Hi guys:
I had occasion for the first time today to open and examine the guts of my second hand:
Antique Sound Labs AQ1003dt 30watt PP EL-34 amp.

This amp uses ultra linear tap on the OT primary.
The OT has both 4 ohm and 8 ohm taps on the 2ndary.

The question is that I noticed what looks like a non-factory mod.
The 8 ohm taps have a 0.12uf cap in series with a 10 ohm resistor to ground.

This is not on the 4 ohm tap.

Why do you suppose someone needed to add this mod?
Wouldn't it affect the frequency feeding an 8-ohm speaker?
And should I remove it since the print does not reflect it?
Wooops... I just found a different version of the print than the one I had at the bench.. AND yes it is shown here....
but only on the 8 ohm (total winding ) tap.
This print shows the cap as .047uf ... but there is a 0.12uf in there instead.
What is the purpose?

Could be an attempt at doing something to the feedback. It would tend to shunt RF out of the feedback, that *might* be useful, but what I would wonder what that does beyond the original feedback circuit. Are you sure it's an erzatz modification? That's where the feedback circuit would normally be connected.

But it is going to have an effect on either the 4 or 8 ohm taps. The 4 ohm tap is just taking half of the same winding, to if you are shunting it off the end, it will do something to the 4 just like it does the 8.

If it's mine, I remove this circuit and put it back to the original. If you want to filter something, filter the input to something that the OT can actually handle (which is the essence of the Van Alstine "Super 70" modification).

Brett

p.s. Aha! Now that you have a schematic, it looks on casual glance that there *is no feedback*. This might keep the thing from oscillating or might be there to ensure there is always a load so it doesn't go apesh*t with no load.

Sorry Brett.. thanks...
I edited my OP because I found a later print that DOES show it.. it's NOW shown in my OP.

What do you think?

AND.. I now see your updated reply too... lol
So that RC series arrangement CAN be a load?
If so how many ohms? and would it conflict with in parallel to the speaker 8 ohms?

ARF as a chatroom! I like it. I suppose you can confirm that the amp doesn't actually connect the center tap of the OT to both ground and B+ like it implies on the schematic?

Brett

They didn't dot the ground connection of the centertap. Call it a typo?

But a whole power amp with ZERO feedback other than the screen taps? Not something we see every day. Bet you can't pull those 12AU7s out with your bare fingers after they've been running a half hour. Even with 'tube fingers'.

Zero end-to-end feedback, anyway. It's not unheard of, there can be enough "local feedback" to keep it stable if you know what you are doing. I would hesitate to be too critical without doing a lot of analysis (which, frankly, despite my generally helpful predispositions, I am not going to do). There's no way to figure out what that output filter would do without including the crossover of the speaker. I am not too sure that the schematic is accurately drawn enough to get the right answer, anyway. Found one whopper right off the bat. It kind of looks like it has too many parts.

I have spent endless hours futzing with tube amps, mostly happily, and have seen a lot of pretty strange things work to some extent. So I won't jump to any conclusions. My general observation:

you don't need any funny topologies, SET, etc, a straightforward push-pull Williamson-style amp, properly implemented, cannot be significantly improved upon

you want feedback but it has to be done correctly
you want an input filter to limit the bandwidth ( great insight by Frank Van Alstine, related to previous)
you can't get more than about 30 watts out of it without some other compromises (also related to previous)
given the above, don't waste too much time with anything more macho than an EL34/6CA7
if you want more than 30 watts, bridge two of them together

I don't spend a lot of time on this sort of thing anymore, despite my previous obsession with it. Want a nearly-perfect 30 watt tube amp, go buy a Ultimate 70 kit from FVA and put it in a clapped-out ST-70. If you want a ~75-80 watt tube amp, go get two Ultimate 70 kits, two clapped-out ST-70s, and bridge them. If you want more than 80 watts, get a hearing aid.

Brett

Yeah that's funny ... lol ... It's an artifact from an earlier drawing where the B+ line passes on down to the LEFT channel print previously shown below the top print.... which they have decided now to not draw... so they cut it off... But that B+ doesn't show a termination there on ground as it "appears".. since there is no RED connection dot.

I think this is the right answer....

The corner frequency of that circuit is up around 100KHz---I think it help keep the amp from becoming a combination RF transmitter and space heater....

Its pretty high, all right. We don't know the "L" in the RLC circuit but it's not filtering much at audio frequencies.

Brett

Hio Brett:
It's also very interesting that it doesn't use global feedback. The earlier print I have showed it in use on earlier models.

