Mic Pre 12AY7 + 12AU7

[Emmathom]

This tube pre is concerned by this new thread (sound samples included) :
https://groupdiy.com/threads/mics-blind-test.87738/page-4

 

[rock soderstrom]

John Broskie has done some interesting designs with MOSFET outputs for tube circuits.

Do you have a specific circuit in mind? I always get a bit lost in the zillion circuits that John Broskie offers.

 

[rock soderstrom]

Two further questions about the pre amp circuit:

What is the function of the two 9.1V Zener diodes in the output?

Is the output impedance balanced? How high is the output impedance?

 

[Emmathom]

Do you have a specific circuit in mind? I always get a bit lost in the zillion circuits that John Broskie offers.

Must ask @k brown...

 

[Emmathom]

Two further questions about the pre amp circuit:

What is the function of the two 9.1V Zener diodes in the output?

Is the output impedance balanced? How high is the output impedance?

9,1v*2 protect the output by setting the maximum out voltage at 18v for the following stage (sound card or mixing desk, etc)
My audio card is 20dbu max = 22v ptp max (I took some headroom with 18v)
https://sengpielaudio.com/calculator-db-volt.htm

The output is unbalanced (but there's more than enough output level and cables will be 50cm max. so...)
The output impedance : @thor.zmt who designed most of this circuit should be able to answer...

 

[thor.zmt]

What is the function of the two 9.1V Zener diodes in the output?

It is a soft clipper ahead of the AD input. It will start crushing peaks fairly hard at around 9.6V peak, or around +18dBu.

The advantage of this circuit is that there are zero timeconstants, the way tops are crushed resemble more the way tape saturates.

Of course, to simulate tape saturation properly would need a pre-empasis with boosted LF and HF before the zener limiter and de-emphasis after. So it's not a tapd saturation emulator, just a soft clipper.

Is the output impedance balanced?

Of course.

How high is the output impedance?

The design was specific to interface with the Soundcard @Emmathom uses, which is 10k Input impedance and +22dBu for 0dBFS.

There is a 12dB output pad (which in effect replaces the commonly used output transformer) which deliberately uses the 10kOhm input impedance of the following ADC input as part of the Pad.

The this works is that the Tube Preamp delivers appx +31dBu into the 12dB pad for the onset of the zener limiter.

Instead the use of a 10k:600R output transformer would have been suggested, this is in effect a "workaround" that keeps the load impedance on the circuit high while implementing a 12dB attenuation.

It's not ideal, but the best workaround for not having an output transformer I can come up with.

So normally the 12dB pad is engaged, the output impedance is intentionally 30k to form a 12dB pad with the 10k input impedance of the ADC, getting the relative levels as using a 10k:600R output transformer.

Thor

 

[Emmathom]

Of course.

Really ? I thought it was unbalanced...

 

[thor.zmt]

Really ? I thought it was unbalanced...

Balance referrers fundamentally to "balanced impedance".




For most of the Audio range the output impedance of the Tube Mic Pre is balanced.

What the circuit does not do is to provide a balanced signal.



A 4:1 output transformer should be added, where size/weight/budget allows and the pad dispensed with.



Thor

 

[Emmathom]

Ok unbalanced signal output but impedance balanced...

 

[rock soderstrom]

It is a soft clipper ahead of the AD input. It will start crushing peaks fairly hard at around 9.6V peak, or around +18dBu.

The advantage of this circuit is that there are zero timeconstants, the way tops are crushed resemble more the way tape saturates.

Of course, to simulate tape saturation properly would need a pre-empasis with boosted LF and HF before the zener limiter and de-emphasis after. So it's not a tapd saturation emulator, just a soft clipper.

I had already thought that, thanks for the confirmation. I will try that out!

Of course.

Ditto, thanks for the confirmation, again.

The design was specific to interface with the Soundcard @Emmathom uses, which is 10k Input impedance and +22dBu for 0dBFS.

Same here, most of the time it goes straight into the audio interface, but sometimes a low impedance vintage compressor needs to be fed, so I'll add an optional 10k to 600R OPT. But most of the time I prefer the direct out, I find that OPT, especially the cheaper ones, don't necessarily add really good things to the signal, the really good ones are almost inaudible.

There is a 12dB output pad (which in effect replaces the commonly used output transformer) which deliberately uses the 10kOhm input impedance of the following ADC input as part of the Pad.

The this works is that the Tube Preamp delivers appx +31dBu into the 12dB pad for the onset of the zener limiter.

Instead the use of a 10k:600R output transformer would have been suggested, this is in effect a "workaround" that keeps the load impedance on the circuit high while implementing a 12dB attenuation.

It's not ideal, but the best workaround for not having an output transformer I can come up with.

