EMI RS124

[anvl]

Really beautiful. Congratulations !!!!

 

[thomasdf]

Thanks ! I'm on the 2nd unit now, as well as finalizing the files for front panel silkscreening !
I would like to add a true bypass but the front panel is beautifully simple. Maybe a rotary with "in / bypass / linked" ... but I have to think of how to do that with a simple mechanical switch. Another useful addition might be an input pad.

I was thinking of having the panel engraving and unfill painted so each unit was numbered 1 & 2 and the hold positions were red, like on the original unit... but since it will be silkscreened, I will lose the numbering and red color. All black... sob.

I have discovered a color paper that can be thermo transfered via laser and would allow for various colors ... but I have a friend who's good at silkscreening so it will be faster and probably more beautiful.
I'll keep the idea of engraving for my next build, a STA Level with vintage iron.. and an aluminium front panel plate haha but that's for another time, another thread maybe

Cheers and thanks for the kind words !

edit : don't pay to much attention to the crappy WIP and overly compressed jpg. I am using vector files of course.

edit 2 : for some reason, with my 1mA DC Simpson meter, I ended up using a 45R value to get them to zero. Since it's pretty close to the original resistor value, I am confident the threshold is not all the way off.

 

[thomasdf]

Maybe I've been too focused on this and I am losing some common sense, but maybe someone can help me decipher this :

I used a pair of 200R 2W potentiometers to zero my prototype meter (1mA DC Simpson meter) and ended up with a value of around 45R to do so. When I measure the value of the pots, I get 45R. When I use a 39 or 43 or any other close value resistor the meter pegs all the way to the right. I tried replacing the resistor with another pot (100R this time), I zero the meter, measure the pot ... and around 40R again. I replace the pot with a resistor, the meter pegs all the way again ...
I've seen on the forum that EMRR recommends a 220R resistor for a 1mA meter, tried with a 220R resistor, the meter pegs all the way again.

I am sure of my connections wether they are made with solder or croc clips ... regarding the resistors, I have tried with 3/4W but haven't tried with larger wattage... could this be related ? The resistors does not get hot, at least not that I can sense with my fingers.

 

[Rob Flinn]

The wattage will have nothing to do with it. You can easily calculate the power dissipation P =(V*V)/R where v is the voltage across the resistor. I had a similar thing where I used trimmers to dial in the meter. I used a 200uA meter and with the 39R (from memory) recommended resistor it would not zero, which is why I opted for trimmers. When I measured the trimmers when I got the meter to zero they measured, yes, you've guessed it ....39R ! I just left the trimmers in.

 

[JMan]

This is hurting my brain as I try to understand it. In both of your cases, a fixed resistor of a particular value caused the meter to peg (or prevented you from zeroing it), but a trimmer/pot dialed in to the exact same resistance worked without issue? What could explain this? Or am I just misunderstanding something...

 

[thomasdf]

Rob, I am sorry for you ... but I glad I am not the only one confronted to this black magic haha !

JMan : you nailed it and actually summed it up way better than me. 100% accurate description of the problem

 

[freshtapescent]

When I use a 39 or 43 or any other close value resistor the meter pegs all the way to the right.

I’m a little buzzed, so do let me know if I’m talking out me rear end.

Are these near-pick resistors from the box right into the breadboard, or actual Ohm readings of resistors you took with the same meter you measured the pot?

Even precision resistors have a tolerance. If you grabbed a perfectly normal pair of of 5% 47R, you could have as low as 44 Ohms and as high as 49 Ohms. Though I don’t think this is your problem, you’ll get back to this so don’t forget,

If you were just trying to trim a stock meter, You should only need one pot, connected as a variable resistor to replace R8. Connect one side of the meter to the pot’s wiper and the other side of the meter to one end of the pot. Leave the other side of the pot disconnected!

On the Abbey Road schematic, the 6BC8/6ES8’s cathode currents balance through the 100ohm cathode balance pot, and their combined current flows through 130R R25 in series with the 34Ohm meter shunt resistor R8. The R25/R8 divider relationship along with the stock 200uA meter resistance provides the right amount of quiescent current for the tubes during no-gain reduction.

To my eyes, you need ≈ 50R from cathode balance pot + 130R + 34R= around 200Ohms in total from cathode to ground to get the TUBE to work right with a 200uA meter.

For a 1mA meter, to still guarantee you’re setting the right amount of current for your TUBES, you need to maintain that same resistance to ground, but scale your divider accordingly to provide an order of magnitude MORE current for your 1mA meter. If you’re only making R8 smaller to get more current, I think you’re also allowing more current through your tube and chasing your tail.

From the wiper of the cathode balance pot, hang your 200R 2W pot’s full resistance to ground, with your 1mA meter hanging off the wiper to ground. Now zero your meter, pull the pot, measure from wiper to cathode end to replace R25 and wiper to ground to replace R8, and select measured resistors from the box until you find ones that work.

