Hi y'all, newbie question. In a Vari-Mu compressor, such as a 436 or ba6a. How are the diodes used to rectify the output/sidechain for compression turned into negative voltage to be imposed onto the grid of the input remote cut off tubes and attenuate? I feel like im missing something. How does the diode plate, which in my mind pass a positive voltage.. how does that become an increased negative voltage for the grids? I am clearly confused.
Same way as properly biasing a high gm tube. E88CC for example, Vba100V,, + 9V on the grid, 680 ohm resistor into cathode, current adjusts itself to 15mA (14.2-15.8mA) voltage over cathoderesistor is 10.2V (9.656-10.744V), grid is negativ to cathode by 1.2V (0.656-1.744V).. So, with selfbias, and by choosing a large enough cathoderesistor you can have both, a positive voltage at the grid and a negative gridvoltage in relation to the cathode.
Last edited:
Thanks gorgon53! That makes sense to me. If the signal is louder, isn't that making more positive voltage come out of the diode tube section? How does that result in gain reduction. In the 436 it marks voltage swings as low as -30V at the grids during significant gain reduction, thats if I am understanding that correctly. How do you get -30V from the diode plates? it looks like its noted as -0.5V bias with no signal.
Get some rest! It is the last page of this PDF. (Page 6).
http://www.dvq.com/hifi/images/436c.pdf
@JMFahey gave a great description on how this circuit works here:
https://www.diyaudio.com/community/...ut-introducing-distortion.375752/post-6754231
But my green brain can't wrap around the concept of how exactly the negative rectified voltage is created.
http://www.dvq.com/hifi/images/436c.pdf
@JMFahey gave a great description on how this circuit works here:
https://www.diyaudio.com/community/...ut-introducing-distortion.375752/post-6754231
But my green brain can't wrap around the concept of how exactly the negative rectified voltage is created.
I agree (neither JMFahey nor that PDF explains it), looking at that schematic, an increasing signal level should cause the 6AL5 plates and the grid voltages to become more positive, causing the input triodes to amplify MORE.
There are surely many people who would know how to explain this on http://groupdiy.com, they've been building reproductions of classic compressors for many years.
There are surely many people who would know how to explain this on http://groupdiy.com, they've been building reproductions of classic compressors for many years.
It might be helpful to look at V3 as if it were a conventional full wave power supply rectifier pair. Except with the diodes in reversed polarity. It works the very same way, but makes the opposite polarity (negative) voltage. Then, P3 sets a threshold and P2 sets an amount, of compression.
All good fortune,
Chris
All good fortune,
Chris
Ok hmmm so then, the vacuum diode, is only 'on' when the voltage at the cathode of the diode is negative, making the plate voltage negative/passing current from plate to cathode during the negative portions of the signal? Maybe the direction I assumed tube diodes allowed conventional current to flow was backwards? We're getting warmer.
Electrons only go one way in a vacuum diode, but Benjamin Franklin didn't know about electrons, so "current" goes the other way. An abundance of electrons is "negative", and a dearth of electrons is "positive". The absence is just as real as the presence (which is actually not a lot of reality, but) so it's really really easy to overthink this stuff. Best to just nod your head appreciatively and soldier on. Arf!
All good fortune,
Chris
All good fortune,
Chris
It's always better to post the full schematic to get a full understanding.
The higher the AC output the higher the rectified DC voltage and in this case it's a NEGATIVE DC voltage that gets sent back to the grids of the compressor triodes, two halves of 6BC8. This negative voltage changes the bias current of 6BC8, which is a semi-remote cutoff tube, and the lower the current the lower the gain and, voila, automatic gain reduction! This up and down changes dynamically also changes plate voltage which creates a lot of distortion so having PP circuit you get cancellation in the PP output transformer. It's the main reason why most vari-mu circuits are PP circuits.
Where's PRR when we need him?!
The higher the AC output the higher the rectified DC voltage and in this case it's a NEGATIVE DC voltage that gets sent back to the grids of the compressor triodes, two halves of 6BC8. This negative voltage changes the bias current of 6BC8, which is a semi-remote cutoff tube, and the lower the current the lower the gain and, voila, automatic gain reduction! This up and down changes dynamically also changes plate voltage which creates a lot of distortion so having PP circuit you get cancellation in the PP output transformer. It's the main reason why most vari-mu circuits are PP circuits.
Where's PRR when we need him?!
OMG I can't believe I missed this, I was thinking of the diode like an "active device" like a triode and some positive voltage must be driving the plate to make it conduct. In this circuit it's the cathode being actively driven NEGATIVE that makes it conduct! I don't think I would have made that mistake with a semiconductor diode symbol, but whatever.
Electrons go from cathode to plate. Conventional current goes from plate to cathode when (in this case) the cathode becomes more NEGATIVE than the plate.
And, importantly in this circuit with large resistances in both cathode and anode paths, the only voltages that matter to the diodes are the voltages appearing at their terminals. They are "birds on a wire", caring only for their relative, instantaneous, anode-cathode voltages. That's all they know or can ever know. Maybe this is obvious, but it's maybe best to spell it out for future students.
All good fortune,
Chris
All good fortune,
Chris
So just replace the tube diode with a silicon diode symbol in your mind. A diode is a diode, whether it is made of vacuum or silicon.
Thanks everyone. While I have you here, what keeps the grids of V1 at -0.5V when there is no compression / signal?
Hi, P3 set a threshold/ratio ( it's most often linked with vari mu, exception being Urei 176 - for 'no sidechain amp' kind of vari mu- and the fairchild 660/670 -with sidechain amp this one). It's a bit different than a treshold on a solid state unit though as those ( vari mu) are gbt ( gain before treshold, like a 1176 too) so input/drive level is a part of the treshold outcome ( and is often more dramatic in effect than the treshold pot by itself...).
P2 doesn't set an amount of compression but a time constant ( release time).
The 'no sidechain amp' kind of vari mu are 'slow' ( mainly because of impedance concern in the sidechain circuit) and were availlable with 'longish' attack time ( wrt solid state counterpart or vari mu with sidechain amps), in that case r12 set the attack time constant.
If i remember correctly this is the bias of 6al5 cathode ( which vary between 2,3v and 20v with p3 'treshold') which set it at -0,5v output. But i might be wrong.
Last edited:
Universal Audio 176 circuit below with sidechain amp being the output amp after the interstage transformer. It also uses 6BC8 like the Altec 436.
Yes, the outamp is used to drive the sidechain, it's not dissimilar to the 436 and it's variants (most 50's/60's 'studio compressors' are build along this principle including EMI RS124 - famous because of the Beatles and it's heavy use at Abbey Roads), even most recents units are 436 variants like Chiswick Reach/Thermoionic Culture 'phoenix',...).
Your schematic is DaveP redrawn.
Even if in the 176 the out amp seems beefier it's still limited in out power.
It's the 'non sidechain amp' family.
The 660/670 have a full tube 20w capacity amp to drive the sidechain resulting in massive drive capability ( enough to reach sub millisecond attack/10 seconds release from one small peak despite having to drive 8 grids in //). It's a 'real' limiters, not an aesthetic tool as are most compressors of this time.
Your schematic is DaveP redrawn.
Even if in the 176 the out amp seems beefier it's still limited in out power.
It's the 'non sidechain amp' family.
The 660/670 have a full tube 20w capacity amp to drive the sidechain resulting in massive drive capability ( enough to reach sub millisecond attack/10 seconds release from one small peak despite having to drive 8 grids in //). It's a 'real' limiters, not an aesthetic tool as are most compressors of this time.