SWR Studio 220 Capacitor Question

[g1]

  Best to stick with the value that is in there.  If they are running at a voltage too close to the rated voltage, then you should go with a bit higher voltage rating.
And higher temperature rated caps (105C instead of 85C) should last longer.

[Roly]

Any suggestions for a "correct value" to use as replacements? There is probably a formula to figure out how to determine proper cap size in the power supply but I'm not sure how to do that....thanks.

mmmm ... yeah.

You can take the rail voltages (+/-55V), or nominal output power rating if known, and the output impedance, and work out the maximum RMS output current, then select your filter caps to have ripple current rating higher than this ('tho you might find this an elusive figure for most caps).

For example the ST-220 is specified as 220 watts RMS @ 4 ohms.

P = I²R

(P/R)^0.5 = I

(220/4)^0.5 = 7.4 amps rms


http://au.mouser.com/Search/Refine.aspx?N=12013589

[Enzo]

Crap.  I stopped by my shop recently and meant to take a thumb drive of all my schematics.  Unfortunately I took it back out of my pocket and put it back to look something up, then left it there.   I am sure there is a schematic for it.  Perhaps using the correct value will allow the caps to fit on th board instead of keeled over like that.

Before just throwing new caps at it, you might check to see if the cap is really the source of the hum.  If only one of the caps measured funny, then presumably the two supplies will have differing amounts of ripple.  Do they?

[g1]

Oops! Sorry, when I said go with the values in there, I forgot these were not the stock parts.   
The supply schematic attached says they were 4700uF for this model.

[Hawk]

I will respond in more detail to these posts soon as I've been very busy but wanted to say thanks a million for the power supply schematic!! Talk about helpful! I was struggling with that. :dbtu:

I do have some time:

Roly, thanks for the mouser link but what temps/load life would you put in? Without all filters I can't "apply filters".

Enzo, specifically, to test for ripple current, using a voltmeter set on AC, where would you put your test leads? (Black to ground, red to?). I'm thinking at the positive end of the furthest supply cap. Is that right? I've read that you don't want more than 2vpp ripple?  I have a scope, so wondering if I would do the same thing--probe on far side of scope, scope ground to chassis ground?

Having said that I realize I should also test for dc voltage on coupling caps...Thanks!

[Roly]

what temps/load life would you put in? Without all filters I can't "apply filters".

I think you can select "any", but I don't use the company myself, just included it to show an example of ripple current rating.

to test for ripple current

We need to be clear that Enzo is talking about ripple voltage which is across the capacitor sitting on the supply voltage; I was talking about the ripple current which is through the capacitor and a function of the load current.

Using your CRO you will need to select AC input coupling, CRO ground to amp ground, then probe as you like.

We can't just say "2V max" because the ripple voltage will also depend on the loading on the supply.  When there is no load there won't be any ripple at all, at idle a little, and moving on up to full output the ripple voltage will get larger as the caps are discharged more by the amp between being refreshed via the rectifier.  This ripple voltage will be a sawtooth, so the most sensible way to measure/quote it is peak-to-peak.

[Hawk]

Thanks Roly. So when you say: "When there is no load there won't be any ripple at all, at idle a little, and moving on up to full output the ripple voltage will get larger as the caps are discharged more by the amp between being refreshed via the rectifier"

Do you mean the speaker load and actively playing the bass as I watch my scope? I suppose I could inject a signal into the amp with speaker attached and probe from there...definitely easier hands free...make sense?

[Hawk]

At the top of the power supply stage schematic, on the +/- 15 power supply, we have two IN4004's, one on the neg and one on the pos. rail. If we already have a neg and positive voltage coming off the rectifier why do we need these diodes? Voltage dropping? And then what is the purpose of the zener diodes at the far end?  Thanks for your patience!

[J M Fahey]

That simple supply creates +/- 15V out of main rails, which are much higher, between +/-6- to 75V .

Voltage is dropped through power resistors R1/R2 and voltage is stabilized at +/- 15V using Zeners 1N4737 , which must be 7V5 each .

I guess that the purpose of the 1N4004 diodes is the following:

in many amps, there's a harmless but mildly annoying noise when amp is turned off, a few seconds later you can hear a "squeeeeeiiiiikkkk" type sound, clearly an oscillation.

I guess that when amp is turned off, main rail voltage drops quickly, preamp rails do not because Zeners keep them stiffly at 15V, in a way "n isolating" them from main rails ... until main rails drop below 15V (they can still produce quite a few watts at that level, over 15W  ) and isolation disappears, R1/R2 are very low value and 15V filter caps are puny, only 100uF ; at that point preamp is directly coupled to main rails which are still feeding a loaded power amp ... the conditions are set for some oscillation or motorboating.

For a couple seconds only, of course, until fully discharged.

The added diodes become reverse biased once main rails drop below 15V and prevent that.

Just an educated guesss, of course.

[Hawk]

Thank you. On the pos. leg of both 6800 uf cap I see 40mv peak to peak sawtooth waveform,. So far I don't see the sawtooth anywhere else.

With bass unplugged, Gain at 6, Master at 6, the hum increases into a loud buzz when I turn the Treble to 10. When I crank the Bass to 10 it gives me a very a low "bassy" hum. Any thoughts? Oscillations?

[Roly]

Thank you. On the pos. leg of both 6800 uf cap I see 40mv peak to peak sawtooth waveform,. So far I don't see the sawtooth anywhere else.

With bass unplugged, Gain at 6, Master at 6, the hum increases into a loud buzz when I turn the Treble to 10. When I crank the Bass to 10 it gives me a very a low "bassy" hum. Any thoughts? Oscillations?

