Bohrok,
you are correct in saying that OS will have a dc off-set if the cap is removed, keep in mind that you should use a matched source impedance and not a sine generator of zero impedance as value default would suggest. If you do not match the input impedance then it is like tying the input to ground under dc conditions.
you are correct in saying that OS will have a dc off-set if the cap is removed, keep in mind that you should use a matched source impedance and not a sine generator of zero impedance as value default would suggest. If you do not match the input impedance then it is like tying the input to ground under dc conditions.
In your simulation file you are bypassing the input cap. Take the bypass out and you will have DC offset.
/Martti
Take the 2 batteries out and leave the caps ..you will have 3mv offset , about the same as the real one before trimming "offset - 1k". My input imp. would be 820R + 33k + the imp. of the LTP base. The jumpered caps were just to lower the noise floor to see the -120-140db harmonics (keen showed me this). to elaborate further on the -46mv batteries , that is the voltage reading at the caps with them IN the circuit (not bypassed) at 0v input.. batteries were substituted to bias the LTP accordingly.
OS
Last edited:
Take the 2 batteries out and leave the caps ..you will have 3mv offset , about the same as the real one before trimming "offset - 1k". My input imp. would be 820R + 33k + the imp. of the LTP base. The jumpered caps were just to lower the noise floor to see the -120-140db harmonics (keen showed me this). to elaborate further on the -46mv batteries , that is the voltage reading at the caps with them IN the circuit (not bypassed) at 0v input.. batteries were substituted to bias the LTP accordingly.
OS
Thanks for your patience and valuable information! The DC offset was my fault as I did not notice you had another bypass on the feedback cap.
/Martti
Salas, that circuit is very interesting. I take it you are the originator? And is the second leg of the LTP used for positive feedback?
- keantoken
- keantoken
kean
I quite agree that circuit is interesting.. will try to simulate it....just to see and understand how it works...🙂
I quite agree that circuit is interesting.. will try to simulate it....just to see and understand how it works...🙂
Would be interesting to see some sims of this JLH LFET.
The results i got so far are mitigated using the original
schematic.
Seems that its design angular stone was large slew rate
rather than extreme linearity.
Worth adding that it is H2 thd optimised in some way....
The results i got so far are mitigated using the original
schematic.
Seems that its design angular stone was large slew rate
rather than extreme linearity.
Worth adding that it is H2 thd optimised in some way....
Last edited:
Salas, that circuit is very interesting.
kean
I quite agree that circuit is interesting..
Would be interesting to see some sims of this JLH LFET.
....
What are you cats talking about?
I was talking about that circuit back in October on Naggys original thread, but no one was interested about it. Why all the interest now?
http://www.diyaudio.com/forums/soli...plifier-i-have-ever-heard-48.html#post2329087
http://www.diyaudio.com/forums/soli...plifier-i-have-ever-heard-48.html#post2329087
What are you cats talking about?
I think they have been drinking their bathwater - I cannot find a schematic either.

Nico
What are you cats talking about?
http://www.diyaudio.com/forums/soli...s-improvements-stability-121.html#post2385543
Seems that Nico somewhat abused of some hard 'sky ..
Or was it some french Bordeaux..?....
Last edited:
Salas, that circuit is very interesting. I take it you are the originator? And is the second leg of the LTP used for positive feedback?
- keantoken
Its John Linsley Hood. Yes it uses both positive feedback and not so usual compensation. http://www.diyaudio.com/forums/soli...plifier-i-have-ever-heard-50.html#post2329455
This TMC nested feedback how does that behave under clipping....???..does it go bazurka..?? (like Vikings did after mushrooms)...Or does it give nice and rounded shoulders to the clipping..??
My computer blew up yesterday, so I'll be spending a few days getting everything back under control.
Interesting question how TMC behaves under clipping. The voltage dependence of Miller compensation means that the circuit will become many times faster when the output voltage stops moving during FET reverse bias. At the same time OLG will drop dramatically, lowering OLUG, and presumably increasing phase margin. So it appears we'd have to build it to be sure.
- keantoken
Interesting question how TMC behaves under clipping. The voltage dependence of Miller compensation means that the circuit will become many times faster when the output voltage stops moving during FET reverse bias. At the same time OLG will drop dramatically, lowering OLUG, and presumably increasing phase margin. So it appears we'd have to build it to be sure.
- keantoken
keantoken,
I'm totally confused by your last post. What circuit are you talking about? Please provide a link that works. What is OLUG? Why can't you do a simulation rather than build the circuit?
Rick
I'm totally confused by your last post. What circuit are you talking about? Please provide a link that works. What is OLUG? Why can't you do a simulation rather than build the circuit?
Rick
OLG is open-loop unity gain in frequency.
My computer broke, so I have to set it up again, reinstall and all that stuff. In any case, we have two factors, one increasing stability and the other decreasing it, and the combination will determine how the circuit responds to clipping. I don't want to trust the models here because many things can be different when simulating clipping behavior.
- keantoken
My computer broke, so I have to set it up again, reinstall and all that stuff. In any case, we have two factors, one increasing stability and the other decreasing it, and the combination will determine how the circuit responds to clipping. I don't want to trust the models here because many things can be different when simulating clipping behavior.
- keantoken
My computer blew up yesterday, so I'll be spending a few days getting everything back under control.
Interesting question how TMC behaves under clipping. The voltage dependence of Miller compensation means that the circuit will become many times faster when the output voltage stops moving during FET reverse bias. At the same time OLG will drop dramatically, lowering OLUG, and presumably increasing phase margin. So it appears we'd have to build it to be sure.
- keantoken
I had to pleasure of clipping the blameless TMC amp ,it survived , my beta enhancement device got warm , but I got the same waveform as predicted by LT.
Slight saturation/sticking on the positive peak .. "overshoot" on the bootstrapped rail (negative). On cordell's NFB thread , some suggested that the Vbe multiplier would collapse with the onset of clipping , but I found this would be compensated for by the 22uF Vbe cap. In other words , no issue .. no smoke ... the amps will survive. With the higher rails of the goldmund clones and the calculated gain of the stage , one should never even encounter clipping with a normal source. On my 75V rails , a 2V rms signal will only produce a 77v rms OP (110v p-p) . I had to run my voltage module on 43V rails to produce clipping.
OS
So the amplifier can survive even if driven into clipping....🙄 Question is...will the attached loudspeaker..??
Simulation shows some ringing and a very spiky TMC current... but how that translates into reality is not really predictable...
Simulation shows some ringing and a very spiky TMC current... but how that translates into reality is not really predictable...
Hi MiiB,
In my opinion one should design an amp so that it cannot be driven into clipping. Nowadays with digital sources, it is much easier because the output cannot (should not) exceed 2V and the gain should be chosen that your amp remains unclipped.
In the old days one had to put up with tuner, head-amps, tape decks, etc. outputs that could range anywhere between 100 mV and 1V and you would have to choose a higher gain so that they guy who happens to have a source of 100 mV could have the full benefit of the 30 watts that the amp could provide.
Kind regards
Nico
In my opinion one should design an amp so that it cannot be driven into clipping. Nowadays with digital sources, it is much easier because the output cannot (should not) exceed 2V and the gain should be chosen that your amp remains unclipped.
In the old days one had to put up with tuner, head-amps, tape decks, etc. outputs that could range anywhere between 100 mV and 1V and you would have to choose a higher gain so that they guy who happens to have a source of 100 mV could have the full benefit of the 30 watts that the amp could provide.
Kind regards
Nico
- Home
- Amplifiers
- Solid State
- Goldmund Mods, Improvements, Stability