ruffrecords
Well-known member
A Neve tube schematic. Not sure which one.
Cheers
Ian
Cheers
Ian
RCA BC-6B Console Project
- Thread starter tablebeast
- Start date
Help Support GroupDIY Audio Forum:
systemtruck
Well-known member
Thanks! Yeah i finally found that and ultimately figured out that it doesn’t have the transformer topology I’m working with, so I’m figuring to focus on ones that are closest and work from there...
The Collins 212B is pretty close to the setup i have sketched, minus pan, but it also has some interesting factors…
1- It uses the 600:600 T attenuator for volume.
2 - It DOES have the balanced bussing I’d mentioned. The T-attenuator only applies to the positive phase though. I don’t know enough about attentuator circuits to know why this works. But anyways, the balanced signal probably helps greatly with bus noise. That is awesome and very appealing to attempt. BUT… the Collins program amp is balanced the whole way through !! Kinda crazy to see. So the noise won’t be canceled out anywhere within the console…it will only be canceled out whenever said signal is finally made unbalanced again… so…. Back in the computer i guess. But one could just use an unbalanced Program Amp and the transformer at its input will cancel the bus noise.
3 - The summing resistor, 560 R, for any of the 7 channels, is disconnected from the summing bus when disengaged from that bus. It doesn’t hang with the bus, it moves with the preamp/attenuator. There is a second bus, which is Audition. When the Program bus is NOT used from the channel, a 1200R resistor is strapped across +/- signal path in its place. When the Audition bus is NOT used from the channel, a 2700R resistor is strapped across the +/- signal path in its place. I think the center position is just off/mute.
4 - The preamp does not have a volume control. It’s just a preamp. You adjust your level with the T-Attenuator after the transformer output from preamp. Makes total sense when you don’t care about fine tuning the way your preamp drives.
I’m going to think this over some more. Maybe my block diagram of values does work ok enough with gain losses and crosstalk and loading? Not sure yet. If anybody has thoughts about the values i chose, i would love to hear. But I’m also going to consider the Collins balanced summing approach. If there is a way to slap some panning into this Collins summing design, maybe that is the pot of gold at the end of the rainbow because that would probably sound so good. I would probably go ahead and use an unbalanced Program Amp so that the signal gets cleaned right away.
ruffrecords
Well-known member
One thing to bear in mind is that balanced busing complicates the pan control. A regular unbalanced pan needs a dual gang pot. Strictly speaking, a balanced pan needs a 4 gang. There is little if any benefit in balanced busing in a small mixer.
Cheers
Ian
Cheers
Ian
MidnightArrakis
Well-known member
This mixer looks to me to be like an old-style -- radio broadcast -- mixer. Maybe for "Radio Caroline"??? HA!!!.....
/
ruffrecords
Well-known member
Strangely enough, virtually all the UK pirate radio ships used American AM broadcast consoles simply because at that time there was no local radio at all in the UK, just the three national BBC stations.
But I bet Rupert was aware of them and built his along similar lines.
Cheers
Ian
systemtruck
Well-known member
Alrighty, I’ll keep it simple (unbalanced).
I just have to decide if my drawn up block design of fading, panning, and passive mixing will function desirably enough.
Or, if I need to try something completely different for levels and panning, and then do the passive summing somehow isolated from that.
I've completely rebuilt 2 of those, modifying both for LCR and direct outs, one I built new outboard power supplies.
I like the performance of the original design overall and the clients are very happy with the results.
I started a PCB design for the cards...that was for quick service and long term as some of the cards did have some trace damage. Gotta go back in the archives to see if I finished that PCB design........good luck!
I like the performance of the original design overall and the clients are very happy with the results.
I started a PCB design for the cards...that was for quick service and long term as some of the cards did have some trace damage. Gotta go back in the archives to see if I finished that PCB design........good luck!
systemtruck
Well-known member
Here’s a new attempt at a passive circuit, with a comparison to the former attempt.
Mind you, this is 1 of 8 channels.
The two changes are the use of a 600 ohm Bridged T attenuator ( inspired by the Collins 212B @emrr ) instead of a potentiometer. As you can see from the second (old) example, the 5K potentiometer introduces a whole ton of variance of the overall load on the preamp output transformer. Anywhere from 5000 R, to 694 R. Even if i add something like a 1K or 1.2K slug to the transformer output like i did in the newer attempt, the load still varies from mid 400’s to around 1500’s. So it’s a lot of load variation on the preamp, not something i want to have part of the sound.
The Bridged T attenuator contributes 600 ohms to the whole passive summing/panning circuit but never changes, so it is much easier to tame. Adding a 1K slug to the transformer secondary gives us a new total variation of 584 R to 665 R.
