9V Electric Mistress retrofit with MN3007

[DrAlx]

Use this one for the reason that anotherjim stated.

http://www.banzaimusic.com/CD4049-buffered.html

I have ordered lots of these buffered "B"chips from Banzai, and they are not quite as advertised !!!.
They are actually M74HC4049B1.
Larry spotted that the max supply according to the data sheet is 7V.

Having said that, I have used these chips as clock buffers on 5 different EM3207 type builds, all running a little under 9V supply and never had a problem with them.  I am guessing that is because everything is going at relatively low frequency for that chip (i.e. clock is only around 1MHz), and nowhere near the upper limit for the chip, so power dissipation is small.  I don't know what will happen if you run them with more than 9V supply (e.g. for a MN3007 type build).

Even though the unbuffered "U" chips that I have tried didn't keep the waveform square (easily seen using a scope) I could not hear much difference between the "U" chips and the "B" chips.
I have also noticed that the "U" chips are used in at least 2 other flangers (the Current Lover EM clone,  the Hartman Flanger (another EM clone)).
I don't know if there is some other version of the "U" chip that does a better job of keeping the waveform square.  I doubt it somehow.

[Fender3D]

...I don't know if there is some other version of the "U" chip that does a better job of keeping the waveform square.  I doubt it somehow.

Come on guys...
it's barely 1MHz, does it really need a buffer?
Have you ever try WITHOUT?

Anyway, you can use 2 mosfets one N and one P (or 2 transistors) per clock phase:
clock output at gates N's source to GND, P's source to Vcc and drains to BBD

[Mugshot]

i currently have on bench an old Deluxe EM with a busted SAD1024. i etched the daughterboard and followed the connections to the wiring to the main board. however, i get no output; if any, there was only very faint signal and hum.

coupla questions:

1) i checked the MN3007 datasheet and saw pin1 connected to VDD and pin5 to GND. http://www.experimentalistsanonymous.com/diy/Datasheets/MN3007.pdf
the retrofit board has these the other way. what gives? is it because pin5 expects negative voltage? my head is so torched at the moment.

2) still with vintage DEM board, do i have to add decoupling caps and bias network (47K + 100K trim) to the input of the MN3007? do i also have to decouple the output of the retrofit board?

3) is there anything else i'd need to change?

[armdnrdy]

If you look at the data sheet that you posted the link to, you'll see that the supply voltage is listed at -15V. for VDD.
Basically....Pin 1 has to be more negative than pin 5.
Out of all of the stomp box schematics that I have seen, The MN30XX series is rarely powered with negative voltage.

[anotherjim]

I think there's a difference in wanting squared up edges and driving load.
Buffered or Un-buffered, the last stage (the only stage in the un-buffered), is the same -  so it's drive capability should be the same. The Buffered will be better to square up a sloppy incoming clock, but may be no better in coping with the capacitive load of a BBD.
Also, some designs use 1 inverter driving 2 in parallel per clock line, so an un-buffered chip is given some help from the extra stage.

What might be considered when adding CMOS buffers, is that the load on the power supply is most felt during the switch in level - for a moment, both MOS transistors of a complementary pair are "ON" presenting a very low resistance across the power supply. With a buffered chip, 3 pairs in each inverter are on during the switch. Now put them in multiple - all switching at the same time. Well, consider that it could be doing you more harm than good

[12Bass]

I seem to recall someone testing a triple buffer BBD clock arrangement and finding that it was actually worse than using only two in parallel.  On my SAD1024A A/DA build I have fairly healthy local supply bypassing.

[Mugshot]

If you look at the data sheet that you posted the link to, you'll see that the supply voltage is listed at -15V. for VDD.
Basically....Pin 1 has to be more negative than pin 5.
Out of all of the stomp box schematics that I have seen, The MN30XX series is rarely powered with negative voltage.

ah i see. thanks for that.

2) still with vintage DEM board, do i have to add decoupling caps and bias network (47K + 100K trim) to the input of the MN3007? do i also have to decouple the output of the retrofit board?

3) is there anything else i'd need to change?

[Thomeeque]

 Hi,

 we need exact schematic of your DEM first - if you don't have it, please match your vintage DEM with one of these (preferably by board number e.g. EH-5150-B)..

 T.

Btw. what you mean by "no output" - are you talking about output from the retrofit or from the whole effect? Because even with non working SAD or retrofit you should get clean signal at the DEM output.

EDIT: Hmm, actually, if you are sure there was SAD1024 (not RD5106A) it is probably safe to say that you don't need to add any extra parts (or extra mods on the DEM board), just fit retrofit there instead of SAD and adjust trimmers. But if you are not getting clean sound at the out now, problem lies elsewhere, not in BBD line.

