Expected output power from SF Pro Reverb?

[Brian Swatton]

I've an old silverfaced Pro Reverb in for repair. I've not had to work on one of these before, but have repaired some other valve amps, including Fenders.

In my initial tests with a 1Khz sinewave I'm only achieving about 16W into 4 ohms before compression, 18W into 8. I was expecting about 40W, or is that just for odd music peaks?

The closest circuit diagram to what I have is for the AA1069, but the 4k7 and 1k in the HT chain are 10k and 2k2 in this unit. The chassis bears the stamp "A 19671"

A word from a Pro pro would be appreciated

[Enzo]

AA1069 is a chassis type - a revision level if you prefer. A19671 is the serial number of the amp.

This is a guitar amp, not a hifi, put a meter across the load and turn it up until it doesn't go any higher, and calculate power from voltage on the meter and impedance. See if that comes closer to your expectations.

Some techs demand power output ratings only be made with no distortion. But some amps are not distortion free at ANY level, so I suppose those would have zero output. I look at amp power ratings as a means to get an idea how loud it will be when played normally. And that will be anything by clean.

If you still can;t get over a 14 watt figure, then we have trouble.

[Brian Swatton]

That's brilliant Enzo thanks.

I know what you mean about not being hi-fi. Though in my electronics training I was taught to measure power for a clean sinewave before clipping/compression, I'm also a guitar player.

Edit: I can indeed achieve 40w rms though of course it's heavily clipped by then.

Thanks again

[MWJB]

Put 20mVAC @ 1000Hz on the input grid, bright switches off, all controls 50% rotation, reverb & tremolo switched off at foot switch, tremolo speed at full clockwise rotation.

You don't say what dc voltages you have? This would be the first thing to do.

[Brian Swatton]

Hi MWJB,

Thanks for the response. I'm not actually trying to diagnose a fault right now, rather taking a measure of current conditions before I try to reproduce the problem (losing clean headroom after 20mins or so of playing loud).

I have done some quick checks around HT chain and bias currents, nothing seems too out of place. The caps are very old, but read ok for capacitance, one has a bit more ESR than I'd like to see, but is doing it's AC bypassing job. I will look for replacements for these anyway, just because of their age.

Thanks anyway

[pdf64]

losing clean headroom after 20mins or so of playing loud

That old chesnut can be a tough one to resolve, to the customer's satisfaction.
Hopefully the regular bench checks will reveal an obvious problem.
New 6L6 can help, the Winged=C= brand are recommended.
However, bear in mind that speakers dissipate power, heat up, and their voice coil resistance (and presumably impedance) increases, thereby acting to reduce the amp's power output. So better speakers may be needed.
Check how hot the magnets get after heavy use; they can get to too hot to keep your hand on for more than a few seconds.

[MWJB]

"Thanks for the response. I'm not actually trying to diagnose a fault right now, rather taking a measure of current conditions before I try to reproduce the problem (losing clean headroom after 20mins or so of playing loud)."

It's likely that this problem exists because of current conditions but only manifests itself after 20 minutes...in other words if everything was fine regarding current conditions, there wouldn't be a problem after 20 minutes. So it would still be good to know dc voltages & bias current.

If the electrolytic caps are more than 20years old then they need to go, we're talking about the power supply caps in the dog house, cathode bypass on the circuit board and bias electrolytics. You probably won't find 70uf/350v caps that easily, use 100uf, likewise the bias electrolytics may be funny values like 70uf/70v, use 100uf/100v.

[Brian Swatton]

That old chesnut can be a tough one to resolve, to the customer's satisfaction.
Hopefully the regular bench checks will reveal an obvious problem.
New 6L6 can help, the Winged=C= brand are recommended.
However, bear in mind that speakers dissipate power, heat up, and their voice coil resistance (and presumably impedance) increases, thereby acting to reduce the amp's power output. So better speakers may be needed.
Check how hot the magnets get after heavy use; they can get to too hot to keep your hand on for more than a few seconds.

Hi, thanks for the input.

6L6's and rectifier have been changed by another eng apparantly. I did about 30 min playing with it the other day to see if the problem showed (no, but don't think I was really loud enough, can go louder today). I had wondered about speakers myself, and using temperature as an indication - I measured only 7'C above ambient for both.

