EMI RS124

[letterbeacon]

So now I'm working on the sidechain part of the circuit. The audio stutters and clicks on the faster recovery settings, but is fine in the slower ones and on hold. I assume this is because V1 isn't balanced properly.

I can't seem to tune out the clicks when I use the neon balance circuit - they get slightly quieter, but not by much. When I balance the plate voltages using a DMM on V1 pins 1 & 6, I still get the clicks and stutters on the faster recovery settings.

I've tried two different 6BCBs and a 6ES8 but I get the same issue.

Should I be able to tune out the clicks with the balance circuit? The dip in volume is very slight at the moment...

 

[Rob Flinn]

I couldn't get rid of the neon clicks completely, but I don't notice any artefacts when running program material through either of mine.

 

[letterbeacon]

I couldn't get rid of the neon clicks completely, but I don't notice any artefacts when running program material through either of mine.

Did you install R8? I think as it's up by the meter I might have wrongly assumed it was a shunt for the meter, but now I look at it, it looks like it could also affect V1.

I don't have a 39r resistor to hand, so I'll have to get one with my next component order.

 

[Rob Flinn]

Did you install R8? I think as it's up by the meter I might have wrongly assumed it was a shunt for the meter, but now I look at it, it looks like it could also affect V1.

I don't have a 39r resistor to hand, so I'll have to get one with my next component order.

R8 in series with R25 bias the 6bc8 or ecc189.    Depending on the meter you use the R39 will not necessarily work. I had a random meter from my pats bin that didn't line up, so I bought a pair of 200uA meters for the job & even with those wouldn't bias or zero, so  I used 2 trimmers to achieve the correct bias & zero the meter.  If you go back through the thread I detailed what I did.

 

[letterbeacon]

Ok great thanks, will have a read.

I’ve got a 100uA meter in there at the moment with a trim pot allowing me to shunt enough to get the meter to ‘zero’ (or read 100uA).

 

[Winston OBoogie]

Heya guys,
sorry I haven't been around here much to be a 'cheer-leader' but,  I'm happy to read your unit(s) seem(s) to be working better now.
Regarding the 39R resistor, this value will only work to zero the meter if your own meter is as original spec.
I haven't read back on Rob's method to calibrate things but I'm sure it's exactly as is needed i.e.  just find a value that works for the meter in hand.

For the neon trim etc.  I was never able to trim out all the thumps, regardless of valve used.  But the best position would remove most of the lower frequency portion of the oscillation thumping and what remained when things were in balance was more a 'tick, tick, sound  rather than a 'thump, thump'. 
There should not be any artefacts when using the unit on the fastest release.  I'd go as far as to say that the first few positions are where you'll end up using it.  I found the slower settings all but useless myself!
I also thought it was at its best when there was a fair amount of compression happening rather than just tickling the meter.  Just use that  "Hold" feature to prime the unit before use to stop the initial suck down.

I'm interested in Rob's impressions of using the 6ES8 over the 6BC8 which is also the way I did it.
This valve change isn't written on the schematic but there is a  note in the EMI archives written by Mike Bachelor  indicating that this valve gave less distortion, which is absolutely true.  There is also an overall response difference which is what I'm interested in hearing opinions about.    Note that not all Abbey Road RS.124's had this change but, at one point, quite a few of the original 32+ in- house units that I know about did. 

Anyway, onwards and upwards
 

 

[letterbeacon]

Thanks WoB, that's really useful!

I've substituted R25 for a 500r trim pot and R8 for a 100r. This has allowed me to dial in the 1.1v needed for the cathode bias and make sure my 100uA measures full scale without any input.

Trouble is, I'm still getting artefacts. I'd describe it as more really fast audio dropouts. When I'm playing music through it it happens wherever the input pot is. When I play a 1kHz sine through it, it occasionally does the audio dropout thing, but when I crank up the input pot so the meter reads 20uA the dropouts stop.

