Gyraf G7 60V mod Official thread

Since there are several capsules that might work only, or at least safer, with 60V/120V polarization voltages, how about putting up some clear, dummie proof info on how to do it.  I've searched as much as I could on this and there are many versions and none seems to be fully confirmed. 

To start with, one thing would be to build the PSU as is and test the voltages before changing anything. There is a good chance it would give a lower voltage. Question is, why?  If construction/all parts are right and good quality, shouldn't it give 80/160? 

I've replaced the 10k with 47k before the pattern select voltage divider. this yielded around 123 volts for a couple of Oliver mod CMV 563 builds.

Not sure how it effects the usefulness of all the pattern voltages because these were cardioid builds. But IIRC the 60 volts for the capsule  back plate was correct.

hitchhiker said:

I've replaced the 10k with 47k before the pattern select voltage divider. this yielded around 123 volts for a couple of Oliver mod CMV 563 builds.

Click to expand...

BTW,  I just finished a PTP Matador version of the elam 251 and this adjusted G7 supply yields 119.9 volts with the 47K.  The same supply configuration gets exactly 120 volts attached to a G7 mic.

Cheers!
Lance

Thanks Lance!  (@elam: 199V ??)

I started to read about how these things work and it is getting a bit confusing

The supply starts with 135V at the second tran and then rectified + ripple cap.  The schematic writes 180V here.  Theoretical formula (Vac X sqrt2) gives 190V but that is only true with perfect diodes and an infinite capacitor.  So lets keep 180. 

We have a voltage divider made with 3 resistors:  10k, 100k, 100k.  The switch can change the "middle" of the divider so we get once: 
180 x (100/100+100+10) =  85 V  and once  180x ( 100+100 / 210) = 171 V       

I imagine once the mic is hooked, these voltages will drop some and come close to the 80/160  ?

If 10k is replaced with 47k, the voltages become:  73 V  and  146 V

If this is still too much, a 68k would give    67 V and 134 V

-----

For all patterns to work, I understand we also need these worked on inside the mic.  There is a 470k - 470k divider in there. 

If we reduce the second resistor to 300k, we get 70V instead of 80.  But there is also a 10k in line in there (upper right corner), I don't know if this should be factored?

Ooops!
The 199 volts was a typing error, I meant to type 119 volts.

  IIRC my supply is giving 177 volts before the 10k etc voltage divider.

Interesting info here:  http://groupdiy.com/index.php?topic=51320.0

I guess the best way is to:

1.  raise the 10k as hitchhiker suggested.  47k gives 129V but in the mic there is a 10k in series.  That will lower it more, depending on current draw.  At 1mA, to 119V.  But I imagine draw is even lower so the 47k could be a bit higher. 

2.  Depending on what we get above, the lower 470K from the microphone should be lowered.  282k gives a perfect 60.  300k gives 62V which is probably more in line with the divider from the psu. 

Wait!  I forgot this also lowers the HT so the tube might run unhappy.    What if we split the + rail like in the drawing below?  One thing though, it will leave both pattern voltages with less filtering.  In the original, I see the 8 pattern gets the second cap filtering, but the cardioid does not.  Perhaps a third filter cap can be added *after* the switch?