Modular Synth Panels

I recently completed construction and initial testing of the CGS tube VCA
using a 1SH24B tube.  I built it in MOTM-format using a panel and
bracket from Bridechamber.

The version I built is different than the original circuit.  In order
to avoid damaging the PCB from excessive unsoldering, I did not first build
a stock unmodified version for comparison purposes.

This is taken from posts I made to the cgs_synth and tubesynthdiy
yagoogroups where I went into more detail as far as my reasons for
doing things.  Since no one posted any corrections, I'd thought I'd
go ahead and post it here.

 From what others had said, there were three issues:  the output was
noisy, there was a "high-pass effect" or "loss of bass", and there
was a "thump" during the attack of the EG when used as a VCA.  This
is a list of changes designed to address those issues:

I eliminated the two signal diodes, the two LED's, and their 1.5K
resistors. I connected a 1.2K 1/2W resistor between Pin 1 of the
cathode-heater and ground.

I added a a 47uF axial electrolytic cathode bypass capacitor between
pin 2 of the tube (connected to the negative rail) and ground.  Do
not connect it to the ground side of the cathode or the in-rush
current could burn out the heater.

I replaced the TL072 with an OP275.  I eliminated the stock 100nF
bypass caps, and underneath the PCB, added two .1uF ceramic caps with
heat shrink tubing going from ground, with the capacitors as close as
possible to the power pins on the socket.  I eliminated the 47pF cap
on the output, and replaced it with a resistor lead.  I did not add
the gain resistor.  While 330 Ohm is the minimum specified resistor
for output protection, I replaced it with 1K (outside the feedback
path) to have the same output impedance as MOTM.

I increased the two 220nF coupling caps to 470nF.

I decreased the CV input resistor to 50K so a 0-5V EG it has about the same
range as the Gain knob.

I wired up the SPDT bias switch -- using properly shielded cable -- to two
settings:

For a "clean" setting, I added a 20K multi-turn trimmer to the
board.  There is an area with some writing without any traces to the
right of the -Ve hole for the stock 330K resistor, where I drilled
two extra holes.  The trimmer is wired as a rheostat between -Ve and
the switch.

For a "distortion" setting, I connected a 475K resistor between
ground and the switch.

The suspression grid was connected to the negative rail.

I used Bournes conductive plastic 50K linear pots for all four knobs.

While it likely varies from 1SH24B to 1SH24B, the tube I have
installed is more than a bit microphonic.  Clicking the bias switch
is enough to produce ringing.  So I stuffed some anti-static foam in
between the tube and the board to minimize vibration.

I think that's it.


These modifications seem to have worked very well.  I don't notice
any noise on the output.

On the clean setting, with a roughly 10Vpp input potted all the way
up, the distortion is barely noticeable.  Three things are happening
as the value of the variable resistor is adjusted lower:  the grid
becomes more negative, the level of the AC signal decreases, and the
frequency of the high-pass filter increases.  As such, there is some
loss of bass, which rolls off gently at the lowest audible
frequencies.  With or without a signal, the grid bias is equal to
-15V, due to the low impedance.  I would suggest using a 100K
multiturn trimmer instead -- that way you could adjust a "clean"
setting with the input potted down a bit for slightly better bass
response (although the high-pass effect of the original circuit was
due to a lack of cathode bypassing, as the corner frequency of the
stock values of 220nF and 330K in parallel with the grid resistance
was below the audible range).  You also might want to try even larger
values for the coupling caps.  I used AVX BQ  yellow box caps, which
go up to 1uF, but you should double-check the lead spacing.
http://avx.com/docs/catalogs/bf-bq.pdf

The distortion setting produces obvious distortion, and can produce
oscillations and other noises at various input and feedback settings.
The low frequency response is well below the audio range (so the module will
just about work as a depth control for a fast LFO).  The AC voltage
on the grid is about 3.3 RMS.  With a signal the grid measures around
-17.6 VDC (it varies with its frequency), and without a signal the
bias is around -14.1V.

While the LPF in the CV circuit is the same as the stock circuit, the
cathode bypass cap improves the transient response of the tube
itself.  It seems likely that rapid changes in voltage on the control
grid (eg. an envelope with a steep attack) would cause a sudden
demand for plate current enough to deplete the cathode, which then
bounces back causing an unwanted noise.

There is a bit of a pop or click with a fast envelope, although it is
not any worse than most solid-state VCA's.  Changing the signal grid
bias, and switching between the two settings, also effects its CV
response.  As such, it is possible to all but eliminate any bleed,
such that there is only a boost at the initial transient for punchy
basses and leads.  I found this while experimenting with different
values.  It would increase the distortion from the "clean" setting,
but it would still be a very nice feature.  I might add that if I can
find a suitable SP3T switch.

Regardless of the signal level on the grid, or the bias switch
setting, the output of the tube, and therefore the module, is a bit
below 1V RMS, as it seems that is the saturation limit of the tube.
So while the signal on the grid is much higher for the distortion
setting, the difference results in just that -- distortion.  In fact,
the "clean" setting measures just a bit more, although the
"distortion" setting sounds louder to to the additional harmonics.
Now 1V RMS is only about 2.8Vpp.  Although it is enough on the
board, I might change the gain and feedback resistors to 2K
(resulting in about 5.6Vpp) if it seems too low to work well with my
other modules.

For those building this in other formats that might have he extra
space, putting a bias rheostat on the panel would be a nice feature.
Using a reversing attenuator for the second input/feedback control is
another option, as negative feedback is also useful.

So with these modifications there is very little noise, sufficient
bandwidth, and it works fine both as a "regular" VCA and a
distortion/waveshaper module.

I am open to questions and comments.