Custom console rebuild

Dmichel123 said:

Mostly Blues, pretty much all early roots stuff. We've had a couple cumbia bands in the studio and we do some jazz and R&B. We're in Hayward, CA. Bigtone Records. We recorded a couple tracks on Kim Wilson's (of The Fabulous Thunderbirds) newest release, "Blues and Boogie".

Click to expand...

Cool. I have actually heard of you. I see you are already lucky enough to own a nice RCA mixer.

Cheers

Ian

The owner is Jon Atkinson, an incredibly talented guy. I do most of the repairs, but I also engineer and mix a lot of projects. I engineered and mixed the band live on 'Worried Life Blues' on Wilson's new disc.  That was about 2 years ago, before we got the RCA BC-3C. Our mixer at the time was a Altec 342B with the output tubes removed and the phase inverter modified to a cathode follower. I'm pretty sure the B+ voltage was way too high for the 12AX7s, but we had no issues using it like that for over a year!

UPDATE: Console is working great. Thanks for the tips, guys.

: So I realized that the mixer tube in the console is a 12AX7 and not a 12AU7.  Changing that tube in the simulation netted me ~14dB of gain. However, the simulation is showing a much earlier high frequency rolloff than I was seeing with the 12AU7. Is this a function of the 12AX7 SPICE model being "better", more accurately modeling miller capacitance? In the SPICE circuit, I can change the mixer tube bypass cap to a .0047u and extend the HF response at the expense of ~7dB(moves the -3dB point to 18kHz from 6.5kHz) We haven't recorded with the console yet or compared it to our existing console/pres, only hooked it up to an interface and spoke into a mic. If we find that it feels restricted in the upper frequencies, would this smaller bypass cap be an acceptable "fix"?  I want to say that our RCA BC-3C has some HF shaping going on in or around the "program" amp, IIRC.

I am also considering "upgrading" the tubes in the output section. I have some 12BH7 and ECC88 laying around, along with many other tubes.

Any input is appreciated.

Thanks,
Danny

The earlier roll of is due to the Miller effect which now has a  lower the turnover frequency due to the higher gain of the valve. The reason it is so low is due to the 250K bus feed resistors which form an RC filter at the grid with the Miller capacitance. Try changing the resistors to 25K in your simulation and you should see a dramatic improvement.

Cheers

Ian

ruffrecords said:

The earlier roll of is due to the Miller effect which now has a  lower the turnover frequency due to the higher gain of the valve. The reason it is so low is due to the 250K bus feed resistors which form an RC filter at the grid with the Miller capacitance. Try changing the resistors to 25K in your simulation and you should see a dramatic improvement.

Cheers

Ian

Click to expand...

Ian,

Simulation results:

12AX7 250k mix Rs, 47uf Ck:          74dB gain,    HF -3dB point 6.5khz, LF -3dB point 14hz
12AX7, 25k mix Rs, 47uf Ck:            71.4dB gain, HF -3dB point 8.5khz, LF -3dB point 32hz
12AX7, 250k mix Rs,  .0047u Ck:  67.4dB gain, HF -3dB point 19khz,  LF -3dB point 13hz
12AU7, 250k mix Rs,  47uf Ck:        60dB gain,    HF -3dB point 70khz,  LF -3dB point 13hz

These simulations were done with one channel cranked and all the rest turned down. I don't feel like the 25k mix resistors gain me enough HF for the additional load they place on the mic pres. Simply changing the mix tube to 12AU7 looks like the best and easiest way to get more HF... Unless that tube model is not modeling Miller capacitance at all.

Thanks in advance
Danny

Those results are strange. The improvement idue to the 12AU7 is simply because it has a lower gain. Its mu is 17 compared with 100 for the 12AX7 which is a ratio of about 15dB and as expected the 12AU7 gain is about this amount lower than the 12AX7.

The fact that the reduction of the 250K resistors to 25K makes only a small difference could because the source resistance is the 250K  resistor plus other impedances that are effectively in series with it, namely the 100K anode resistor of the input stage so you still have 125K of series resistance. However, if you work out the 3dB point with 125K and typical Miller capacitances you get a value around 12KHz which is twice what you observe. I suspect this is because the two anodes are connected together which doubles the Miller capacitance.  Can you try a sim with the FX anode disconnected.

Cheers

Ian

Ian,

With FX rtn triode removed, -3dB point is 12.8khz.

Dmichel123 said:

Ian,

With FX rtn triode removed, -3dB point is 12.8khz.

