Old Rubycon SH-series capacitor from the 1980's.

[JDW]

I'm replacing the capacitors in a SONY power supply used inside an old Apple HD20SC hard drive enclosure.  These were released in 1986 and mine seems to have been built in 1987 or 1988.  It worked before recapping and still works after recapping but there's a short beep (squeal?) that occurs at the time I switch on power.  I traced the trouble to a single Rubycon cap.  My replacement for the stock Rubycon SH-series 10V 2200uF capacitor is a Nichicon UHE1C222MHD1TO. It didn't make that squeal sound when I had the stock caps inside. And when I remove that one Nichicon and put back the Rubycon (only that 1 cap -- all the rest are my replacements), the beep goes away. Viewing the 5V rail on the scope shows a tad bit more noise (ripple) with my Nichicon replacement, but I believe the problem is that my replacement's ESR is too low, causing either a phase shift or unwanted oscillation.

I measured the Rubycon cap with my DE-5000 meter. At 120Hz it shows an ESR of 0.2-ohm. My replacement Nichicon is much less than 100-ohm -- so low the meter won't show the value at 120Hz.

Here's a photo of the stock Rubycon cap:



It measures D=13mm, H=26mm.

Do any of you have an old datasheet for the Rubycon SH-series? I can't Google up anything.

The problem is I am not finding any 10V 2200uF replacements on Mouser that would fit D=13.5mm(max), H=29mm(max) and have sufficiently high ESR. The ones that fit all have a quite low ESR specs. I don't have a schematic of this power supply, so a datasheet on the Rubycon would help a lot. Any ideas or suggestions would be appreciated.

Thank you.

[OwO]

add a resistor

[PKTKS]

In 80's and 90's -pre internet and  practically pre the networked "PC" ...
 (XT/AT in rare cases would have a modem if someone would paid the phone bill...)

There were no "datasheets" as we know it today.

"MANUALs" were the only source of those prints and unless
you hard copy them you would have to buy one (very expensive
large size bibles..)

So *UNLESS* you can find the "SCANNED" version of one
of those old stuff... or one single page hard copy..
chances are we have no clue about that.

Even founding some sheets they are hardly comparable
to some recent today almost machine based generated sheets.

Another era another tools.. those pages are part of other class
of tools.

I do have some large "stuff" crammed in folders (papers)
which is hard to replace hard to find and hard to get rid of...

digital stuff is another level of problem..
they do not suffer dust problem but you do need backups.

Paul

[jfiresto]

In 80's and 90's -pre internet and  practically pre the networked "PC" ...
 (XT/AT in rare cases would have a modem if someone would paid the phone bill...)

There were no "datasheets" as we know it today.

"MANUALs" were the only source of those prints and unless
you hard copy them you would have to buy one (very expensive
large size bibles..)....

Huh? You just rang the local sales office and they would happily send you datasheets, or complete data books, if they liked to provide data in aggregate, to encourage people to design in other parts.

Rubycon may have an archivist who can send data for discontinued parts.

[PKTKS]

Huh? You just rang the local sales office and they would happily send you datasheets, or complete data books, if they liked to provide data in aggregate, to encourage people to design in other parts.

Rubycon may have an archivist who can send data for discontinued parts.

May be in some very specific areas with such representatives.

But not "around the world" and due to costs probably not for free.

Even my  CMOS (National) and TTL databooks costed me a little
fortune those days...  hard to obtain and very expensive.

Paul

[jfiresto]

... Even my  CMOS (National) and TTL databooks costed me a little
fortune those days...  hard to obtain and very expensive....

If you were in Brazil back then, did you try the sales office in Sao Paulo?

I found National was always very helpful in answering requests. They would sometimes even send a sample or samples, when I just asked about a datasheet, if they thought it might help the evaluation and design.

[PKTKS]

IN early to mid 80's Brazilian electronics industry was still active...

today no longer the case, practically all stuff is imported.

But those days there was interest from the representative sales
office to promote their stuff so although quite expensive the
literature was indeed available.

Never free. Several reasons..
 +PLUS nasty just beyond insanity brazilian taxes...

Nevertheless I managed to have at least Philips, National
Texas and some Intel and Zilog stuff..

Still carefully stored although paper is something hard to conserve

Paul

[shakalnokturn]

No datasheet sorry...
I'm assuming this cap is the first after the rectifier on the rail that has the regulation loop? (Opto)
As suggested the quick way round would be adding a little series resistance.
You could also consider modifying or adding soft start, but I bet you want to leave this collection piece as close to original as possible... (Do the caps even need replacing?)

