Why does this use mis-matched bias feed resistors?

[Jennings]

Can someone help me understand what's going on here, please?


I'm looking at the schematic attached, trying to understand the fixed bias approach.  The fixed bias seems to be deliberately fed unequally to the EL81 output tubes.  There's a 270k resistor (R14) to B3 and a 470k (R15) to B4. 


Ultimately I'm wanting to refurb this amp and re-purpose for guitar, using a standard 8 Ohm speaker on the 10V OT tap. 

[sluckey]

I'm looking at the schematic attached, trying to understand the fixed bias approach.  The fixed bias seems to be deliberately fed unequally to the EL81 output tubes.  There's a 270k resistor (R14) to B3 and a 470k (R15) to B4. 

The bias will not be affected by those unequal resistors. Voltage at the output grids will be the same. Those unequal resistors are probably meant to help provide equal signals from the PI. Now look at those PI plate resistors... 39K and 390K! Really odd.

I would put a test sine wave signal into the amp and check that the PI plates are delivering equal signals to the output tube grids. If so, accept the PI as is. If not equal, I'd rebuild the entire PI to a more common circuit as found in a Fender, Marshall ,Vox, etc.

[Jennings]

Ah, I see...that makes sense.  I hadn't thought to look back at the PI...Thanks    I'm tempted to rebuild the PI as you advise, and might alter the EL81 to cathode bias instead then.  Could go the hole hog I suppose and ditch the ECC40 in favour of rewiring for an EF86. 

[HotBluePlates]

Can someone help me understand what's going on here, please?

... Those unequal resistors are probably meant to help provide equal signals from the PI. Now look at those PI plate resistors... 39K and 390K! Really odd. ...

An unbalance-here to offset an unbalance-there.

The "tail resistor" of this diff-amp is only 1kΩ.  That's not large enough to develop a large drive signal at the cathode of the section whose grid is connected to ground.

To make up for the lesser-drive into that 2nd section, they made the plate load resistor a lot larger.  That also necessitated making the grid-reference of the next stage larger as well.  All to extract the most gain possible from the side whose cathode is driven with a weaker input signal.

[PRR]

...check that the PI plates are delivering equal signals...

Meter/scope loading will upset this way-high-impedance design. If they "had" to raise the offside grid resistor to a half-meg to balance it, probing with a 1Meg instrument will give it a limp.

[Jennings]

The service manual makes for an interesting read...sounds like they're quite proud of their PI arrangement, and efforts to achieve a balanced feed to the output tubes 
https://nzvrshome.files.wordpress.com/2018/12/Philips-EL6401-amp.pdf



I've got a number of random NOS EL81 tubes floating about here, but I don't have a tester sophisticated enough to match them...so it would be useful if I either make the fixed bias arrangement adjustable, or convert to cathode biasing.  Any advice or opinions on which way to go with that?  Normally I'd just cathode bias for ease, but since the tap and arrangement is already in place for fixed bias, I'm tempted to explore modding that.

[sluckey]

Either biasing method is fine. Just depends on how sophisticated you want to make this little amp. In fixed bias you could even use a separate bias pot for each tube to balance the tubes' performance. Requires 1Ω resistors for each cathode.

For cathode bias you could use a separate cathode resistor for each tube then measure the voltage BETWEEN the cathodes (one probe on one cathode, the other probe on the other cathode). Then experiment with different tubes until you get a pair that measures zero volts between cathodes. Zero volts means the two tubes are conducting equally. Or you could do something like I did on my Vox AC15. See page 2 of this pdf...

     http://sluckeyamps.com/VAC15/Vox_AC15.pdf

[Jennings]

Thanks for the advice and suggestions on options...much appreciated  .  The amp arrived today, and both EL81 tested well on my Sencore checker, so I'll presume they're reasonably paired for now as the amp hasn't got any pre-existing bias mods or signs of change.  Which means phase one will be to renew the filter caps and anything that is worn, broken or wildly out of spec to get a working base amp.  I'll then see how it performs and to guide any tweaking/changing around the PI and preamp. 


On the bias approach...perhaps I should rebuild the fixed bias approach initially, add the trim pots etc, and replace the OA55?  I've redrawn the original schematic accordingly, and highlights the changed bias components in the dotted orange box.  Also added the suggested EL81 cathode resistors.  Oh, and moved the main fuse into the live side rather than neutral.  Does what I've drawn look/sound sensible initially?

[sluckey]

There is no high voltage center tap shown in the original schematic. If there really is a CT, DO NOT CONNECT TO GROUND as shown in your redrawn schematic.

Your wiper fail resistors are shorted and doing nothing. Whenever the wiper is connected to one end of the pot as shown in your revised schematic they are not needed.

[Jennings]

Thanks Steve...sometimes I can't see the wood for the trees    Corrected schematic attached.