Old Capacitors and ESR

[trrriple]

Hello!

I'm working on restoring an old power designs 6010 power supply. I've recently acquired a DE-5000 to help with this.

I've replaced the bulk caps already but I notice the supply has a measurable 120 Hz output ripple that's pretty large, large enough that my multimeter even picks it up. I haven't found any schematics for this meter so I'm playing it by ear, and it's mostly a learning exercise for me.

There is a 310 uF Sprague 601D1059 cap across the output (or very close to) that I'm considering replacing also. I took it out and tested it's ESR. I found a datasheet for 601D caps and it appears all the ESR specs are measured at 120 Hz which was a surprise to me because I thought ESR was measured at 100 kHz.

Anyways when I measure at 100 kHz my DE-5000 gives me an OL on capacitance and nothing on ESR. I understand capacitance isn't something you measure at a high frequency anyways but that ESR is what made me go search for a datasheet. At 120 Hz the capacitance is okay at 330 uF but the ESR is .1 Ohms which to me is very suspicious.

Is this some sort of cap with a super low ESR? Is this cap just junk? I'm new to this so I appreciate any help! Thank you!

[bob91343]

It's not a case of good or no good.  ESR certainly can affect ripple but unless you know the parameters, you don't know the acceptance level.

Try shunting the existing capacitor with a new one and see what it does to the ripple.  Is the ripple out of spec?  Ripple normally will increase with heavier load but should be very small with no load.  There also may be other components that affect ripple, in particular the input filter capacitor and any 'speed up' capacitor in the control circuit.

[trrriple]

The ripple is high regardless of load. The ripple spec on the supply itself is 1mvpp and it's way-way over that. I don't have an exact number at the moment I'll have to get it later.

I'll try the shunt suggestion too. The biggest cap I have floating around is only 100 uf so quite a bit smaller but I'm guessing it should at least have some effect if this is the right direction.

I did replace a largish electrolytic near the transformer already that I assumed was the input cap but I'll have to probe around to be sure. I also replaced a shorted diode but that didn't seem to help the issue. Where might I find a speed up capacitor? There are a bunch of smaller caps on the board of course that I haven't touched yet. Everything is axial leaded so there aren't really any tell-tale signs of failed caps that I've seen.

[TimFox]

Manufacturers usually report ESR at either 120 Hz (or 100 Hz) for mains-frequency rectifiers, or higher frequency (such as 100 kHz) for switching power supplies.
That large capacitor is probably above resonance at 100 kHz, and should be measured at 120 Hz.  What is the Sprague spec for 120 Hz ESR?  Note that the reactance of 330 uF at 120 Hz is 0.25 ohms, so an ESR of 0.1 ohms will have a small, but measurable, effect on the filtering.
A good freeware tool for designing conventional power supplies is 'PSUD2' from DuncanAmps.  https://www.duncanamps.com/psud2/index.html
If the capacitor in question is across the output terminals, it probably is not the cause of your ripple.  Either there is a fault in the feedback amplifier (a "speed-up" capacitor is one from the output directly to the amplifier, bypassing the resistive divider), but more likely a fault in the rectifier or main filter capacitor.  An oscilloscope (connected very carefully) is the best way to trouble-shoot this.
Vishay now makes the old Sprague 659D series.  Your part number seems incorrect (or obsolete), since there should be only 3 digits after "659D", giving the capacitance.  In the current Vishay data sheet, at 120 Hz, the maximum ESR value for 330 uF capacitors ranges from about 0.5 ohms (35 V rating) down to about 0.3 ohms (200 V rating).  see  https://www.vishay.com/docs/42039/601d.pdf

[trrriple]

What do you mean by large capacitor? I assume over some amount of capacitance is when we can start saying they are large.

In this case this is a linear supply so it sounds like the caps are going to need to be measured at 120 Hz? As for the sprague spec, it's a bit hard to determine because I can't find the datasheet for this particular cap due to it's age. What I found was here: https://www.vishay.com/capacitors/list/product-42039/

601D337F100FS1 seems similar enough to the cap I have and it's ESR is rated max .33 Ohms @ 120 Hz.

