OT impedance and break up relationship

[bob p]

If you want to talk theory, here's an Operational Characteristics plot for the General Electric 6L6GC. It shows the relative distortion levels for 4k2 loading vs. 2k1 loading.




As you can see:
a. Plate current will be higher (though not as much as you might expect, since the slope of the plate current line is low.)
b. Screen current is lower (again, not a whole lot of slope on the curve)
c. Power output won't change much (still in the flat portion of the power curve)
d. Distortion is higher: moves up from 9% to 11% under the stated test conditions.

Granted, these test conditions are at Eb = 250, so they won't exactly match your situation, but they should give you leg up in understanding that the load impedance might not deserve all of the blame that it's getting in this thread. If anything, the data shows that the 6L6 is very tolerant of a 50% mismatch to the downside. Of course, we've always known this.

If you really want to address this answer form a theory standpoint, it would help to take some real world measurements from the unit under test and draw the load lines rather than guessing about what's going on.

[Enzo]

Or scope the thing to see what the distortion looks like and where it arises in the circuit.

[bob p]

Or scope the thing to see what the distortion looks like and where it arises in the circuit.

There you go being practical Enzo.

I suggested signal tracing at the end of post #8 as the pragmatic approach to the problem, but that went over like a lead zeppelin. Since this is the Theory area and he asked for theory I thought I'd post the charts ...

[cjenrick]

a Fender breaking up too early now i have heard everything,

measure the ratio of the OPT. you can either use a signal generator or inject a signal while the amp is on,

disconnect the CT of the transformer if using the generator across the brn and blu leads,

once you have your ratio, square that number and multiply by 2 ohms, this will give you your plate to plate Z.

wire the speakers for 8 ohms and see what happens, series - parallel, might sound good,

does it have the bias balance control instead of the voltage adj? if so you may have the tubes running too mushy with no way to fix it, mod the bias circuit to BF specs.

maybe convert the inverter back also.


so what is going on with the load to the power tubes? well, when you switch from a 2 ohm sec to a 4 ohm, you are decreasing the turns ratio, lets say you have a 3K pri, 3K/2=1500 Z = 38.7:1 and 3K/4=750 = 27.3:1 so the Bassman will reflect 27/38= 70 percent of the 2 ohm load to the tubes or about 2K instead of 3K which means a bit more distortion at the low end, but maybe not enuff to cause the breakup you are hearing. and distortion goes down since you are generating twice the NFB voltage with the 4 ohm wind.


so before you drop 200 bucks, maybe get the scope out see what is shakin,

and feel the speakers to see how loose the cones are, sometimes they be flappin sick,

[Randall]

It is putting out 39 clean watts with the 4 ohm OT into an 8 ohm load. That's the best I can do at the moment.

Guy just got the amp recently. It sounds great to my ears, just like a BFSR should sound, it breaks up. He actually brought me two BFSR's, the other one is a lot less original and is in a homemade cab, but it has the proper OT, even the correct year, so I am going to swap them. That's why this is in the Theory forum, not Maintenance and Repair, I wanted to understand how the mismatch effected break. Now I know, not that much. But, he is concerned with the amp's value, so we will put the proper OT in it, and will tell him maybe a BFSR is not the amp for him.

Thanks for your input.

[Jazz P Bass]

FYI here is the Super Reverb RI schematic that includes AC & DC voltage measurements.

Super Reverb RI.pdf

[Dave H]

It is putting out 39 clean watts with the 4 ohm OT into an 8 ohm load. That's the best I can do at the moment.

If that's the case you could just rewire the speakers series/parallel for 8 ohms and call it done

It would make the OT primary 8k and the 6L6 data sheet has examples at 6k6 fixed bias (and 9k cathode bias) so it's not far off.
I've noticed on the scope that an upward mismatch as above is softer clipping with less crossover distortion than a downward mismatch.

[Enzo]

Are you maxing at 39 watts by feeding the input jacks or by feeding the power amp directly at the phase inverter?

And one last time: did you ever determine where the "breakup" is ACTUALLY occurring?

[bob p]

I wanted to understand how the mismatch effected break. Now I know, not that much...

According to that chart the move from 9% to 11.5% is an increase of 28%. That much of a change is significant to a hifi guy for sure, maybe not so significant to a guitar player, and probably not at all significant to a super reverb player as the THD from the tube is going to be swamped by speaker distortion.

Enzo: Are you maxing at 39 watts by feeding the input jacks or by feeding the power amp directly at the phase inverter?

Enzo makes good points about how you test distortion. Injecting at the PI can give you a very different answer than injecting 100mA at the input, especially if you turn up the treble control.

Enzo: And one last time: did you ever determine where the "breakup" is ACTUALLY occurring?

In another thread we were laughing at "techs" who proudly state that they don't use oscilloscopes and pronounce amps as being good solely by listening to them. I think that it's imperative to signal trace every amp to verify that there aren't any unrecognized problems. It would really suck to be the guy who missed an obvious problem by skipping that step, and mistakenly told the customer that the SR may not be the amp for him. Just trying to verify you've already covered all of your bases...

[Randall]

Enzo, it was a signal generator into input jack. And no, I did not spend time looking for break up, because at this point, I'm not sure this amp really has a problem, except for the wrong OT. This customer is kind of high maintenance, the kind who keeps you on the phone for 20 minutes telling you all about everything, and nothing. So, he just got the amp, so I am starting to think it sounds like a good BF should sound. I'm not sure he really even knows. Anyway it sounds pretty darn good when I play it.

And as for wiring the cab to suit the wrong transformer, that dog won't hunt. It may well work and sound just fine, but if someone is going to pay dearly for a super clean BF, they aren't going to want to hear it has the wrong OT and the speakers are wired differently then Leo did it. Just the way that goes.

[g1]

Are the speakers stock (8 ohm), and is it actually wired as a 2 ohm load?

[Leo_Gnardo]

if someone is going to pay dearly for a super clean BF, they aren't going to want to hear it has the wrong OT and the speakers are wired differently then Leo did it.

B-b-b-b-b-but ..... Stevie Ray did! (cue synchronized group face-palm now)

[bob p]

And no, I did not spend time looking for break up, because at this point, I'm not sure this amp really has a problem...

I don't mean to flog a dead horse, but I think you've got the logic backwards:

Looking for break up when signal tracing is an excellent way to confirm whether or not the amp has a problem.

Not looking for break up / not signal tracing is an excellent way to not really know whether or not the amp has a problem.

I prefer the first technique over the second. It's a free and easy way to find the answer. I like to perform easy tests that rule out problems, thereby eliminating guesswork and saving time.

[bob p]

B-b-b-b-b-but ..... Stevie Ray did! (cue synchronized group face-palm now)

That's the difference between a player and a collector; players accept solutions that work and sound good. Anal retentive vintage gear collectors fuss over things that don't always matter.

[dstrat]

One thing I have not seen brought up: Isn't a Bassman transformer made for bass frequencies or wider bandwidth?

which amp would distort more a Bassman on 10? or a SR on 10?

distorting too much is very subjective, i am sure you would get 50 different answers from 50 guitar players.

Randall , should you not get your customer to demonstrate to you what he thinks is too much distortion?