Page 1 of 4

Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 11:45 am
by ilya
I'm researching the pure discrete mic preamp circuits and I think I'll settle on either Transamp (by Paul Buff) or Cohen type topology. Looking at the output stage, I often see the floating source output that resembles D. Pontis work (from an old HP journal).
I'm thinking to replace this stage with a single THAT1606 line driver. I know that it needs a low inpedance drive, but as far as I can see, both TA and Cohen have op-amp driven differential outputs, so this should be good enough for THAT1606. Am I correct? Or are there any objections on doing it the way I want?
Thanks!

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 12:02 pm
by JR.
+1 for a THAT corp output driver. Don't discount the merit of using a THAT corp or TI off the shelf mic preamps IC. They don't suck.

Back decades ago I rolled my own mic preamps because we couldn't buy standard chip sets that quiet. Some of the favorite low noise bipolar devices are no longer made. While there are probably suitable alternates.


JR

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:01 pm
by mediatechnology
I think using the 1606 is a great idea.
By using the differential input of the 1606 you can realize a common mode rejection ratio that's typically 60 dB.

TransAmp Documents: viewtopic.php?f=12&t=114&p=894
Pontis Output: viewtopic.php?f=12&t=878&p=10276

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:03 pm
by ilya
Thanks, JR.
Yes, I know about their mic preamp chips. Although, we did a comparison between different preamps in our studio, and the 1510 type preamp was the total looser sonic-wise. I have no idea why that happened. It surely measured just fine on the bench. Maybe because of the lack of "color" (other preamps were API and Neve type, so they had input and output transformers). This is a reason why I don't want to use chips and am looking for a discrete solutions instead.

Wayne, thanks for maintaining the archive of very interestin and rare documents. I'm using it a lot in my hobby musings)
Yes, 1606 came to mind when I saw a ton of trims that need to be adjusted for this and that CMRR. And price wise 1606 is not that expensive. Maybe even cheaper if I factor in a cost of a couple of good multiturn trimmers and precision resistors...

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:06 pm
by JR.
ilya wrote: Tue Dec 29, 2020 1:03 pm Thanks, JR.
Yes, I know about their mic preamp chips. Although, we did a comparison between different preamps in our studio, and the 1510 type preamp was the total looser sonic-wise. I have no idea why that happened. It surely measured just fine on the bench. Maybe because of the lack of "color" (other preamps were API and Neve type, so they had input and output transformers). This is a reason why I don't want to use chips and am looking for a discrete solutions instead.
Sorry I am not a fan of uncontrolled listening tests, while life is pretty much just that.

If you want to have some fun, do some null testing between your favorite preamps and each other or perhaps even the reference input with preamp wiring inverting.

JR

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:08 pm
by mediatechnology
What was the load on the 1510 ICs during the tests?
Was it buffered?

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:21 pm
by ilya
mediatechnology wrote: Tue Dec 29, 2020 1:08 pm What was the load on the 1510 ICs during the tests?
Was it buffered?
It was pretty much the circuit from the datasheet. I think the output was buffered and sent to 1646 driver. This was quite some time ago, so I can't remember the exact details.

John, I know that you're not a fan of uncontrolled listening tests. Neither am I. But that session was the only one we had, and there was no time for setting up a proper ABx test. However, nobody knew the exact inside circuits of the preamps. They were just preamp1, 2 and 3. And all engineers (IIRC 4 people including me) didn't like the 1510 pre.

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:23 pm
by JR.
ilya wrote: Tue Dec 29, 2020 1:21 pm
mediatechnology wrote: Tue Dec 29, 2020 1:08 pm What was the load on the 1510 ICs during the tests?
Was it buffered?
It was pretty much the circuit from the datasheet. I think the output was buffered and sent to 1646 driver. This was quite some time ago, so I can't remember the exact details.

John, I know that you're not a fan of uncontrolled listening tests. Neither am I. But that session was the only one we had, and there was no time for setting up a proper ABx test. However, nobody knew the exact inside circuits of the preamps. They were just preamp1, 2 and 3. And all engineers (IIRC 4 people including me) didn't like the 1510 pre.
If you still have it, maybe you can dig deeper and figure out why... like I said, null testing can be revealing.

I stopped trying to improve circuits that we can now purchase off the shelf with great performance. They would have made my life easier last century.

JR

Re: Cohen/Transamp output stage

Posted: Tue Dec 29, 2020 1:26 pm
by mediatechnology
I ask because the 1510/1512 can't directly drive low impedance loads at elevated level and the datasheet doesn't even show an output build-out resistor.

Re: Cohen/Transamp output stage

Posted: Wed Dec 30, 2020 2:39 pm
by Gold
JR. wrote: Tue Dec 29, 2020 1:06 pm If you want to have some fun, do some null testing between your favorite preamps and each other or perhaps even the reference input with preamp wiring inverting.
A null test won’t tell you anything about the dynamic behavior of the transformer. Without quantifying the dynamic behavior you get an incomplete picture of what the preamp is doing in the wild.