PureAnalog M/SLevel Vari Mu Compressor (manual + schematic)

[chrispbass]

--Todays episode is bought to you in conjunction with UTM Industry--

You are live in the lab with Chris P Bass (well kind of...just go with it!)

So the output transformer has arrived, and very nicely made it is. Weighty with a good quality feel about it...thank you UTM



So last time, I was calculating plate currents, and this was really in regard to cross referencing the schematic and for me to calculate dropper resistors as I've gone off piste with regards to the PSU.

V3 = 1.725mA
V4 = 8.65mA
V1/V2 = 3.95mA

So adding those up and doubling them we get about 28.6mA
The meter will draw a further 1mA plus a smidge for the detector circuit giving say 29.6 or about 30mA.

Looking at the schematic for the PSU, Z measured a voltage drop of (259.8 - 245) across the 500 ohm resistor giving 14.8/500 = 29.6mA which is pretty darn good!!

I decided to build the psu on a separate board just so I'd got plenty of room to allow for testing etc.



Ive got 3 stages of filtering on my PSU and running regulated DC heaters. The block connectors bought off the pcb were handy to help dial in the voltages.

Initially, i just connected up the heaters to test these first, and one of the 6SS7 didn't seem to be getting warm. Inserting a 0.1 ohm resistor in the supply and measuring the voltage drop confirmed this. We have a duffer! I've got a LM338 regulator and with the heat sink, its barely getting warm. The bridge rectifier is getting a bit hot. Its rated for 4A, but may use something beefier when it comes to building the unit properly.

Initial B supply of 245V is on the pcb going to the block connector on the left to get the B supply for the 6SS7 and also the 24V for the detector. I used a trimmer rather than the 22K resistor, as I didnt have one to hand, but also helps with dialing that in.

I'm a smidge off, but more than close enough with my hotch potch of dropper resisitors, and happy to report that my measured voltages pretty much tally with Z's measured ones, so anyone planning on building this can be confident and go with what Z has on the schematic as your reference.

Before the output Tx arrived, I'd used a couple of 150 ohm resistors to compensate for the primary resistance to V4.




This was so I could check voltages etc and tbh, I'm glad I did as although I try to be very careful, I did slip with a probe causing a tiny short. This took out one of the 150 ohm resistors (not with a bang, just open circuit), but it could have been the primary of the transformer had it been connected!! Lucky!!

So with voltages correct and everything connected up, one last check to make sure there was no DC where there shouldn't be, and connected the output transformer.




All good, so time to fire up Mr signal generator and see if its amplifing. Using the multimeter, shows its amplifing. Using the scope verifies this although the output waveform shows its crapping out when upping the input level!! Also, doesnt seem to be compressing!!.

After a bit of head scratching I decided to check the detector output which is the -ve DC control voltage to the 6SS7. Measuring at the centre tap of the input Tx showed no voltage. Disconnecting it, bought it back so must be a short to ground! I checked resistance from the centre tap to ground and there was indeed a short circuit. Although I couldnt actually see a physical connection, I re-soldered and now all good. Re-connecting the control voltage to the centre tap confirms this. Unit now compressing and with a nice clean waveform straight through

Very happy and more testing to do of course. Also, there will be a case of matching V1 and V2.

Going back to the PSU, I'm thinking that the 150V shown on the schematic is probably a reference for the 15dB point. I will try to confirm this.

Happy Sunday all

 

[zzzzz]

@chrispbass legendary work. This rules. Your build is probably definitely going to be better and (physically) cleaner than the original!

 

[chrispbass]

@chrispbass legendary work. This rules. Your build is probably definitely going to be better and (physically) cleaner than the original!

Thank you Z

Much work still to do regarding testing and of course the all important listening tests. I haven’t done anything else with it since the last update, but some more parts have now arrived, so I can simplify my parallel dropper resistors and hopefully get those voltages dialled in nicely.

For anyone fathoming this circuit out (myself included), one thing to note is how much voltage swing there is in the circuit when reaching and exceeding ‘the threshold’ of the 24v DC reference in the detector circuit. This is audio level of approx 17V RMS which when multiplied by 1.41 will give a peak voltage of 24v (AC) to ‘overcome’ the 24V DC to begin compression. After this point with increasing input, the -ve DC control voltage starts to rise and thus affecting gain with regards to the 6SS7. If your meter is dialled in at FSD, you can confirm this, as just as the meter starts to indicate gain reduction you should see approx 17V RMS at the cap/diode junction. With increasing input, you can see the steady state voltage of 89.5V rise (across the 20k resistor) to over 200 volts as the current increases. This will also cause a voltage drop across the 500 ohm resistor.
Visually (no measurements yet), but at about 150V, the meter is sitting at a point which would pretty much show 15dB of compression (referencing pics of the original unit) so almost proof of why 150V is shown on the schematic, even if not particularly scientific at this stage.

Thinking ahead though (walk before you can run Chris!), a few mods I can potentially see are the addition of a pad to the output (10/20dB), poss pot to replace the 56k resistor as an adjustable attack control and perhaps the inclusion of a HPF (100-120Hz). I’m thinking that a few switchable cap changes could accomplish this as there is a CR filter formed by the 100nf/270k. May be peeing in the wind on that one…thoughts?

I’ll add more updates when I get a mo.

 

[KrIVIUM2323]

The pot in place of the 56k might work or not, it'll depend ( can induce a 'thump' which can be interesting on some instrument, probably less on a bus or master).
100nf/270k is where sidechain hp indeed takes place. Lowering to 0,047nf, 0.022nf or even 0.001 should do.

I think your comment about the 150v/15db reduction gain makes sense too. This is how it was given on Altec's 436/438 manual ( some value inside a 'box' are for 15db reduction on the schematic) .

 

[chrispbass]

The pot in place of the 56k might work or not, it'll depend ( can induce a 'thump' which can be interesting on some instrument, probably less on a bus or master).
100nf/270k is where sidechain hp indeed takes place. Lowering to 0,047nf, 0.022nf or even 0.001 should do.

I think your comment about the 150v/15db reduction gain makes sense too. This is how it was given on Altec's 436/438 manual ( some value inside a 'box' are for 15db reduction on the schematic) .

Thanks K

Yes, the 15dB ref is also shown on the original Gates schematic that Rock posted on page 1.

At this stage, It’s obviously wise to get it working correctly stock, but yes, ultimately I suppose mods will depend on intended use.