LFO of 9V Electric Mistress

[tiago razera]

Hi all.

It's been a long time since I last posted here (2003 I think). This forum is awesome!!

I managed to find some MN3010 (dual-stage 512 BBD) here in Brazil and decided to build the 9V Electric Mistress, using the layout from Markus W. I did some tweaking on the PCB to match the pins of the new BBD and other little modifications and got it working almost perfectly.

My only complain is the LFO. It seems to stay too much time in the high end of the wave. Instead of going WEOWOEOWOE it goes WEEEEEEEOWOEEEEEEEE. Don't know if I made myself clear.

I checked all components values and connections and it's ok. Tweaking the clock trimpot only change the start point of the wave, but the waveform seems the same.

It is supposed to be like this?

As a side note, I plugged the effect on a 15v supply, and the effect became more proeminent, the LFO will go higher and lower. I liked better this way.

Thanks to everybody, this place rocks!!!

[Thomeeque]

It is supposed to be like this?

Hi!

Not able to help much at this moment, but I understand exactly what you mean (before clonning EM I've experimented with my own flanger design and during the process I did found that perfect sweep course is absolutely essential for perfect flanger) and I can responsibly say that it's not supposed to be like this - sweep course of EM is IMO perfect (check e.g. http://www.youtube.com/watch?v=qC7KEXo9XRo).

Tweaking the clock trimpot slightly influences this course - extreme settings criple it little, but if you are let's say between 20-80% of range it should be alright.

Diodes and transistor around the clock have influence as well, which exact types do you have there?

LFO it-self should generate pure sharp triangles at LM324 output and then slightly rounded triangles later (thanks to 1uF caps on the way /= shape will be dependant on speed as well/), you can check it if you have an oscilloscope.

T.

[tiago razera]

Thanks for the response Thomeeque.

The transistor is a 2n3905 (a lower gain 2n3906). The diodes are 1n4148 as per schematic and 1n4007 on the LM311 loop.
The caps 1uf and 33uf are electrolytic, couldn't find tantalum.

Tiago

[Thomeeque]

The transistor is a 2n3905 (a lower gain 2n3906). The diodes are 1n4148 as per schematic..

What schematic do you refer to?

..and 1n4007 on the LM311 loop.

That could be the problem, 1N400x never worked well on the LM311 loop for me (and oldschoolanalog Dave as well IIRC) even some schematics have it there, try to put 1N4148 there instead. 1N4001 should be in Base of the transistor (in series with 62k resistor).

The caps 1uf and 33uf are electrolytic, couldn't find tantalum.

That should not matter..

T.

[tiago razera]

This schematic: http://www.diystompboxes.com/DIYFiles/up/9V_Mistress_project_file_v1.pdf

Sorry for the confusion, the 1n4148 is in the IC loop and the 1n4007 on the 62k resistor, just like the schematic.

I'm at work now, will double check the components values later.

Tiago

[tiago razera]

Checked all components, everything is alright.

I have made some modifications as follows:

- 13k substituted with 12k
- 30k substituted with 33k
- TL074 instead LM324

I probed the signal directly at the output of the MN3010 and got a vibrato sound, but at the lower part of the oscillation it becomes mute, like it was misbiased for that moment. And the vibrato isn't very proeminent, nothing close to the pitch shifting that flangers can do.

To me it seems the LFO isn't sweeping the voltage high and low enough. I can only get a flanger sound when the clock trim is set to 0 resistance.

As I used a MN3010, which is a dual 512 using in parallel, I didn't do any buffering between the CD4013 and BBD. May this cause the problems?

Tiago

[oldschoolanalog]

Try an LM324 first and see what happens.

[tiago razera]

Try an LM324 first and see what happens.

Just did it. Same results

[Lurco]

A 330k resistor shunting the 33k will make it 30k. This may shift the DC level of the LFO towards the right direction?

