Would this "half circuit" work?

[buildafriend]

What you see here is half of the 1st pre amp from the Rob C Akai M-8/M-7/Roberts mod. This circuit is typically followed by a 12AX7.

It's a single EF-86.

I was thinking that this might make a decent low gain mic pre all by it's self.. I was about to just try it out since I have all of the parts just sitting in front of me. 

[abechap024]

I have 4 of these, the Akia mods that is. Should work like you have it.  I'm also curious to learn more about them, but anyone see also any easy improvements that could be made to this circuit?

[sahib]

It follows the standard circuit that Mullard had. I had 8 modules. May be I should have kept them.

Out of topic. I kept the knobs but sold the VUs to a gentleman in Germany. He had the most amazing Akai reel to reel collection I have ever seen. His workshop/office was crammed with M series and of course all working. That's what he specialised on. Only Akai.

[Henke]

First of all, you don't wanna leave that 0.15 cap floating, you should tie it down with a resistor to ground before the output jack.
This circuit has very limited drive capability and requires a very high load impedance for decent performance.

[ruffrecords]

The one disadvantage of this circuit is its relatively high output impedance - about 250K. SO you need to feed it into a high impedance such as a HI-Z DI input.

Cheers

Ian

[buildafriend]

First of all, you don't wanna leave that 0.15 cap floating, you should tie it down with a resistor to ground before the output jack.
This circuit has very limited drive capability and requires a very high load impedance for decent performance.

Ground the R7 side or the output side of the 0.15uF cap?

[buildafriend]

The one disadvantage of this circuit is its relatively high output impedance - about 250K. SO you need to feed it into a high impedance such as a HI-Z DI input.

Cheers

Ian

So it's really best when paired with the gain stage that is usually found next to it? (12AX7)

[Henke]

First of all, you don't wanna leave that 0.15 cap floating, you should tie it down with a resistor to ground before the output jack.
This circuit has very limited drive capability and requires a very high load impedance for decent performance.

Ground the R7 side or the output side of the 0.15uF cap?

The output side! At least 1Meg and make sure you feed it into a Hi-Z DI-input, a typical line input will be way too heavy a load.
As ruffrecords said the output impedance is impractically high, so consider adding a cathode follower at least. And on another note i'd suggest moving the level control to the output side of the 6267 for better noise performance.

[abechap024]

Sweet. I am getting excited about this.  I have my 4 modules and have plans to rack them up all together in a different chassis.

Curious about the output impedance, would a step down transformer help with this (say 4:1)? Obvious answer is yes, but then again I'm still learning about toobs!

Also if I did use the 12ax7 as a cathode follower, could a use one side of the 12ax7 for one channel CF and another side for another channels CF or would the cross-talk be a possible issue?

Interesting about putting the volume control after the EF86, would the 500k work there or something higher like 1M? I like the idea, do you think a pad should be added (would the ef86 stage possibly overload or is the 250V rail large enough?)

Thanks for being patient with these rather primitive questions. It does help a lot!
Abe

[ruffrecords]

The one disadvantage of this circuit is its relatively high output impedance - about 250K. SO you need to feed it into a high impedance such as a HI-Z DI input.

Cheers

Ian

So it's really best when paired with the gain stage that is usually found next to it? (12AX7)

Yes, or a cathode follower, for example,  which would provide a low impedance output.

Cheers

Ian

[ruffrecords]

Sweet. I am getting excited about this.  I have my 4 modules and have plans to rack them up all together in a different chassis.

Curious about the output impedance, would a step down transformer help with this (say 4:1)? Obvious answer is yes, but then again I'm still learning about toobs!

But probably not enough. A 4:1 transformer will reduce the output impedance by 16 times (assuming it has enough primary inductance  - but that is a different discussion) to around 15K which is still not really low enough.

Also if I did use the 12ax7 as a cathode follower, could a use one side of the 12ax7 for one channel CF and another side for another channels CF or would the cross-talk be a possible issue?

Crosstalk would not be an issue but a 12AX7 cathode follower would have limited drive capability even though its output impedance would be nice and low.

Interesting about putting the volume control after the EF86, would the 500k work there or something higher like 1M? I like the idea, do you think a pad should be added (would the ef86 stage possibly overload or is the 250V rail large enough?)

Depends on the ratio of the M67 transformer. If this is 1:10 then the total gain will probably be in excess of 60dB. You might think that would overload easily but an EF86 can probably swing near to 60V rms at is plate - that's +37dBu which means the input needs to reach -23dBu to overload it. You can always at a switched 20dB pad on the transformer primary for hot sources. A volume control at the output would definitely improve the noise performance

Cheers

Ian

[buildafriend]

Sweet. I am getting excited about this.  I have my 4 modules and have plans to rack them up all together in a different chassis.

