EM3207 (v1.1) - MN3207 based EHX Electric Mistress (9V) clone

[Thomeeque]

But now I've got a new problem: in the EM3007 I can't understand where pin 1 of the IC2 goes to. Is it VCC or GND? Because the PCB board I made from this image http://thmq.mysteria.cz/em3207/build/EM3007_v1.1/EM3007_v1.1_PCB_mirror_1200DPI.png
shows that pin 1 is connected to...nothing.
And the schematic shows it connected to...GND? Shouldn't this create a short?

T.

[ricbox]

But now I've got a new problem: in the EM3007 I can't understand where pin 1 of the IC2 goes to. Is it VCC or GND? Because the PCB board I made from this image http://thmq.mysteria.cz/em3207/build/EM3007_v1.1/EM3007_v1.1_PCB_mirror_1200DPI.png
shows that pin 1 is connected to...nothing.
And the schematic shows it connected to...GND? Shouldn't this create a short?

T.

This slip will be the cause of my retirement from the DIY world.

haha, just joking! Thanks ;-)

[Thomeeque]

  You're welcome.

Btw. there's jumper-only layout EM3007_v1.1_jumpers_1200DPI.png provided, worth checking IMO.

T.

[ricbox]

Yep, in fact I missed some. I don't want to be much of an arse, but I really need big help on this.

During my first run, after soldering the jumpers, everything went bananas and smoke came out of the pcb. Still didn't know what damage did, probably the 2N5087 (Q2) got fried so I changed that.
Now I've got a signal that distorts first and then gets cleaner and cleaner and flanged 'till my ears hurts, and then returns to distorted.
I made a clip to prove it, with all the pots to the max and the trimpots set casually (RT3 to the MAX I believe), the filter switch up

http://soundcloud.com/riccardo-lomazzi/clip-em3007-not-working

I really don't know where to begin set it right.

[Thomeeque]

Yep, in fact I missed some.

Well, question now is what else did you miss.. go thru your build part by part and check carefully everything.. check soldering, PCB traces, component values, pinouts.. if you are able to make sharp high-res photos (both sides of the PCB, wiring, more angles) maybe we could help.

From the sound sample we know that all function blocks (LFO, clock, BBD and rest of the audio path) work, just not properly. Modulation is very shallow (with RANGE and RT3 at max it should be really huge), check primarily C17 value (isn't it too small?). About distortion check primarily Q1 pinout/orientation (it's known bug in the EM3007 layout, now with big red warning sign). How did you adjust bias voltage?

And you should of course find out why it did burn first time, there's no way how could faulty Q2 cause this if the rest was alright.

If you will not succeed, follow instructions here.

RT3 to the MAX I believe..

I don't want to be much of an arse either, but please know next time

T.

[ricbox]

From the sound sample we know that all function blocks (LFO, clock, BBD and rest of the audio path) work, just not properly. Modulation is very shallow (with RANGE and RT3 at max it should be really huge), check primarily C17 value (isn't it too small?). About distortion check primarily Q1 pinout/orientation (it's known bug in the EM3007 layout, now with big red warning sign). How did you adjust bias voltage?

So, Q1 was ok, because inverting it made the distortion disappear, but even the flanging effect.

• I removed the distortion by playing with RT1 (it's on half).
• I changed the ICs, to be sure it wasn't one of them that got smoked.
• Rotating RT2 to the max makes a very loud ears-breaker whistle
• Rotating RT3 doesn't do anything
• Rotating POT3 actually does, and it kinda flange the signal while I rotate it
• And oh, POT2 isn't set as a reverse-log. I mean, I just put it as it was.

[oldschoolanalog]

Serious troubleshooting demands accurately posted voltages. As Tomas requested. Please do so. 
That being said;
>>>>Rotating POT3 actually does, and it kinda flange the signal while I rotate it<<<<
This statement leads me to believe it is in "filter matrix" mode. Check your switch (SW) for correct wiring and function.
Please review and adhere to the Build Notes in the pdf. Tomas spent a lot of time and effort on this document. If followed carefully (along with reading this entire thread) there should be no issue too large to prevent successful completion of this flanger.
As far as flangers go, this is as simple as it gets.
We await your V's...

