MagneQuest/Peerless Forum

Welcome! Need support, you got it. Or share your ideas and experiences.

Return to MagneQuest/Peerless Forum


Message Sort: Post Order or Asylum Reverse Threaded

EXO-173 progress

148.240.100.162

Posted on May 8, 2004 at 10:30:39
keto
Audiophile

Posts: 1006
Joined: September 29, 2001
Thinking it might be interesting to modify the Pimm 120mA CCS in the output cathode of a PP2A3 circuit, I added another dial on top of the chassis, to vary (balance???) current between the two output halves.

"Is this the volume control?" "Nooooo!!!!!!"

This scenario is avoided by locating the trimmer between the 2 output tubes and a VR tube, and, should someone still think its the volume control by mistake, no problem, because its really a 250R pot with a 100R across each leg. I think a one-turn pot is good for this application, because it gives a good visual of nominal "center".

Though the 250R (really about 100R) is a nasty pot, which I'm assuming is adding a funky quality to the sound, it does the job: once warmed up, a stable null of 0.7mV humm (TJ2.5V300Bs) is within its range.

It seems like a really good idea to have separate 2.5V/2.5A filament windings, so that the output cathode can be coupled or de-coupled to a specific degree. The immediate benefit is a reduction in noise. It may be that a coarser and finer pot in series (a la Pimm PP47), which would de-couple via a higher impedance, might be even better. There might be a turning point, where the cancellation of the filament noise becomes less effective, or there might not. A 100R effective pot seems like a good minimum, though.

In addition to humm being louder with just the 120mA sink (connected via a 25R resistor from each of the four filament pins) and wanting to be able to adjust the current between halves, I've also been wondering about reducing the "intrinsic compact circuit".

This where my slope gets really slippery.

With an EXO-173 connected to a PP output with separate cathode bias for each half, the phase-inversion is done entirely by the EXO. When the cathode biases are connected, there's a coincident phase-inversion: that of the "compact" circuit (see link).

My first thought was that it would be interesting to reduce the "compact" effect, by decoupling the two halves a little. Not sure if this is what really happened, but the amp sounds even better now, somewhat decoupled. And is quieter.

Since my other amp is set up with the "compact" circuit, direct-coupled between SE and PP stages, I've been staring at that configuation a little. My latest pseudo-observations are that

1) such a circuit has a hugely increased and assymetric input capacitance; and

2) the "second" half will have a quiescent operating point that is Vg volts high and cool with respect to the "first" half, and this along the nominal plate load slope. Think 20% less output and no deeper into class A.

If true, two strikes against using it in the output of a PP2A3. Better done in the previous stage with 1) a lower input capacitance and 2) plate CCSs; these, to minimize the two effects, respectively. I've got a 6SN7 driver lined up to give this a try, in amp number two.

That's the latest anecdotal report from the tropics, where mango season is in full swing.

--keto

 

Hide full thread outline!
    ...
Re: EXO-173 progress, posted on May 10, 2004 at 19:05:27
Sector-7G
Audiophile

Posts: 11847
Joined: May 14, 2002
Just get Mike to send you a 'four terminal' 173, with its CT split into two terminals. I have taken a similar Pp grid choke and AC grounded the CT with 6.8 uF Auricaps( one per leg ) and DC to a 50k pot, it is a fixed bias dealie.
regards,
Douglas

 

Re: EXO-173 progress, posted on May 11, 2004 at 15:53:44
keto
Audiophile

Posts: 1006
Joined: September 29, 2001
Hey, Douglas. The 6.8uFs ground to the output cathode? and the 50K ct to earth?

Revisiting my own EXO-173, the WW pot in the shared, PP2A3 cathode is a very bad idea, methinks. At least it sounded that way. A current source (or bypassed resistor) and shared filament secondaries is better, provided you have a means of adjusting current, which is what your "dealie" would do, right?

My perceptions are definitely on the move, and I look forward to reworking the EXO173-loaded (the "reference") of my two amps.

A simple solution would be shared cathodes, and following Gary Pimm's lead, measuring exact DCRs of the two OPT primary halves, at operating temperature (not voltage!) and adding a precision resistor to even them up. Then with a well matched pair of tubes, you'd be very ballparky wrt to DC offset, alternating between the pair to see which configuration is quietest. Quietest in my universe means DC offset null, which may be completely wrong. I don't have a way of measuring it otherwise, or the interest in carefully and extensively probing and recording high voltages.

Or, how about a 60mA CCS (or appropriate bypassed resistor) on each of the two filament potentials, to null filament humm with a high impedance "center tap"?

Thanks for the explanation of the 4-tap item. --keto

 

progress, posted on May 11, 2004 at 17:17:47
Sector-7G
Audiophile

Posts: 11847
Joined: May 14, 2002
Hey-Hey!!!,
The DHT amp I have on the boards will run 300B's but much else is similar. Two filament windings, a hum pot on each cathode and a current sensing resistor from wiper to ground. Say ten Ohms or so. Fixed bias, supplied through the split CT, PP grid choke, to each power valve.

5687 input tube, configured as a common cathode diff amp, feeding a PP, common cathode diff amp of two 12B4A's, loaded reisitively, and feeding the ends ov the grid choke the AC signal.
regards,
Douglas

oh yes, MQ-431 output Iron, Hammond 300BX power Iron, with the 10 Hy/500 mA choke, configured as a choke input supply, B+ in the conservative 310-320 V range for 80 mA per valve. CCS-es by DN2540

 

yet another comment on filament windings, posted on May 15, 2004 at 22:16:05
keto
Audiophile

Posts: 1006
Joined: September 29, 2001
While I can't completely separate all the factors involved, its pretty clear that with a CCS or bypassed resistor in the output cathode, a common filament winding sounds and measures best. I also think I understand why Gary Pimm uses an 11-ohms resistor from each filament potential to current sink, since I'm now using a pair of 12-ohms in that position (instead of 25-ohms) and it sounds the best yet. Have you had any luck dialing things in, as described?

 

it's just a sack of bits now, posted on May 16, 2004 at 03:49:15
Sector-7G
Audiophile

Posts: 11847
Joined: May 14, 2002
Soooooon though, very soon. Nothing has jumped out of the Long Grass to bite this idea in its tender flanks and I have been bashing it about for several months now.

Its not actually mine, and a bit of distance has gotten in the way of a marathon build session between me and the parts buyer.
regards,
Douglas

 

Page processed in 0.027 seconds.