Mcintosh auto bias arrangement

[Mars Amp Repair]

Just thought I'd post this auto-bias arrangement on this Mcintosh tube audio amp. Interesting arrangement...They of course are not considering what would happen to the transistors if/when the tubes short. I suppose they just expect the end user of such a high end audio gear would periodically just replace the tubes...RIGHT!...
I just converted a Traynor YBA-200 from the autobias circuit to a manual adjust. Transistors connected directly to power tubes esp in musician's gear just don't mix well...glen

[nickb]

I don't think there is any auto bias here. I just see a couple of darlington pairs used as emitter followers. Yes, the transistors could be rather vulnerable.

Now the screens fed from the opposite plate IS interesting.

[Mars Amp Repair]

well, after staring at this for a while, there has to be some voltage division from the -400V supply to create the -62V on the power tube grids. The outermost drive transistors (Q4 & Q4) collectors is supplied from the cathodes of the power tubes.

As the voltage on the cathodes changes, it would invariably change the operation of the two 'drive' transistors. I believe this is some type of bias tracking system but haven't quite figured it out yet. It may end up to be the equivalent of 'cathode bias combined with negative bias'. I've seen some arrangements like that.

Also, I imagine the winding on the output transformer that the KT88 cathodes go to is some kind of hum-bucking system as it is also terminated at the filament ground.

I just realized they are using the 'bias' on the power tube cathodes to offset the DC level of the filaments in the amp. As I recall this moves the filaments close to the cathode dc wise to also reduce hum or the effect the filament might have on the hum/noise of the tubes.

Tube amps that are designed for High end HIFI applications do weird things to make them 'Ultra-linear' and ultra low distortion.

Just got off the phone with a tech at McIntosh. In the past they have been very helpful & actually been excited esp when working on the older 70's era stuff...this amp is not very old but this time the tech behaved as if I were wasting his time & was actually rather rude. He claims that there is no 'auto bias' but could not answer how they account for the wide differences in the transconductance of tubes. It was very much like when you ask a 'techie' about something he doesn't know the answer to & he just acts annoyed as he doesn't want to reveal that he is knowledgeable. I asked with enthusiasm if they had a operation explanation of the circuit & he flatly replied 'NO'.
Oh well, might have just been a bad day for him.

Anyway, the circuit is interesting to pick apart...thanx, glen

[nickb]

The divider is R21 + R31, R15.

I *think* the purpose of the collectors attached to the the cathodes is to dramatically reduce the collector to emitter voltage swing. That is turn will reduce the distortion due to the Early effect in the driver transistors. The OPT cathode windings are simple negative feedback direct to the cathodes. I'm come across this before ( Quad amp maybe?) in HiFI amps.

[Helmholtz]

The divider is R21 + R31, R15.

This can't be directly responsible for the -62V at the grids. I think it's the emitter current of Q5/Q6 which drops the voltage across R22/23, R24/25.

[nickb]

This can't be directly responsible for the -62V at the grids. I think it's the collector current of Q5/Q6 which drops the voltage across R22/23, R24/25.

The voltage is most certainly set by that network. -400* 121/ (121+332+332) = - 61.656051 at the first base. The grids will be about 1.2V less i.e -62.8V. Of course the current to develop that voltage drop flows primarily through the transistors but the voltage is set by the base.

[Helmholtz]

Yes, voltage divider sets base voltage and base voltage defines emitter voltage. Seems I misinterpreted.

[nickb]

What type are the output tubes? I'm still thinking about the screen grid to opposite plate arrangement. With say a nearly 400V swing the screen will go to +880V. A bit less due to the cathode feedback but still so high as to make you really sit up and take notice. The screen current would be large with just a 220ohms to limit it.

Are the 2.2uH inductors in the plate circuit really that or ferrite beads? Inductance can cause all manner of unwanted effects. A ferrite OTOH just absorbs HF and so is rather taming.

PS: What does it sound like if you plug your guitar in a turn it all the way up. I really want to know!

[Leo_Gnardo]

What type are the output tubes? I'm still thinking about the screen grid to opposite plate arrangement. With say a nearly 400V swing the screen will go to +880V. A bit less due to the cathode feedback but still so high as to make you really sit up and take notice. The screen current would be large with just a 220ohms to limit it.

Are the 2.2uH inductors in the plate circuit really that or ferrite beads? Inductance can cause all manner of unwanted effects. A ferrite OTOH just absorbs HF and so is rather taming.

PS: What does it sound like if you plug your guitar in a turn it all the way up. I really want to know!

I was guessing the output tubes to be 6550 or KT88. We'll find out soon enough I'm sure.

McIntosh put plate inductors into many of their classic tube power amps. I was always happy with my 240's, I'll have to warm 'em up one of these days for a test run.

What do they sound like with a guitar? Can't tell about the model we're discussing here, but I'll bet it's just fine. 50+ years ago Mac tube amps powered the main speakers at Woodstock. And in the early 70's there was the famous Wall Of Sound but that was mostly powered by 300w/ch Mac transistor amps. FWIW I used a 240 as a power section for my bass rig for a while, no complaints. It just got to be cumbersome to carry around, plus I was putting a $$$$ collector's item at risk every time I took it out of the house.