[Thomeeque]

So, Q1 was ok, because inverting it made the distortion disappear, but even the flanging effect.

Probably yes.

• I removed the distortion by playing with RT1 (it's on half).

Aren't you by any chance meaning RT2 when saying RT1 and vice versa?

If yes, you have probably set optimal enough bias voltage this way.. Damn, I should have not been lazy to finish the ALIGNMENT PROCEDURE chapeter

Keep it as it is for now, then later (when the rest is OK and with COLOR at zero) you may try to fine-tune it by ear or by scope for minimal distortion.

• I changed the ICs, to be sure it wasn't one of them that got smoked.

Good. Btw. do you already know why did it smoke?

• Rotating RT2 to the max makes a very loud ears-breaker whistle

If it's RT1 you are talking about and COLOR is at max., than it's normal (it's called self-oscillation and you adjust RT1 to point where it stops /or where you like it/ - another ALIGNMENT PROCEDURE step).

• Rotating RT3 doesn't do anything

Than you have something wrong around (check RT3, R31-R33, D1, C17, Q2, PCB..). PHOTOS, VOLTAGES!!!

• Rotating POT3 actually does, and it kinda flange the signal while I rotate it

Well and does it change depth of modulation (which you have very slow thanks to the next bullet) as well? If no check MODE switch wiring as Dave's said.

• And oh, POT2 isn't set as a reverse-log. I mean, I just put it as it was.

Then yet solder log/exp jumper according to BUILD NOTE 7 (without it RATE pot is not functional and acts as it was set to the slowest rate).

@Dave - thanks, I should have spent yet little more time and effort on the ALIGNMENT PROCEDURE chapter

[ricbox]

Ok, so, I apologize for being such a pain in the ass, but I managed to get all the voltages. I know that

I was so into many different pcb, that I printed this in horizontal reverse, so I had to CAREFULLY invert all the pins on the ICs and double check every component. Don't worry: I'm pretty sure the pins match the scheme.

The cause for the smoke was an incorrect position of the DC lug on the Voltage Supply. It runs at 9V (8.97 to be precise, but come on)

Here's the list.

Q1
E  8,75
B 4,75
C 3,9
Q2
E 8,47
B 7,35
C 1,71

IC1

1 1,67 5 8,7
2 1,42 6 1,7
3 1,19 7 1,4
4  0,00 8 1,5


IC2


1 8,47 5 4,4
2 5,9 6 VAR
3 0,19 7 4,7-5,0
4 0,4 8 0,00

IC3
1 2,44 8 1,7
2 1,42 9 1,7
3 8,64 10 1,6
4 4,0 11 0,00
5 3,9 12 3,8
6 3,8 13 3,9
7 3,6-4,2 14 3,9

IC4


1 0,00 5  8,75
2  2,3 6  5,9
3 1,7 7 8,6
4 0,00 8 8,6

IC5


1 5,4 8 8,6
2 5,4 9 8,7
3 5,7 10 0
4 0,00 11 0,00
5 5,4 12 0
6 0,00 13 0,00
7 0,00 14 0,00

IC6


1 8,6 9 0,00
2 5,8 10 6,0
3 5,9 11 6,0
4 5,9 12 0,00
5 5,4 13 6,7
6 0,00 14 6,0
7 5,4 15 6,7
8 2,8 16 6,0



RT1
1 0,00 (?)
2 0,00 (??)
3 0,00 ()
RT2
1 8,6
2 1,6
3 0,00 ( connected to 27k R27)
RT3
1 2,2
2 8,2
3 8,2 (to diode)


Better if I host a pdf somewhere next time.
I swear I don't know why RT1 is going like this: it seems like it's in short. I fixed the eventual joints there, but it hasn't changed.
Oh, and I almost forgot: the LED bypass doesn't even work. I soldered it right, following the classic True Bypass Scheme, but it doesn't lights up.

