Not-so Mesmerizing Mesmerizer

[zgrav]

Hmmm. I have a semi-working build for the Mesmerizer. The sound through the board without dialing in the effect is OK. The volume pot works. The depth control and rate control change the intensity and speed of the flashing LEDs. I am using an old XR2206 IC that I know is good, and it is working. I get a sine wave output of about a 2V range at pin 2 of the XR2206 that does not modulate the sound when I adjust the rate and depth controls. When I open the connection between pins 13 and 14, the output at pin 2 changes to a triangle wave with a higher voltage output. I do get a pulsing in the sound from the triangle wave, but only at its peak signal level. It also only makes the sound louder and not quieter than the unmodulated signal.

I double checked part values using my DMM for all resistors before they were installed, and confirmed caps matched the values on the build document and what was printed on the board when they were soldered into place. I have gone back over the solder connections and reflowed them from each side of the board.

For the active parts of the circuit, I have swapped out the TL072 and also subbed the 2N5457s with J201s and NTE458s without success. I have not tried to switch out or sub the 2n5458 transistors yet, but that is next on the list.

Any suggestions on isolating things to check is appreciated.

 

[Chuck D. Bones]

Q5 & Q6 are used as voltage-controlled resistors in this circuit. To do that, their gate-source voltage needs to swing from near zero to near Vp. Try checking the voltage on Q5 & Q6 source (center) leads. Should be around 1.2VDC. Try measuring Vp on the 2N5457s. If Vp is at the high-end of the range (-6V) then it's possible that Q5 and Q6 are spending all of their time in the low-resistance end of their range. If you have Q5 & Q6 in sockets, try pulling one out while you have music playing thru the pedal. If the sound changes, then that JFET is basically stuck on. If that turns out to be the case, then it's time to start screening JFETs.

 

[Robert]

If that turns out to be the case, then it's time to start screening JFETs.

Aside from the recently discovered XR2206 issue (which I do not think is the problem here), this is the other common problem I'm aware of.

I've been bitten by this a couple times myself. Definitely suspect Q5 and Q6.

 

[Chuck D. Bones]

JFETs seem to be a common theme on most of the Troubleshooting posts of late. They're not always plug-n-play.

 

[zgrav]

Thanks to both of you. I did get a couple of the cheap XR2206s from Amazon, and they were bad as expected. I am returning them through Amazon along with posting a critical review.

 

[stevtron]

Mesmerizer doesn't work

No tremolo efx, the XR2206 LFO is veeery slow, even with max potentiometer position … What are the correct values for VDD,VCC and VREF ? I checked every part and soldering . It seems the charge pump does not work correctly. No 18V , no matter where I measure … On Vcc only 1.67 V, on Pin4 of...

forum.pedalpcb.com

... had some problems too, maybe that helps .

 

[zgrav]

What voltages should I be seeing on the charge pump? I have 8.69 on pins 1 and 8, 6.4V on pin 7, 4.07V on pin 6, and -4.02V on pin 5.

 

[Chuck D. Bones]

Pins 5 and 7 are AC waveforms, there is no point in measuring them unless you have a 'scope. Pin 6 is a control input; the voltage on that pin is not specified in the datasheet. What voltage do you get on the cathode of D7? If it's between +16.5V and +17V, then the charge pump is working.

 

[zgrav]

Thanks for that info. VCC is a solid 16.76, VDD 8.68, and VREF 4.3. Readings for the 6 transistors are as follows:
Q1: 1: 10.28, 2: 2.06, 3: 0.0
Q2: 1: 16.76, 2: 10.28, 3: 7.99
Q3: 1: 10.15, 2: 2.3, 3: 0
Q4: 1: 16.76, 2: 10.15, 3: 7.99
Q5: 1: 0.65, 2: 1.26, 3: 0.47mV (with socket empty, varies from +20mV to -35mV )
Q5: 1: 1.25, 2: 1.26, 3: 0 (with socket empty, varies from +30mV to -25mV)

 

[Chuck D. Bones]

Everything looks good except the voltages on Q5 and Q6. Pins 1 & 2 should both be about 1.25V. Pin 3 should swing +/- 600mV when Depth is dimed. Hard to measure pin 3 accurately with a DMM, better to use a scope. You should get the same readings for Q5 & Q6 with the 2N5457s in or out of the sockets. Are you sure about the values of R14, R15, C10 & C11? Did you measure Vp for Q5 & Q6?

 

[zgrav]

CDB -- thanks for the reply. I'll check it out and see.

 

[zgrav]

Update -- part values were correct but I reflowed the connections for the resistors and swapped the two 1 uF capacitors to change them from tantalum to mlcc. The schematic does not show those as polarized, but the template on the PCB is for polarized electrolytics.

After that I was getting comparable voltage readings for Q5 and Q6, and I tried a bunch of different JFETs in the sockets. I found a couple of J201s and one 2n5457 that worked well, a few others that worked a little (including some nte458). At this point I think the pedal sounds like it is supposed to, and the rate and depth controls both have good ranges.

I did not measure these for Vp, but will do so. Is this an effective way to do it? https://viva-analog.com/jfet-characterization-technique-using-only-9v-battery-and-dmm/

 

[Chuck D. Bones]

Yes, I've tried that method and it works well. I play it safe and put a 1K resistor in series with the battery for both measurements. That way, there is no chance of damaging the FET if I get the leads swapped by accident.

If you're talking about C10 & C11, they're shown as polarized on the schematic I have.

 

[zgrav]

I guess the tantalum caps would have been fine for c10 and c11, but I was uncertain since the schematic and parts list just referred to 1 uF caps. I had followed the PCB when I put the parts in, but had second thoughts when things were not working.