Current Lover Madbeans Flanger

[Josemitejam]

Hi all!

So I just put together the Current Lover from Madbean and I have no effect at all when the pedal is turned on. The LED lights up though. I get a clean signal while on and off. I checked the IC voltages and it looks pretty good, except for a few differences here and there (posted below). I did some audio probing and all the caps and resistors work only up to C15 and R21, then they stop. The volume trim is the only trim working.

Here's the ic readings:

IC 1

1. 4.8
2. 4.8
3. 2.4 ...... should be 4.18
4. 0
5. 4.2
6. 4.5
7. 4.5
8. 9.3

IC2

1. 9.3
2. 4.6
3. 2.4......should be 4.85
4. Varies
5. 0
6. 4.6
7. 7.1.....should be 4.2
8. 7.1......should be 4.2

IC3 and IC4 are spot on

IC5, Spot on except for pin 7 was at 0....should be 7.8

IC 6 and 7 are spot on


Could it be a bad MN3007 chip???


Three more things that I should say:

1. D4 is for an LED but the directions show the LED hooked up down by the foot switch, so I just left it empty.

2. I accidentally soldered sockets in at 3 of the trimpot sites, so I just plugged the trimpots into the sockets, I would assume that this is totally fine.

3. I drilled initially for the effects loop but drilled poorly so I have a busted dpdt in one of the drilled holes just to fill the space. It's not in use!

Here's some links for pictures. I highlighted all the resistors and caps that are giving me a signal through an audio probe in the schematic picture.

https://imagizer.imageshack.com/img923/1992/LW5ohB.jpg

https://imagizer.imageshack.com/img924/9949/Jnrb5U.jpg

https://imagizer.imageshack.com/img922/1139/yRMggu.jpg


Here's the link to the Madbean Current Lover pdf:

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwie4cGw0fTvAhUwhuAKHYYxA1YQFjAAegQIAxAD&url=https%3A%2F%2Fwww.madbeanpedals.com%2Fprojects%2F_folders%2FFilterMod%2Fpdf%2FCurrentLover_2015.pdf&usg=AOvVaw07_ezy01vjcfFk6vrBXld-


Thoughts? Suggestions?

Thanks!!!


Joe

[Josemitejam]

As I'm audio probing it more it seems that all the pins are are giving me a clean signal on the MN3007 chip Except pins 2, 5, and 6. 5 goes to ground, so I'd assume I should hear nothing. But 2 and 6 go to the next part of the circuit, the 4049 chip. The problem seems to be right in this area. Why is it that I can get good voltage readings from the rest of the chips even if when I audio probe I'm getting absolutely no sound after this point?

Could it be a faulty mn3007?

[Josemitejam]

Also, in the directions it says that it "should run on 12vdc" but it also later says that this pedal is "for 9v operation".  Seems to be a bit conflicting...perhaps it runs better at 12v? Could it be that it needs more voltage?

[njkmonty]

where did you get the mn3007 chip from?

[Josemitejam]

From smallbear - does it seem like the chip could be the culprit?

[njkmonty]

SMALLBEAR  is pretty reliable, it was the obvious first place to look as many fake chips out there

[DrAlx]

BBD bias is wrong (i.e. IC2 pin 3 voltage). Have you tried varying the bias trimpot setting? BTW, R10 is totally unnecessary (a design error IMO) and can make it impossible to set the bias correctly. I would just short out R10.

[DrAlx]

Even with bad BBD bias you should be able to audio probe and hear audio at BBD pin 3. So take BBD out of socket and probe pin 3 on the socket. If no audio there then work back from that point towards the circuit input and see where audio first disappears.

[thomasha]

Pin 2 and 6 of the BBD should have the clock signals. If you had an oscilloscope it would show something at higher kHz values.
You will not hear anything with an audio probe.
If there is no clock signal, there will be no output, but the voltage that you are reading seems OK.

It looks to me that the bias at pin 3 is wrong.

BBD Bias is normally closer to half of the supply voltage, and only works for a narrow band. I don't think it will work at 2.4v. That also explains why the output pins have the wrong bias. One you correct the input bias, the output will correct itself.

