"Another Poor Man's" Fairchild 660/670

[nielsk]

Did you happen to try this with the 175 output section only? (no extra 12AX7 / DC Threshold stage)

 

[rotheu]

Did you happen to try this with the 175 output section only? (no extra 12AX7 / DC Threshold stage)

No, I didn't try that. I really like to have the DC Threshold. It changes the sonic character of the compression a lot,
depending on the settings. You also get different compression ratios all the way from 2:1 compression to 30:1 peak limiting.
And it's a very easy to build stage and it doesn't add much to the cost.

But don't rush to build it yet, because I'm gonna be updating the SC amp based on feedback and recommendations from
some people here, on the forum. So I'm working on it. Just waiting to buy some tubes (it's tough, I lost my job in December ).

 

[THE BAD DOGS]

Hello,

VERY NICE PROJECT!

I want to do this project and have some questions :

* Is there a bom list?
* On your pictures it seems to be a parrallel circuits with 4 tubes on each, one with 6bc8 and other with 6ba6 and a knob to switch between? Do you think you use this solution at least or you only stay with 6bc8 design?
* In case of building a "monster" with 2 parrallel circuit with each kind of tube, is there anything to change in the schematics after (you were talking about 6v6 output)?
* What about the "sum and lateral" linking mod instead of stereo linking mod?

Best regards,

Jean-Christophe

 

[rotheu]

Hello,

VERY NICE PROJECT!

I want to do this project and have some questions :

* Is there a bom list?
* On your pictures it seems to be a parrallel circuits with 4 tubes on each, one with 6bc8 and other with 6ba6 and a knob to switch between? Do you think you use this solution at least or you only stay with 6bc8 design?
* In case of building a "monster" with 2 parrallel circuit with each kind of tube, is there anything to change in the schematics after (you were talking about 6v6 output)?
* What about the "sum and lateral" linking mod instead of stereo linking mod?

Best regards,

Jean-Christophe

Hi, Jean-Christophe

* Is there a bom list?

Everything is on the schematic, if you have any specific questions, please, feel free to ask.

* On your pictures it seems to be a parrallel circuits with 4 tubes on each, one with 6bc8 and other with 6ba6 and a knob to switch between? Do you think you use this solution at least or you only stay with 6bc8 design?

I did it just to A/B them. I'm staying with 6BC8.

* In case of building a "monster" with 2 parrallel circuit with each kind of tube, is there anything to change in the schematics after (you were talking about 6v6 output)?

Pretty much everything has to change, including the SC. 6BA6 needs much bigger SC amp, that's why I used 6V6 (actually 6BQ5, which is 6V6 equivalent).

* What about the "sum and lateral" linking mod instead of stereo linking mod?

That's possible, you just need split windings on the transformers.

BTW, I got some tubes, so in a few days I'm gonna be posting some updates on the SC Amp.

rotheu

 

[veermaster]

Rotheu, what is the 1Meg Resistor+Diode+10nf Cap in your Timing Network for, is it really needed?
Thanks, Emre

 

[rotheu]

Rotheu, what is the 1Meg Resistor+Diode+10nf Cap in your Timing Network for, is it really needed?
Thanks, Emre

Hi, Emre

1Meg Resistor+Diode+10nf Cap are needed for stereo linking to isolate two timing networks. Without them your time constants will change when linked. For single channel you don't need them.

rotheu

 

[veermaster]

Thanks for your answer!
emre

 

[SUPERMAGOO]

nice project!

pcc189 have tried to use instead of 6bc8?. some time ago I adapted the circuit Altec-436 to be used with pcc189
with good results.

 

[rotheu]

nice project!

pcc189 have tried to use instead of 6bc8?. some time ago I adapted the circuit Altec-436 to be used with pcc189
with good results.

