Bandmaster AB763 low output w/ weird distortion (scope shots attached)

[Randall]

I don't understand where the positive potential comes from. I usually think of a cathode as being negative, or pretty close to it.

[Enzo]

Negative? In common gain stages the cathode sits at a positive volt or two. In these tail pair PIs, 100v is quite common.

Look at the circuit, the 100k plate resistor drops 200v, that means 2ma of current - Ohm's Law. double that because ther are two. 4ma because they share the same cathode circuit. The current through the plates and the tube must also flow through the cathode. The cathode is 22k to ground plus a few extra ohms. V = IxR = 0.004 x 22,000 = 88v Pretty close to the amp voltage. Actually using your exact numbers I get 101 volts. The current through teh circuit all flows the same direction, so that 101 volts will be positive on the top end of the 22k.

[Tom Phillips]

...Is that input signal 10mV RMS or p-p?...

RMS and the exact value is not critical. ~10mV is used as a test signal because as you follow it through the stages of the amp it does not cause any of the stages to clip.

...Voltages at pin 4 on the 6L6's are basically the bias voltage and symmetrical ...

Nope. The bias voltage is applied to the grid on pin 5. Pin 4 is the screen and, in the Bandmaster, the voltage will be (Edit: Should be) just a little less than the plate voltage.


Let's reserve final judgement until you can test the amp with the 4 Ohm load. In my experience the old Bandmasters would produce about 35 Watts at the onset of visible clipping at the output. That would be about 12 VRMS across the 4 ohm load.

[J M Fahey]

Thanks for the info Juan Manuel, I thought an 8 ohm load would be fine but I'll make a 4ohm one and re-test. When I set the maximum output volume (as shown in my scope picture), I turn it up as you describe to see the heaviest clipping, and lower it slowly until the sine wave isn't clipping at all. That's when I take my RMS measurement to calculate the output wattage, not at 10-15% clipping as you describe. I thought output power was measured with an undistorted sine wave?

It is
But then there´s Guitar Players, Marketing, hard to meet expectations, plus Tubes very real self imposed current/power limiting.
Plus historical tolerance for high Tube distortion numbers.

Combine that and expected and often quoted tube power ratings are hard to meet ... unless accepting quite a bit of clipping.

Personally I have *rarely* measured a 100W Twin, 100W tube Marshall, 50W Bassman, even less absolutely incredible 60W Deville, etc. , without quite a lot of clipping, that is.

Tom Philips said:

Let's reserve final judgement until you can test the amp with the 4 Ohm load. In my experience the old Bandmasters would produce about 35 Watts at the onset of visible clipping at the output. That would be about 12 VRMS across the 4 ohm load.

Original Marshall JTM45, built with bargain-of-the-week surplus Radio Spares "universal" transformers, were closer to 30W undistorted.

Besides, just check datasheets, output power for EL85, EL34, 6L6, etc. was often rated at around 5% distortion.

Here´s a couple *official* power ratings, from serious companies:

Fender HRD (which "might" be excused because "tubes distort anyway"):


Cheating Fender SS amplifier, Princeton 112 , where it is *easy* to get 0.5% distortion or less just before clipping, so it´s unjustified to quote any higher:


Even less justified "as serious as can be" mainstream High Tech Semiconductor company, here they display unclipped power rating as low as 0.02% so no excuse to display just besides it *really* clipped 10% distortion: TDA2050 power rating:


besides, Digital scopes disply a somewhat funky image by themselves, so not easy to find the exact no clipping point, that´s why I asked you to find the real gross clipping points and then lowering drive, but still allowing some clipping, which as you see is an industry standard anyway.

Can´t find them now, but some old Kustom schematics said something like: "rise drive until you get 20V RMS (100W into 4 ohms) ... then if you wish lower it until you get a pure sinewve ... around 18V RMS"
Quoting fully from memory, but that was the general idea, and they acknowledged some 75W RMS ... yet nobody ever complained that old Kustoms were not loud ... quite the contrary.

[J M Fahey]

I don't understand where the positive potential comes from. I usually think of a cathode as being negative, or pretty close to it.

No, the grid is the one more negative than cathode, so cathode is more positive than grid.

And it´s a relative voltage.

