BBD sanity check

[imJonWain]

I need to rubber duck with some humans.

Background:
I agreed to fix a friend's early 80's Micro Chorus.  Some quick tests showed the the SAD512 was probably kaput.  Only clean output of the pedal and I could follow a signal on my scope to the chip input and the output was only very low DC and clock noise. LFO + astable and everything else works fine.  Rather then pay for a R5106 or SAD512D he asked if I could mod the pedal to use a more common or available part.  Cool,  I'll turn this into a learning exercise since I don't usually mess with chorus or flanger type effects.

So I ordered some 3207s, 3102s, and a few 4046s to experiment with.  Plan is to build a daughter board for the new bbd chip and whatever clock modifiers I need (x2 or x4 if I use the 3102).


Thing is I haven't been able to get it to the 3207 to output anything besides clock noise and DC.  I'm sure it's something silly I am missing.

I initially tried using a 4070 to split/buffer the uChorus astable output and run that to the 3207 clock pins with resistors for vgg.  It worked great for creating two opposite phase clocks and was able to drive the 3207 but I was getting clock noise on the output without any signal.  I figured since the 3207 is ideally wanting some dead time between the clock pulses maybe it didn't like how fast the rise/fall times were (they crossed). I had been hoping to use a low pass and the XOR to double the frequency but couldn't get it stable enough anyway.

So today I setup the 3207 + 3102 on the breadboard and drove the 3102 0x1 pin directly from a bread boarded astable + LFO setup (exactly the same as the MXRs).  It makes a lovely clock at .5f as one expects going into the 3207 with a small dead space between pulses, sweet.  I setup a simple 47k to GND and 47k/22nF low pass at the combined outputs of the 3207 so I can check for output.  I ran a wire from the chip socket inside the pedal to the breadboard (~100mm) and ran a sine gen to the pedal input ( tried 300hz up to 1kHz).  Still no output besides some clock noise ( which the LPF does kill off).  For kicks I disconnected my astable from the 3102 and and fed it a 300kHz clock signal from a waveform generator.  That should make the delay the same as the SAD512 @ 75kHz.  Same, no output (as I expected).


What am I doing wrong?  I am gonna go over the whole setup tonight but wanted to have a sanity check with everyone here.


Have I killed two of these things already?  Are they "that" sensitive?

Is the 3207 that sensitive to noise/wire length at the input?  I'd imagine I'd just get a sloppy output.

I see various 3207 based chorus schematics online where the input bias comes off a BJT emitter follower.  Could that be it?  Does it demand a particular bias or no output?  It's right around 4V as is.

Is the 3207 input more sensitive to frequencies then the sad512? I'd expect just a sloppy output but mn3207 based pedals do seem to have a much narrower input filter then the MXR does when values are plugged into the OKAWA online calculator.

In case I was being terrible at measuring I hooked up a amp + guitar to the whole setup and was still only getting a clean output.


Thoughts?  I'm sure it's something ridiculous but wanted to bounce ideas around.

Thanks!

[DrAlx]

3207 datasheet has graph of bias voltage against supply. For 9V supply, the bias needs to be somewhere around 5.5V.

I wouldn't worry about clock waveforms. If clock waveforms cross poorly then all FETs can be on at same time and signal will shoot through BBD without delay, but you will still get a signal out.
So lack of output signal is either bad bias voltage or bad/fake 3207. Where you get the BBD from?

[imJonWain]

I was planning to try a seperate bias supply, I'll try that.

They are from Cabintech so they should be real.

[Fender3D]

That's my schematic
If you search thoroughly right there where you found it, I posted a schematic adapted for MN3007, tested and working.
You may further adapt it for your 3207...

[imJonWain]

Ha, I didn't make the name connection!  That adapted schematic you posted works decently as is?  I tested it on the breadboard and noticed the astable waveform was farting out trying to drive the mn3207 clocks so I didn't bother.  It worked w/o input bias adjust?

[Scruffie]

Ha, I didn't make the name connection!  That adapted schematic you posted works decently as is?  I tested it on the breadboard and noticed the astable waveform was farting out trying to drive the mn3207 clocks so I didn't bother.  It worked w/o input bias adjust?

That was my finding too, it was nowhere near square enough even without load and gave a distorted output.

The micro chorus has a bias adjustment trimmer? You should jumper the 3 series diodes from the input op amp though, the Reticon series wanted a lower bias voltage.

[Fender3D]

You can't get away from bias adjustment if you want a decent performance, but that's the easiest setting in chorus and flanger...

http://www.wizardinside.it/foto/schemi/m_chorus_schem.pdf schematic
http://www.wizardinside.it/foto/schemi/m_chorus_PCB.pdf PCB

C14 and 16 might as well be 22pF

Unless clock isn't too much high in frequency, it works nice.
Lowering C14 and C16 helps...

[imJonWain]

Thanks for the input all!  Seems to have been a bias issue all along...  I had it too low at first and then too high.  I'm using the cool audio 3207 with VCC=9V and mine output garbage above ~4.6V bias and below 4V.  Fair enough.  XOR gate works fine too it seems.