I didn't have a chance to check out pcc189, how does it compare to 6bc8?
I used original UA176 (great compressor!), built UA175 based compressor with 6bc8, modified Altec 436B (amp circuitry and timing components) with very good results. Also in this project 6bc8s are working very well.
What will be the advantage of using pcc189? Thanks.

rotheu

 

[SUPERMAGOO]

I did this experiment several years ago.
here in argentina I could not find 6bc8. schematics with 7es8/pcc189 having suspected that making a few changes might work in a prototype Altec 436.
sound seemed a little strong and fast vari-mu comp.
After a time I could find 6bc8, and modified the circuit again.
I felt better with pcc189.
I am no engineer or mathematical justification for why it would be a better pcc189.
All I can say that I liked to my ears. and also two very good comps to be used as gyratec and ear660.
I apologize for my bad English. I use google traslator.
cheers.

 

[djn111]

Hi,

I like to have a go at this, but some specific questions:

- the input pot. lin or log
- the meter a real VU or mA
- the 200 balance pot is to balance the u variantion in the both up en down tubes right, so a zero volt DC offset.
- how to calibrate
- how did you organise the power supply recetifier tube/solid state, CLC,300V,RC, 240V, if so can you help me with the current of the first to supply voltage, and how did you reg the 100v, current is changing with the settings, I was thinking 0C3 but the current with a grid off 0V, Ua 100V means 4x 0.36mA = 144mA way to much for a 0C3.

Dirk Jan

 

[SIXTYNINER]

 

[rotheu]

Hi,

I like to have a go at this, but some specific questions:

- the input pot. lin or log
- the meter a real VU or mA
- the 200 balance pot is to balance the u variantion in the both up en down tubes right, so a zero volt DC offset.
- how to calibrate
- how did you organise the power supply recetifier tube/solid state, CLC,300V,RC, 240V, if so can you help me with the current of the first to supply voltage, and how did you reg the 100v, current is changing with the settings, I was thinking 0C3 but the current with a grid off 0V, Ua 100V means 4x 0.36mA = 144mA way to much for a 0C3.

Dirk Jan

The input pot is log 50K

The meter is real VU

The 200 Ohm balance pot is to balance both sides of the tube
I'm thinking to build something similar to EMI Altec 436 calibration thing, basically a pulsing low frequency signal rectified and used as CV. You just turn the balance pot until you get maximum cancelation of the feed through clicks.

I will go with clean solid-state regulated PS. Since I didn't build it in the chassis yet, I don't have a dedicated power supply. I'm running it from "Lambda" regulated power supply, which gives me 300V regulated, then I have a transistor (TIP50) based regulator with zener diodes at the base as voltage reference to regulate it down to 240V. After that just a 20K(2W) resistor and 100V zener shunt for signal amp (100V regulated), but this is temporary. One channel (signal+side chain) draws only 60mA. You can just regulate parallel feeds after the rectifier as long as you have at list 90mA per channel available (just to be safe).

how did you reg the 100v, current is changing with the settings

You regulate the voltage. The regulator will keep the voltage constant, current draw will change.

 

[rotheu]

If you like to go with tube regulated supply, there is a lot of info here

http://www.audioxpress.com/magsdirx/ax/addenda/media/bicknell2890.pdf

 

[Matthew Jacobs]

Many thanks rotheu for sharing this project on the forum.

I'm thinking of making this my next "big project" and my first P2P project.

I'm thinking of doing a stereo unit. I wondering if anybody can point me to a Solid State PSU that could do the job. Something like the G9 PSU or is it not powerful enough. Also, how can I get the various voltages this baby needs, 100v, 240V & 300.

Thanks in advance.

J

 

[syn]

Rotheu

          thank you very much, for this excellent project. How about W rating of the resistors and V rating of the caps?
Thank you once again.
Best

 

[rotheu]

Rotheu

          thank you very much, for this excellent project. How about W rating of the resistors and V rating of the caps?
Thank you once again.
Best

Hi,
All resistors are 0.5W, electrolitic caps are 450V, coupling caps are 400V, timing network caps are 25V.

 

[MeToo2]

Thanks very much for writing this up.   ;D

I'm no expert, but I've been reading about vari mu compressors the past few days, and I've been lurking on other sites for a few months.

Was wondering why no one seemed to have thought of putting in an impedance buffer to drive the side chain capacitor charging, so that it could charge more quickly to reduce attack times, whilst not unduly loading the output of the main stage: well you've obviously have gone several steps further.