[JDW]

 Fluid filled electrolytic capacitors do not have eternal life, folks. They don’t last forever. I also see evidence of leaked fluid. And even when you don’t see evidence of leaked fluid, it’s very well may be that some of the fluid has dried out through the years, especially in power supplies where there’s constant heat. So yes, after 30 years, these capacitors do indeed need to be replaced.

 I don’t have a schematic for this power supply, and I don’t think one is readily available because I cannot Google it up.

I actually started soldering three 1/8W .5 ohm resistors in parallel to give me 170m-ohm of resistance, but I found when I tried to solder that cap and resistor in there wasn’t enough space. It comes dangerously close to other high-voltage components. I’m also not sure of what the wattage spec should be on such an in-line resistor, although three 1/8 wide resistors in parallel could be enough. I just don’t know.  But it doesn’t matter when there’s not enough space.

 None of you suggested a high ESR replacement capacitor, so I guess you don’t know?

[Yansi]

A classic example of "fixed until broken completely".

Why have you changed the caps, if it was working fine? Age isn't enough argument to replace them.

[shakalnokturn]

Not saying JDW is a case of fix until broken... Sometimes components do need replacing sometimes it's even worth replacing them before time because you know that if you wait long enough you'll be having a whole lot more fun getting the thing working again, if possible. (Nichicon SMD electrolytics on Mac classic mainboards are one example.)

It's the "replace all the electrolytics" approach that some people have that I can't agree with, often there's no point, if it's to tear out perfectly good enough caps to throw in budget far east Yaego-cheapo replacements that won't last 1/2 as long as the old originals remaining lifetime it just seems a waste.

Don't know of specific high ESR electrolytics, I'm sure some vintage Tektronics people do, it's a problem that's mentioned regularly WRT to old Tek equipment.
Maybe a Parametric search on Mouser, Farnell, RS... will give you an option to filter only the highest ESR models?

[JDW]

Not saying JDW is a case of fix until broken... if it's to tear out perfectly good enough caps to throw in budget far east Yaego-cheapo replacements that won't last 1/2 as long as the old originals remaining lifetime it just seems a waste.

Maybe a Parametric search on Mouser, Farnell, RS... will give you an option to filter only the highest ESR models?

Thank you for being more understanding than some inexperienced people who have never dealt with vintage Mac motherboards destroyed by leaked capacitor fluid which ate through traces!

My replacements are mostly Nichicon, 5000hr or 10,000hr and of the highest quality I can find. So swapping best-quality modern caps for those 30+ years old is not a waste at all.

I did a parametric search on Mouser but I cannot find anything that will physically fit.  Another problem is, without any kind of data on my stock Rubycon SH-series capacitor, my parametric search is somewhat in vain.  Consider that most datasheets only specify MAXIMUM ESR.  I don't know that spec on my Rubycon, and guessing isn't good.  I only know what I can measure, and it's 0.2-ohm according to my DE-5000, @120Hz.  I cannot measure higher frequencies, unfortunately.  (My DE-5000 fails to do that on most large capacitance devices.)  And again, I lack the physical space to add a series resistor.  And on top of that, I'm not really sure what wattage rating would be best even if there was space for such a resistor.  Do you?  (No, I don't have a schematic.)

[JDW]

I usually share my recapping knowledge with others in the form of a replacement capacitor list, so a readily available replacement capacitor that works well is critical and is why I started this thread in the first place.

Using my DE-5000 meter, I made the following measurements today spanning several minutes for each test to ensure the most accurate measurement, all at 120Hz because higher frequencies produced no results on my meter...

Stock Rubycon 2200uf 10V capacitor (SH-series, 105°C) (inside Sony PSU in Apple HD20SC):
• θ = -75°
• Q = 3.7
• R = 0.1Ω (only the tenths place was shown, so it could be 100, 150, etc.)
• D = 0.270

Nichicon 2200uf 16V capacitor (UHE1C222MHD1TO, 105°C):
• θ = -88.6°
• Q = 45
• R = 0.0Ω (only the tenths place was shown, so it could be 20mΩ, 60mΩ, etc.)
• D = 0.022

LCC 2200uf 25V capacitor (new but 16.5 years old, 85°C):
• θ = -87°
• Q = 20
• R = 0.0Ω (only the tenths place was shown, so it could be 20mΩ, 60mΩ, etc.)
• D = 0.049

Note that D is Dissipation Factor, aka tan δ, which is almost always shown in capacitor datasheets even when ESR is not.

