M2 is next in que for me. Looking into Cinemag and Electrprint transformers. Run the transformer into a phase splitter/PLH or phase splitter into circlotron. NO feedback all the way through. Way more complex than Nelson's version, but may be interesting. Looked into Vbe multiplier as another optin after transformer for output bias, but i suck at bjt circuits.
The phase splitter would come after the transformer to present it with higher Zin. i believe Vbe multiplier can be used in similar fashion to tL431 after trasnfromer as well. All this is assuming wiring it as an autoformer for increased bandwidth. Havent tried it yet, just info gleaned from reading and processinging through weak mind. If you leave input Jfets in F4, then i dont think there is any issue, just plug in step up transformer. Not knowing much about transformers, i think the output wants to see a high impedance. Phase spitter idea came from Sy's impasse. It needs to see equal load on drain and source to work properly, so circlotron might be best/only route over PLH output. MAy sit down and try to do the math this weekend, but i doubt ill get to it. It is possible that none of this works
PS. Ilike non feedback circuits as you might can tell
PS. Ilike non feedback circuits as you might can tell
so , let's just ignore posted questions
Never the less, that was a genuine question.
Nelson also makes a similar remark in his SIT paper.
I am guessing he is referring to source resisors, but I am clutching at straws really.
Put a 550 ohm R in the output of your B1 and call it a 600 ohm source? Put no R in the output of your B1 and call it a 50 ohm source. What can't you understand.
My sig gen is a 50 ohm output. If I don't terminate the ouptut in 50 ohms(add an extra 50 ohm cause the next equipment input Z is 20k) the output amplitude will read double what it should.
All this source R stuff adds up. If you have input C you would be figureing out what bandwidth is possible. If you have low input C you don't care... This has been a reoccurring theme in N.P. stuff for years
ZenMod buddy, what thread do you think you are in? It's not!
My sig gen is a 50 ohm output. If I don't terminate the ouptut in 50 ohms(add an extra 50 ohm cause the next equipment input Z is 20k) the output amplitude will read double what it should.
All this source R stuff adds up. If you have input C you would be figureing out what bandwidth is possible. If you have low input C you don't care... This has been a reoccurring theme in N.P. stuff for years
ZenMod buddy, what thread do you think you are in? It's not!
1:Will the BA-3 front end run on +-30v?
2:Can Conrads biggest heatsink take 150w of heat?
Conrad Heatsinks - Products
The MF35-151.5
2:Can Conrads biggest heatsink take 150w of heat?
Conrad Heatsinks - Products
The MF35-151.5
From experience the most you can push one of those heatsinks is 130W, but I try and keep it below 110W. Just do monoblocks ie 4 heatsinks
Ok,I guess it is possible to keep the bias down to not reach more than 130w/heatsink.
Will this affect the sound negative?
Is there a heatsink that can take 150w?
Ok,I mean "take 150w" and be in the temperature range of 50 maybe 60degres celsius.
delta T s-a of about 30C degrees then?
The Derating Factor (DF) for that will be about 0.5 ref 75Cdegrees.
Rth s-a = Delta T / P * DF = 30 / 150 * 0.5 = 0.1C/W
That is a mighty big heatsink.
BUT,
one must consult the manufacturer's derating table for their specification method.
Each term in the calculation will have a very significant effect on the actual dissipation.
The Derating Factor (DF) for that will be about 0.5 ref 75Cdegrees.
Rth s-a = Delta T / P * DF = 30 / 150 * 0.5 = 0.1C/W
That is a mighty big heatsink.
BUT,
one must consult the manufacturer's derating table for their specification method.
Each term in the calculation will have a very significant effect on the actual dissipation.
Last edited:
I have already done extensive testing with these heatsinks. No need to do the calculations.
Any where between 110W and 130W and you will be ok.
Regarding sound quality. If you have at least 10W of class A operation (assuming BA2 output) then the rest is just icing on the cake.
So I would not be too concerned about it for a push pull output stage (source follower).
If it was a single ended output stage (BA1) then I would definitely want more biasing overhead.
Just buy 4 of those heatsinks though, it opens up a whole range of options that you will want to explore sooner or later.
Last edited: