varimu idea

[abbey road d enfer]

sidechain is fed back from half of the OT sec as before. as the sidechain is fed back after the PP OT, the sidechain is free from even harmonics 

Why would it be so? In most vari-mu compressors rectification is full-wave.

i wonder if the sidechain circuit will load the OT in some strange way

Indeed it will. This is an issue with many vari-mu compressors that rectify the signals coming from the anodes of the output stage; since the voltage/current requirements are bound to be identical, the loading via an additional winding is bound to be the same. This additional winding would be a high-impedance one.

I don't see any significant advantage in your topology, and indeed, PRR's comment rules.

 

[buildafriend]

There is not enough gain in a 1-stage amplifier to give substantial compression.

What are you trying to do?? Save a buck? I would assume many of the cheaper legacy designs were as cheap as the designer felt he could get away with; and that he probably had more think-time on the issues than you do.

Plagiarize!!

Why not just strap a pair of 6V6's in a push pull configuration to the output and make some other small adjustments?

 

[5v333]

hi prr!

substantial compression? how much is that?
your absolutley right about saving a buck here!!! instead of 3+ transformers
and alot of extra tubes per channel and big mains transformers, i was looking forward to more simple and cheaper approach for my own use, most intended for mastering. alittle bit of sound but not too much and only using 1-4db GR mostly at 4-16dbu output. this idea felt sooooo simple and cute!

hi abbey!

im guessing mostly here!! but signal with higher rms due to more harmonics should result in different/higher recitifed DC. taking sidechain directly from the plates would contain the full thd signiture of the amplification. but do you mean that the fullwave rectification cancels harmonics? or at least even.

a pair of transistors could buffer the signal between the diodes and the windings. prob need another threshold circuit then.

 

[abbey road d enfer]

im guessing mostly here!! but signal with higher rms due to more harmonics should result in different/higher recitifed DC. taking sidechain directly from the plates would contain the full thd signiture of the amplification

The rectifier needs the same voltage as that available on the anodes - probably even more actually. You can't avoid this voltage to exhibit the same distortion as the audio output, unless you use a separate low-THD side-chain, which is contrary to your idea of simplicity. 

but do you mean that the fullwave rectification cancels harmonics? or at least even.

Not at all; on the contrary, the rectifier introduces distortion. that's why the better designs use a buffer between the output and the rectifier.

a pair of transistors could buffer the signal between the diodes and the windings. prob need another threshold circuit then.

If you want to add SS circuitry, there are many things you can do. I know at least one designer who grafted a SS side-chain to a Fairchild gain cell.
[/quote]

 

[5v333]

my writing might be sloppy, im sorry if im not making my statments/questions in a correct manner.

abbey


yes i agree, the thd of the output will be part of the sidechain signal.
however after the OT the even harmonics will be cancled, so wouldnt that mean that the sidechain will also behave differently if we grab it of the secondary instead of directly from the plates.

grabbing the sidechain from the secondary instead from the plates will not generate very high negative DC but could it still be useful? maybe not when tracking vocals but for bus/program processing.

im doing some calculations with the 6386 as reference. i need to do more but so far it looks like having  a q of
Vgs-4
Va125
Ia8
Rl (one tube) 10K
input 1.2Vrms (4dbu)

i get about -1.6vdc which throws gm down a factor of about 2.
i need to see if that actually means a gain factor of 2 aswell but if that means that 6db GR is avaliable then im not that pessimistic. atleast not when the simplicity and economic gain is to our advantage.


hey buildafriend.

what about the 6v6?
is that something from the prr varimu that i cant seem to find but am very interested in!

 

[DaveP]

I kept this question and answer from PRR, it may be what you are referring to.

"> I guess like matching 6L6's

6L6 has a bent-enough Gm curve that you could build an effective vari-Gm stage around a pair. You may not be able to get 30dB clean reduction without heroics; anyway two 6L6 is pretty heroic already (idle ~200V ~180mA). But the tubes are readily available on Friday night."

His original  varimu thread started in May 2003, I believe.

DaveP

 

[abbey road d enfer]

yes i agree, the thd of the output will be part of the sidechain signal.
however after the OT the even harmonics will be cancled,

You need to reassess your understanding of how a push-pull stage and a transformer work. Due to the strong coupling between windings, the waveforms at both sides of the primary and at the secondary are identical, only the amplitude and polarity are different.

so wouldnt that mean that the sidechain will also behave differently if we grab it of the secondary instead of directly from the plates.

The only meaningful difference would result from different voltage. For example, if the voltage to the rectifier was derived from a separate secondary identical to the primary, there would not be any significant difference except for some loss due to the added DC resistance and leakage inductance.

grabbing the sidechain from the secondary instead from the plates will not generate very high negative DC but could it still be useful? maybe not when tracking vocals but for bus/program processing.

Typically, the rectified output voltage would be about 1.8Vdc for a nominal +4dBu.  This may be good for a few dB of GR, but I would think that would be challenging in terms of finding the adequate op point and getting stable and repeatable operation.

 

[buildafriend]

I kept this question and answer from PRR, it may be what you are referring to.

"> I guess like matching 6L6's

6L6 has a bent-enough Gm curve that you could build an effective vari-Gm stage around a pair. You may not be able to get 30dB clean reduction without heroics; anyway two 6L6 is pretty heroic already (idle ~200V ~180mA). But the tubes are readily available on Friday night."

His original  varimu thread started in May 2003, I believe.

DaveP

Hm..  ;D Thanks Dave. This is such a great looking building block for making improvements. Maybe with a higher plate supply and some 6L6 and power supply adjustments this could get fun. I always loved KT88 too, it might be completely overkill though. In addition to changing the input valves from triodes to tetrodes or pentodes it still couldn't hurt to add some kind of push pull output stage to help reduce potential thumps. One step at a time.. I feel as if I'm left with no choice but to experiment now. With a week off of work coming up and my lab finally back in full effect let's see where this go's.. new thread eventually.

All the best,

 

[5v333]

You need to reassess your understanding of how a push-pull stage and a transformer work. Due to the strong coupling between windings, the waveforms at both sides of the primary and at the secondary are identical, only the amplitude and polarity are different.

alright something is missing in my view on pushpull stages. ill have to revisit the texts on the subject. thanks!!

well if its the same thd signiture at the plates as at the secondary windings then yes of course, its better to pick the sidechain from the plates via a pair of buffers!!

and yes you are all right that it was pointless!! 


i wonder if the buffers and rectification could be the same devices by the way? something like 2 pnp emitter followers with a variable bias voltage that sets the point of conduction. and at the same time acting as high impedance buffers..?

 

[abbey road d enfer]

i wonder if the buffers and rectification could be the same devices by the way? something like 2 pnp emitter followers with a variable bias voltage that sets the point of conduction. and at the same time acting as high impedance buffers..?

I think the b-e junctions would zener when reverse-biased, but maybe a MOSFET would do; I'm not sure, I would need to think about it, or melt some solder...