[DLC86]

Hi everybody
I need your opinion 'cuz I want to add a wet level control to this pedal. do you think it's better to add the pot in series with the output mixing resistor as a variable resistor, as a voltage divider or put the entire effect circuit in parallel with the dry line using a line mixer?
Thanks in advance

[Thomeeque]

Ok, so, I apologize for being such a pain in the ass, but I managed to get all the voltages. I know that

I was so into many different pcb, that I printed this in horizontal reverse, so I had to CAREFULLY invert all the pins on the ICs and double check every component.

You have.. You did.. WHOAAAAA!!! ... OK, it took a while, but my heartbeat is back and eye-ball's diameter goes back to normal But now I definitely MUST see it!!

Don't worry: I'm pretty sure the pins match the scheme.

Well if you have flipped legs of all ICs to back without breaking them and the IC's top (side with labels) is now bottom, it should be OK. But you must rotate transistors as well plus pot's will have reversed function (if directly soldered to the PCB from the top).

The cause for the smoke was an incorrect position of the DC lug on the Voltage Supply.

Well, I'm trying to find out what impact could this event have for the circuit itself which I still don't have enough info for ("an incorrect position of the DC lug on the Voltage Supply" could mean lot of things, I need to know what happened "electrically" - shortcut on DC jack, reversed polarity on EM3007's DC pins..), but we can forget about it for now.

It runs at 9V (8.97 to be precise, but come on)

Feel free to be precise (at least two valid digits with proper rounding applied, e.g. 9.0V here).

Here's the list.

OK, where to start.. First, IC pin numbering goes in anti-clockwise circle, e.g.:



So all your right-side IC pin numbers have to be flipped vertically. But OK, I'm able to decode, just for the next time (and maybe to clear some of you confusion if you have tried to map between schemo and layouts)..

Second, at some point you did flip Q1 pins C and E (C is at VCC so it must be like 8.7V, 3.9V is good value for E).

Third, I see that I was wrong with my theory that you are interchanging RT1 and RT2. Long story short adjust RT2 to get V_Bias = 5.2V (measure V_Bias at IC1 pin 7 - normally you should measure it at pin 3 of IC2 but you have some mess there). 5.2V is proper bias for MN3007 at VCC = 9V according to datasheet where it must work somehow, you may fine-tune it later. As you have V_Bias (and V_ref) so low now, that's not surprise that you have distortion there (at 1.7V there's not much of swing and used op-amps may not act linear in this area as well).

Fourth, many of your measured values are very hard to believe. E.g. IC6 pin 6 at 0V when directly connected pins 9 and 11 are at 6V, IC1 pin 7 at 1.7V when directly connected IC2 pin 3 is at 0.19V (and there must be connection here because we have heard flanging), 8.2V at RT3 pins 2&3 when directly connected pins 5, 6 and 8 of IC4 have 8.7V, 2.8V at IC6 pin 8 (should be ground) etc. Even some of it may be the real problem, most probably you did just make errors when getting voltages..

Setup V_Bias to those 5.2V and take new measurements at IC2, RT3, Q2 (identify E and C legs yet by some other connected part).

Do new set of measurements at IC3 as well, there must be huge swing on pin 7 (rectangles, you must wait a while as you have probably slowest rate set, it's period should be around 6 seconds) and fluent up/down at pin 8 (triangles). For all pins with swing post lowest and highest value.

Better if I host a pdf somewhere next time.

Whatever.

I swear I don't know why RT1 is going like this: it seems like it's in short. I fixed the eventual joints there, but it hasn't changed.

There should be no DC voltage after C8 and C10, so values at RT1 pins seems correct.

Oh, and I almost forgot: the LED bypass doesn't even work. I soldered it right, following the classic True Bypass Scheme, but it doesn't lights up.

Is this first pedal you are building?

T.

