VC-ADSR Upgrade wierdness

[Pav]

Hi Tony,
hope all is well with you.

Just converted vcos and mididacs over to Yarns voltages, and in testing i experienced weirdness with my ADSR-VCA
CV input Upgrade combo. Weirdness has been there since i built them. I have never felt the attack, decay works as they should in normal mode,
yet they do effect the looping as expected. At full clockwise attack knob audio dissapears when using vca with an oscillator. Back off, it returns

The one with The CV upgrade - has never appeared to have any effect ..so i investigated the upgrade pcb specifically thinking
it might be a quick fix , a power or wiring gremlin. Power was reaching TL074s and sock4 board wiring to pcb and to 4way A-D S-R molex
thingy had continuity.

I put a fixed 3V dc voltage in turn into each of the A,D,S,R sockets and measured the Attack-Out (etc) voltages sent to the main adsr board
(at bat42 and 220 R resistor junction). All 4 approximated 0v full CCW to 12.x full CW. Note - I did not have power on the main board at this point not realizing there was an Attack-In (etc) that might or might not be important.

The puzzle for me came when i fed the 3V into Attack Socket , with all knobs fully clockwise.
As i rotated Attack knob it changed the voltages output for Decay,Sus,Rel as well!!!

Similarly if I fed 3v into Decay socket, rotate Decay and Attack, Sus, Rel all change in step with it !!
Similarly tests for Sus and Rel. A short ?

Surely that cannot be right ?
Hopping you may have some ideas.

Cheers
Pav

[Synthbuilder]

I put a fixed 3V dc voltage in turn into each of the A,D,S,R sockets and measured the Attack-Out (etc) voltages sent to the main adsr board (at bat42 and 220 R resistor junction). All 4 approximated 0v full CCW to 12.x full CW.

A +3V input to any input should give a range of output voltages from -0.3V to +3.1V as you move the CV input pot from CCW to CW. However, this test can only be done with the ADSR module not connected, or connected and powered up. It is not valid to test the outputs from the Upgrade board when the ADSR board is connected and not powered up.

No input should affect the other channel's output voltage, except when the link is fitted between pins 2 and 3, and for this case only, the attack CV is normalised to the decay CV input.

If they are cross infecting each other, one possible fault is that 0V (pin 2 on the MTA156 header) is not connected to the Upgrade board.

[Pav]

Thanks for the reply
todays testing.
Today I had both ADSR-VCA and upgrade module powered on the bench.
I had removed the normalization link between 2-3 to avoid confusion previously.
I turned all knobs to min on the ADSR-VCA itself to min.
I plugged in 4v to each cv input socket n turn and took readings at the AD SR OUT header pins and got 0.x - 4.x volts for all 4
suggesting to me that the upgrade board is working.
I will recheck for cross infection but it may have gone now i am properly powering it.

Just to confirm , that no plugs in cv sockets means the knobs will have no effect as they are tied to ground ?

Turning to the ADSR-VCA board itself,
On taking readings at the AD SR IN pins and wiggling the ADSR-VCA knobs - i get no voltage change on the Attack-In pin at all.
The others I do ..eg 0-11v for decay. I now Need to explore further.

Lastly my adsr logic needs validation - i keep second guessing myself.

For D,S,R I have to have the ADSR-VCA knobs between 12pm and max to see ANY change on scope. Does that make sense. ?
Decay knob appears to change the speed at which it reaches peak, which if normalized to attack I might expect but we are not normalized.
Is not that Attacks role?

As the CV voltage is turned from max to min , the decay time increases and slows the time it takes to get to peak.
when sustain level = max = peak level , then i would not expect to "see" a voltage drop on scope. (
But When Sustain significantly less than max, I would expect to see the voltage decrease from peak to sustain level, after a time set by delay..
and I do not.. ( I set slow /fast switch on slow to help ). I am holding gate open btw trying to see this.

Im going to put the scope on a moog mothers envelope ..as that is what im comparing sound wise.

Sorry such a long note.

[Synthbuilder]

I plugged in 4v to each cv input socket n turn and took readings at the AD SR OUT header pins and got 0.x - 4.x volts for all 4 suggesting to me that the upgrade board is working.

Yes, this looks good.

i get no voltage change on the Attack-In pin at all.
The others I do ..eg 0-11v for decay.

You should get 0 to 10V from attack, decay and release. And 0 to +5V for the sustain. Check the wiper of the attack pot on the ADSR main board to verify you're getting the correct voltage there.

I have to have the ADSR-VCA knobs between 12pm and max to see ANY change on scope. Does that make sense. ?
Decay knob appears to change the speed at which it reaches peak, which if normalized to attack I might expect but we are not normalized.
Is not that Attacks role?

All the main four pots on the ADSR main board should make a difference. The module is designed to allow for precise attack, decay and release times - the scale is exponential - so you may not see the difference if the scope is set to a long time base. You'll hear it though.

As the CV voltage is turned from max to min , the decay time increases and slows the time it takes to get to peak.

So the attack time is being controlled by the decay input? If so, this is not correct. Decay time is the time it takes for the envelope output to fall from the peak to the sustain level. Decay CV should have no effect on attack time unless the normalised link is in place and the Attack CV pot turned away from the central position.

when sustain level = max = peak level, then i would not expect to "see" a voltage drop on scope.

Indeed. The sustain level when set to max should be very close to the maximum peak of the attack phase. There is a small difference (tens of mV only) due to offsets in the op-amps used and the speed of the comparator, U5a.