Edit: Just saw your edit. Good to know I'm probably looking in the wrong spot. I'll do some more digging and see if I can't figure out where this speed up cap is supposed to be as well as ensuring I didn't screw up the input cap replacement.

Thanks!

[mawyatt]

Manufacturers usually report ESR at either 120 Hz (or 100 Hz) for mains-frequency rectifiers, or higher frequency (such as 100 kHz) for switching power supplies.
That large capacitor is probably above resonance at 100 kHz, and should be measured at 120 Hz.  What is the Sprague spec for 120 Hz ESR?  Note that the reactance of 330 uF at 120 Hz is 0.25 ohms, so an ESR of 0.1 ohms will have a small, but measurable, effect on the filtering.

Z at 120Hz for 330uF is 4 ohms not 1/4 ohm!!

Best

[TimFox]

Oops--I stand corrected.  I forgot the "1/" operator on my calculator.  The 0.1 ohm measured ESR will be negligible compared with a reactance of 4 ohms, but still measurable on a DE-5000.

[trrriple]

I'm pretty sure I got the part # right. Here's some pictures.

[TimFox]

From the full set of labels, that is a custom part Sprague made for Power Designs.  "310 uF" is a non-conventional value for such a capacitor.
It should be close to a 330 uF part with the same voltage.

[trrriple]

Here's a couple pictures of the supply itself. Yes those big caps were totally separated from their leads

I've replaced those and their little brother who's hiding under the transformer.

[mawyatt]

One reason you get poor or incorrect readings with the DE-5000, or any LCR meter, is the impedance of the device under test at frequency. With an impedance of 4 ohms at 120Hz the DUT can be effectively measured, however at 1KHz you are looking at ~50 milliohms which is difficult, of course at 100KHz this is 100 times smaller!! On the flip side trying to measure say a 330uH inductor at 120 Hz which has a Z of ~250 milliohms, this is towards the lower impedance range, whereas at 1KHz the Z is ~2 ohms and a more reasonable impedance is available for the instrument to accurately measure, and so on.

Anyway, these LCR meters have voltage and current ranges where the measurements are reliable, and ranges where they aren't. Generally just a quick DUT impedance estimate will allow one to place the measurement frequency in a more favorable spot for better results.

Agree that a quality 330uF with low ESR and proper voltage rating as a replacement should work, however these are across the output so may effect regulator stability. So a proper load transient test should revel any issues regarding stability.

BTW have fond memories of those old Power Designs, we had a bunch in our labs back in the day!! Gald to see you are restoring this and keeping these old devices around a few more years

Best,

[trrriple]

Good to know thanks!

Yeah I have this fella I'm trying to get working correctly again, and a TW347D. I'm just a big fan of how well they were put together and I think easier for me to understand how they work!

[TimFox]

With respect to your question about self-resonant frequency on "large" capacitors.
I could not find data on the 601D capacitors, but this note from Nippon Chemi-Con has a graph (Fig. 16) for a 470 uF/35 V unit that shows a broad resonance at about 100 kHz that is swamped by its ESR.
https://www.chemi-con.co.jp/en/faq/detail.php?id=29BDA8D
I didn't have any comparable 601D or other axial lead capacitors, but I checked an unused Panasonic 330 uF/50 V radial lead unit (about 25 mm long and 10 mm diameter, probably smaller than your 601D), without bias voltage, on my DE-5000 and got similar results to your 601D.  Wherefore, I would conclude that your capacitor is good.
120 Hz:  Cs = 316.2 uF  ESR = 0.1 ohms
1000 Hz:  Cs = 304 uF  ESR = 0.02 ohms
10 and 100 kHz:  OL

[strawberry]

ESR 0.1R is normal for non water based electrolyte capacitors 330uF/63V https://eu.mouser.com/datasheet/2/427/VISH_S_A0002931203_1-2568370.pdf
In good test equipment they use special hermetic high quality capacitors. That's why it measure OK after decades.

high frequency(45kHz) SMPS capacitor capacitance(uF) can be smaller. Low ESR capacitors