Techincal Question about Caps

[effigy]

Ok, here's the deal...

I've got a very nice 1969 Solarus which works and sounds great. Even though it seems fine right now, I'd still like to replace the two 50uF 150VDC caps since they're nearly 40 years old. The problem is that I can't seem to find any new, used or NOS caps that match these values. The closest that I can find is a 47uF 160VDC or a 100uF 160VDC. Would it be safe to use either of these, or should I keep searching until I find an exact match? Or, if neither of those are suitable, what would recommend?

Sorry if this question seems stupid, I don't know much about electronics. Although, I have helped a friend replace the filter caps in his Fender Twin, so I know the proper procedure/precautions to take such as draining them and whatnot.

Thanks,
Brian

[JoeArthur]

Official disclaimer - I believe these two caps are the ones for the bias supply filtering as that is the only place I can see those values used.

Yes, either of those values would be appropriate for replacement.

You can always replace a cap with a lower voltage rating with one of a higher voltage rating.  The difference between a 150 volt and 160 volt rating is nothing to worry about. Just don't go the other way.

The 47 mfd is also an appropriate replacement for a 50 mfd.  Caps, like most other components are always rated with a tolerance and I have seen some caps with anywhere from 20% to 80%.  Believe me, those 3 microfarads are easily within tolerance differences.

Since this is a bias supply, even the 100% difference of 100mfd wouldn't be noticed by the circuit - simply because the rectification is provided by a solid-state diode.  Tubes rectifiers don't like too much capacitance in a filter component, but solid-state usually doesn't care.

Personally, given the choice, I would go with the 100mfd over the 47mfd, simply because the bias supply is a half-wave rectifier circuit and the more filtering the better.

But then I wouldn't replace them just because of their age. 

Electrolytics given regular use will far outlast the "normal time between replacement" recommendation provided by most techs - a recommendation based upon the flatness of a tech's wallet more than fact.

If you need to replace bias supply electrolytics, then the most obvious symptom would be an increase in 60hz hum in the amp - with the input shorted (normally no guitar plugged in) and the volume control maxed.  60hz is pretty close to the lowest b-flat of a bass guitar.  The next most obvious would be an inability to achieve the bias voltage required by the output tubes.

If these two symptoms do not exist, then you do not need to replace them... except of course unless it just makes you feel good about it. 

[effigy]

Thanks JoeArthur!

Yes, there is a bit of 60 cycle hum. All of the other caps have been replaced by the previous owner except for these two, so I figured that I might as well replace these as well.

----------

Here's another question for anyone who might know...

Usually there's a production date written on the inside of the chassis, but mine doesn't have a date, instead it's marked "HMP" (possibly HWP) or, if you look at it up-side down, it looks like JMH. Does anyone know what this means? I'd like to think that I have "James Marshall Hendrix's" amp, lol, but I know that's not the case.

Thanks for your help,
Brian

[rick.heil]

Oh if only.....not sure!  This could be fun!
(cue the detective music)

[Soundmasterg]

I wouldn't replace a bias supply cap with anything larger than a 100uf, because the time constant could be too excessive for that cap to be resupplied with voltage, and it could cause a situation where the bias is lagging behind the plate voltage, and could cause eventual tube failures/runaways. You can find more info about this in some of the amp books around, like Kevin O' Connor's TUT series.

Greg

[JoeArthur]

Can you be kind enough to provide a more specific reference where Mr. O'Connor states a concern about bias supply caps larger than 100mfd?

All I seem to be able to find is a concern about the size of the grid leak bias supply resistance, due to tube design values rather than anything to do with charging time constants. 

Thanks!

[Soundmasterg]

JoeArthur, I had to check again and its not in O'Connor's stuff that I saw it like I thought. Its in Dave Funk's book, and a couple years ago there was a discussion about it on Ampage that I think I saved. I'll have to see if I can find it.

The basic idea is that if its much larger than 100uf, the cap takes so long to charge that the bias supply lags behind the plate supply. Its not really much of a high current node there, so thats why so many people just used a 50uf and a half wave circuit. You get some hum reduction going to a 100uf and a full wave circuit, and you'll get a slightly more stable bias supply by going to the 100uf, but beyond that its very much a case of diminishing returns.

[JoeArthur]

JoeArthur, I had to check again and its not in O'Connor's stuff that I saw it like I thought. Its in Dave Funk's book, and a couple years ago there was a discussion about it on Ampage that I think I saved. I'll have to see if I can find it.

The basic idea is that if its much larger than 100uf, the cap takes so long to charge that the bias supply lags behind the plate supply. Its not really much of a high current node there, so thats why so many people just used a 50uf and a half wave circuit. You get some hum reduction going to a 100uf and a full wave circuit, and you'll get a slightly more stable bias supply by going to the 100uf, but beyond that its very much a case of diminishing returns.

What I don't get is the reasoning behind why it would take longer for a larger than 100mfd cap to charge.  And before you get talking about time constants, which I do understand, I don't see the resistance aspect of it which would cause it.

Simply because power supplies in solid-state equipment use much larger capacitance in the power supply without encountering problems in charging rates - and I'm talking 15,000 mfd and higher.

I do see where using a smaller cap would cause the problems you are talking about - because if it is small enough it wouldn't be able to hold a sufficient charge during the discharge cycle before the next charging cycle... but not the other way around.

And a power supply is a power supply - there is nothing magical in using it for bias or B+.

One thing I did find in one of the TUT books was a 10,000 mfd cap used in one of Mr. O'Connor's low voltage bias supplies.

And I would believe Mr. O'Conner over Dave Funk any day of the week.

Anyways... if you can find something I'd be interested in reading it.