[Thomeeque]

Hi everybody
I need your opinion 'cuz I want to add a wet level control to this pedal. do you think it's better to add the pot in series with the output mixing resistor as a variable resistor, as a voltage divider or put the entire effect circuit in parallel with the dry line using a line mixer?
Thanks in advance

Depends on the control range and let's say total "fidelity" you want to achieve (original output mixer is complex passive circuitry, changing mixing resistor value influences even frequency characteristics, proper de-emphasis, output volume and impedance). If you are not sure, start with trying e.g. 100k pot instead of R18 and see (hear ). If using external mixer remember that you should apply de-emphasis to all signals after IC1B output.

Good luck, T.

[oldschoolanalog]

@ ricbox: Just a thought, FWIW. Please consider the option of etching a correct board and starting over. This way you could build it and test the sections as you go. Also, any troubleshooting will be easier for all parties involved. Consider this a learning experience and move ahead. 
@ Tomas: I feel your pain... 

[Thomeeque]

@ ricbox: Just a thought, FWIW. Please consider the option of etching a correct board and starting over. This way you could build it and test the sections as you go. Also, any troubleshooting will be easier for all parties involved. Consider this a learning experience and move ahead.  

On the other hand first EM3007 v1.1 *Mirror Edition* sounds prettty tempting

@ Tomas: I feel your pain...  

Thanks, T.

[DLC86]

Depends on the control range and let's say total "fidelity" you want to achieve (original output mixer is complex passive circuitry, changing mixing resistor value influences even frequency characteristics, proper de-emphasis, output volume and impedance). If you are not sure, start with trying e.g. 100k pot instead of R18 and see (hear ). If using external mixer remember that you should apply de-emphasis to all signals after IC1B output.

Good luck, T.

Thank you, I'll give it a try as soon as I have 5 spare minutes...
but I had another (maybe stupid) idea: what if I use the pot to change the gain on the transistor at Bbd output? Will this avoid those side effects?

[ricbox]

Ok, so, I apologize for being such a pain in the ass, but I managed to get all the voltages. I know that

I was so into many different pcb, that I printed this in horizontal reverse, so I had to CAREFULLY invert all the pins on the ICs and double check every component.

You have.. You did.. WHOAAAAA!!! ... OK, it took a while, but my heartbeat is back and eye-ball's diameter goes back to normal But now I definitely MUST see it!!

Don't worry: I'm pretty sure the pins match the scheme.

Well if you have flipped legs of all ICs to back without breaking them and the IC's top (side with labels) is now bottom, it should be OK. But you must rotate transistors as well plus pot's will have reversed function (if directly soldered to the PCB from the top).

The cause for the smoke was an incorrect position of the DC lug on the Voltage Supply.

Well, I'm trying to find out what impact could this event have for the circuit itself which I still don't have enough info for ("an incorrect position of the DC lug on the Voltage Supply" could mean lot of things, I need to know what happened "electrically" - shortcut on DC jack, reversed polarity on EM3007's DC pins..), but we can forget about it for now.

It runs at 9V (8.97 to be precise, but come on)

Feel free to be precise (at least two valid digits with proper rounding applied, e.g. 9.0V here).

Here's the list.

OK, where to start.. First, IC pin numbering goes in anti-clockwise circle, e.g.:



So all your right-side IC pin numbers have to be flipped vertically. But OK, I'm able to decode, just for the next time (and maybe to clear some of you confusion if you have tried to map between schemo and layouts)..

Second, at some point you did flip Q1 pins C and E (C is at VCC so it must be like 8.7V, 3.9V is good value for E).

Third, I see that I was wrong with my theory that you are interchanging RT1 and RT2. Long story short adjust RT2 to get V_Bias = 5.2V (measure V_Bias at IC1 pin 7 - normally you should measure it at pin 3 of IC2 but you have some mess there). 5.2V is proper bias for MN3007 at VCC = 9V according to datasheet where it must work somehow, you may fine-tune it later. As you have V_Bias (and V_ref) so low now, that's not surprise that you have distortion there (at 1.7V there's not much of swing and used op-amps may not act linear in this area as well).