Click to expand...

OK so the doubling of plate capacitance was making a difference. Makes more sense now.

Even so, I think you are right to stay with the 12AU7 for that stage.

Cheers

Ian

We are going to use the console as it sits for now (with 12AU7 mix tube), and I am going to start drawing up the circuit for the next round of mods. I should have a new schematic and plenty of new questions soon!

Thanks for all your help
Danny

Here is a preliminary schematic of the next revision of the mic pres. This mod will need three 12BH7 (or 12AU7 or 6CG7 or ECC88?) and leave me with an unused triode to play with in the mixing or output stages.

I would go for 12AU7 or ECC88 because they both have a lower plate resistance than the 12BH7 and the 6CG7. ECC88 has lower distortion than th 12AU7 so that would be my first choice,

Cheers

Ian

Ian,

I found some OTs that I think may work for the preamps. They are UTC/TRW W-787s.

7800/4800:600/150
2W 40mA DC primaries
Listed frequency response is
300-10k, I'm assuming they will do much better than 300hz when used well below 2W, if I've learned anything from CJs tear down posts.

New schematic.

Output tube changed to ECC88
Output transformer ratio changed (5.66:1 ratio,  4800:150 on UTC W-787)
ADDED EQ! Modified UA 610 feedback scheme, using inductors I have on hand.

Once again, any input is greatly appreciated!
Danny

Unfortunately the 6DJ8 has a maximum working anode voltage of only 130V. Even the 6922/E88CC has an absolute working maximum of only 250V which probably explains why you never see single ended transformer output stages using the 6DJ8.

For the 12AU7 the dc anode voltage can be as high as 300V

For the 12BH7 it is 450V

For the 6CG7 it is 330V

Cheers

Ian

The level control cannot be in its present position because it is inside the feedback loop.

Cheers

Ian

Ian,
I've reverted back to 12BH7 for now. I was only using the 6DJ8 SPICE model because I don't have an ECC88 model. My plan is to put Daven or similar attenuators after each OT, so losing the gain knob shouldn't hurt too much. Maybe I'll do your old Neve trick and use 1k pots after the OTs.

I've done some more thinking and schematic drawing on this console. What concerns should I have about Mic pres with ~40dB fixed gain? The circuit that's in them now has about the same gain, I'm thinking it won't be an issue.

Also, I'm trying to think of the best way to implement inserts and direct outs, while keeping a balanced mix bus. I think separate, switching TRS sends and returns would be the "best" way to do it, any tips or opinions on this are welcome. If I'm imagining it right, on top of having the ability to insert effects into channels separately, I would be able to record multitrack to tape or digital from the send/direct out and then mixdown from the tape machine or interface into the return jacks (which would feed directly into the balanced mix bus). I plan on adding 3 or 4 additional line level inputs to the mix bus(new AUX/FX return will need to be 600ohm balanced, will be no mixer tube in next program amp ).

I was looking at a UA 1016 circuit for the line/program amp. Fully balanced from front to back, looks good! THEN, I started toying with the idea of adding compression to the line amp. I looked at a few circuits and soon found that the UA 175 is basically a UA 1016 with a vari-mu stage and sidechain tacked on. 

Any input is greatly appreciated.
Thanks
Danny

More info/questions:

Anybody have experience or opinions on using these UTC W-787s as OTs with 12BH7 or 12AU7? 

https://www.ebay.com/itm/UTC-W-787-Audio-Tube-Plate-to-Line-Transformer-Pri-7600-4800-Sec-600-ohm-Split-/183002880652?

I have these transformers that I hope will work for the line/program amp input and output:

UTC input from RCA BTE-10B exciter, T101 on schematic. Is this 150/150:15k/15k?
http://www.bobgroome.us/Schematics/RCA%20BTE-10B%20Exciter.pdf

Stancor PCO-150 output. 10k:600/150/16/8/4 . Same as used in Fairchild 660 sidechain amp.

I will need to build a new power supply before the next set of mods. I want to have a decent idea of how many tubes I want to add and their heater current requirements, so I can build a PSU with a little current to spare in case of future mods, etc. The new PSU will be external. Moving the PSU out of the console chassis will give me a lot of room for the new line amp and should help with noise(console is already pretty quiet!).

I should have a few new schematics to post soon.

Thanks
Danny

We're expecting some vintage output transformers in the mail today. I'm gonna build a channel out with added 12AU7 triode+OT+EQ and test. I'll report back with my impressions/findings.