You should check, as already said, around pin 3. Check if C4 is shorted, or if the trimpot was soldered correctly. On one side of R10 you should have VC, and on the side of the trimpot it should be a little lower. See how high you can get the bias at pin 3.

[Josemitejam]

Thanks for the responses! Here's what I've found after following up:


I get audio at pin 3, with both chip in and chip out.

I adjusted the bias all the way clockwise and I now get a much better reading at pin 3 of 4.3v!

Pins 7 and 8 are still high though at 7.6v. 

On one side of r10 I get 9.2v and 4.8v on the other side of r10, then I get 4.8v again the collector of the bias trim.

I can't find any shorts (although the solder did climb up the pins a bit on the 3007 socket).


Some more thoughts on it:

1. ic1 pin 3 is at 2.4 and should be at 4.18, is this enough to be an issue? Maybe the problem stems back to around here? I checked the area for soldering and values and it all looks good.

2. There's some confusion about what the power supply should be, in the directions it says that it "should run on 12vdc" but it also later says that this pedal is "for 9v operation".  It can also run on as high as 15v. Seems to be a bit conflicting...perhaps it runs better at 12v? Could it be that it needs more voltage??


Thanks again! One day I'll know enough to pay it forward to others here!!

[DrAlx]

R10 is stopping you from setting bias to correct level. It is a design error and not needed. Short it out on the back of the board, and you will be able to set bias to proper level and that should fix things. 

Another design error is R30 of 15k. If you look at the original 9V EM on which the CL is based you will see they used 39k. The CL designer lowered it to make the sweep wider, but the amount by which you can lower it depends on the supply voltage. The lower the supply voltage, the less you can lower it.  Make it too low and you will have the VCO freezing at the top of the flanger sweep giving dropouts in the sound. That problem has been reported many times here.  I would just use 39k like the original EM if you are on 9V supply.

Madbean should update their build document, as these same two issues (not being able to set the bias, and getting dropouts in the sweep) keep getting mentioned by people who build this circuit.


Answer to first question:  IC1 pin3 is wrong and should be half your supply voltage. Check voltages at IC7B pin 5 and 7 and on the voltage divider made up of R38 and R39 since they determine what you will get as IC1 pin 3.  Check you got correct resistor values for R38 and R39.  Circuit will probably work to some extent even with the IC1 pin3 voltage being as low as 2.4, but that low voltage means there is a build error.  It needs fixing.

Answer to 2nd question:  My guess is the circuit was originally developed and made to work at a higher supply voltage than 9V, and when doing that the above two problems (bias and dropouts) might not have occurred. I would say the circuit is only suitable for 9V operation if R10 is lowered sufficiently (or shorted out) and R30 is increased sufficiently.

[Josemitejam]

Ok here's the update - a bit of progress made:

Changed r30 to 39k

Replaced r10 with a jumper

Was able to tweak bias so that ic2 pin 3 is at 4.8.....which is correct!

Ic2 Pin 7 and 8 are still at 7.6....they should be at 4.2 

R38 and R39 Look good

ic7 pins 5 and 7 are accurate

Ic1 pin 3 is still low at about 2.4, I also just noticed that when I read it with multimeter there is a lot of noisy static from it (as well as pin 5).



Still no effect though. I'm going to try to pick up a 12v power supply tomorrow to see what happens.

Thank you for all that info DrAlx! Any further advice is super welcome!!!

[DrAlx]

Don't set bias voltage by trying to copy some value you have seen reported elsewhere, but rather by varying the bias trimpot until signal passes through the BBD.  It is not clear that you actually tried to do that.
As already mentioned above there is a narrow range of bias voltages that let signal through the BBD and you want to be near the middle of that range.
The MN3007 datasheet has a graph of typical bias voltage for different supply voltages, and for a 9V supply the graph shows bias under 4v, so its possible you could have a bias level that is too high with 4.8V.

BTW page 2 of the build doc mentions shorting out R10, so they at least addressed one of the two problems I mentioned.

Clock signals to BBD are probably OK.

