12AT7 substitution

triode · 3025

0 Members and 1 Guest are viewing this topic.

Offline triode

  • Full Member
  • ***
    • Posts: 77
on: October 07, 2017, 08:32:55 AM
Hi,

I have some 6211, 5965, 6829, and 6414 tubes.
Would be these OK to use as 12AT7 substitution in Stereomour?
I understand that these tubes close to 12AT7 and generally OK as substitute, but Stereomour has C4S loading the tube.
Would these tubes work OK with C4S?

Thanks in advace



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19319
Reply #1 on: October 07, 2017, 01:33:27 PM
They will plug in, glow, and pass sound.  In the original Stereomour, we have the TL431 regulator presenting 2.5V of bias under each driver cathode, and additionally we have a C4S set for 3.6mA of plate current.  These two fixed parameters will establish what the plate voltage will be, and we'd sure like to see it at about 1/2 of the available B+ on the driver plate.  This is especially true for tubes like the 5965, which sound great when they are run around 200V, but have some non-linear areas at low plate voltages.

With the 5965 variants, I don't think you'll even see 150V on the plate in the stock circuit.  This adds some thermal stress to the MJE5731As and, if low enough, can reduce driver headroom enough to limit amplifier power.

What I would most recommend would be rewiring for 4V of bias instead of 2.5.  This would involve removing some jumpers and putting 4 resistors onto the driver PC board.  Another way to get to 200V is to raise plate current, but you'll roast the C4S board by the time you throw enough current through to bring the plate voltage back up. 

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline triode

  • Full Member
  • ***
    • Posts: 77
Reply #2 on: October 07, 2017, 05:21:04 PM
Thanks Paul

How low C4S can handle on plate voltage? Would 100V-150V too much for the transistor to take? would heatsink clip on MJE transistor helps?

Note that my Stereomour is configure for 45 output tube, so should has slightly lower B+

Thanks
« Last Edit: October 07, 2017, 05:27:14 PM by triode »



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19319
Reply #3 on: October 08, 2017, 07:47:06 AM
The C4S circuit places just about all of the B+ to plate voltage difference on the MJE350 at the plate current.  This is 388V-200V=188V, and 188V*0.0032A=0.6W of dissipation.

Without a heatsink, the MJE5731A will see about 65C of temperature rise per watt of dissipation.  This will put the transistor at about 100C with 1W of dissipation in a reasonable warm chassis, so this is the realistic limit without a heatsink.

The clip on heatsink that we added in Steremour 2 will rise about 50C with 3W dissipated, and the MJE5731A will have an extra 10C of thermal resistance from junction to case at 3W, so we can safely run considerably more current.

To get to 200V on the plate with a 5965 and 2.5V of bias, you would want to run 10mA.  This would dissipate 1.9W through the MJE5731A, which isn't the end of the world with the heatsink, and dissipating 2W on each plate of the 5965 is within acceptable limits for the tube.

Since you have the amp running with 45's, you could aim for 8mA and accept a plate voltage closer to 175V.  Drawing that kind of additional voltage through the power supply may mean that those 1.2K resistors may need to be nudged down a bit to bring things back into line.

Rebiasing the driver stage to 4V will require two 2.5 (or 2.49K) resistors and two 1.5K resistors, preferably 1/8 or 1/4W size. 

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline triode

  • Full Member
  • ***
    • Posts: 77
Reply #4 on: October 08, 2017, 09:38:10 AM
Thanks Paul,

I think I'll stay away to 5965/6211 tubes then, and keep the original circuit instead so I can still use regular 12AT7.

I did plug in some 5965 and test the voltage in the anode side of 5965 and shows around 100-130V (depend on the tubes).
I also tried 6414/6829 and they give 150V voltage in anode. Aside from headroom and how it sounds, would that acceptable and not put too much strains to the C4S?

Thanks again



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19319
Reply #5 on: October 08, 2017, 09:40:54 AM
Yes, the C4S won't overheat till you get down around 100V.

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man