Official C12 Clone - Build and Support Thread

[chunger]

Both PSU types will continue to be available.

I will make head-to-head recordings to validate the new PSU.  FWIW, there are differences between our PSU and the original circuit.  Ours is full wave rectified.  The original was half wave rectified.  Reasonable improvements were made but the overall scheme retained.

The reason we are moving forward with the fully regulated PSU is about 90% of the build problems encountered so far relate to the passive B+ section of the PSU and, there should be much more adjustment range in the new setup for experimenting with different tubes, etc.  If Matador was not confident that he could design a system that is as clean or cleaner than the current passive B+, I don't think he would move forward with the design.

If the new fully regulated PSU design proves on the bench and in the studio to perform as good or better than the passive setup, I will encourage less experienced DIY people to build the new one to reduce the chance of build failure.  Again, both will continue to be available.  Nothing is being phased out.

 

[Melodeath00]

Both PSU types will continue to be available.

I will make head-to-head recordings to validate the new PSU.  FWIW, there are differences between our PSU and the original circuit.  Ours is full wave rectified.  The original was half wave rectified.  Reasonable improvements were made but the overall scheme retained.

The reason we are moving forward with the fully regulated PSU is about 90% of the build problems encountered so far relate to the passive B+ section of the PSU and, there should be much more adjustment range in the new setup for experimenting with different tubes, etc.  If Matador was not confident that he could design a system that is as clean or cleaner than the current passive B+, I don't think he would move forward with the design.

If the new fully regulated PSU design proves on the bench and in the studio to perform as good or better than the passive setup, I will encourage less experienced DIY people to build the new one to reduce the chance of build failure.  Again, both will continue to be available.  Nothing is being phased out.

I appreciate the reply, chunger. Thanks! I guess we are using the term "cleaner" as in a general improvement to the circuit. I have no idea how "clean" a vintage PSU would be, but I imagine we are capable of building cleaner PSUs these days.

What exactly does "full wave rectified" mean compared to "half wave rectified," and what effect does it have?

 

[Matador]

The new fully-regulated PSU design is quite a bit more flexible than the previous design:

1) It is agnostic of tube type.  Once set for 120V, you can run any tube type and the supply will adjust itself to maintain 120V.  Currently different tubes of the same type require adjustment of the B+ trimmer, and switching to a different tube type (e.g. 12AT7 or others) require removing and changing filter resistors
2) B+ supply is adjustable from 120V up to 245V
3) It provides true fixed bias, or a low impedance negative supply
4) It can accommodate both "positive" and "negative" heater supplies, all from a fixed regulator type (to switch requires only a jumper be moved)
5) Isolated grounds - there are three supplies that each have their own "ground" plane (think B-, HEATER-, and BIAS+).  This means you can float the heater supply up, etc.  Or you can strap all three together, it's your choice.
6) Optimized BOM with fewer component values that need to be ordered

The B+ supply is good up to about 10mA, and the heater up to 12V, so it's even possible to use this with quite a few tube preamp projects as a nice clean, low-noise tube preamp supply.

In any case, as Chunger said the intention is to offer both, so anyone can pick which variant they want. 

 

[Category 5]

The new fully-regulated PSU design is quite a bit more flexible than the previous design:

1) It is agnostic of tube type.  Once set for 120V, you can run any tube type and the supply will adjust itself to maintain 120V.  Currently different tubes of the same type require adjustment of the B+ trimmer, and switching to a different tube type (e.g. 12AT7 or others) require removing and changing filter resistors
2) B+ supply is adjustable from 120V up to 245V
3) It provides true fixed bias, or a low impedance negative supply
4) It can accommodate both "positive" and "negative" heater supplies, all from a fixed regulator type (to switch requires only a jumper be moved)
5) Isolated grounds - there are three supplies that each have their own "ground" plane (think B-, HEATER-, and BIAS+).  This means you can float the heater supply up, etc.  Or you can strap all three together, it's your choice.
6) Optimized BOM with fewer component values that need to be ordered

The B+ supply is good up to about 10mA, and the heater up to 12V, so it's even possible to use this with quite a few tube preamp projects as a nice clean, low-noise tube preamp supply.

In any case, as Chunger said the intention is to offer both, so anyone can pick which variant they want. 

Awesome!  please make a PSU only kit available.  In the mean time can you peruse my post from late last night and give your thoughts.  If I need to make a component order I'd liked to try and get it in with another order tonight.

Thanks a million guys!

 

[Matador]

Awesome!  please make a PSU only kit available.  In the mean time can you peruse my post from late last night and give your thoughts.  If I need to make a component order I'd liked to try and get it in with another order tonight.

Thanks a million guys!