[Pav]

You should get 0 to 10V from attack, decay and release. And 0 to +5V for the sustain. Check the wiper of the attack pot on the ADSR main board to verify you're getting the correct voltage there.

Wiper Voltages on the adsr (v4 btw) with weirdness are
Attack 0-13v
Decay 0-11v
Sustain 0-8v
Release 0-11v
pots are Taiwan Alphas.

I did compare the voltages with a working v5 adsr (without upgrade board) and
good one meets your stated voltages 0-10v with sustain 5v.
So i paired this good one with the upgrade board, and i can now really hear the
effect of cv inputs as they should be. (I think...I hope.. yes...as they should be

So that leaves me a challenge to fix the v4 adsr.
I mention the versions as that might or might not mean different voltage levels at
the wipers. But the scope shows me the good adsr attack is a drop negative that releases to 0 volts
wheras this one is a positive atttack and release to 0v.

[Synthbuilder]

v4 btw...

Ah, right. Issue 4 and earlier are very different to the issue 5. The change is caused by the move away from the THAT VCA chip to a discrete version of it.

Attack 0-13v
Decay 0-11v
Sustain 0-8v
Release 0-11v

That is what I would expect from an issue 4. So those are working OK.

So that leaves me a challenge to fix the v4 adsr.

It might be worth swapping out U2 (the 4052) on the main board to see if that improves things.

[Pav]

Ill give the U2 swap a go. need to check if i have one spare. this morning i was puzzling on the direction of its use 4-->1 or 1-4.

[Pav]

update -4pm: U2 swapped. Not absolutely convinced it fixed anything but the oscilloscope is now seeing looping and gated looping modes, attack an decay working and we see "sharkfins". Jut proves what i originally thought. In Normal mode, I used oscilloscope to set sustain level to 1v. This visually exposes the original issue. Decay is changing the time it gets from 0v to sustain level 1v, I never see the peak voltage 5v reached with attack knob at minimum or maxim or in between.

[Synthbuilder]

this morning i was puzzling on the direction of its use 4-->1 or 1-4.

It's two sets of four inputs to one output, although only three inputs are actually being used for each switch.

On the schematic the top switch sets the level to which the envelope generator moves towards, either 15V (for the attack), 0 to 8V (for the decay phase and eventual sustain level), and 0V for the release. The bottom switch sets the gain of the VCA which controls the speed of the lag generator. The higher the voltage the slower the lag generator's time constant. It switches between the attack, decay and release pots (or the three CVs when used with the upgrade board).

[Synthbuilder]

U2 swapped.

Make sure it's a 4052 and not a 74HC4052.

Decay is changing the time it gets from 0v to sustain level 1v, I never see the peak voltage 5v reached with attack knob at minimum or maxim or in between.

Possibly a fault with U1. Or maybe the circuitry around U4b (pins 5, 6, 7). Maybe even D3 around the wrong way.

[Pav]

D3 right way round , I pulled it anyway and replaced it - no joy.
I was thinking U1 before i read your note , but swapping that - no joy.
U4 pin1 reflects the adsr out voltage on the scope when i gate , and pin 7 has a voltage changing on pin 7, as the voltaget increases to sustain level at the speed set by decay.
Resistors around U2 and U4 all have resistance that reflect their values and markings..R14,R21,22,,R23,R41.

Can i just pick your brains to see if i understand the role of the 4051 (from datasheet) . as my working assumption is that to get
an attack I need both 9 and 10 pins high as these connect pin 11( attack voltage) with common out pin 13 that feeds the THAT
at the same time pin4 (15v less resistor drop) connects with common out pin3 that after a buffer goes to THAT.

I see when i gate the adsr logic is pin 9 high pin 10 low which is coincidently decay voltage at pin 15 to common out p13
and sustain voltage pin2 to common out pin 3.

this is where i reach my limitations .. I suspect to go though the logic stages some circuitry has to step through the logic steps
starting with 1 1 , 1 0, 0 1,0 0 for ADSR cycle to complete, but we are missing the first step.

This is why I am impressed with your knowledge, and wish we had started in my teens.

[Pav]

SOLVED

got up early this morning and put my logic and improving schematic reading to the test.
sadly my ability to spot a dry joint is declining - eyes are going!

In normal mode, pin 10 U2 will only go high if NOR gate U1 pins 5 an 6 stay low when gated.
-pin 5 was going high removing D3 section of the circuit from the equation.
- shorted pin5 to set low and behold attack is back.

for pin5 to go low either Nor gate pin 8 or pin 9 needs to go high so tested these while gating
pin 9 is tied to ground in Normal mode by the on-of-on switch pin2...so pin 8 stays low.
i gated with a temp wire link to +15v to pin 8, Lo Attack returns!

I then first checked Q3,and Q4 , and stupidly replaced Q3 even though it was working.
so that left C14, D4,R33,R12 suspect. Ruled out D4 - already replaced it.
Turned pcb over with a view to first reflowing the three remaining. C14 joint was poor..
and i should have spotted that. Reflowed C14 dodgy joint -et voila.
I now have Attack phase back in Normal Mode.
The wierdness I had heard was in effect an A-R envelope but to only whatever the Sustain level was set.
If Im sounding smug -apologies.

Thanks again Tony for the help and explanations.
Ill be first in line if a new Oakley module is posted, euro or otherwise.

[Synthbuilder]

Thanks again Tony for the help and explanations.

I'm glad you got it working.