None of this is refined yet, just looking at approaches. But i think the stepped Bridge T attenuator is a nice solution because it eliminates most of the load variation.
I think i would build a 20 step 1dB increment rotary switch along with a 20dB + 40dB similar Bridged T pad below it (if load permits), so that’s 60dB of attenuation. If that’s too nuts, i could do a 20 increment 2dB rotary and no pad and be happy with 40dB of mixing adjustment.
It’s a bunch of work to build EIGHT of these, but i have already wanted to use them anyways for recall purposes. It would be killer to also have incremental adjustment of panning and preamp gain, but that is now 24 stepped knobs haha. Those aren’t necessary for loading design, so won’t make the cut.
EDIT: i think I have the math wrong with the bridged attenuator.. i think i need to need to consider its load directly down to ground instead of its load across it and redo all my math. So this means I’ll probably need a much higher attenuator value. But the point is the same.. this is a constant impedance and eliminates the potentiometer load variations.
Mind you, this is 1 of 8 channels.
The two changes are the use of a 600 ohm Bridged T attenuator ( inspired by the Collins 212B @emrr ) instead of a potentiometer. As you can see from the second (old) example, the 5K potentiometer introduces a whole ton of variance of the overall load on the preamp output transformer. Anywhere from 5000 R, to 694 R. Even if i add something like a 1K or 1.2K slug to the transformer output like i did in the newer attempt, the load still varies from mid 400’s to around 1500’s. So it’s a lot of load variation on the preamp, not something i want to have part of the sound.
The Bridged T attenuator contributes 600 ohms to the whole passive summing/panning circuit but never changes, so it is much easier to tame. Adding a 1K slug to the transformer secondary gives us a new total variation of 584 R to 665 R.
None of this is refined yet, just looking at approaches. But i think the stepped Bridge T attenuator is a nice solution because it eliminates most of the load variation.
I think i would build a 20 step 1dB increment rotary switch along with a 20dB + 40dB similar Bridged T pad below it (if load permits), so that’s 60dB of attenuation. If that’s too nuts, i could do a 20 increment 2dB rotary and no pad and be happy with 40dB of mixing adjustment.
It’s a bunch of work to build EIGHT of these, but i have already wanted to use them anyways for recall purposes. It would be killer to also have incremental adjustment of panning and preamp gain, but that is now 24 stepped knobs haha. Those aren’t necessary for loading design, so won’t make the cut.
EDIT: i think I have the math wrong with the bridged attenuator.. i think i need to need to consider its load directly down to ground instead of its load across it and redo all my math. So this means I’ll probably need a much higher attenuator value. But the point is the same.. this is a constant impedance and eliminates the potentiometer load variations.
ruffrecords
Well-known member
Both of your schematics are a bit of a compromise but that is normal with all engineering.
A bridged T fader is meant to be driven by a 600 ohm source and loaded by one so strictly speaking the transformer secondary should include a series build out resistor to achieve this. One reason this sort of technique was later abandoned is that,with a 600 ohm source and an identical load, you lose 6dB of level in the pursuit of constant impedance. Your scheme avoids this and you end up with a variable load. A good transformer should not have a problem with this. You can probably dispense with the 1K slug.
The variable load of the pan pot in the regular pot version means the pot law changes depending on the pan setting. It tends to mean you get more attenuation than you expect as you move down from fully up. In practice it is not really a problem because most people set the level by ear. Just something to be aware of. To minimise this you can increase the ratio of the fader to pan pot. At the moment you are at 2:1 (10K:5K). I tend to use 5:1 with a 10K fader and 50K pan pots and 47K bus feed resistors.. You could achieve the same or better simply by lowering the value of the fader pot.
I think you are nearly there.
Cheers
Ian
A bridged T fader is meant to be driven by a 600 ohm source and loaded by one so strictly speaking the transformer secondary should include a series build out resistor to achieve this. One reason this sort of technique was later abandoned is that,with a 600 ohm source and an identical load, you lose 6dB of level in the pursuit of constant impedance. Your scheme avoids this and you end up with a variable load. A good transformer should not have a problem with this. You can probably dispense with the 1K slug.
The variable load of the pan pot in the regular pot version means the pot law changes depending on the pan setting. It tends to mean you get more attenuation than you expect as you move down from fully up. In practice it is not really a problem because most people set the level by ear. Just something to be aware of. To minimise this you can increase the ratio of the fader to pan pot. At the moment you are at 2:1 (10K:5K). I tend to use 5:1 with a 10K fader and 50K pan pots and 47K bus feed resistors.. You could achieve the same or better simply by lowering the value of the fader pot.
I think you are nearly there.
Cheers
Ian