I am highly suspecting C4.  Can you remove it and try the loaded and unloaded readings of all of those points as before?  Unloaded voltages seem very low, which implies leakage current.  The bleeder resistor R6 is sized to bleed away (120V/470K) = 0.25mA, but your unloaded measurements indicate 40V lost across R1 (91K) which is 0.5mA, which is 0.25mA too much.  R4 is the only cap not protected by a series resistor and would suffer the effects of a B+ short the most.

 

[chunger]

I hope to make PSU kits will be available separately. . . it seems there is a surge of interest in DIY tube mics of various types and this particular PSU package (based on the economical Chinese case, power transformer, and mains wiring) is a very nice option for one of the toughest parts of any tube mic build.

I'll likely order loose power supplies from Alctron to bundle with the PSU pcb's.  I am moving towards making both passive and regulated options available separately in full kits and bare PCB's.

 

[Melodeath00]

The new fully-regulated PSU design is quite a bit more flexible than the previous design:

3) It provides true fixed bias, or a low impedance negative supply

Matador, is this to imply that the original PSU PCB design does not provide "true fixed bias," or am I reading this wrong? I'm still very new to this whole electronics thing . Many thanks!

Also, I second the motion to have PSUs available separately on their own. I may be using an alternate body for my C12, but will still need a PSU. Thanks!

 

[Matador]

Matador, is this to imply that the original PSU PCB design does not provide "true fixed bias," or am I reading this wrong? I'm still very new to this whole electronics thing . Many thanks!

Indeed:  the original C12 design is not fixed bias, despite the cathode being "grounded".  I think I answered it previous in the thread if you care for the details.

 

[Matador]

For any interested in alternate build configurations, I am building another mike with a slew of other BOM and stuffing options to show everyone what is possible with this design.

If you are interested in seeing a build for this, please "like" this post and I'll take photos and generate docs for this config.  But in short:
[list type=decimal]
[*]Will use Dave Thomas's C12-type capsule.  This is not backplate-correct to AKG but is very close to the overall sound for a fraction of the price
[*]Convert the design over to use a 12AT7 tube - this involves a few resistor changes in the PSU and a change to the plate resistor.  NOS grade 6072A tubes are getting harder and harder to find, however there appear to be many high-quality medium-mu triodes to try in this design.  For mine, I have a 6201 as well as a few NOS 12AT7 GE's to try
[*]Using silver mica caps in the mike front-end (C10, C11, and C13).  These are much more robust to soldering/heat than the styrene's and should be similar sonically.  Probably much more forgiving for first-time builders as they tolerate IPA washing a lot better than styrene's as well.
[*]Switching to a 2uF metallized polypropylene output cap
[*]Cinemag CM-2480 output transformer
[/list]

 

[chunger]

Pitchers!!!

Really sensible options posted by Matador.  I am very interested in the specific resistor values used for 12AT7 application because premium 12AT7's are a lot cheaper.  I have a NOS Mullard and Siemens that I would like to try in circuit.  We tested the cinemag 2480 head to head with the AMI T14 and they were really close.  It was only with direct side-by-side evaluation that differences were detected.  I really wanted to like the cinemag better because it is about $40 cheaper, but in this application, with the "premium" mic configuration that we were testing (NOS GE 6072A tube/Tim Campbell capsule), the AMI T14 was a little more transparent/open but had less bass.  Cinemag was much prettier looking though and the build quality of the Cinemag fantastic!  Differences were NOT by a large margin and different ears and different applications could very well favor the cinemag.  Perhaps a better match to the K67 based budget capsules as well.

At the end of the day, the "budget build" should come in at around $500 and the "premium" build around $1000.  That's a big difference.  Hopefully we can put some sound clips up side-by-side once Matador is done. . . assuming the mic doesn't disappear to a paying customer as soon as it rolls off the bench.

 

[Melodeath00]

Matador, is this to imply that the original PSU PCB design does not provide "true fixed bias," or am I reading this wrong? I'm still very new to this whole electronics thing . Many thanks!

Indeed:  the original C12 design is not fixed bias, despite the cathode being "grounded".  I think I answered it previous in the thread if you care for the details.

Ah yes, I remember now in the Tube Biasing thread you explain the C12 is not really true fixed bias. However, my question is, is this new PSU design different in the fixed bias portion from the older design? I wasn't sure if you were listing "true fixed bias" as a new feature, or just as a general feature of the new PCB. Thanks!

 

[Matador]

Ah yes, I remember now in the Tube Biasing thread you explain the C12 is not really true fixed bias. However, my question is, is this new PSU design different in the fixed bias portion from the older design? I wasn't sure if you were listing "true fixed bias" as a new feature, or just as a general feature of the new PCB. Thanks!

