Looking for constructive criticism on my first design

[pebmrb]

euro21:
thanks for the circuit.I see now.So the 5k resistor can help with the ringing at frequency extremes?

Osvaldo de
Banfield:
thanks for the link.I will have to study those circuits.

Plenty of good info here,glad I posted.This circuit seemed unique to me,so I thought I better get some opinions from those with more experience than myself before doing a final build..
Any other opinions on this circuit?

[MelB]

Thanks for posting. Never thought of connecting to the cathode. I have some hammond 126B's coming. With everything I'm learning here looks like interstage transformers are a very good idea! Keep us posted on what you do and how it sounds!

[Wavebourn]

Hint: add one more resistor in series with anode load and bootstrap it by capacitor from cathode of the cathode follower.

It is not a criticism, just a suggestion what else to try.

[pebmrb]

MelB
Thanks for the encouragement.I hope the thread continues to provide some scrutiny,I have much to learn too.
As far as sound quallity goes:In it's current breadboard state: rats nest wiring,jumper cables,and with average OPTs,it is sounding very good.
I think with a little fine tuning,and actual solder joints instead of alligator clips,I can get it to an even higher level.

[pebmrb]

Waveborne:

So I put another resistor in series with the 33k plate resistor of the gain stage,and connect this voltage divider to the cathode of the cathode follower stage via a capacitor?

[nazaroo]

Don't waste those beautiful and very useful transformers like this.

First, you can clean up the Output section by an order of magnitude,
by simply adding only two more tubes, and a handful of caps and resistors.

What you get:

(1) 99% of PS hum and noise is rejected at the Anode of the output tubes (PSRR to max).

(2) Double the power output. The extra headroom is precious.

(3) Most of the non-linear distortion gone, through Push-Pull operation.

(4) True balanced output from the Interstage, without a splitter stage.

(5) No chance of DC flowing through the secondary of the interstage transformer, for 100% headroom and purity.

(6) NO DC or PS noise flowing through the Output transformer, meaning NO core magnetization at all, and 100% headroom out of Output tranny.

(7) Output tubes operating in a better range for maximum tube life.

(8) Jaw dropping uncanny clarity of the stereo staging; close your eyes and actually kiss Holly Cole.

Key points:

Split main PS voltage over the 5998 and the bottom tube, with the top 5998 getting 100 volts or more, and the bottom seeing hopefully about 250 if you've got it.

Set up the idle current in top and bottom tubes to achieve the DC idle at the bottom tube's plate to about 250 from ground.

You will never go back to SE again.

[Wavebourn]

Hi Nazaroo;

I don't like such bias of output tubes. I believe they do not like it as well. And I don't know why you need extra capacitors in top tubes of output stage.

Originally Posted by pebmrb

Waveborne:

So I put another resistor in series with the 33k plate resistor of the gain stage,and connect this voltage divider to the cathode of the cathode follower stage via a capacitor?

You may split it, like 27k at the bottom, 6k2 at the top.

The result will be almost like CCS load.

[nazaroo]

Originally Posted by Wavebourn

Hi Nazaroo;

I don't like such bias of output tubes. I believe they do not like it as well. And I don't know why you need extra capacitors in top tubes of output stage.

Greetings Wavebourn:

do you refer to upper or lower tubes?
I left the bias of the tubes to the OP.
I have already assumed they (original circuit) are wrong, without looking closely.

As for the uppers, well, I have had great results with 5998s,
and much poorer results with 6AS7s.
I didn't bother to analyze why, since I already found a tube I liked.

The caps allow the top tube to function as a PS regulator.
This along with the xformer blocking cap keeps the audio signal
isolated and across the primary.

I also agree with your assessment of the input stage:
I think you can get better mileage out of a Mu-type CSS stage, or SRPP.

[nazaroo]

PS: I left them out in the diagram, but 5998s should always have 1k grid-stoppers to block parasitics and retain stability.
Don't alter this value.
See the 'Purple Cow' for examples.

[nazaroo]

Hmmm... lets clean up the details for this circuit:

(1) If the Interstage transformer has no center-tap, (which would be connected to ground),
then you need a voltage-divider across the secondary to ground (see new diagram). The value of these resistors should be high enough to prevent excessive grid-current, but not so high as to induce instability in the (self)-bias D.C. circuit.


(2) You probably don't need the bypass-caps in my configuration, because the backward output-impedance seen by the speaker-load through the xformer is rather different now.

(3) With the lower voltage across the RS242, you'll have to recalculate the cathode-resistor, and perhaps its wattage.

(4) Depending upon available voltage at B+, you'll want to drop 10 to 50 volts across the load resistor between the upper and lower tubes. At 20-25 mA, that would be about 500 ohms to 2.5 k R.

(5) As mentioned, you need a 1k Grid Stopper for each triode in the 5998, and you should probably add one for the RS242 as well.

(6) The Resistor-values for the 5998 CSS circuit will be set by operating-point chosen. Start with 20-25 mA and your available voltage as a guide. Test it with the expectation of finding about +250-300 volts (to gnd) at the bottom of the 5998 cathode resistor.

(7) The BIAS for the 5998 can be referenced/checked against the specsheets for this tube. If you are substituting a 6AS7 or similar tube, check those datasheets.

(8) The blocking cap for the output primary doesn't need to be over-large; start with 10 uF polyprop and add a HF bypass cap.