Fourth, many of your measured values are very hard to believe. E.g. IC6 pin 6 at 0V when directly connected pins 9 and 11 are at 6V, IC1 pin 7 at 1.7V when directly connected IC2 pin 3 is at 0.19V (and there must be connection here because we have heard flanging), 8.2V at RT3 pins 2&3 when directly connected pins 5, 6 and 8 of IC4 have 8.7V, 2.8V at IC6 pin 8 (should be ground) etc. Even some of it may be the real problem, most probably you did just make errors when getting voltages..

Setup V_Bias to those 5.2V and take new measurements at IC2, RT3, Q2 (identify E and C legs yet by some other connected part).

Do new set of measurements at IC3 as well, there must be huge swing on pin 7 (rectangles, you must wait a while as you have probably slowest rate set, it's period should be around 6 seconds) and fluent up/down at pin 8 (triangles). For all pins with swing post lowest and highest value.

Better if I host a pdf somewhere next time.

Whatever.

I swear I don't know why RT1 is going like this: it seems like it's in short. I fixed the eventual joints there, but it hasn't changed.

There should be no DC voltage after C8 and C10, so values at RT1 pins seems correct.

Oh, and I almost forgot: the LED bypass doesn't even work. I soldered it right, following the classic True Bypass Scheme, but it doesn't lights up.

Is this first pedal you are building?

T.

First of all, THANK YOU so much for being supportive. You don't expect it that much from the direct creator of a project, this is a nice surprise.

And no, it's not my first build: I previously made a BOSS CE-2 Clone, and I think you know well how hard it is to bias it. But I made it!
This one is giving me problems, particularly beacause I haven't understood well the different sections of the scheme. Not blaming your faults, you did a monstrous job with this pcb.

But I realize that projects like this need patience and constance, two things that I had when I built the CE-2, but I don't have right now. I'm really, not in the mood (women problems lol). I won't be asking for help anymore, I think I'm gonna give it up. Maybe I'll take another look at it someday.

Thanks again for your help.

[Thomeeque]

Maybe I'll take another look at it someday.

I strongly hope so..

but I had another (maybe stupid) idea: what if I use the pot to change the gain on the transistor at Bbd output? Will this avoid those side effects?

That seems clever, unfortunately this pot would affect amount of feedback ("color") adjusted by the COLOR pot as well.

Btw. feedback brings another complication maybe (depending on what exactly you are up to again): even "dry" signal path R16-C10 contain the feedback signal, so it's not completely dry (it's completely dry only when COLOR is at zero).

T.

[DLC86]

That seems clever, unfortunately this pot would affect amount of feedback ("coor") adjusted by theCOLOR pot as well.

Btw. feedback brings another complication maybe (depending on what exactly you are up to again): even "dry" signal path R16-C10 contain the feedback signal, so it's not completely dry (it's completely dry only when COLOR is at zero).

well, i figured it out looking at the schemo again..
Anyway yesterday I tried several solutions, firstly a 100k and a 250k lin pots in series with r18 and both worked only in the last quarter of turn, maybe it would have been needed a rev log taper. anyway I ended up with a 100k pot set as voltage divider (with lug 1 to ground) placed between 2 100nF caps before r18 and it works perfectly for my needs, now I can achieve those Wall tones where the mistress is set lower in the mix. The only issues are a very little interaction with the color pot and changes in the overall volume but it doesn't bother me since I have the old version of the PCB with the booster on the output.

Thanks for your help and for this great pedal once again.

P.s: If it's not clear what I did I can draw a little schematic

[Thomeeque]

 Great, congrats!

anyway I ended up with a 100k pot set as voltage divider (with lug 1 to ground) placed between 2 100nF caps before r18

If you mean like this:



you could have wired pot there directly without those caps, there's no DC to be decoupled at this point.

Maybe you could post some sound sample of your favourite setting compared to the same setting with "full" wet, so others have idea..?

Cheers, T.

[DLC86]

Yes, exactly!
I've added those caps just because my electronics knowledge is almost totally empiric : when I was working on the echobase (that has a similar arrangement in that section) i remember that the level pot was way more interacting with the feedback without the decoupling caps, so I thought it was the same here. Anyway I'll even try without those caps.
About the samples, I'll se If i can do'em tomorrow afternoon. 
Cheers