Yes, it is different.  It is low impedance and regulated from a separate rectifier.  It's level doesn't change as a function of tube current like in the C12 design.

 

[Matador]

12AT7 Conversion - The Microphone

Here are some details around the mike changes needed to run a 12AT7 tube.

First, let's look at the load line for the 6072A and we can gather why AKG picked some of the components they did.

So a 100K plate resistor across a 120V means that the maximum current is 120V/100K or 1.2mA.  We can draw the load line from the 120V, 0V point to the 0V, 1.2mA.  The idle current @-1V bias is about 0.7mA:  if we want to fake this load to test the PSU we need a 120V/0.7mA = 171K load resistor.  180K is close enough.  The Q point at -1V looks to be in the middle:  plate voltage idles at about 55-60V, idling current looks like about 0.7mA.  Bias resistance of -1V bias at plate current of 0.7mA implies a 1.42K "cathode" resistor (really just R4 in the PSU).

For gain:  we can see from the load line, a 0V to 2V swing on the grid equates to a 60V swing on output, which is gain of ~30dB.

So let's try to replicate this with a 12AT7 tube.  Another tube chart is in order:

So we need about 60V swing with 2V input to match 6072A gain.  Load line intersecting at 2.5mA gives approximately this.  120V / 2.5mA implies a plate resistor of 47K.  Q point at -1V looks to be in the middle:  plate voltage idles at about 55-60V, idling current looks like about 1.1mA.

So to generate this bias a cathode resistor of -1V/1.1mA = 900ohm is implied.  Easily settable with R3 (EDIT: fixed this should be R3, not R4), or can be an external resistor for those wanting to try true "cathode bias".  A 0V to 2V swing on the output causes a 75V swing on the output, which is a gain of about 31dB.  Understandable since the 12AT7 transconductance is higher.

Output impedance of the head amp is half the stock C12 design (due to lower plate resistor).  Transient response is probably better.

 

[chunger]

So, what I'm reading here is. . . to use the 12AT7, we change R17 in the mic from 100K to 47K, reset the bias via trim pot (R4) in the PSU to -1V and we are done?  EDIT:  NO!  See below.

I should include a 47K resistor in the parts kit and modify the build documentation to reflect this option  ;D

 

[Matador]

Bah Chunger posted before I could get to the PSU part.  :-[

12AT7 Conversion - The PSU

So the two big changes from above (from the PSU perspective) are:  a) the idle bias current draw increases from 0.7mA to 1.1mA, and b) the "cathode" resistance decreases from 1.4K down to 900ohms.  "B" is not an issue because we have R3 which can easily accommodate this change.  With the default 91K values of R1 and R2, the additional 0.4mA of current will drop cause an additional 72V drop in B+, which means there would be no way to get 120V out of the PSU.  So some re-calculations are in order.

Let's assume that R4 is set for half, which is 50K.  We have about 270V coming out of the rectifier diodes and we want 120V out of the PSU, which is a 150V drop across R1+R2+R4.  Total current should be 1.1mA + 0.25mA (through R6) which is 1.35mA.  150V drop at 1.35mA implies a total resistance of 112K.  Since R4 provides 50K, then R1+R2 needs to equal about 62K.  R1 = R2 = 33K sounds pretty good to me.  This gives an adjustment range of about 160V down to about 50V.

12AT7 Conversion - Summary

Microphone:  Change R17 from 100K down to 47K
PSU:  Drop R1 and R2 down to 33K.  Adjust R3 down to about 900ohms, then tweak with the actual tube.
Test load: decreases from 180K down to (120V/1.35mA) = 88K.  91K would work as a substitute for a real tube for PSU testing.

 

[micaddict]

 

[Category 5]

Hey Matador.  I haven't had the chance to trouble shoot the psu since your last response but did notice I a pic I took of it that the secondary of the transformer is labeled 175v instead of 200v like my first one. 

Anyone else see this?  Is 275 out of the diode bridge about right?  Maybe they just mislabeled it. 

 

[Matador]

Hey Matador.  I haven't had the chance to trouble shoot the psu since your last response but did notice I a pic I took of it that the secondary of the transformer is labeled 175v instead of 200v like my first one. 

Anyone else see this?  Is 275 out of the diode bridge about right?  Maybe they just mislabeled it.

Probably mislabeled:  200VAC should give about 280V after the rectifier diodes.  275V seems really close to that.

 

[Aaronrash]

This is great news. Any estimate on when the new PSU boards will be availible?

 

[micaddict]

Being dual triodes, both the 6072/12AY7 and the 6201/12AT7 can be had with balanced sections/triodes, cherry picked by the supplier. Obviously, for some applications this is more important than for others. Still, it is suggested that those with balanced triodes have better odds for low noise and microphonics, regardless of application.
Any feelings about this?