smoke (lab labs weren’t all on, so hard to tell). I unplugged it right away, investigated a bit and turned it back on. The smell seemed to go away and the supply *seems* to work (after cleaning the switches), though I don’t think the oven light is ever turning off (hard to tell because it’s flickering, though I assume it would cycle on and off for seconds or minutes).

So I’m suspecting something went toasty in the oven. Thoughts?

[c4757p]

I began to smell something burning, but not in the normal Magic Smoke sense. This smelled like, I dunno, it reminded me of of those toy cap guns that took those rolls of paper caps?

Hmm - two very distinct smells that aren't quite the same: cap guns and overpowered potentiometers. But that was my first thought. Is there (or was there) anything shorting a potentiometer?

[timb]

I began to smell something burning, but not in the normal Magic Smoke sense. This smelled like, I dunno, it reminded me of of those toy cap guns that took those rolls of paper caps?

Hmm - two very distinct smells that aren't quite the same: cap guns and overpowered potentiometers. But that was my first thought. Is there (or was there) anything shorting a potentiometer?

The back panel cal pot *may* have been shorted, I'm not 100% sure, because after  the whole smell incident I started trying to get the thing in regulation; apparently I didn't have the nut loose all the way or the securing nut wasn't tight and the whole pot was turning. At this point I had the bottom cover off anyway and noticed it. I'm not sure if it was already turned and I moved it more or what. But it could have been shorted on the bottom cover. (Though it seems to work fine.)

Should the oven be physically warm to the touch? It feels warmer than ambient, but no more so than surrounding parts. (I imagine it's well insulated so it shouldn't be.)

The oven light never going off does bother me a bit.

Oh, picture!



There's a modern cap soldered across one of the metal Sprague caps on the auxiliary board.



Not sure what's up with that.

I did manage to find a date code on one of the Spragues; 17th Week, 1965


Sent from my Smartphone

[timb]

The new transistors came in for my 2020B today. Hooked them up with some jumper wires for testing and they work perfectly! Thanks @SeanB, that 2N3773 was a good call. =D

Tonight I'll get them bolted and soldered on for real. I've even got some genuine Sil-Pads to mount them with! (Fuck Mica and the greasy horse it rode in on.)

Finally, I've designed a *really* cool replacement for the broken panel meter. I think you guys will really dig it.


Sent from my Smartphone

[motocoder]

Does anyone know wher I can get a copy of the manual for the 2020 (not the 2020B). I have one that seems to start out reasonably stable, but after sometime, maybe correlated with the oven heating up, don't know yet, the stability drops and it starts to drift a bit. Hoping I can find a schematic and board layout for it.

[David Hess]

Not sure what's up with that.

It is a common way to repair a circuit that has a worn out aluminum electrolytic capacitor without removing it.

[timb]

Not sure what's up with that.

It is a common way to repair a circuit that has a worn out aluminum electrolytic capacitor without removing it.

Hmmm, yeah, I assumed it was there to repair the cap it was strapped over; but if you don't remove it (or at least one leg of it), how do you even test it to know if it's bad? (I've got a decent cheaper LCR meter that can do in-circuit testing, but it doesn't really work well on bulk caps unless they're out-of-circuit most if the time.)

Also, removing the bad cap prevents it from leaking on stuff.

It just seemed sort of lazy and bodgy to me.

But, perhaps it's more common than I knew and I just hadn't run into it much. =)

Either way, I think I'm going to replace all the AEs in the unit; I can't imagine they'd be very stable after 40 years?


Sent from my Smartphone

[David Hess]

Not sure what's up with that.

It is a common way to repair a circuit that has a worn out aluminum electrolytic capacitor without removing it.

Hmmm, yeah, I assumed it was there to repair the cap it was strapped over; but if you don't remove it (or at least one leg of it), how do you even test it to know if it's bad? (I've got a decent cheaper LCR meter that can do in-circuit testing, but it doesn't really work well on bulk caps unless they're out-of-circuit most if the time.)

Testing is easy for a power supply input capacitor.  Just measure the ripple voltage.  If you know the load current, which you will in a power supply because you can set it, then a measurement of either the peak to peak ripple or the slope of the ripple can be used to determine the capacitance.

The general rule is 8200 microfarads per amp/volt assuming 60 Hz power and full wave rectification.

Also, removing the bad cap prevents it from leaking on stuff.

It just seemed sort of lazy and bodgy to me.

But, perhaps it's more common than I knew and I just hadn't run into it much. =)

I do not recommend the practice myself for the reasons you identify.   It is handy as a temporary measure and for testing though.

Either way, I think I'm going to replace all the AEs in the unit; I can't imagine they'd be very stable after 40 years?

I usually replace them all or at least all of the similar ones if one goes bad or has become marginal.

[timb]

Ahh yeah, totally didn't think of measuring the ripple on for the filtering caps; that's a great tip!

I'm also wondering about all the ceramic disc caps on this as well; I've heard tales of tin whiskers and such. 50 years (not 40 as I stated before; 1965 date code) is a long ass time! Half a decade...


Sent from my Smartphone

[David Hess]

I'm also wondering about all the ceramic disc caps on this as well; I've heard tales of tin whiskers and such. 50 years (not 40 as I stated before; 1965 date code) is a long ass time! Half a decade...

Ceramic capacitors tend to fail with a short which would be apparent.

Tin whiskers are unlikely unless pure tin plating or lead free solder was used.  In a power supply, there are a lot of areas where the currents are high enough that any whisker would vaporize anyway.

[macboy]

...
So after testing R17 and R18 with makeshift Kelvin probes, they’re both exactly at 0.75Ohms, despite R18 being a bit deformed. However, after testing most of the resistors on the board, I did find R34 (the matched resistor to the LM399 reference) is measuring 3K~, instead of the marked 7.32K. Though, it doesn’t seem to be affecting accuracy? I might still replace it; unfortunately I can’t find any 0.3W replacements on Mouser (can only find 0.125W ones in-stock). So I may have to resort to using an SMD version with a little adapter board I’ve got laying around.

7.32 k can be made with 8.2 k || 68 k, both are E24 series.

[SeanB]

Anecdote on 2n3773's. I once needed a pair of replacement pro Electron devices ( can't remember exact part but it was a BDYxx series one) so put in an order with stores for 2. A month later I get a call from the head stores, asking how many I would need in the next decade. They were apparently a superseded part, and while the original manufacturer ( now part of the Airbus consortium) still had the masks and the process equipment, they would be a special order. Cost was $100 each, and there was a minimum order quantity of 10k units, with a 11 month lead time from confirmation and payment up front. I was asked if there was a replacement instead for them. So looked in the "transistor bible" and on the page in particular was listed 30 or so equivalents, among stem was the 2N3773 listed as "Improved specs". Checked in stores, and they were both there, in stock ( around 30k on hand) and available for us to order. Placed the order for 4 cases, as they would also replace 2N3055 transistors in any application as well.

When they arrived I used 2 as needed and stuck the rest in stock, looking like they would be a decade or more till finished. Little did I know that next month in would come the first of a line of Tower ferroresonant UPS units, which each either used 50 of them, or a $5k transistor brick. New batteries ( all 4 car batteries), 50 new transistors ( or scrap it if it blew the big silicon block and keep for spares) and run for a week with a 2 bar heater as load after repair. Some got a big chunk of heatsink with the transistors added as replacements. Horrid with the emitter sharing resistors being 10cm lengths of 0.5mm copper wire soldered to 24 emitters, and commoned on an insulated bolt. Transistor 25 was the driver for the darlington thus made.