Help needed with Neutron Filter

Started by TheLoneRoger, June 13, 2020, 06:35:05 AM

Previous topic - Next topic

TheLoneRoger

Sooo, I built a Neutron Filter using the PCB from General Guitar Gadgets, and after a few teething problems, I got it working, mainly due to stuff I found out here, so thanks.

Then I foolishly left it plugged in overnight, and in the morning when I tried it, it just squeaked and died. Turns out one of the batteries was fully discharged, but even after replacing both batteries with good ones I can get nothing out of it.
I've checked all the connections and components - all good - and frankly, if it had been working, and hadn't been touched overnight, then the only thing that could have happened is for a component to have failed, right?

What I get is nothing at all. Not so much as click when I connect the batteries.
My question is, where do I start with troubleshooting this?

When building it, the options/deviations I went for are:
VTL5C3 vactrols
Two battery power (I tried the single battery charge pump method, but didn't like the whine)
Replaced Rx with a 500k pot + 560R in series in order to get both 'up' and 'down' filtering working, as per threads here.
Couldn't get a 15uF NP cap for the output, so used 10uF and 4.7uF in parallel
Likewise 100k pot instead of the 150k
And I followed the second wiring diagram, i.e the one published separately rather than the one in the original PDF.

Any help greatly appreciated - for the brief time I had it working, I loved the sound this thing makes!

http://www.geofex.com/PCB_layouts/Layouts/neutronpub.pdf
http://www.generalguitargadgets.com/wp-content/uploads/neutron-tbp.pdf

Kevin Mitchell

If you think something could have blown or had gotten jostled you'll want to post voltages per our debugging procedure.
Here's some details from my build back in April that may help you;
Quote from: Kevin Mitchell on April 01, 2020, 10:30:49 AM
Finally got mine together after designing the template months ago.


Some quick notes on my build
-Homemade PCB with the GGG project files (most recent Neutron Documentation)
-Current wiring is a little different than the photo.
-2 unmatched (or didn't bother) VTL5C3 optos
-The 7660 caused some serious whine. I swapped it with a LT1054 - no trouble at all now.
-DP3T (on,on,on) toggle instead of the Mode rotary switch
-1MC for Gain
-1MB with a 180K resistor wired between lugs 1 & 2 for Peak, gives ~150K audio taper. Works well from what I can tell.
   -Geofex PDF shows both pots wired backwards
   -GGG shows only Peak wired backwards


Hope that is at least useful!
-KM
  • SUPPORTER
My apartment looks like an imploded RadioShack.

TheLoneRoger

Thanks, yes, I saw your post and it was that that persuaded me to replace Rx with a much bigger pot, which got it working initially, so thanks for that.

And I had intended to post some voltages, and here they are.
(Note, that I have now changed back to the single battery/charge pump method, and put a couple of chokes in places of the jumpers, as suggested in the text, in the hope that that will kill the whine - when I get it working)



Kevin Mitchell

#3
I had raised RX to crazy values in a desperate attempt to get my homemade optos to work. I could only dial in the up sweep or down sweep but not both. With the VTL5C3 I used a 10k trimpot and everything came together.

Check for shorts.
Post a picture of both sides of the PCB if you feel you should.

Do you have an audio probe? Trace the signal and see about finding a problem area. Hopefully it doesn't stop at the optos.

-KM
  • SUPPORTER
My apartment looks like an imploded RadioShack.

TheLoneRoger

Interesting - I had to go up to a 500k pot before I could get both 'up' & 'down' working with VTL5C3's - with a 10k pot it would only work in one direction.

I don't have an audio probe, but what's bothering me is the complete lack of any sound - like I say, not so much as a pop when I connect the batteries up. Feels like there should be a clue in there somewhere, but it's eluding me...

Kevin Mitchell

Photos and audio probe. You can use the output jack's wire - just unsolder it from the board - make sure ground is still connected. Follow the schematic you'll understand where the signal is going and where you may be having problems.

You can also remove the optocouplers and add an LED to see if your lfo is working at all (a visual of what we'd observe with an oscilloscope).

Test your optocouplers in case something had fried their LEDs.

-KM
  • SUPPORTER
My apartment looks like an imploded RadioShack.

TheLoneRoger

Well, this is embarrassing!

I'd used an old stomp switch, which I suspected might not be right, and when a replacement arrived, I wired it back in the wrong way!

Needless to say, it's all working now, though there is still that annoying whine, so I've ordered an LT1054, and I'm hoping that'll cure it. Might try your trick with the 1m pot/180k resistor for the peak control too.

Thanks for all your help - it was tracing the audio that led me to to realise what I'd done, so thanks again - I'm sure I've learned some valuable lessons in the process!

Kevin Mitchell

That's great! Glad you've nailed it down. Definitely making the right moves with the LT1054 and the improvised 150k pot.
Consider my notes on the mistakes in the wiring diagrams - if you hadn't already. I was a bit surprised that the documents haven't been edited by now.

-KM
  • SUPPORTER
My apartment looks like an imploded RadioShack.

TheLoneRoger

Thanks.
I can confirm that the LT1054 has eliminated the whine, and that the 1M/180k workaround for the Peak control is an improvement.

I'm also thinking of replacing the up/down DPDT with a 3PDT so I can switch in fixed resistors for RX instead of using a pot, having measured what seem to be optimal values on the pot, which in my case seems to be 500k for 'up' and about 115k for 'down' (+ 560R).

Just curious - what would be the advantage of using matched vactrols, and how would you go about matching them? Is it simply a matter of plotting a few input voltages against the measured output resistance?

Kevin Mitchell

My guess is that matching devices used as variable resistors for filter circuits all have the same principle. Think of it like a phase 90. The closer matched the jfets, the more pronounced the filter is. But when it comes to optos they're likely close enough to not notice drawbacks.

I think you should check out your used optos. If you and I are using the same devices then RX shouldn't be that off. There may be something else going on.

-KM
  • SUPPORTER
My apartment looks like an imploded RadioShack.