Maestro Filter Sample/Hold Original Layout, Problems

[karis12]

I normally don't ask for help regarding circuit builds, but I've been working on this for 24 hours straight, and the best I could do was get the envelope filter side working 

The deets:

*Circuit is a combination of GGG and TonePad
   - one JRC13600D dual OTA, one single op amp (TL071), and one dual op amp (TL072)
   - three 2N5457 FETs, one 2N3906 PNP, and one 2N3904. Diodes are 1N4001
   - no Decay, Resonance, Attack and Direction controls. I chose to include only the original Range and S/H Speed + F-S/H switch.
   - B100K and B1M (linear taper) potentiometers
   - LT1054 charge pump to provide +/-9 volts bipolar power to circuit
*Original layout on a perfboard, designed to fit a 125B

The issues:

*Despite checking the schematics many times, I need to flip the orientation of Q5 and Q6 to get the Filter side working
*S/H side is silent, even as I adjust the corresponding trimpot
*Range control's response is concentrated in the first half of the pot range with a B100K

How crucial is the layout to the design of this circuit?

Futzing around with it for hours, trying to find wrong component placements etc., I WAS able to *sort of* work the S/H, but instead of a cool step filter effect, I was greeted with a weird thumping sound whose speed was controllable by the S/H Speed knob, and the filter affected by the bias trimmer. During that time, I couldn't work the envelope filter part of the circuit. Almost as if I did the layout wrong in such a way that one side "hates" the other.



For people who have DIYLC, here's the file: https://ufile.io/e0phm

Thanks for your time.

[karis12]

UPDATE:

*My Q3 seems to have the Source and Drain pins reversed on my layout. I insulated the transistor pins and twisted the legs to correct this, but so far it's not helping me to solve any of the issues.

*Replaced the JFET input op amps (TL071/TL072) with a 741 and a 1458. This seems to make things better, but I still can't put both chips in and have either half function - setting the bias trimpots to work with the S/H side "displaces" the appropriate settings for the Filter side, and vice versa.

*I found a way to get random Sample Hold using any transistor: I clip an alligator clip to the emitter side of the noise transistor, and make sure it has contact with my skin somewhere, while dialing in the trimpots. I get random filtering sounds on the S/H mode this way, especially when I make contact with my guitar's ground points with my hands. I don't know if it means you can get the noise by tying a 2.2 megohm resistor to ground from the emitter side, or if my body is functioning as some sort of antenna, but somehow it works in my case. Very inconvenient though, and I'm sure this shouldn't be the correct function of the circuit.

Any opinions at this point would be much appreciated.

[karis12]

UPDATE 2:

*Used a battery for the circuit, and found out that with it, I could no longer induce random voltages in the noise transistor with my trick from the previous update.

*It seems that with lower input voltages I get some more low range from the envelope filter side

Should the 10k trimmer affect the envelopec filter side of the circuit? If not, this might provide a clue as to how I can troubleshoot this thing.

[Kipper4]

Show us the schematic please.

You can label the wires with txt in diylc.
and use the blue jumpers for clarity.

[karis12]

Show us the schematic please.

You can label the wires with txt in diylc.
and use the blue jumpers for clarity.

Thanks for the reply Kipper4. Unfortunately I am unable to produce a schematic for this, except the ones I mentioned (GGG and TonePad versions).
GGG: http://www.generalguitargadgets.com/pdf/ggg_fsh_sc.pdf
TonePad: http://www.tonepad.com/getFile.asp?id=96

Aspects from the GGG:
1. Two 100uf capacitors for the +/-8V output only; none for the +/-9V
2. Single op amp for the envelope filter portion of the circuit
3. Range and S/H Speed controls + switch only; no mods
Aspects from the TonePad version:
1. No PNP transistor on the charge pump
2. Sample/Hold portion of the circuit uses a dual op amp for the noise transistor amplification and LFO
3. One LM13600 instead of two CA3080s

I noticed that the 10k trimmer (R27 on GGG schematic) affects the sweep of the Range control of the Envelope Filter side. Reading up on the subject a little further, MadBean's version seems to indicate that it should not be the case, which leads me to think that I may have wired mine incorrectly (VR2 on my image).

[karis12]

UPDATE 3:

This is day 4 of this project, and here's some things I noticed:

*The circuit powers up with the negative 9 volt voltage pin of the charge pump at +0.6 or so, which causes the thing to go silent. I need to pull out the chips and transistors and plug them in again in order to get it working with the correct negative voltage.
*I'm very confused with the pinouts for the FETs. GGG and TonePad seem to have conflicting layouts on this, where the Source and Drain pins are reversed, notably with Q5 and Q6. Can anyone explain how exactly this should function?

Again, thanks for the time to read this, and take care.

[nocentelli]

*I'm very confused with the pinouts for the FETs. GGG and TonePad seem to have conflicting layouts on this, where the Source and Drain pins are reversed, notably with Q5 and Q6. Can anyone explain how exactly this should function?

Again, thanks for the time to read this, and take care.

In the manner in which the JFETs are used for Q5+6, the source and drain pins can be used interchangeably. Very crudely, applying a varying voltage to the gate will vary resistance between source and drain (like a voltage controlled variable resistor: This is how JFETs are used in phasers etc in place of of optocouplers, for example). It should not matter which connection is drain and which is source, as long as the gate is connected as per schematic.

[karis12]

Thank you for that. Still, it doesn't explain why my circuit starts up with the negative voltage at +.6 or so. Any possible explanation for this? Perhaps an emergency shutoff feature of the charge pump?

