Strange PX25 Circuit Diagram

slowmotion · Post by **slowmotion** » Tue Jun 18, 2013 5:54 am

Just for fun check out the WE 87 amp:

http://www.moviemice.com/we/87_amp/diagram-148.jpg

They seem to have the same connection as the Osram amp.

Paul Barker · Post by **Paul Barker** » Tue Jun 18, 2013 7:35 am

Excellent. Whatever it does it was not a freak use of it by Osram.

Post by **Nick** » Tue Jun 18, 2013 7:56 am

Is it not just making better use of the small caps they had, the grid is bypassed with a corner of 0.8Hz, if it was just the 4uf cap, then it would be more like 40Hz, So gain is constant down to 1Hz

JamesD · Post by **JamesD** » Tue Jun 18, 2013 2:23 pm

Paul,

I'm not sure where you get the 370Hz from - can you run through the calc so I can see what I'm missing as with 530ohms cathode resistor I get more like 80Hz...

Thanks

James

slowmotion · Post by **slowmotion** » Tue Jun 18, 2013 2:38 pm

Would the two caps in effect be in series?

JamesD · Post by **JamesD** » Tue Jun 18, 2013 3:02 pm

I don't think so from the perspective of a cathode bypass cap in the output signal current loop.

Working from Lynn's ETF presentation and signal current loops and his explicit examples.
The 2uF is the input signal current loop bypass cap and the 4uf is the output signal current loop bypass... at least as I see it!

James

Post by **Nick** » Tue Jun 18, 2013 3:10 pm

so I can see what I'm missing as with 530ohms cathode resistor I get more like 80Hz...

I was too idle to go back and read the previous page, so I guessed at 1k, so I agree with 80Hz, maybe a little more from the rk in parallel.

Paul Barker · Post by **Paul Barker** » Tue Jun 18, 2013 7:33 pm

JamesD wrote:Paul,

I'm not sure where you get the 370Hz from - can you run through the calc so I can see what I'm missing as with 530ohms cathode resistor I get more like 80Hz...

Thanks

James

firstly sorry I made one mistake in my calculation I have 168 hz. My bad. The rest you will have to put me straight. Here is my calculation.

I think I am about to learn something. I haven't read Lynn's papers and no doubt you will inform me where I am in error.

I used the cathode by-pass formula which is:

the resistance to be bypassed is named r
the cathode resistor is named rk
the other resistances involved are called rz and those are made up of (ra+rp)/(mu+1)

1/rz +1/rk=1/r

So from the Osram Data sheet
rk = 530
ra = 1265
rp = 3200
mu = 9.4875

rz = (1265+3200)/(9.4875+1) = 425.74
r = 1/((1/425.74)+1/530) = 236 ohms

capacitance in farrads = 2pifr
in microfarrads = 1,000,000/(2pifr)

Capacitor value is 4, I get to the f by applying the formula in a spreadsheet. Ordinarily we insert desired frequency and spreadsheet tells me cap value. I had to work it around about way, here is the formula:

1,000,000/(6.283X168X236)=4

How do you do it?

Post by **Nick** » Tue Jun 18, 2013 8:28 pm

Yep that looks right, I ignored the cathode resistance, so your 160Hz is about right.

Paul Barker · Post by **Paul Barker** » Tue Jun 18, 2013 8:52 pm

so that would mean it was never intended as a cathode bypass cap, it has another function.

Post by **Nick** » Tue Jun 18, 2013 9:35 pm

Yes, but I don't believe the valve needs a cathode bipass cap in the normal way. The normal reason for the bipass cap is to prevent degeneration between the cathode and grid. The 2uf cap will do that. But as James said, that cap is in the output loop. I am not certain the cathode resistance equation is valid as the grid is decoupled.

Paul Barker · Post by **Paul Barker** » Tue Jun 18, 2013 10:04 pm

Yes that is why I said I was about to learn something and that I hadn't read Lynn's papers. the only reason I didn't read them is the dyslexia. I don't churn through long swathes of the writen word. I prefer presis or direct to the point materials like RDH3. I prefer people who can read text easily to provide the to the point items.

steve s · Post by **steve s** » Tue Jun 18, 2013 11:31 pm

My Books say the grid should be decoupled from the anode circuit on the px25, not sure why

Paul Barker · Post by **Paul Barker** » Wed Jun 19, 2013 7:45 am

well I looked at the pictures of the ETF presentation and cannot see one which is as per the Osram Diagram . Conventional SE with bypassed cathode resistor is the only output loop which is the same.

All I can sumise is as Nick said the bypass of the grid resistor some how bypasses elements of the output reistances which normally contribute to the effective cathode resister little r. So under this unique condition therefore the valve and load impedances no longer play a part. I am not saying this is so, I am saying this would explain the small cap.

The only way to prove it is to build the output current loop with that value cap but without the grid resistor with it's bypass (i.e. go from IT to ground as normal these days, and test the FR. Then build with the insertion of bypassed high grid resistance (or bypassed grid choke which would be better from dc resistance affecting grid current issues perspective). Test that FR.

If then we find the FR is the same in both cases, nothing else is going on. If the FR is not at all right, we have a tone control. If it is pretty good we have a distortion control mechanism as the WE connection and the ultrapath are but in a different unique way which has laid unnoticed all these years.

vegard · Post by **vegard** » Fri Jun 28, 2013 10:07 pm

I was searching for filament chokes , and stumbled across this :
http://www.ax84.com/static/rdh4/chapte12.pdf

on p 538 (Attached a snapshot of it) the circuit at question is described as
grid Circuit decoupling. I read the text but must admit I do not understand
what the outcome is. It follows; I don't know if this information answers Your question.

audio-talk

Strange PX25 Circuit Diagram

#16

#17

#18

#19

#20

#21

#22

#23

#24

#25

#26

#27

#28

#29

#30