After recapping, when I use the stock Rubycon cap, all is well and there is not beep/squeal shortly after I switch on power, and the peak-to-peak ripple on the scope shows as it did prior to recap -- about 50mVp-p.  The ESL spikes are 100mVp-p max.

When I use the Nichicon cap, the ripple is about 70mWp-p and ESL spikes exceed 100mVp-p, and there is a beep/squeal sound shortly after I switch on power.  I don't see a buzzer on this PSU, so I think the sound is coming from one of the two transformers, as per what my ear is telling me.

I then thought it best to try a new but very old cap I got when I bought my JAMECO electrolytic radial capacitor kit back in April 2003 (more than 16 years ago).  That cap too is 2200uF but 25V and rating at only 85°C.  It's marked LCC and I cannot find any data sheet for it.  I doubt JAMECO had these kits in stock for more than 1 year, and I doubt the cap was made more than a year before the kits were assembled, so I would guess the age of the cap could be as old as 18 years.  But again, it was new, never used, and always stored in a location where people are (i.e., no long term temperature extremes or excessive humidity).  With this old LCC cap, there is no beep sound, and the scope measurements are exactly what I see with the stock Rubycon cap.  This is interesting because if you look at my measurements above, the LCC still has much better test results than the Rubycon.  The LCC's Dissipation Factor is more than twice the Nichicon's though, so perhaps that is the key here?  Even so, it's still much lower than the 0.270 of the stock Rubycon.

In any case, the problem is easily solved for me by using either the stock cap or the LCC, but if I share a capacitor replacement list with others, the issue of that one cap is a sticking point.  Sure, I could just advice using the stock cap, but what about cases where that cap is bad?  And so I am trying to find a suitable replacement 2200uF 10V (or higher) cap that will physically fit (13mm x 26mm) and produce a stable output voltage.  I think the beep after the switch-on is telling me of some instability, and the higher ripple voltage of the Nichicon replacement seems to bear that out.

Any further thoughts on this would be appreciated.

Thank you.

[fzabkar]

Would adding a series inductor prevent the power-on squeal?

Searching for a capacitor with inferior ESR specs just doesn't seem right.

[exmadscientist]

As someone who faces similar search troubles when designing electrolytics into new designs, I feel your pain. Many engineers do not understand the value of ESR in a bulk capacitor, and I have gotten some strange looks indeed when saying that a part can have too low an ESR and I need to find something higher. Manufacturers aren't immune, either: new parts advertise their lower and lower ESRs, and I still search out the old cheap ones. I know, I know, I could just add series resistance (and have indeed done so when I haven't had space for an electrolytic), but there's just something so elegant about how many problems a single humble aluminum electrolytic capacitor can solve.

One tip that can help in the search is to remember that ESR and ripple current are inversely related, particularly HF ripple current. (Ultimately the ripple current limit is set by heating, caused by ESR- and dissipation-related losses.) Thus if a part doesn't specify its ESR, but has a lower ripple current rating than comparable capacitors, you can infer that it is likely to be a good choice when you want ESR. Sorting by ascending ripple current on Digi-Key quickly turned up the EEU-EB1A222 as a reasonable candidate for you. (There may be better choices; that's just the first one I found.) As well, physically smaller capacitors are often forced to have higher ESRs due to their construction, so don't limit yourself to a single size if smaller parts might fit.

(Aside: those measurements look a bit fishy. In particular, I am suspicious of those high Q values for electrolytics; you may want to check your meter. But that has no bearing on your main point.)

[JDW]

As someone who faces similar search troubles when designing electrolytics into new designs, I feel your pain....
Sorting by ascending ripple current on Digi-Key quickly turned up the EEU-EB1A222 as a reasonable candidate for you. (There may be better choices; that's just the first one I found.) As well, physically smaller capacitors are often forced to have higher ESRs due to their construction, so don't limit yourself to a single size if smaller parts might fit.
(Aside: those measurements look a bit fishy. In particular, I am suspicious of those high Q values for electrolytics; you may want to check your meter. But that has no bearing on your main point.)