So normally the 12dB pad is engaged, the output impedance is intentionally 30k to form a 12dB pad with the 10k input impedance of the ADC, getting the relative levels as using a 10k:600R output transformer.

I'm still chewing on that a bit, but thank you in the meantime.

Generally a clever design as far as I can see so far. I will borrow the output solution for my new iteration of a another moded GATES SA70 with 6J7/6C5 toobs.

I have very little space in the box and have therefore thrown out the original planned (additional) tube solution for the output.

I'm curious to see how that turns out...blasphemy, my first hybrid. I'll burn in tube hell for it.

 

[k brown]

Do you have a specific circuit in mind? I always get a bit lost in the zillion circuits that John Broskie offers.

One example:
https://www.tubecad.com/2007/01/blog0094.htm

And:
https://wtfamps.com/2018/02/07/class-a-fet-power-buffers/

 

[thor.zmt]

I'm still chewing on that a bit, but thank you in the meantime.

Simple, the audio interface has 10k balanced input impedance on line ins, with, as mentioned nominally +22dBu for 0dBFS.

In my case this 10k resistor is used together with 30kOhm series resistance to make a "pad" to lower the output from the tube preamp.

The reason is simple, this tube preamp is meant as "electronic crayons", that is strict neutrality is not always desired.

If we directly drive the Audio interface we need to lower the gain control and input to the 12AU7 by 12dB (and thus lower H2 by 12dB and H3 by by 24dB) for given -xxdBFS signal.

In the "cardboard cutout" variant of our "electronic crayon" (per Eddie Chilettie the Altec 1566 Tube Pre) we would have an output transformer stepping down the signal 12dB.

A 12AU7 stage with an unbypassed cathode resistor is pretty "clean" to start with, so to get some colour in this pre we can first of all bypass the input cathode resistor and we can use the output pad. So there is a fair bit of adjustment in the pre to get different levels of "colour".

Perhaps, one day @Emmathom could make a pair of sound sample's, one with the "clean" setup (input cathode no bypass, direct out without pad and gain suitably adjusted) and a second in "colour" mode, input cathode bypassed and output pad.

The follower as is draws some 5mA or so, suggesting we should limit the current swing to ~ 2mA P-P which is ~+20dBu into 10kOhm. Of course, our follower can swing more current, but it starts to distort progressively more. Which is a non-issue as we have our limiter which will distort a lot more.

With pad the load impedance increases and we can handle more level for the same distortion level from the follower. So we keep the tubes harmonics dominant.

Generally a clever design as far as I can see so far. I will borrow the output solution for my new iteration of a another moded GATES SA70 with 6J7/6C5 toobs.

Go for it, depending on the output impedance of the preceding stage watch how much capacitance the buffer FET has.

I'm curious to see how that turns out...blasphemy, my first hybrid. I'll burn in tube hell for it.

Pick the right FET, no problems.

It's not really blasphemy, just smart use of new technologies.

Otherwise you could never use anything except tube rectifiers for tube circuits. And maybe using indirectly heated tubes will send you to DH Tube hell?

Thor

 

[Emmathom]

Perhaps, one day @Emmathom could make a pair of sound sample's, one with the "clean" setup (input cathode no bypass, direct out without pad and gain suitably adjusted) and a second in "colour" mode, input cathode bypassed and output pad.

For sure will do. no pb.
But my first tests showed that -25dB at max. position is too high value. I think should go for -18db so I must do the mod. before the test.
my Son came in to visit me until the 5th of July and I'll record these samples after that...

 

[Emmathom]

Comparison K.byp & unbyp.

As promised, I recorded acoustic guitar + voiceover thru the same mic using a splitter in my Quad tube pre.
- one is K.unbyp. (without pad)
- the other is K.byp (with -25dB pad)
Attached is a pict of how the pot levels are set to match same recording level : left is K.unbyp. (no pad) and right is K.byp (-25dB pad).

6AK5 K bypassed

6AK5 K unbypassed

 

[thor.zmt]

Comparison K.byp & unbyp.

As promised, I recorded acoustic guitar + voiceover thru the same mic using a splitter in my Quad tube pre.
- one is K.unbyp. (without pad)
- the other is K.byp (with -25dB pad)
Attached is a pict of how the pot levels are set to match same recording level : left is K.unbyp. (no pad) and right is K.byp (-25dB pad).

6AK5 K bypassed

6AK5 K unbypassed

Thanks, it will be interesting for anyone planning to build something similar...

Also, who can hear the difference and describe it?

Thor

 

[Emmathom]

Personally, and by really listening, I perceive a very slight saturation on the K. byp track.
So a pleasant touch of saturation (but very light)