Curious about the output impedance, would a step down transformer help with this (say 4:1)? Obvious answer is yes, but then again I'm still learning about toobs!

But probably not enough. A 4:1 transformer will reduce the output impedance by 16 times (assuming it has enough primary inductance  - but that is a different discussion) to around 15K which is still not really low enough.

I might just give it a shot with a 4:1. Any recommendations as to which one might be a smart selection?

I'll also be adding an output volume pot. 100K Log/audio taper.

If this "fails" then I'll throw the cathode follower on a second turret board. I have 2 turret boards that were going to become part of my BA-2...

[abechap024]

Crosstalk would not be an issue but a 12AX7 cathode follower would have limited drive capability even though its output impedance would be nice and low.

Interesting. Thanks for shedding some light on this. These pres sound surprisingly good as-is and probably sound better a lot better once the are interfaced correctly!

So, do you think there will be enough drive to drive a 10k type line input with the 12ax7 CF?

Thanks again!

PS see attached Schem, planning on attaching the CF next....
Best,
Abe

[abechap024]

Heres an attempt at a CF circuit, does this look like it will work? Any ways to improve upon it? I'm kinda shooting in the dark!


Thanks.
<pic removed to avoid confusion>

[ruffrecords]

Heres an attempt at a CF circuit, does this look like it will work? Any ways to improve upon it? I'm kinda shooting in the dark!


Thanks.

Not as it stands. The circuit will have nearly the entire HT supply volts across the tube and about a couple of volts or so across the cathode resistor. Tube dissipation will therefore be high and output swing severely limited.

You should aim for about a quarter to half the supply volts at the cathode to maximise the available output swing. You will be lucky to get a quiescent current of 2mA in a 12AX7 so the cathode resistor needs to be around 62K in total. At 2mA and 150V on the plate the 12AX7 needs a negative bias of a round 0.5 volts so a cathode bias resistor of 1K would be about right. Even so, because of the low quiescent current, the drive capability into a 10K load is limited.

What you really need is a tube running at a much higher quiescent current. The good old 12AU7 would be a much better bet. You could run this as a CF with a quiescent current of 10mA and a grid bias of around -3V which would give you a decent drive capability into a 10K load.

Cheers

Ian

[Henke]

Heres an attempt at a CF circuit, does this look like it will work? Any ways to improve upon it? I'm kinda shooting in the dark!


Thanks.

As ruffrecords said, a 12AU7 / ECC82 is a much better choice for this task. With your 280V B+, Rk could be around 12k-15k (rated for at least 3W) in series with a bias resistor around 270-330 ohms under the cathode. Put a large value resistor from between these resistors to the grid and cap couple the wiper-grid connection. The 12AU7 has a cathode-heater voltage maximum of 180V, so heater elevation won't be necessary.

[abechap024]

I might just give it a shot with a 4:1. Any recommendations as to which one might be a smart selection?

An Edcor 4:1 is an obvious choice...I got a few kicking around I'm going to try it and see how it works...

If this "fails" then I'll throw the cathode follower on a second turret board. I have 2 turret boards that were going to become part of my BA-2.
..

Right on, thats what I'm going to do...seems like a better option, because these aren't very high gain to begin with. I think they have plenty of gain now, but the 4:1 might make it a little too quite for some sources.

As ruffrecords said, a 12AU7 / ECC82 is a much better choice for this task. With your 280V B+, Rk could be around 12k-15k (rated for at least 3W) in series with a bias resistor around 270-330 ohms under the cathode. Put a large value resistor from between these resistors to the grid and cap couple the wiper-grid connection. The 12AU7 has a cathode-heater voltage maximum of 180V, so heater elevation won't be necessary.

Thanks Guys! OK i think i got it, though i could've very easiily goofed it up. does look about right? Thanks!


<pic removed because incorrect, check later post for more correct circuit>

[Henke]

Thanks Guys! OK i think i got it, though i could've very easiily goofed it up. does look about right? Thanks!

Getting there, although you should get rid of that 250k anode resistor! And your 100 ohm grid-stopper should be soldered right at the grid pin, sorry that i left it out in my previous explanation. 1uF for the wiper cap may be somewhat overkill as the CF input impedance is *very* high due to the bootstrapping effect in this bias arrangement.

[ruffrecords]

Can you tell me what you use for drawing schematics? They are very nice looking. Remind me of the ones you used to get in Wireless World many years ago.

Cheers

Ian

[Henke]

Oh and make sure to include an input termination resistor across the secondary of your input transformer in addition to the zobel network. The transformer will behave better and the EF86 stage will be less prone to oscillate when your mikes are disconnected.