Grounding and bussing in a high-Z mixer

[emrr]

I got to have a look at the tube Neve again yesterday, and what it told me was 'no worries'; it works just fine. It's one of the pair he had built originally to kick off the company. The Neve co. owns the studio unit and trotted it out at the last AES con. This one would be the outside broadcast console.

The one I'm talking about is pictured here; you can click on it and get an enlarged view:

http://rupertneve.com/company/history/1960/

It says there 1961, but the filter caps have 4/62 dates on them (! they're not mine to change). It is totally hi-Z between the mic input and program output xfrmrs. The external case seen in that pic is long gone, so it's essentially a solid front panel with completely exposed guts on all other sides. The amps are all built on solid metal L brackets, so there's some isolation shielding between amps.

It's ten channels (11 counting echo return), and has 12AX7 channel output stages that drive 250K pots in parallel with 500K monitor pots.

The 250K pot sweepers feed a panning pot (added later by ?) which is a dual 250K pot array; pair of 250K linears to ground with 68K trim resistors from signal top to sweeper, signal leaving via sweeper. Hope that's clear.

There's also a 500K pot for echo send hanging off the channel fader sweepers.

470K mixing resistors are used from the sweeper outputs of the channel pans, echo, and monitor sends.

The L and R busses have all the 470K's placed together at the input tube of the group amps. The monitor and echo busses simply have the 470K's hanging directly off the pots with what looks like #20 wire stringing the pots together.

Absolutely no shielding is used for any of this bussing, or any of the routing from channel related pot to pot. It has hi-Z bass and treble EQ at preamp interstage, which is laid out on the front panel away from the channel blocks, and there's no shielding there either. The wire runs for the group and master faders are shielded, along with the mic and DI inputs and echo return input. The echo return is unbuffered, being a 100K pot feeding another pan array feeding the group inputs through 220K mixing resistors placed at the group input grid.

It does use 12 VDC for the channel tubes, and 6VAC for the output amps. No shielding on any of the 6VAC.

As you might guess from the pic, it was originally set up as a three group mono mixer, with a four channel group and (2) three channel groups. The B+ feeds are still wired per the original three group branching, even though we're mixing in modified stereo now.

Whoever actually built this for Mr. Neve sure knew what they were doing, or got really lucky with the layout. Same for the later modifications.

 

[NewYorkDave]

It is significant, I think, that the mixing resistors were placed right at the input of the booster amps for the critical L/R busses, whereas bus bars were used for the (presumably less critical) echo and monitor busses.

I assume you're familiar with THIS. I can't vouch for its accuracy or origins. Someone else sent me the material and I just assembled it into a PDF for convenience.

 

[emrr]

I didn't know that was out there; I have a paper copy that's a little further down the copier line. It must be for the studio mixer, because the one pictured has a lot of circuit differences, which I was hired to suss out. The only thing I see that is exactly the same is....nothing. Much looks very similar, but there's more differences than similarities. It should be pointed out that the resistor assembly located at the input of the L / R booster amps is a later modification. There's no telling where the original three group mixing resistors lived. The blue pan pots are a major modification, and since the channel faders had to be rewired to feed them its entirely possible that the mixing resistors hung off the channel pots originally. The beige vertical panel on the right side was a later SS oscillator added in the holes where the original third group master lived before being removed in favor of stereo. The panel with the two VU meters was originally one large meter. This is about to turn into a major detour; it is a nice example of the original question though.

 

[Jim Zuehsow]

Dave,
I posted an article from 1954 on building a 4 channel tube mixer. There might be something of interest in it. You'll find it here:

http://groupdiy.twin-x.com/index.php

 

[NewYorkDave]

Thanks, Jim. I've collected a lot of articles/papers related to tube-era mixers, large and small, but that's a new one for me.

 

[Jim Zuehsow]

I'll have to dig up one of the other ones written by George Augsburger where they use a pentode for the mic pre, but what they do is ground the plate and take the output from the screen. It's suppossed to be lower noist that way. A lot of the old techniques seem to have been lost somewhere along the way.
Jim

 

[NewYorkDave]

Are you referring to THIS article, perchance?

Just yesterday I received a couple of DVDs from a friend. I believe a scan of that article may be on one of the DVDs, along with a bunch of other great stuff. But I don't have a DVD reader at home--and due to the weather, it's likely I won't have access to one till Monday at the earliest. Talk about delayed gratification :?

 

[Jim Zuehsow]

I don't recall the title, so I have to page thru a bunch of mags to find it. It's a good excuse to refresh my memory and burn up a couple hours.

 

[NewYorkDave]

I was able to find a way. Look at that thread again :thumb:

It must not be the same article you're remembering, because he really doesn't go into any detail about the preamp circuit.

RE: grounding the plate and using the screen as the "plate", works fine as long as you're careful not to exceed the screen's dissipation rating (which is unlikely in a low-level input stage). The plate becomes a shield enclosing all other electrodes in the tube. I've seen it illustrated in only a couple of places, one of which is RDH4 (in the chapter on mic preamps/mixers, if I recall correctly). It may have not been a widely-used technique even back then.