When the dropouts are occurring with the sine wave, the control voltage is jumping between about -4.5V to -3.8V, but then the dropouts stop when I have the input pot so the control voltage measures -9V.

There's really a good case to be made here for using a matched pair of resistors on the secondary and then substitute, say, a mono 2K5 log pot pre the input transformer.

I'm tempted to try this. Do you mean replace the pot on the schematic with 2 matches 47k resistors between the secondaries and the CT and then a 2k2 log pot between the + and the - before the input transformer?

 

[letterbeacon]

Yes the circuit is exactly the same as the one you posted apart from an extra dropping resistor in the power supply after the rectifier circuit and before R13 and R14 to bring the voltage down a bit.

I just tried another input pot in case that was more closely matched but got the same issues. I don't have a 2k4 pot on hand at the moment so that will have to wait for the time being.

 

[Rob Flinn]

Not sure what brand of pot you're using but I normally use Alps Blue Velvet for Vari Mu's like this.  The tolerance 2wafer to wafer" on their spec sheet is no better than any other brand, but I've never experienced the issue you're having ony any Vari Mu's I've built, & I have built quite a few over the years.  I have had other brands like OMEG where if you used them as a stereo volume control one channel would kick in way before the other.

 

[Rob Flinn]

2/ Reading through the changes you made to R25 and R8:  you have 100R in // with the meter for R8, and a 500R trimmer for R25 yes?  I think the trimmer should be across the meter, and a 130R as per the schematic for R25 myself. ]

I couldn't get the bias correct with a 130R at all.  From memory once I had 1.1v on the wiper of the balance pot by adjusting the 500R, the 100R to trim the meter to zero made little difference to the overall bias.

 

[letterbeacon]

Thanks for the tip on the Alps, Rob. Mine's just an Alpha at the moment I think.

And thanks Winston for your list of things to try, I'll give them a go later this afternoon when I'm back in front of it.

2/ Reading through the changes you made to R25 and R8:  you have 100R in // with the meter for R8, and a 500R trimmer for R25 yes?  I think the trimmer should be across the meter, and a 130R as per the schematic for R25 myself.

I originally built it with R25 as a 130r fixed resistor and R8 as the 100r pot and I was getting the dropouts. I measured 1.18V on the wiper of the balance pot when I had it like this. When I swapped R25 for the trim I was able to dial in 1.1V as per the schematic, but that didn't seem to change anything.

 

[Rob Flinn]

OK got it.    Just out of curiosity, do you know what the 500R trim value ended up as?

Oddly it was nowhere near with a fixed 130R, but measuring the trim pot it's actually at 133R ............  So I would guess if I put a 130r in it should be pretty damn close.

 

[Winston OBoogie]

Oddly it was nowhere near with a fixed 130R, but measuring the trim pot it's actually at 133R ............  So I would guess if I put a 130r in it should be pretty damn close.

Haha, OK well,  I'm glad that sort of stuff doesn't only happen to me then!  ;D

 

[letterbeacon]

Yep, that could be a problem for sure, I've seen some pots that would require dissassembly to line up wafers so, good tip on the Alps.

I just soldered up a 2 deck rotary switch with 10x 4.7k 1% resistors on each deck in place of the input pot, but sadly I get the same dropouts.

Things to try:
1/ The diode's threshold bias derived from the output valve cathodes, does that measure correctly  (3.4V) ?

I measure 3.2V on the cathodes of V2/

2/ Reading through the changes you made to R25 and R8:  you have 100R in // with the meter for R8, and a 500R trimmer for R25 yes?  I think the trimmer should be across the meter, and a 130R as per the schematic for R25 myself.

As I mentioned above, this is how I had it when I first built the unit and I first noticed the dropouts.

3/ What happens  if you just use the standard values 130R and 39R, forget the meter being off calibration?  Still drop outs with lower signals?

I don't have any 39r resistors to hand, but I'll hopefully get some in the week.

