Roger Mayer RM57 build?

[dirty1_1garry]

Here is my quick LTspice emulation. I used 2N3391A spice model for all NPN's.
Less sidechain control voltage (Vsc)  = more compression.
So... in theory compressor compress 8)
Output 200Hz sinus looks pretty distorted, so may be not ideal transistor models +  same model everywhere, what could be non optimal.

 

[Spencerleehorton]

Are you able to try q2-q3 to  2n5088
All other NPNs to 2n3707, see how it looks then?

Also was there a reason you used 200hz rather than 1k?

Some of those voltages might come in useful.

I’ll try a bc550c for the VU meter, I’m on version 2 now so will check everything tomorrow.

 

[dirty1_1garry]

Are you able to try q2-q3 to  2n5088
All other NPNs to 2n3707, see how it looks then?

I need to check  2N5088 spice model, same as 2N3707.

Also was there a reason you used 200hz rather than 1k?

I'm using 200Hz for emulation compressions because that's necessary to check time characteristics (100ms at minimum) and in the same time it's necessary to check whats going on with wave shape and 1kHz becomes too frequent on 100ms.

 

[Spencerleehorton]

Also your attack and release looks different.
I was getting 5v at slowest release so at 10meg.
I’ve got attack as 100k log and release says 2.2M lin but is Lorlin starting at 470k up to 10meg in 11 steps.
Could you also include the driver amp?

 

[Spencerleehorton]

Also q11 to 2n5088

 

[Spencerleehorton]

Im trying BC184C today as they seem to have about the right HFE 350-400.

getting similar voltages as per LTspice but different on emitter of your Q4, I'm getting 6.8v there, all other voltages are virtually the same in that part of the circuit.
Will test other voltages and see where it differs.

 

[dirty1_1garry]

here is an output amp.
15dB gain, headroom up to 6.8Vrms = 18dBu

 

[dirty1_1garry]

second emulation with corrected VT's.
preamp have pretty low headroom.

 

[Spencerleehorton]

interesting that your getting flat bottom half of waveform as well, as that is what I'm getting?

could you post a picture of Q4 emitter waveform please.

also could you try submitting 2N3391 for BC184C see how it looks as bit different HFE on BC184C

 

[dirty1_1garry]

and here is corrected preamp for better headroom. R18 changed from 47K to 33K

 

[dirty1_1garry]

interesting that your getting flat bottom half of waveform as well, as that is what I'm getting?

see preview image.

could you post a picture of Q4 emitter waveform please.
also could you try submitting 2N3391 for BC184C see how it looks as bit different HFE on BC184C

give me a couple minutes)

 

[Spencerleehorton]

ah ha!!, R18 i knew something was wrong here, just didnt know which way to go!!, will change and look at waveform!!!

 

[dirty1_1garry]

also could you try submitting 2N3391 for BC184C see how it looks as bit different HFE on BC184C

Swaped 2N3391A to BC550C (witch is close to BC184C). Absolutely the same result, same voltages.

Found that changing detector input pot from 100K to 1R improve RM output waveform and make it more symmetrical.

 

[Spencerleehorton]

changed R18 from 47k to 50k trim then trimmed down and have lovely symmetrical waveform now!!
just need to wait for some FETs to arrive then i can get some compression going!!!
thanks for all your help so far dude.

 

[Spencerleehorton]

are you able to test with 2N5461 FET or other p channel FETs? J175?

 

[Spencerleehorton]

One pcb i have trimmed R18 (47k) to 15k and the other pcb i have trimmed to 20K, might be better to leave it as trimmer in case of tuning later.
PCB A with 15k as R18 has BC184C as the 3 transistors for the preamp, PCB B has 20k and has 2N5088 as the 3 transistors.

 

[dirty1_1garry]

changed R18 from 47k to 50k trim then trimmed down and have lovely symmetrical waveform now!!
thanks for all your help so far dude.

yap, yap, yap! glad to hear that!)

are you able to test with 2N5461 FET or other p channel FETs? J175?

it all depends on the availability of their models, will check

Additionally I can say that I've tried to replace all NPN's (preamp and detector) to BC550 and... no significant changes occurred. So you can try BC184/BC550 everywhere.

 

[dirty1_1garry]

Here is 2N5461 emulation. It has a significant lover threshold, start to compress at much lover levels.

 

[Spencerleehorton]

Yes I found the 2n5461 compressed at lower levels and ended up squashing it to nothing!!
I’ll wait for some 2n3820 to arrive!!!

Is there a way to adjust for the compressing of the 2n5461 at lower levels?

 

[Spencerleehorton]

Could you try a J175?