In that amp, 22k is NOT the biasing resistor but the "long tail" one; the real biasing one is the very normal 470 ohms one: cathodes are connected to its upper end, grids to the lower one (through reference resistors), voltage *difference* is caused by the roughly 4mA going through it.
So top will be 0.004*470=1.88V higher than bottom so grids will be 1.88V more negative than cathodes.

Maybe we should ask them , but I guess 12A(T/X)7 will be *happy* with that

[TigerAmps]

Hi - This discussion seems to have two moving parts: The amp does not work properly and a discussion about rated output and distortion.

Focusing on the amp working (taking the Enzo approach of isolation):

1) Are the caps (power supply and pre-amp) original? Are they around 50 years old (based on being a black face amp)? If so, I would definitely replace them all!
Aged caps can cause some strange issues; I actually prefer when they flat out fail or leak their guts out. It takes the guess work out of the mix!
The last 1972 Bassman 100 I opened, every cap in the power supply ruptured and leaked it's guts!

2) I would also try a different output transformer. Just remove the 3 primary wired and tack in a known working OT.

3) disconnect the pre-amp from the PI and check the signal coming out of the pre-amp. Is it clean or distorted?

4) inject signal into the PI. Is the output distorted?

5) measure the idle current in the power tubes. Is it reasonable, low, high?

6) Check the grounds on the power tube sockets (and all the grounds) - re-solder them. Re-tighten the ones with the nuts and star washers.

7) Then there is always freeze spray on resistors and caps (usually the last resort unless there is crackling or other strange noises).

The signal is usually the easy part of old Fender amps; Tracking down the hum and bad grounds is usually the more time consuming part.

[xtian]

Shot-in-the-dark dept.: OP reports 10v p-p AC ripple at B+. I'm used to seeing ~3v. So, @waspclothes, have you tried clipping in a fresh 47uF cap across your reservoir to see if it solves your issue?

[Enzo]

10v p-p is 5v peak, and since it is ripple, the "RMS" value of 5v peak is about 3.5v. SO his 10v p-p is about the same as your 3v.

[waspclothes]

Okay, I had a chance to try a few more things:

4 ohm load - resulted in a 6.4VRMS output (10.2 Watts)
100mVRMS input signal - output still clips at the same spot, results unchanged.
extra 47uF at B+ - no change in output

[alexradium]

check the OT,without power tubes, inject 10VAC on primary,and measure voltage on secondary with load.
Are you sure tubes are good?
Do you know that main spkr jack is shorted if nothing is connected?

[waspclothes]

check the OT,without power tubes, inject 10VAC on primary,and measure voltage on secondary with load.
Are you sure tubes are good?
Do you know that main spkr jack is shorted if nothing is connected?

Hi Alex, I only have this function generator ( Velleman HPG1 1MHz Pocket Function Generator ) at my disposal - I'm not sure if it would provide enough current to test the OT? I will try though. I have an 18VAC wallwart, maybe I'll try that too and post results.

As for the tubes, yes these are brand new tubes - tried multiple pairs to rule that variable out.
Yes, I'm plugged into the "speaker" jack so I assume the jack isn't shorted to ground. The wiring looks to be original, but will check. I did check the 820R and 100R resistors in the NFB section to make sure nothing was funky there, and it all checked out - so did all the resistor values in the P.I.

[alexradium]

Hi Alex, I only have this function generator ( Velleman HPG1 1MHz Pocket Function Generator ) at my disposal - I'm not sure if it would provide enough current to test the OT? I will try though. I have an 18VAC wallwart, maybe I'll try that too and post results.

As for the tubes, yes these are brand new tubes - tried multiple pairs to rule that variable out.
Yes, I'm plugged into the "speaker" jack so I assume the jack isn't shorted to ground. The wiring looks to be original, but will check. I did check the 820R and 100R resistors in the NFB section to make sure nothing was funky there, and it all checked out - so did all the resistor values in the P.I.

you can inject your signal generator in any hifi or guitar practice transistor amp and connect the out to primary monitoring voltage with meter

[waspclothes]

you can inject your signal generator in any hifi or guitar practice transistor amp and connect the out to primary monitoring voltage with meter

Okay, I ended up using an AC wallwart to inject a 60Hz signal into the OT, here's the scope shot:



Input is 14.9 VRMS and output is .546 VRMS.

[Jazz P Bass]

I do believe that your OT is toast.

[waspclothes]

I do believe that your OT is toast.

Thanks JPB, what sort of ratio should I have seen when I tested that OT? - for future reference.