[karis12]

UPDATE 4:

I found another difference between the two circuits I wanted to ask about.

GGG's layout for the envelope follower portion has a 4.7uf capacitor tied to the +9V rail, immediately followed by a 100k resistor to ground. The output is sent to the Filter-S/H switch lug.

On the other hand, TonePad's version, which includes the Up/Down switch functionality, has the same 4.7uf capacitor's positive lead going straight to ground, followed by the Decay potentiometer plus 10k resistor, which I guess is an analogue of the 100k on the former. This is followed by another op amp stage that is part of the Up/Down function, before the output is sent to the switch.

Is this another case similar to the Tube Screamer, where the capacitor "doesn't care" where it's being referenced to?

[Kipper4]

quick question
Is IC5 an lfo? (ggg)

The charging cap C6 (ggg) discharges depending on the size of the cap V the R19 to ground.

R19 is replaced with the 1M decay pot in the Tonepad version, since its wired as a "variable resistor"
This allows you to control the rate of discharge of your charge cap.

The charge cap is often refered to as a C5 in some threads..... trivia.

The up down function (tonepad) is a phase swap
the first half of the up down switch (LH) is changing the gain of the envelope detector.. Using a parallel resistance. 1k2//22k

the other half of the switch (RH) is swapping from inverting to non inverting input IC2b
I'm not sure what op amp that is but I assume since there's less resistance in the +input it will win out in a comparator arrangement.

I'm still not sure I'm getting my head around this circuit.
I've never built it. Calling Dino.

[karis12]

quick question
Is IC5 an lfo? (ggg)

The charging cap C6 (ggg) discharges depending on the size of the cap V the R19 to ground.

R19 is replaced with the 1M decay pot in the Tonepad version, since its wired as a "variable resistor"
This allows you to control the rate of discharge of your charge cap.

The charge cap is often refered to as a C5 in some threads..... trivia.

The up down function (tonepad) is a phase swap
the first half of the up down switch (LH) is changing the gain of the envelope detector.. Using a parallel resistance. 1k2//22k

the other half of the switch (RH) is swapping from inverting to non inverting input IC2b
I'm not sure what op amp that is but I assume since there's less resistance in the +input it will win out in a comparator arrangement.

I'm still not sure I'm getting my head around this circuit.
I've never built it. Calling Dino.

Thanks for putting up with me Kipper4. Yes, GGG's IC5 is the LFO for the Sample/Hold side.

My charging cap is labeled C10 in my layout.

I went ahead and tied the positive side of the 4.7uF cap to ground instead of +9V and it didn't change the function of the Envelope Filter. It still didn't do anything regarding my -9V starting up positive, though.

I'm just about ready to give up on this build, but I've always wanted one since I found the effect on my digital multifx unit (no longer with me) almost 10 years ago. Having been able to work the S/H side somehow, through some weird methods, is preventing me from shelving it, due to that amazing step filter tone. It's really unique and inspiring.

UPDATE 5:

*Obviously, taking out both op amps from the circuit prevents the weird "-9V to +0.6V" phenomenon from occurring, which means that either I have made some wiring mistakes to induce such weird behaviour from the circuit, or the LT1054 simply cannot handle the load demanded by the ICs and transistors being powered altogether (AFAIK unlikely).

*I still can't produce the random voltage from the noise transistor in order to activate the step filter effect, without me having to touch the emitter leg with a precision flathead screwdriver. I've auditioned nearly all of my available BJTs and I still couldn't find the right one.

[digi2t]

UPDATE 5:

*Obviously, taking out both op amps from the circuit prevents the weird "-9V to +0.6V" phenomenon from occurring, which means that either I have made some wiring mistakes to induce such weird behaviour from the circuit, or the LT1054 simply cannot handle the load demanded by the ICs and transistors being powered altogether (AFAIK unlikely).

*I still can't produce the random voltage from the noise transistor in order to activate the step filter effect, without me having to touch the emitter leg with a precision flathead screwdriver. I've auditioned nearly all of my available BJTs and I still couldn't find the right one.

I've never built it. Calling Dino.

At this point, I'm sorta in the same canoe....

First things first... draw up a schematic of what you're trying to do. Far be it from me to claim any sort of "expertise" in this things, but as an old truck driver told me years ago, "Son, it all starts with a road map.". Sit down with the cut and paste software of your choice, and start... well, cutting and pasting. Then come back and show us (and by "us" I mean the folks that are WAY smarter than me) what your road map is. It might don on you right away that there is an intrinsic problem somewhere, or someone will point it out for you.

From my first glance, I wondered if the 1054 could carry all that load. The 1054 can cover a lot of ground, but if one side of your supply is returned to normal by removing components, either a) there's a short somewhere, b) faulty/shorted component, or c) there's too much load for the 1054 to support on that side. Personally, I'm not a hugefan of charge pumps (though I have used them in the past), but if this is on the breadboard, I would start by building a robust 7809/7909 supply on the side (or 317/337, my personal fav), and use an AC supply. This will guarantee enough current to both sides. Once you have a sure power supply, then you can start troubleshooting the actual circuit itself. Start at the base, then work your way back up.

As far as s/h circuits go, the only one I've built to date is this one using the LF398;



which worked like a charm.

[karis12]

I'm shelving this pedal build for now, but thank you for everyone who took time to read and reply to my posts. As far as I'm concerned, there shouldn't have been any issue regarding my layout, and might probably be an issue related to the charge pump. Next time I try and put this together, I'll look into other powering options, or choose low-current devices, and see where that takes me.