The only other odd thing I see with the CL is how they set Vgg voltage (BBD pin 4). Typically this would be set to about (1/15) of the supply voltage which comes to 9V/15 = 0.6V. I think earlier version of CL used a diode connected between that pin and 0V so it was using the diode drop voltage as a reference. All I see in latest schematic is a capacitor at pin 4 which is not the way I would hook up that BBD. You don't report a pin4 voltage. If you can't manage to set a bias voltage that works, and find pin 4 voltage much larger than 0.6V then try grounding pin 4 (shorting out  C5) and see if that helps.

[Josemitejam]

Thanks for the reply! So I have tried both setting the voltage to the approximate setting that the build guide shows, but I've also slowly rotated the bias while listening for flange and I hear nothing. The voltage will certainly change when I hook up my multimeter, but I hear no flange.

I also took out C5 and wired it to ground, but pin 4 is still at 0v.

How small of a window would it be to get the signal? Are we talking within the .10v range...for example 3.80v - 3.90v will allow signal to pass? Or wider?

IC2 pins 7 and 8 are high no matter how I adjust the bias, they're always between 6.85v - 8.25v.

Here's another picture in case it reveals anything.

Thanks!

[Slowpoke101]

Check R13. It is meant to be 47K but you may have the wrong value there.

Also, you are measuring about 2.4V on pin 3 of IC1. Note the voltage is applied to this pin via a 1M resistor R4. I suspect that your meter has an input impedance of 1M which may be loading the voltage measured on pin 3 down. R4 and the meter would make a voltage divider which would cause the meter to display half the voltage applied to R4. You have measured VB as about 4.8V so half of that makes 2.4V. However pin 1 of IC1 has the correct voltage on it so pin 3 should also correct be when you are not measuring it.

Regarding pin 4 of the 3007. I agree with DrAlx, it just seems very strange. I'll check some other schematics and see what I find.

[DrAlx]

The bias window is less than a volt. If you turned the trim pot slowly you would not miss it.

+1 on what slowpoke says about pin3 IC1.

Yes you put in 47R instead of 47k for R13. Well spotted slowpoke!


EDIT: I have shared the original 2011 CL build document here:  https://1drv.ms/b/s!AvrH61utWEtEjkvGbzjT3OeANniq?e=CG4k8Z

You will see it has a diode setting the Vgg voltage.  The BBD should work even with a Vgg of 0V, but it's optimal at about 1/15 of the BBD supply.
MN3007 datasheet simply specs stuff for 15V so quotes a 1V figure.

For the latest CL version, maybe the Vgg pin is quite happy with simply being left to float with whatever voltage is on the cap, but personally I would have made sure that pin had a DC path to a known voltage.  Once you fix that 47R problem you will likely have a working flanger.  You might want to see if shorting/unshorting C5 makes any difference to the sound.

[Josemitejam]

Awesome!! Can't wait to get back to it tonight! I'll let you all know what happens!

[DrAlx]

The 2011 PCB could take either the MN3007 or the MN3207 by using appropriate jumpers on the PCB.  A diode was used to set Vgg only in the MN3207 builds.

With a 9V supply, the MN3207 has an optimal Vgg of 14/15 of 9V (i.e. 0.6V below 9V) but will be OK with Vgg set to 9V.
So the diode in that case is forward biased and sets Vgg to be a diode voltage drop (~0.6V) below the supply voltage.

On the other hand, the MN3007 has an optimal Vgg of 1/15 of 9V (i.e. about 0.6V) but will be OK with Vgg set to 0V.
The build instructions say that if using the MN3007, the diode and capacitor on pin 4 should be left out and the connections where C5 would go should be shorted together. In other words, force pin 4 to be 0V.

I guess when going to the 2015 PCB, they left out the diode but for some reason put C5 back in instead of simply forcing a Vgg of 0V.
Maybe any residual charge on C5 will leak away and give a Vgg of zero anyway, in which case you may find it makes no difference whether C5 is shorted or not.
My gut feeling though is that its better to short out C5 so that pin 4 is forced to 0V instead of having it floating near 0V.

[Josemitejam]

Alright it works beautifully now! Great catch slowpoke! And great information DrAlx!

I'll be playing with adding that C5 back in to see what happens and then seeing how higher voltage affects it (12v).

Time to jam first though!