[Mugshot]

Hi,

 we need exact schematic of your DEM first - if you don't have it, please match your vintage DEM with one of these (preferably by board number e.g. EH-5150-B)..

 T.

Btw. what you mean by "no output" - are you talking about output from the retrofit or from the whole effect? Because even with non working SAD or retrofit you should get clean signal at the DEM output.

EDIT: Hmm, actually, if you are sure there was SAD1024 (not RD5106A) it is probably safe to say that you don't need to add any extra parts (or extra mods on the DEM board), just fit retrofit there instead of SAD and adjust trimmers. But if you are not getting clean sound at the out now, problem lies elsewhere, not in BBD line.

thanks for the reply, man. it's this one, http://www.metzgerralf.de/elekt/stomp/mistress/images/deluxe-electric-mistress-v3-schematic.gif

i have changed the transformer to 220V and confirmed proper voltage (thru the 15V regulator) is supplied to all parts of the circuit.

with the retrofit board detached, i probed the thing: there is nice output on the 4558's inverting input (pin2), but considerably lower volume on pin1, and an even lower output on pin7 (very faint). adjusting the gain trimmer does nothing, so as with the bias trim. the balance trim i think is useless with the retrofit board, and maybe i can already remove it or short it out.

with the retrofit on, im getting loud hum on the 3904 emitter.

[Govmnt_Lacky]

Posting some voltages (as in the Debugging thread) might help identify the problem. Show voltages without AND with the adapter board 

[Thomeeque]

thanks for the reply, man. it's this one, http://www.metzgerralf.de/elekt/stomp/mistress/images/deluxe-electric-mistress-v3-schematic.gif

 OK, this circuitry requires something at place of SAD actually even to pass dry signal. IC1a derives DC bias voltage from the SAD output (see 47k+5uF at positive input) - without SAD output it is biased to 0VDC, it will not pass audio signal (or very poorly).

with the retrofit board detached, i probed the thing: there is nice output on the 4558's inverting input (pin2), but considerably lower volume on pin1

 It sounds fishy - you should not hear much on negative input (right, guys?) and at output you should hear at approx. equally* loud signal as on the IC9a output.

 Post voltages of IC1 pins at least for starters..

 T.

* IC1a gain depends on frequency partially, unity gain is around 300Hz

[DrAlx]

Just looking at how that very first op-amp (IC9a) is biased.
There is a DC path from pin 3 to the output of the other op-amp (IC1a)
So if IC1a is not biased correctly than neither is IC9a.
So the BBD output is needed to set the bias for both op-amps.

[Thomeeque]

Just looking at how that very first op-amp (IC9a) is biased.
There is a DC path from pin 3 to the output of the other op-amp (IC1a)
So if IC1a is not biased correctly than neither is IC9a.
So the BBD output is needed to set the bias for both op-amps.

 Wow, right, totally missed that!

 So to check audio path without SAD and retrofit inserted connect pins 2 (or 15) and 6 (or 12) of SAD socket.

 Btw: check pinout of retrofit's Q1, I recall some issues with Eagle package I was using those days, 180° rotated to usual 2N3904..

[Fender3D]

If you subbed the schematic at page 1 on this thread, you may try lifting Q1 and connect MN's pins 7-8 directly to gain trimmer.
I did this in my micro chorus retrofit and it works... I guess 10K is too much load for MN, maybe 47K will work better...  again maybe you'll have to lower the 160K resistor to restore a better wet/dry balance
This way you should be able to feed a proper bias with the trimmer on board.

Should the above not work then you should decouple input and output of the schematic above by placing 2 caps (one to MN pin 3, the other at Q1 emitter), 1 resistor at Q1 emitter (10K-100K) and a gain trimmer + 100K to MN's pin3; lift the 47K resistor going to the gain trim and connect it at board bias trimmer's wiper (trimmer near IC1b).
This way, board trimmer will feed bias for op-amps and retrofit bias trimmer will feed bias for BBD.

It might be not a "perfect" hack but it should work, try and report, we may then check if other mods are needed to work properly.

[Mugshot]

thanks for the replies guys. i havent seen any vintage DEMs on this thread using the retrofit board to compare the results, hence the clarifications. i followed your suggestions on the first page for the connections, btw.

i will be posting voltages tonight guys.