[Dave W]

LM324 V's from my 9V EM w/3007 retrofit. Range full CW (max), Speed full CCW (slowest), Regen CCW (off). All V's to the nearest .1 V.
1) 7.5                 8] ~2.8 - ~5.1
2) 1                    9) 4
3) 1                    10) 4
4) 8.8                 11) 0
5) ~2.3 - ~5.6     12) 4.4
6) 4                    13) 4.4
7) .6 or 7.5         14) 4.4

Hope some of this helps...

[tiago razera]

Thank you all for the answers.

A 330k resistor shunting the 33k will make it 30k. This may shift the DC level of the LFO towards the right direction?

Did it, same results.

LM324 V's from my 9V EM w/3007 retrofit. Range full CW (max), Speed full CCW (slowest), Regen CCW (off). All V's to the nearest .1 V.
1) 7.5                 8] ~2.8 - ~5.1
2) 1                    9) 4
3) 1                    10) 4
4) 8.8                 11) 0
5) ~2.3 - ~5.6     12) 4.4
6) 4                    13) 4.4
7) .6 or 7.5         14) 4.4

My voltages pretty much matches the above.

Some things I'm learning.
Every schematic with MN3000 series I saw had a resistor of about 47k to +9v at the output of the chip. The EM SAD1024 had the opposite, a resistor going to ground. Removed that and the effect became much more proeminent.
Yet it won't do the pitch shifting on extreme settings, and if I turn back the feedback pot halfway the effect is barely noticeable.

For a 512 stage MN3010 it is necessary to buffer the clock? I'm starting to think that this is the problem.

Thanks all.

[Dave W]

Looks like it's not the LFO.
Check this out for info on retrofitting an MN30XX for an SAD1024 in a 9V EM:
http://www.diystompboxes.com/smfforum/index.php?topic=78270.0
Make sure your Vdd, Vgg (note R2 & R3) & Gnd are hooked up properly. Also, that 3904 at the output is necessary. You may or may not need to buffer the clock depending on high you want to clock it.

[tiago razera]

I've read all the thread about the retrofitting, maybe I missed something.

Also, that 3904 at the output is necessary

There's another purpose for this besides being a buffer?

Make sure your Vdd, Vgg (note R2 & R3) & Gnd are hooked up properly.

The Vdd and Vgg of the chip are connected to the ground and GND to the +9v. R2 and R3 i put 220k instead of 200k, but as a voltage divider it must not be the problem. Is this correct?

You may or may not need to buffer the clock depending on high you want to clock it.

With the actual clock frequency of the schematic (with 47pf cap on the LM311), it's necessary to buffer?

Tiago

[oldschoolanalog]

From the retrofit thread (thanks Tomas!):
>Output buffer made by common-collector Q1 voltage follower is added to compensate (approx. 5 times) bigger output impedance of MN3007 compared to SAD1024 - it seems to be necessary here, because following EM's passive feedback/output network creates relatively big load to BBD output (contrary to A/DA).<

220K should not be a problem. The "10% rule" & all that...

Considering the much larger clock line capacitance of the "MN" BBD's, it's probably a good idea to buffer the clock lines.

Just a thought. Why not build the verified (by many) 9V EM w/3007 retrofit and save the near impossible to find MN3010 for repairing old A/DA's? That is one chip that really is "rarer than hen's teeth/unicorn foreskin/etc."

[tiago razera]

Obviously I missed some things from the retrofit thread... 
Thank you for pointing me out.

I etched the 9v EM board some 3 years ago, there was no other layout than the SAD1024. If I had seen the retrofit thread before I wouldn't even try with this one. But, a lot of learning in the course.
Strangely the only BBD I could find here in Brasil was the MN3010. Found two MN3207 too, but it's in use for other effects.

Meanwhile I tried to make this thing work I've finnished an Ibanez FL301, a RAT and etched a board for the UltraFlanger from Hollis.

Will do the MN3007 board adapted to the 3010, since it's too much modifications to the old one.

Thank you all for your help.

[tiago razera]

Just to end the topic.

Added the 2n3904 buffer on the output of the BBD as mentioned on the great retrofitting thread and suddenly the mistress comes alive!!!

Great liquid flange sounds, exactly what I was looking for (Gilmour).

Thanks Dave, oldschoolanalog and Thomeeque for the help.