Thank you for your honest and experienced opinion.  I appreciate the specific capacitor recommendation very much indeed, and found it in stock on Mouser as well:

https://www.mouser.jp/ProductDetail/Panasonic/EEU-EB1A222?qs=sGAEpiMZZMtZ1n0r9vR22RyzzjEKRzqKYIaKulH23zw%3D

DF@120Hz is a fairly high 0.32 (a good thing in this case), and the rather low 540mA Ripple Current spec indicates the ESR is rather high (also a good thing, I think, as it more resembles what the stock cap was probably like back in the day).  And yet it's rated at 105°C and for 10,000hrs, and from a well respected maker of aluminum electrolytic, making that cap a wonderful pick, in theory.  I will purchase it and put the theory to the test.

By the way, here are photos of the PSU inside the HD20SC drive enclosure, and the bottom photo clearly shows the 10V 2200uF Rubycon I've been talking about.  The beep sound, to my ears, comes from the area of those two transformers nearest to the cap:

https://www.icloud.com/sharedalbum/#B0wGQOeMmGzwJPA

And here is the full list of caps showing the stock capacitance, voltage spec, physical size and lead spacing, with my replacement capacitor part number shown in parenthesis at right:

C226: 22uF 35V, D=5.2mm, Leads=5mm -- (Mouser: UHE1H220MDD1TD)
C202: 47uF 25V, D=5.2mm, Leads=5mm -- (Mouser: EEU-FR1E470B)
C222: 47uF 25V, D=5.2mm, Leads=5mm -- (same p/n as C202)
C109: 150uF 400V, H=32mm, D=25.8mm, Leaders s=10mm -- (Mouser: 860021383023)
C110: 4.7uF 350V H=32mm, D=12.8mm, Leads=5mm -- (Same as SONY C164 - Mouser: UPM2G4R7MHD)
C210: 330uF 16V, D=8.1mm, Leads=5mm -- (Mouser: EEU-FR1E331B)
C215: 470uF 10V, D=8.1mm, Leads=5mm -- (Mouser: EEU-FR1E471YB)
C213: 22uF 100V, D=10.2mm, Leads=5mm -- (Mouser: UBT2A220MPD1TD)
C214: 2200uF 10V, D=12.7mm, Leads=5mm -- (Mouser: UHE1C222MHD1TO) <-- the problem guy!
C209: 2200uF 16V, D=12.7mm, H=30mm, Leads=5mm -- (same p/n as C214)

I knew I needed a "16V" 2200uF cap for C209, so I thought it no harm in going with the same cap for C214 too, but that wasn't a good assumption.

Anyway, I took measurements more than once with my DE-5000 to make sure of the accuracy.  It took several minutes for the DF value to settle though.  It kept rising and rising.  But the biggest frustration with this meter is that it won't show me enough ESR info at 100Hz or 120Hz.  The tenths place isn't enough.  I want to see down to 1mΩ!  I can see that on smaller capacitance caps which allow me to crank up the test frequency, but this large capacitance caps don't let me use anything higher than 120Hz.

[The Electrician]

I only know what I can measure, and it's 0.2-ohm according to my DE-5000, @120Hz.  I cannot measure higher frequencies, unfortunately.  (My DE-5000 fails to do that on most large capacitance devices.)  And again, I lack the physical space to add a series resistor.  And on top of that, I'm not really sure what wattage rating would be best even if there was space for such a resistor.  Do you?  (No, I don't have a schematic.)

The DE-5000 has no problem measuring ESR at higher frequencies--you just have to set it to measure Rs instead of Cs and ESR.

Here's a Nichicon 2700 uF measured on a bench meter at 1 kHz:



Here's what my DE-5000 measures at 1 kHz:

[JDW]

The DE-5000 has no problem measuring ESR at higher frequencies--you just have to set it to measure Rs instead of Cs and ESR.

Thank you for the excellent tip!  I'll starting doing that from now on.

By the way, I have a Fluke 8845A benchtop meter with 2x4-wire test leads to take precision resistance measurements.  Is it safe for the meter and the cap to measure ESR on large capacitance caps with the meter in Ohms mode?

[The Electrician]

The DE-5000 has no problem measuring ESR at higher frequencies--you just have to set it to measure Rs instead of Cs and ESR.

Thank you for the excellent tip!  I'll starting doing that from now on.

By the way, I have a Fluke 8845A benchtop meter with 2x4-wire test leads to take precision resistance measurements.  Is it safe for the meter and the cap to measure ESR on large capacitance caps with the meter in Ohms mode?

You can't measure ESR with a DVM--it measures resistance using DC excitation.  ESR is measured with AC excitation.

[JDW]

You can't measure ESR with a DVM--it measures resistance using DC excitation.  ESR is measured with AC excitation.