 

[thomasdf]

I will try that tomorrow or friday, and will report back.

 

[thomasdf]

Thanks a lot for a clear explanation I hooked up the meter accordingly and was able to zéro it with the pot. But with the resistors… COME ON !

I think I’ll stick to the pot since it allows for tweaks after or during warm up, and is probably better since I am building a stereo pair that might go through tube change one day, etc

 

[rock soderstrom]

Thanks a lot for a clear explanation I hooked up the meter accordingly and was able to zéro it with the pot. But with the resistors… COME ON !

uhh, strange!

 

[thomasdf]

I am well into the 2nd unit now, and by building it side by side with the first one I realized I had forgot R13 on the first unit ! 10K feeding B+ to the 47K plate resistors junction for the 6BC8. It sounded nice but I suppose the 6BC8 was under powered in such a config !

 

[freshtapescent]

Ok! Now you're in business. That'll most definitely help the meter work better too.

 

[thomasdf]

Both unit are now finished wiring excepted for the bypass and sidechain link switches (Schadow eyeball on the way). I yet have to turn on the 2nd unit and see that it does not catch fire, as well as checking that the added forgotten resistor does not mess with what I had and enjoyed so far on the 1st unit. I'll probably do that tomorrow ... and will move on the with front panels (DIY silkscreening + drilling ... including the 72mm hole for the VU meter, oh well..)


Edit : you can see the IEC socket holes did not go as planned, I really hate making these ... a friend of mine lend me a 28x21mm hole punch but I went for a IEC + fuse combo ... arf

 

[thomasdf]

The two units are working and sound as expected. I already want to do some modification, because on very loud signals (loud mixes or master levels) I cannot achieve unity gain/loudness. This is because the compression kicks too early and deep for such duties.

It think my best move would be to add a 6dB pad on the input, and raise the threshold.

Tell me if I'm wrong, but my understanding is that in order to adjust the threshold there are two options.
- VK1 style : adjusting the signal going to the sidechain (it uses a stereo 47k pot) or
- adjusting the DC signal being fed to the 6AL5 like in the orig 436
I prefer the 2nd way, as I am not looking for too much options or controls and one I find a higher threshold I'll probably be fine with it. Can I use lower values for R10/R11 220k ?, or should I tap some B+ from the power supply like in the 436 ?

Why do you think EMI ditched that part of the design (appart from simplifying the controls) ?
edit : I see the 436A has a similar design, so EMI did not ditch things, they based their design on the A or B version .. Altec added that control on the C version, obviously.

Thanks !

edit : and one of the units GR meter does not always come back to zero (sits at -5, and then after so GR goes back to 0 sometimes), I'll investigate why in detail later on.

 

[MagnetoSound]

Adjusting the bias on the SC rectifier alters the ratio as well as the threshold. I prefer to adjust this setting according to the program material, and then adjust the amount of compression by setting the input level but you can adjust the sidechain input instead if you prefer.

I’ve never really got to the bottom of why the return-to-zero is so inconsistent in vari-mu compressors, other than the tubes being susceptible to power supply variation, which is rarely fully regulated.

 

[thomasdf]

The thing is I’d prefer if this was a fixed setting, as in the original design, but just a slightly higher (fixed) threshold & ratio.
I think I won’t mess with the 220K resistors since I see it’s common to almost all varimu design using a 6AL5 : Altec & EMI, but also UA175/176, STA Level.. only the VK1 uses a diff arrangement with 18K resistors and a dual 47K pot adjusting the threshold (= amount of signal going from the output to the sidechain)

edit : I added a B+ tap from the power supply, through a 22k pot and a 270K resistor, ala 436C. I need to use the compressors a bit more to fully decide where I want that threshold to stay .. and if I can't reach a middle position I am considering ditching the pot and having a switch toggling between two sets of resistors for either stock setting (really really great on instruments, I mean GREAT) and a more modern "bus" setting where I can reach unity gain and only compress 5dBs or less on CD like levels ...

 

[anvl]

Dear Thomasdf. I followed Altec 436c to have a Threshold control pot in my RS124 p2p circuit and it's working really beautiful.

 

[thomasdf]

Yes, that is exactly what I did. I used a pot inside the enclosure, right now it is set for 9VDC at center lug / 220K junction and it seems to be usable on loud sources as well as individual instruments.

I will change the input pots, they are log now but the responses is not great... I will try linear and rev log, the idea is to get more control in the lower settings of the input knob. Maybe with these two mods I can get away without any pad.

 

[anvl]

I placed Threshold control knob in the front panel. I find it more versatile like this (in combination with 47K dual log pot on input you can manage every situation)