Adding a DI to a 12AX7 pre amp build

[Jeremy.Starseed]

Hi folks 

In finishing up a summing mixer build, one which uses a Pultec MB-1 stage (a 12AX7 & 12AU7 combination design) it struck me as wise to add the possibility to use the pre-amps separately, and to add a DI input on each of the 2 pre's while I'm at it.

Having had no luck searching for sample schematics to draw inspiration from, it seemed wise to ask you if you have a good solution for this (adding DI, bypassing the 300 Ohm input transformer). 

There are a few points to keep in mind; not letting noise in via this added signal path, and using the switch positions wisely...so enough to provide a good challenge and warrant a question.

I'd be grateful for any suggestions.

What say ye, any good ideas of how to proceed?

Note: the included schematic is representative and may not be identical to actual build

 

[johnheath]

This might be a solution for you... it is used in some various designs.

About values on resistors and caps can be discussed for ages, but what you need is a "switching" phone jack to bypass the input transformer.

Best of luck

/John

 

[ruffrecords]

The Pultec MB-1 is unusual in that the grid of the input tube is not at 0V dc and hence both ends of the input transformer secondary are several volts dc above signal 0V. Therefore, at the very least. you will need to ad a dc blocking capacitor on any DI input. Lot's of designs use the switched jack socket approach to disconnect the input transformer when the DI is plugged in but because of the need for dc blocking this does not work with the MB-1. Bottom line is I think you will probably need to use a relay to undertake the necessary switching.

Cheers

Ian

 

[johnheath]

The Pultec MB-1 is unusual in that the grid of the input tube is not at 0V dc and hence both ends of the input transformer secondary are several volts dc above signal 0V. Therefore, at the very least. you will need to ad a dc blocking capacitor on any DI input. Lot's of designs use the switched jack socket approach to disconnect the input transformer when the DI is plugged in but because of the need for dc blocking this does not work with the MB-1. Bottom line is I think you will probably need to use a relay to undertake the necessary switching.

Cheers

Ian

I must confess that I haven't seen the schematic of the Pultec MB1 :/

Listen to Ian

Regards

/John

 

[abbey road d enfer]

Bottom line is I think you will probably need to use a relay to undertake the necessary switching.

More precisely, an SPDT is required. It could  be an additional switch, it could be a swirching jack (not the usual break contacts but an independant set of contacts) or, as you suggested, a relay driven by one of the break contacts. Final decision is subject to front-panel real-estate issues, ergonomics and skills.

 

[ruffrecords]

An alternative would be to modify the MB-1 input circuit so the transformer is also 0V referenced and the grid dc voltage is blocked by a capacitor (I use a similar scheme in my Eurochannel design). Then both the transformer and the DI are 0V referenced and are dc blocked via the same capacitor. Then you can use the the usual jack socket switch to disconnect the transformer when an instrument is plugged in.

Cheers

Ian

 

[abbey road d enfer]

An alternative would be to modify the MB-1 input circuit so the transformer is also 0V referenced and the grid dc voltage is blocked by a capacitor (I use a similar scheme in my Eurochannel design). Then both the transformer and the DI are 0V referenced and are dc blocked via the same capacitor. Then you can use the the usual jack socket switch to disconnect the transformer when an instrument is plugged in.

Cheers

Ian

I completely agree; that is what a designer should do when designing from scratch with DI included from the start.
Some may argue it wouldn't be an MB1 anymore...

 

[ruffrecords]

I completely agree; that is what a designer should do when designing from scratch with DI included from the start.
Some may argue it wouldn't be an MB1 anymore...

The MB-1 with its NFB down to dc was in part my inspiration for the Eurochannel because with a single zero in the open loop it is unconditionally LF stable. I never did quite understand the reasoning behind the odd biasing arrangement other than it is a possible solution to the dc offset problem and it allows the transformer to be connected directly to the grid which presumably improves bass frequency/phase response and there is also some very low frequency boost due to one of the decoupling caps impedance at very low frequency. Maybe this was all intended to tailor the low end response for the transformers available. I don't know. What I do know is that with exactly the same number of components you can rearrange it so the transformer secondary is 0V referenced. The odd thing is I have seen a similar circuit used in the V76 I think so I am wondering if there is some subtlety I have missed.

Cheers

Ian

 

[abbey road d enfer]

Ian

I completely agree; that is what a designer should do when designing from scratch with DI included from the start.
Some may argue it wouldn't be an MB1 anymore...

The MB-1 with its NFB down to dc was in part my inspiration for the Eurochannel because with a single zero in the open loop it is unconditionally LF stable. I never did quite understand the reasoning behind the odd biasing arrangement other than it is a possible solution to the dc offset problem and it allows the transformer to be connected directly to the grid which presumably improves bass frequency/phase response and there is also some very low frequency boost due to one of the decoupling caps impedance at very low frequency. Maybe this was all intended to tailor the low end response for the transformers available. I don't know. What I do know is that with exactly the same number of components you can rearrange it so the transformer secondary is 0V referenced. The odd thing is I have seen a similar circuit used in the V76 I think so I am wondering if there is some subtlety I have missed.

Cheers

Ian
[/quote] I'm exactly in the same train of thought, and I simulated the MB1 and an edted version with the secondary referenced to ground. There is indeed a change in the LF response.
If we exclude the OT, the "true" MB1 has an LF Roll-ON with about +0.5dB@10Hz, +10dB@1Hz, when the modified circuit has some roll-off, depending on the value of the added input cap and the grid resistor. With 100nF and 1.5Meg, the roll-off is ca. -0.3dB@10Hz, -3dB@1Hz.
Anyway, this is moot, since the resonance of the 2uF cap and the xfmr primary and associated LF roll-off  is dominant there.

 

[ruffrecords]

I'm exactly in the same train of thought, and I simulated the MB1 and an edted version with the secondary referenced to ground. There is indeed a change in the LF response.
If we exclude the OT, the "true" MB1 has an LF Roll-ON with about +0.5dB@10Hz, +10dB@1Hz, when the modified circuit has some roll-off, depending on the value of the added input cap and the grid resistor. With 100nF and 1.5Meg, the roll-off is ca. -0.3dB@10Hz, -3dB@1Hz.
Anyway, this is moot, since the resonance of the 2uF cap and the xfmr primary and associated LF roll-off  is dominant there.

I just found the sim of this I made back in 2008. The bass rise is due to the 0.25uF decoupler and its associated 160K resistor. If you increase the 0.25uF to 10uF the response is flat. I am still no wiser as to why they did this because. as you say, the output cap and transformer primary roll off occurs at a higher frequency.

If you look at the V76 first stage you will see it is virtually identical except they use a 10uF capacitor and a 100K resistor. Given the sim results, the V76 version is likely to be much flatter.

Cheers

ian

 

[CJ]

limited amount of current from V2 of the V76,

some needs to drive V3, some needed to drive the fb ladder,  if fb is applied to 500 ohm cathode resistor, it sucks too much current, thus the 12K in the cathode to form a low loss divider?

 

[Jeremy.Starseed]

OP stepping in to announce that the discussion is now well and truly over my head.

Not to imply I won't enjoy trying to figure out and learn something

I like where it went, actually. I have a V76...love it so much!

 

[abbey road d enfer]

OP stepping in to announce that the discussion is now well and truly over my head.

That's the problem when you post in the Drawing Board! Attracted by bait, wild boffins gather. ;D