4/ Correct values for R12 (33K), and C4 (.5uF)?  The release resistor network taken from the cold side (not at the diode) of R12?

R12 has the correct colour bands on it for 33k. I'll pull C4 and will measure it when I next have it out.

 

[Squeaky]

Hoping this is not a hijack. I love the RS124 compressor and it was one of the first I ever built. I would like to build a stereo pair someday.  In contrast to the RS124, I have had endless problems (thumping) with a UA175B that I built.

When building the RS124 I followed the schematic as closely as I could. I used a fancy volume control (Goldpoint) on the input which might have helped? I also had Edcor make me a power transformer to use, I think they now sell it as the XPWR223.

I think the problem I am having with the UA175B (I have tried so many different input valves) might be related to my layout, could this be possible? Could layout be having an impact for the OP? Trying to learn here. I know of blocking distortion from building guitar amps and thumping seems like an extreme example of that in some ways (with a clamp and swell thrown in for good measure) are they all related problems on a spectrum?

I only use the first couple of setting on the RS124. I also added a bridged-T attenuator to the output but rarely use it. I'm not sure I would do so again. You can wind the input way up and still not hear much in the way of artifacts.

 

[letterbeacon]

My suggestion was only to use the stock values including that 39R to rule out  your new values causing issues.  Since it doesn't seem to be the issue, I think, based on Rob's arrived at values, when you're tidying stuff up at the end, you could use the stock R25 of 130R, and then your trimmer across the meter to dial that in.

Anyway, sorry to send you on a couple of wild chases on the other stuff.

Oh don't apologise, I'm grateful that you're willing to give me a hand with this!

Does the drop out stuff seem like it might be "blocking" distortion from the coupling caps into the output valve?  I never had an issue with that in this particular circuit but it's a known possible issue on other units.

I'm not sure about that,  the only example of blocking distortion on youtube that I can find sort of thumps when it cuts out whereas mine just clicks off like a kill switch.

I'll try and post up a recording of the dropouts later today so you can hear what's happening.

The amp appears to be working fine though when I have the diode tube out.

 

[Rob Flinn]

Hoping this is not a hijack. I love the RS124 compressor and it was one of the first I ever built. I would like to build a stereo pair someday.  In contrast to the RS124, I have had endless problems (thumping) with a UA175B that I built.

When building the RS124 I followed the schematic as closely as I could. I used a fancy volume control (Goldpoint) on the input which might have helped? I also had Edcor make me a power transformer to use, I think they now sell it as the XPWR223.

I think the problem I am having with the UA175B (I have tried so many different input valves) might be related to my layout, could this be possible? Could layout be having an impact for the OP? Trying to learn here. I know of blocking distortion from building guitar amps and thumping seems like an extreme example of that in some ways (with a clamp and swell thrown in for good measure) are they all related problems on a spectrum?

I only use the first couple of setting on the RS124. I also added a bridged-T attenuator to the output but rarely use it. I'm not sure I would do so again. You can wind the input way up and still not hear much in the way of artifacts.

Maybe start a seperate thread for this question.  Then if other people have the same issue they will be able to search it easily.

 

[letterbeacon]

Here's a clip to give you an idea of what's happening to the audio:
https://www.dropbox.com/s/dlraucxldqk2d4h/RS124.wav?dl=0

This is with the Recovery switch on 1, and the input pot at about 9 o' clock. At 14 second I crank the input pot fully clockwise and you can hear the audio drops out less, but it's still doing it.

 

[letterbeacon]

Well it was always going to be a stupid mistake...

It turns out that the place where I bought the .5uf actually sent me a 0.05uf. I was going over the sidechain bit and noticed the incorrect cap code on the side. I don't have any .5uf caps to hand so I've put 2x 220nf in parallel with the 0.05uf and it now doesn't click anymore! I've only had a quick play with it but it's sounding great!

Thanks Winston and Rob for sticking with me on this!

 

[Rob Flinn]

Glad you worked it out.  It had to be something silly !