Implementing new Master section on Alice 828 console - help needed

Hi everyone

I have an Alice 828 mixer that I really like although I would love to implement a Discrete summing section and balanced output with transformers.

The desk has 8 channels plus 1 stereo echo return and summing is done originally with 33k resistors, then the Master has a really cool Stereo Limiter, Master fader, makeup gain with IC and unbalanced out.

I would like to get rid of the IC's on the Master and balance the LR outputs without loosing the limiter, by implementing the CAPI 2 Channel ACA & Booster (API style) or 2x Neve BA283 type output stage, both of them provide transformer balanced outputs.
http://capi-gear.com/catalog/product_info.php?cPath=22_46&products_id=403
http://www.jlmaudio.com/neve_b2.gif

Could you please help me out in explaining how this limiter works?
and can you suggest I could I implement the CAPI 2 Channel ACA & Booster (API style) or 2x Neve BA283 type output stage in this circuit?

Thank you so much,
schematic of the Alice 828 Master section with limiter attached

> how this limiter works?

Rude and crude.

Same as most. Peak-catch the output with a diode and cap. Run that to a gain-turn-down.

This one's "turn down" is TR1. Yes, a BJT has a *small* "linear" range as a passive attenuator, signal swinging both ways around zero. I would pencil this as 5mV fairly clean, 20mV getting dirty.

Not all info is here. What are these diodes?

At a guess, output is limited at 1.2V (or more). IC1 has gain of 68K/2K2 or 32. So input level is limited to 1.2V/32= 37mV. Much more than my usual rules for a BJT attenuator.

Inovonics' first product used this attenuator but with a LOT of added parts to reduce distortion (and a far fancier peak-catcher). Noteworthy that they moved to other topologies fairly soon, even for the less dynamic radio limiter market.

CBS used several series diodes as attenuator to get first-order distortion cancellation and added clean range.

_I_ would put your fancy line driver at IC2, and leave this stage strictly alone.

The mix bus is passive and appears to have 9 sources. The bus attenuation is therefore 1/9 (19.1dB) and as PRR says the bus voltage should be no more than 37mV. This means the nominal bus feed level is about 37mV x 9 = 333mV which is about -7dBu. If you use the CAPI ACI you will change this to active summing and the bus attenuation will be 0dB. The signal going into the limiter will therefore be much too large. You could add a 19dB attenuator at the ACI output but that will worsen the noise performance by the same amount.

Cheers

Ian

PRR said:

_I_ would put your fancy line driver at IC2, and leave this stage strictly alone.

Click to expand...

Thank you so much for explaining the circuit PRR.

I would just replace the IC2 output stage for discrete transformer balanced outputs, so the  CAPI 2 Channel ACA & Booster  is out.
I'm thinking of replacing IC2 with the output stage of Neve BA283 card feeding Carnhill VTB1148 transformers (LO1166 equivalent)
It's the BA283AM circuit:



What do you think?

ruffrecords said:

The mix bus is passive and appears to have 9 sources. The bus attenuation is therefore 1/9 (19.1dB) and as PRR says the bus voltage should be no more than 37mV. This means the nominal bus feed level is about 37mV x 9 = 333mV which is about -7dBu. If you use the CAPI ACI you will change this to active summing and the bus attenuation will be 0dB. The signal going into the limiter will therefore be much too large. You could add a 19dB attenuator at the ACI output but that will worsen the noise performance by the same amount.

Click to expand...

Thanks Ian, I understand. If I want to keep the original crude limiter I have to leave the summing as it is.