I do hope you mean the +ve end of the +ve supply cap and the -ve end of the -ve supply cap, otherwise one of them is in the wrong way around!  (not impossible with a somewhat bodgy retro-fit)


All hum (and noise) tests must be done with the input shorted (to sort any internal hum from external pickup), and with the tonestack set to neutral, 5/10, 12 o'clock.

Are they?

[J M Fahey]

Same here.
With bass unplugged, and nothing else I imagine all sorts of horrid things, you need a shorted plug in the input or nothing plugged in the amp input while your description hints at a cable plugged in the amp input, nothing at the other end ("no bass")

[Hawk]

Thanks.

A few things: by shorted you mean attaching a 1/4" jack but with out cable attached? (You're right about the noise being worse with the guitar plugged in. Without anything plugged in the amp, the hum is much better I just want to know that I can reduce the hum as much as possible. Again, increasing either the bass/treble increases the hum considerably. I don't have a new amp to compare it to so I'm not sure how much is too much).


I took the ground lead of the scope and attached it to chassis ground and the scope lead to the positive lead of the cap, then did the same with the other cap. However, now that I look at the schem. I see that the pos of one cap goes to the +ve rail and the neg side of the other cap goes to -ve, so my inclination is to check the -ve as well which I didn't do before. Is 40mv ripple acceptable? Also shouldn't I really be checking at R1/R2 as the caps are supposed to reduce ripple and the result would show at R1/R2? Thanks everyone!

(I have assumed that the caps are wired in correctly but I have will check that out later when I'm home.)

[Roly]

Without anything plugged in the amp, the hum is much better

In most amps the input sockets are arranged to short the input when nothing is plugged in.  Attempting to deal with internal amp hum with an open circuit input picking up all the external noise in the world is a bit like running an air conditioner or heater with all the windows and doors open.

Is 40mv ripple acceptable?

Under what conditions?  With the amp at idle that would be quite acceptable.

Looking at <SWR Studio 220  ps.pdf> there are a couple of filter caps for the main +ve, C1 & C3, and -ve, C2 & C4, supplies.  R1 and R2 come off the same points, (A) and (B).  Generally speaking we assume wire or PCB traces have zero resistance, so what you measure at the caps should be identical to what you measure at R1 and R2.

These are the supplies for the main power amp which is fairly insensitive to hum/ripple.  With the main or master volume turned down whatever hum and noise you can hear will be due to the power amp alone.

The other supplies to check are the +/-15V for the preamps (across the 1N4737A zener pairs) the hum here should be quite low, but similarly the split supplies imply op-amps in the preamp and these are also highly resistant to hum and noise on their supplies; so if you have excessive hum and noise our first guess is that it's getting into the signal path somehow, say due to poor grounding or layout.

So we are not trying to hit a moving target (and bamboozle ourselves) we conduct all tests with all EQ set to flat, 5/10, 12 o'clock, and any Fx off/minimum.

Representative gain and master volume settings would also be 5/10, 12 o'clock.

With the master at 5/10 but the gain at 0/10 you will add the hum/noise contribution of the preamp back end, and with the gain at 5/10 you will add and get and idea of the hum/noise contribution of the preamp front end.

Not too may preamps are totally silent, particularly if they are high gain, but you must remember that hum and noise performance is referenced to full output power level.  This is a high power amp intended to fill an arena and that under those conditions the nearest audience member may be some tens of metres away from the backline and the residual hum and noise quite inaudible.  With such an amp I would expect that its hum/noise up close in a quiet bedroom or workshop would seem quite high.

But you can get an absolute handle on it by actually measuring the hum and noise on the output with the gain, master, and EQ controls all set for 12 o'clock.  This will give you a representative peak-to-peak value which you can compare to the full output peak-to-peak (the +/-ve supply rails, in this case I think +/-60V, and work out in dB just how far down your hum and noise is on full output.  You may discover that it's within a reasonable spec.

dB = 20 * log₁₀(V1/V2)

[Hawk]

Great, thanks. Excellent info.

I'm going to try and drill down here so I understand everything that is being said.

"The other supplies to check are the +/-15V for the preamps (across the 1N4737A zener pairs) the hum here should be quite low"  Here, using my scope  I found some ripple but very low, approx 20mv. Even with a 50 ma 1khz signal injected the ripple wasn't much more. Acceptable?

"With the master at 5/10 but the gain at 0/10 you will add the hum/noise contribution of the preamp back end, and with the gain at 5/10 you will add and get and idea of the hum/noise contribution of the preamp front end." Tried this but no audible difference. A good thing, right?

"But you can get an absolute handle on it by actually measuring the hum and noise on the output with the gain, master, and EQ controls all set for 12 o'clock.  This will give you a representative peak-to-peak value which you can compare to the full output peak-to-peak (the +/-ve supply rails, in this case I think +/-60V, and work out in dB just how far down your hum and noise is on full output.  You may discover that it's within a reasonable spec. dB = 20 * log10(V1/V2)" 
Okay, this really interests me but not quite  sure how to proceed. Do you mean V1 is the 12 o'clock voltage and V2 is the voltage with the gain and master full-on?  With the settings at 12 o'clock, and the speaker output attached to a 4 ohm dummy load, and nothing plugged into the amp, I get a value of 13 mv (using my voltmeter). With the gain and master at full volume I read a voltage of 70 mv.  Am I on the right track before I start play with the log calculation?  Thanks for all your help.

Having listened to and tried to apply everything so far I'd say it's fair to assume that this amp doesn't have a lot of hum, and, at first, a lot of it was coming from my bass plugged into the amp. The controls are dirty so I'll apply some deoxit. Otherwise, at this point, I'd say this amp is doing well.


P.S. How do I insert quotes on this forum?   Also, how can I save all these posts as they are great?