[Brian Swatton]

...
It's likely that this problem exists because of current conditions but only manifests itself after 20 minutes...in other words if everything was fine regarding current conditions, there wouldn't be a problem after 20 minutes. So it would still be good to know dc voltages & bias current.
...

Fair comment, though I was only asking what power output I could expect from this unit.

Biases were a bit low and unbalanced at 17mA and 25mA, I have balanced them, but not increased them. They are now about 20mA. Crossover distortion is quite evident, so I may try increasing the bias a little at a later stage, but I don't believe it's relevant to the problem at hand. Please correct me if I'm wrong there.

I can only remember the HT to the 6L6's was 445v, and bias was 50v, the measurements were made quickly the other day. It is back in the cabinet at the moment. I'll get it out again and make notes of the others in a bit, but probably not for an hour or so as I have other things going on.

[MWJB]

"...but I don't believe it's relevant to the problem at hand. Please correct me if I'm wrong there."

You are very probably right. I just get a bit OCD about order of business when it comes to troubleshooting! ;-)

[Brian Swatton]

You are very probably right. I just get a bit OCD about order of business when it comes to troubleshooting! ;-)

No problem, an engineer has to follow his line of thinking eh.

I measured some relevant voltages :

HT 444v (508 on stby)
6L6's: plates 440v, scrn grids 440v, grids -50v

HT 'C': 412v to :
12AT7 phase splitter: plates 308v & 302v, cathodes 150v

HT 'D': 341v to:
Input 7025's: plates 221-227v, cathodes 1.7v

12AT7 reverb driver: plates 407v, cathodes 5.7v

7025 reverb amp: plates 232v & 229v, cathodes 1.7v

12AX7 Vibrato: plates 1) 260-285v varying 6) 360v, cathodes 1) 1.8v 8) 12.2v


The 12AT7 phase splitter readings are a bit out to what I think have on this smudgy old scanned schematic (which I also know isn't quite the right one), I also noticed that as the units output went into clipping, the distortion could also be seen at the phase splitter outputs. So I tried a new one just in case, with no significant difference.

I've also played it again, for an hour this time, loud enough to be well into clipping, while monitoring output on scope and acV meter, I could hear or see no changes.

[pdf64]

I also noticed that as the units output went into clipping, the distortion could also be seen at the phase splitter outputs

That's perfectly normal, not indicative of a problem. It's because the 6L6 reach saturation when the Vac from the phase splitter tries to drive the grids positive, with respect to the cathode, eg roughtly 50Vac p-p, given that the grids sit at -50Vdc WRT the cathode. At that point, the normally very high grid impedance falls to a 1-2k (think of it as a forward biased diode). This will clamp the output from the phase splitter, which has a much higher output impedance than that.

Re the phase splitter Vdc, check all (ALL) resistor values, as they will drift up, possibly considerably so, you may well find some that are double the nominal.
So maybe the resistors in the cathode circuit have increased in value.

[Brian Swatton]

That's perfectly normal, not indicative of a problem. It's because the 6L6 reach saturation when the Vac from the phase splitter tries to drive the grids positive, with respect to the cathode, eg roughtly 50Vac p-p, given that the grids sit at -50Vdc WRT the cathode. At that point, the normally very high grid impedance falls to a 1-2k (think of it as a forward biased diode). This will clamp the output from the phase splitter, which has a much higher output impedance than that.

Ah right, understood.

Re the phase splitter Vdc, check all (ALL) resistor values, as they will drift up, possibly considerably so, you may well find some that are double the nominal.
So maybe the resistors in the cathode circuit have increased in value.

All cathode resistors measure correct for their colour codes, but have obviously been replaced and are not the same values as the drawings I have. The 2 x 1M to the grids are actually 330k, and, more importantly I think, the 22k is actually a 33k.

I'm still not entirely clear, should I be getting more than 16-18w before clipping? I know rms power need not be undistorted, but I was personally expecting a bit more clean than I'm getting.

[pdf64]

Schematic see
http://www.webphix.com/schematic%20h...1069_schem.pdf
It should give 30 watts, maybe more depending on how well the HT holds up under load.
With the bias adjusted for about 30-35mA per 6L6, the sinewave should be reasonable up till about 10Vac rms, by 11V the peak/s may start rounding off.

[MWJB]

33K for the PI tail is not right, go back to 22K 1W. Whilst you are around the PI check the integrity of the ground from the 100ohm NFB load resistor to brass plate.

Can you confirm the signal voltage that you are applying and control panel settings please?