If you subbed the schematic at page 1 on this thread, you may try lifting Q1 and connect MN's pins 7-8 directly to gain trimmer.
I did this in my micro chorus retrofit and it works... I guess 10K is too much load for MN, maybe 47K will work better...  again maybe you'll have to lower the 160K resistor to restore a better wet/dry balance
This way you should be able to feed a proper bias with the trimmer on board.

Should the above not work then you should decouple input and output of the schematic above by placing 2 caps (one to MN pin 3, the other at Q1 emitter), 1 resistor at Q1 emitter (10K-100K) and a gain trimmer + 100K to MN's pin3; lift the 47K resistor going to the gain trim and connect it at board bias trimmer's wiper (trimmer near IC1b).
This way, board trimmer will feed bias for op-amps and retrofit bias trimmer will feed bias for BBD.

It might be not a "perfect" hack but it should work, try and report, we may then check if other mods are needed to work properly.

cant access the "regular" 9V schem right now, but i think i saved a copy somewhere. i guess i will have to mod the stages of the vintage DEM if the first suggestion doesnt work. i'll try firing up the iron tonight.

[Govmnt_Lacky]

Just want to throw this out there....

You guys know that there is already a fully tested and proven DEM build that runs off the RD5106A BBD (about the same price as the MN3007) and it sounds GREAT!!!

Do a search for "The Gilmour"

Just thought I would throw that out there.

[Mugshot]

here are the voltages i got for the vintage DEM (with the retrofit on, based on this schematic http://www.metzgerralf.de/elekt/stomp/mistress/images/deluxe-electric-mistress-v3-schematic.gif)

78L15 puts out 9.7V from the 11.23V rectified supply.

LM741
1) 0.0V
2) 6.03V
3) 6.03V
4) 0.0V
5) 0.0V
6) 6.2V
7) 8.87V
0.0V

RC4558
1) 0.59V
2) 0.53V
3) 2.23V
4) 0.0V
5) 6.07V
6) 5.85V
7) 5.65V
8.87V

LM311
1) 0.0V
2) 1.95-3.15V (sweeps)
3) 1.6-2.37V (sweeps)
4) 0.0V
5) 8.87V
6) 8.87V
7) 5.8-7.6V (sweeps)
8.87V

LM324N
1) 1.4-6.7V (sweeps)
2) 4.04V
3) 3.3-4.7V
4) 8.87V
5) 1.29-1.65V (sweeps)
6) 1.29-1.6V (sweeps)
7) 1.4-3.3V (sweeps)
3.42-4.6V (sweeps)
9) 4.08V
10) 4.08V
11) 0.0V
12) 1.03V
13) 1.03V
14) 1.03V

CD4013BE
1) 7.69V
2) 1.13V
3) 6.45-8.03V
4) 0.0V
5) 1.0V
6) 0.0V
7) 0.0V
0.0V
9) 0.0V
10) 0.0V
11) 0.0V
12) 0.0V
13) 8.9V
14) 8.9V

2N5087
E 8.2V
B 7.75V
C 2.5-1.7V (sweeps)

RETROFIT BOARD

CD4049UBE
1) 8.95V
2) 8.86V
3) 0.0V
4) 0.0V
5) 8.89V
6) 1.08V
7) 0.0V
8.9V
9) 8.9V
10) 0.0V
11) 8.9V
12) 0.0V
13) 0.04V
14) 0.0V
15) 8.9V
16) 0.12V

MN3007
1) 8.2V
2) 0.0V
3) 0.56V
4) 0.58V
5) 0.0V
6) 8.9V
7) 7.7V
7.7V

2N3904
C 8.2V
B 7.7V
E 7.37V

i tried the first suggestion of lifting the BJT on the retrofit board and tying the MN's out output directly. nope, no dice.

[Fender3D]

If you want 15V from your 78L15, you should feed it with more than 15V...

At a first glance MN's pins 2-6 show you have no clock, and bias voltage is way too low at pin 3

[Govmnt_Lacky]

78L15 puts out 9.7V from the 11.23V rectified supply.

If you want to get 15V from the regulator, you have to feed it AT LEAST near 17.5 to 18VDC. OTherwise... why bother having the regulator in there right? 

[Thomeeque]

If you want to get 15V from the regulator, you have to feed it AT LEAST near 17.5 to 18VDC.

 You can buy slightly more expensive Low Dropout Regulator, e.g. LM2940CT-15 should be happy from 15.5 volts up (but please double check yet, I may be wrong).

You guys know that there is already a fully tested and proven DEM build that runs off the RD5106A BBD (about the same price as the MN3007) and it sounds GREAT!!!

 If it is possible to buy RD5106A, than it is maybe better option (MN3007 is too long BBD to make perfect retrofit for EM)

 T.