EMI RS144

[ruffrecords]

The EMI RS144 was the little 4 input passive pre-mixer used at Abbey Road studios back in the days of the Beatles. "Recording The Beatles" notes that users were warned about interaction between controls because the design had aimed for minimum insertion loss. Users were told to ensure unused inputs were set to zero. With only one input used and the other three set to zero, the loss was only 6dB.

So, I am just trying to work out how they did this as I cannot find a schematic. They used regular Painton faders so these were no doubt bridged T types, So, assuming all the inputs were balanced and more importantly floating, the design could have been little more than four bridged T attenuators wired together much as in the attached schematic from the Audio Cyclopaedia:


Ignore the transformers and the output gain control. I am not quite sure why the output is taken from the middle and bottom of the network, nor why the bottom cold is connected to ground.

The second issue is, if I wanted to replicate it, how to make a reasonably good bridged T attenuator with just a dual ganged pot? It is easy calculate the resistor values required:




But can it sensibly be done with a dual gang pot?

Cheers

Ian

 

[Tubetec]

I think your going to need a dual gang pot with two different resistance values .
That sounds like a custom part order from the manufacturer unless you can rebuild a couple of dual gang types yourself .

Heres how H/P do it on the output of the 200dc

 

[ruffrecords]

Thanks for the example schematic. That is the first one I have seen with actual pot values. I would hope you would be able to use a regular dual 100K pot here and slug one half to "look like" 25K maximum. Time for a little LTspice I think.

Cheers

Ian

 

[Tubetec]

That dual gang pot is of course an Allen and Bradley part made specially for HP .

Obviously the parts need to be different values ,but are they both log law or does one have to be reverse log ?

My HP200cd is away in storage so I cant measure the pots to see whats going on .

 

[emrr]

Not sure what I'm missing, but we have the CAPI/Hairball 'T' attenuators freely available, several others in the EU. Nothingsignificantly different from a bridged 'T', and none available that will have an open shunt at -0dB, so all have some loading loss. All made with a combo of 1K and 500R wafers. 'Close enough' for modern SS, but somewhat less ideal for the purpose at hand, depending on surrounding conditions.

 

[ruffrecords]

That dual gang pot is of course an Allen and Bradley part made specially for HP .

Obviously the parts need to be different values ,but are they both log law or does one have to be reverse log ?

I suspect the laws are neither log, rev log or any other regualr law. LTspice will help with that.

Cheers

Ian

 

[ruffrecords]

Not sure what I'm missing, but we have the CAPI/Hairball 'T' attenuators freely available, several others in the EU. Nothingsignificantly different from a bridged 'T', and none available that will have an open shunt at -0dB, so all have some loading loss. All made with a combo of 1K and 500R wafers. 'Close enough' for modern SS, but somewhat less ideal for the purpose at hand, depending on surrounding conditions.

One thing missing is exactly how did they arrange the four controls such that with three at zero and one in use, there was only 6dB loss.

Cheers

Ian

 

[emrr]

One thing missing is exactly how did they arrange the four controls such that with three at zero and one in use, there was only 6dB loss.

Cheers

Ian

Yep...that should be 12dB. The Audio Cyclopedia is one I've never seen in practice, and does say "series connected - not for critical work'

From Cinema Engineering, see attachment here:

https://groupdiy.com/threads/line-vs-mic-amps-for-mixdown.87954/page-2#post-1157896

I might guess they left out the series isolation resistances? It certainly increases interaction. It would be zero-ish loss into a Hi-Z load, but 6dB into a matching load. Adds up. It only goes to min -6 with real 'T's having an open shunt position and direct zero Ω series connection at -0.

 

[ruffrecords]

Thanks for the Cinema link. I think that is the most succinct summary I have ever seen - but it is basically regular passive mixing. EMI must have done something special to get just 6dB with one channel operating.

I just checked the CAPI web site. Those Bourns T attenuators are $31 each!

Cheers

Ian

 

[emrr]

Thanks for the Cinema link. I think that is the most succinct summary I have ever seen - but it is basically regular passive mixing. EMI must have done something special to get just 6dB with one channel operating.

Yeah! I'd put money on leaving out the series iso resistances which raises the channel interaction! 3 channels set to zero create no additional load at all, any set to the first attenuating step create a parallel load of some amount.

 

[TimG]

Here’s one