Which implies that the "Rs" mode of the DE-5000 is in fact not a typical resistance checking mode used by a DMM but rather a special mode of the DE-5000 which can measure ordinary resistors as well as capacitor ESR due to the fact the DE-5000 uses AC (e.g., 120Hz, 1kHz, etc.).  And despite the fact the DE-5000 uses AC excitation to measure resistance, it's resistance-checking accuracy (for normal resistors, not capacitors) should still be on par with resistivity checking accuracy of high quality DMMs, is that correct?

[The Electrician]

You can't measure ESR with a DVM--it measures resistance using DC excitation.  ESR is measured with AC excitation.

Which implies that the "Rs" mode of the DE-5000 is in fact not a typical resistance checking mode used by a DMM but rather a special mode of the DE-5000 which can measure ordinary resistors as well as capacitor ESR due to the fact the DE-5000 uses AC (e.g., 120Hz, 1kHz, etc.).  And despite the fact the DE-5000 uses AC excitation to measure resistance, it's resistance-checking accuracy (for normal resistors, not capacitors) should still be on par with resistivity checking accuracy of high quality DMMs, is that correct?

After turning on the DE-5000, pressing the "LCR AUTO" button repeatedly gets us to the "Rs" mode.  One more push after that and we arrive at the "DCR" mode.  The "DCR" mode is the same as what a DMM does, but neither the "Rs" mode (AC excitation) nor the "DCR" mode (DC excitation) will have the kind of accuracy that your 6 1/2 digit DMM has.  The accuracy spec of the DE-5000 is given in the user's manual, and it's what one would expect from a low cost handheld device like the DE-5000.

The DE-5000 does quite well for what it is, but it can't compare to a bench LCR meter.

[fzabkar]

ISTM that, if you really want a capacitor with ESR, then choose a capacitor with an ESR of 0 and add an external resistor (as has already been suggested). However, be aware that any internal ESR causes heating, which reduces the life of the capacitor. ISTM that the squeal heard by the OP might be due to inrush current, in which case an inductor may be the solution, as well as providing additional filtering.

[JDW]

After turning on the DE-5000, pressing the "LCR AUTO" button repeatedly gets us to the "Rs" mode.  One more push after that and we arrive at the "DCR" mode.  The "DCR" mode is the same as what a DMM does, but neither the "Rs" mode (AC excitation) nor the "DCR" mode (DC excitation) will have the kind of accuracy that your 6 1/2 digit DMM has.

But when comparing the accuracy of the DE-5000's "Rs" mode versus "DCR" mode when checking the resistance of normal resistors (not ESR), is there a difference in accuracy?  Or will AC excitation show largely the same resistance value of a given resistor as DC excitation?

ISTM that, if you really want a capacitor with ESR, then choose a capacitor with an ESR of 0 and add an external resistor (as has already been suggested). However, be aware that any internal ESR causes heating, which reduces the life of the capacitor. ISTM that the squeal heard by the OP might be due to inrush current, in which case an inductor may be the solution, as well as providing additional filtering.

Thank you for the suggestion, but as I mentioned before, there is not enough physical space to add such.  And wouldn't adding a series resistor (I still don't know what Wattage would be best, by the way) also add unwanted ESL?  Then again, you seem to suggest that added inductance would be a good thing in this case.  However, I don't know what that inductance value should be.  But again, there's no space to add resistance or inductance (separate from what the replacement capacitor provides inside itself), and trying to add resistance or inductance with SMD parts on the bottom (solder side) part of the board would require an undesired amount of cutting.  Please see my previous post where I provide a link to numerous photos of the PCB so you can see what I mean.

[fzabkar]

I see that there is already an inductor at L205, but it's on the output side of the capacitor.

There are two links visible on the underside of the PCB. Would it be feasible to replace them with a single, axial, high frequency choke?

Something that looks like this …

https://cpc.farnell.com/epcos/b82111ec20/inductor-axial-7-0uh/dp/FT01259
https://www.tdk-electronics.tdk.com/inf/30/db/ind_2008/b82111e.pdf

[JDW]

I see that there is already an inductor at L205, but it's on the output side of the capacitor.
There are two links visible on the underside of the PCB. Would it be feasible to replace them with a single, axial, high frequency choke?

It seems you are suggesting I remove L203 and L205, measure their inductance, then replace them with physically smaller axial chokes of the same inductance such that I would then have more physical space to include a series resistor with C214.  Is that correct?

Even so, I still am not sure how to calculate what the wattage rating of the series resistor should be when doing that.  Do you?