I must say that the transformer coupling seems to be a good way to go to avoid problems with any capacitor charging on large / fast transients (hence higher frequencies) whilst preserving the low frequency response. Modern audio transformers seem to be pretty good up to easily 100KHz and beyond and also down to 10Hz when properly loaded.

The fully balanced stages also look very promising for the power supply rejection ratio.  8)

Another potentially crazy 3 ideas for you: feel free to reject on a whim.

1. Would it be worth putting in an insert point between the output of the main transformer (marked "output to 600 ohm") and the input transformer of the side chain?

so that you could filter the side chain input using an external mixer? e.g. for de-essing, creating effects where only the bass triggers the gain reduction etc.

2. Is it worth putting in a 2 way switch to disconnect the side chain output from the centre tap of the input transformer and instead ground it (to temporarily turn off compression, so you can easily hear / monitor the amount of mojo being applied)?

3. By changing the gain of the side chain amp, would it be possible to change the compression ratio in a meaningful way?

[edit]
oops the side chain input pad pot is already there. silly me

Looking at the graph of transconductance and mu v grid volts for this valve (6bc8), it's highly non linear, especially with Eb of 100V. Mu goes from 10-40, but not in anything like a straight line. Suspect your mileage may vary, and the character of the compression characteristics are more down to the individual tubes than the influence of any circuit topology. The original 6386 in the Fairchild has a very different curve shape than the 6bc8, but costs a fortune for anything like a matched pair in one tube. Substitutes like dual matched 6ba6's are used by Manley (see published info at http://www.manleylabs.com/techpage/TBAR.html ) Looking at the curves for a 6ba6 would tend to suggest to me that when used as a matched pair these would be a better substitute for a 6386 than a 6bc8. Especially because of the much larger range of usable (dc) grid bias compared to the (ac) input signal level (40V for 6ba6 compared to just 10 for the 6bc8). But you rejected those tubes on test, so I presume you had good reasons for that. Gyraf suggests the PCC189 which was specifically designed for variable mu agc circuits but not directly in a fairchild type circuit and it requires a different heater arrangement of 300mA constant current. http://www.gearslutz.com/board/high-end/35397-fairchild-670-a.html  http://www.r-type.org/pdfs/pcc189.pdf The equivalent (6es8) are also used in the Langevin Leveline see http://danrudin.net/cgi-bin/download.pl?dir=Langevin/&file=Langevin_AM-5301.pdf Might be worth a try. I think the advantage would seem to be that the small signal transconductance variation is defined over a larger range of grid voltage i.e. 0.01 - 10 mA/V for the grid voltage varying from 0 -to -12 V. That's 60dB. They're also readily available and cheap.
[end edit]

Or are these mods considered heresy when talking about a Fairchild derivative?

Forgive me. I'm a tube noob

 

[lewilson]

I just finished fairchild 660 clone. I  used the Russian version of the 6ba6 tube, the 6K4P. It is a very nice tube. 
  I bought 30 pcs, but the first eight I put in were a good match. At this point its easy to buy a bunch that were from the same batch and so there specs are very close. Plus they are cheaper than a used american made 6ba6. Just a thought.

 

[MeToo2]

I just finished fairchild 660 clone. I  used the Russian version of the 6ba6 tube, the 6K4P. It is a very nice tube. 
   I bought 30 pcs, but the first eight I put in were a good match. At this point its easy to buy a bunch that were from the same batch and so there specs are very close. Plus they are cheaper than a used american made 6ba6. Just a thought.

Good point!

My understanding is that one reason for using the 6386, and lots of them in parallel, was not just to up the current handling, but also to get a better "average" match between the two halves of the long tailed pair. They also wired alternate halves to alternate sides of the circuit to get rid of any systematic manufacturing bias between the halves.

IMVVHO there's no inherent reason why you have to use a true dual triode for this circuit.

JJ electronics claim they match the two halves to within 3dB at 8 separate bias points. But these cost around $120 per tube.

If you can achieve that same balance and gain match by starting with a (large) selection of 6k4p's and selecting on test, that'd also seem to be a very workable solution given their extremely cheap price (saw 10 tubes on ebay for 1 euro each + postage.) You could literally buy hundreds of them for the same price as a couple of 6386's.