Is it possible that since this amp employs "ultra liner" operation... which is closer to Triode than it is to true pentode operation... is there then... less need for feedback?

I have a 2 watt PP 5902 amp (which uses Global feedback) located right next to this EL-34 amp.
.. and I often compare the sound of the two on certain pieces of music I love.
I find that when I hear some pieces of music that sound GREAT on the EL-34 amp.. I quickly switch the speakers immediately to the 5902 amp to compare and find that little 3watt amp sound almost identical ... or at least just as good.
Here's that 5902 schematic:

The ultralinear taps *are* feedback but local to the output stage, of course. I don't understand the circuit a(nd am not going to spend a lot of time trying to figure out, to be honest) but people have successfully, depending on your definition, gotten away with no overall feedback before. For a while it was all the rage and a lot of marketing takes advantage of the "feedback is bad" theory.

There are legitimate designs that use only local feedback for each stage so that each stage it stable unto itself and attains very low distortion. Most of the no-feedback tube designs I have looked at are terrible, of course, like the typical SET amp that gives you no power at 10% distortion. To me, if you are going to spend $1000 on old WE300bs I would expect a lot better than 3 watts and gross errors! Maybe I have missed my calling - should start selling the output stage of a Philco 40-150 in a hogged-out aluminum case and charging $20,000 for it.

Brett

Its a zobel , it is to keep the amp from breaking into high frequincy oscilations when no load is connected to the output .

Thanks:
I just looked this up on Wikipedia:
http://en.wikipedia.org/wiki/Zobel_network

"Zobel networks and loudspeaker drivers
Zobel network correcting loudspeaker impedance

See also Boucherot cell

Zobel networks can be used to make the impedance a loudspeaker presents to its amplifier output appear as a steady resistance. This is beneficial to the amplifier performance.
The impedance of a loudspeaker is partly resistive. The resistance representing the energy transferred from the amplifier to the sound output plus some heating losses in the loudspeaker. However, the speaker also possesses inductance due to the windings of its coil. The impedance of the loudspeaker is thus typically modelled as a series resistor and inductor. A parallel circuit of a series resistor and capacitor of the correct values will form a Zobel bridge. It is obligatory to choose R_B = \infin because the centre point between the inductor and resistor is inaccessible (and, in fact, fictitious - the resistor and inductor are distributed quantities as in a transmission line). The loudspeaker may be modelled more accurately by a more complex equivalent circuit. The compensating Zobel network will also become more complex to the same degree.[3]

Note that the circuit will work just as well if the capacitor and resistor are interchanged. In this case the circuit is no longer a Zobel balanced bridge but clearly the impedance has not changed. The same circuit could have been arrived at by designing from Boucherot's minimising reactive power point of view. From this design approach there is no difference in the order of the capacitor and the resistor and Boucherot cell might be considered a more accurate description."

I had a Magnavox stereo SEP 6BQ5 amp that in its original configuration would oscillate at some high frequency and put out about 70 Vac from the speaker terminal if the speaker was disconnected.

It could certainly have that sort of effect. If so, however, the impedance matching network would have to be specific to a particular speaker model and if you want it to perfectly compensate, each particular speaker. So I would be surprised if you would show it on the schematic for a standalone amplifier.

Brett

So... are we saying that it more likely is there to trap HF oscillations when no load is present?

AND BTW:
Why do you suppose the earlier print showed a version that did employ NFB.. and now there is none?

I've found only rave reviews everywhere about this amp ... but none yet have mentioned the fact that it does not use NFB.

Do you suppose it doesn't require it because in ultra-linear mode it's operating more as class-A .. or closer to A than AB1?
or ... maybe more like a triode than pentode?

what's the conventional wisdom on NFB in ultra-linear.

Or any thoughts here on why this sounds so beautiful without NFB?

A tube amp, especially a higher powered one should NEVER be operated without a speaker attached. Some of 'em will oscillate at supersonic frequencies anyway which may blow out a tweeter, or just cause a degradation of the sound. Since you can't hear them, only way to see is with a scope, although you can measure AC when oscillation occurs. NFB helps some if it is done carefully, but doesn't always do the trick, and Lead dress is extremely critical.

That's why there is usually a small value capacitor across the primary windings, and some other tricks like grid stopper resistors, or in this case the Zobel.

Ah yes.. thank you ... that makes sense...
In the earlier version with NFB the Zobel was not there either...
...so it was added when the NFB was abandoned.