Neutron layout differences

[FarrisGoldstein]

There are two PDFs at geofex for the neutron filter.

This is the full doc, with schematic, layout, and instructions:
http://geofex.com/pcb_layouts/layouts/neutronpub.pdf

There's also a "wiring diagram for the schematic-impaired":
http://geofex.com/FX_images/neutron-tbp.pdf

However, the layout on the second one appears to be different/newer than the one in the full doc. Does anybody know about the differences?

I etched and drilled a board using the first doc, and "option 1", with 2 H11F3s instead of the LED circuit. I also didn't put the Q1 circuit in there (perhaps mistakenly?) because I was just planning to switch the power from the two batteries off-board. However, I'm having problems, and thought the simpler power wiring might help me eliminate some points of failure. But the placement of R20 & Q1 are different between the two layouts.

[waltk]

Yes, there are two slightly different documents/layouts out there.  The main document for both of them say "GEO Design 117" and "Version 1.01d".  If you want to see the version of the main document that matches the wiring diagram on GEOFEX, you can find it on the General Guitar Gadgets site.

Both of the layouts are electrically identical.

If you are planning to use two batteries for the power supply (exclusively), you would hook them up to the three pads above U1 in the main document, and exclude all the "optional power supply" parts from your build - U4, R20, Q1, C15, C16, C17, and C18.

To use batteries, you would connect them in series - with one of the negative terminals connected to the positive terminal of the other battery.  The open positive terminal of the battery-pair is the +9 supply to the board, the open negative terminal is the -9 supply to the board.  The junction of the two batteries is the ground on the board.  You will also then want to switch power in whatever way you think is best.

I'm sure R.G. will correct me if any of this is wrong...

-Walt

[FarrisGoldstein]

Thanks. That's exactly how I have it wired up. Do I need to bridge/jumper anything in the Q1/R20 circuit since I'm not using it?

I'm not using any power switching at the moment, just disconnecting my batteries when I'm not testing: The tip/signal of my input jack is wired directly to lug#3 on my gain pot, and tip/signal of my output jack is wired directly to R21 (which is wired to C14). I'm not using the bypass pad on the board, so C14 is wired directly to the stationary lug on rotary switch. Both jacks are grounded.

I'm using Option 1 for the photocell bit: Two H11F3s on a 14-pin socket (with the center pin pair on the socket unpopulated).

I'm not sure about my choice of pots, and might need some resistors (or wiring differently than in the doc) for it to work at all, let alone sound decent:
Gain pot: C1M (input tip at lug 3, center lug to pad 15 on the board)
Peak pot: C150k (lug 1 to pad 11 on the board, lug 3 to the center position on the rotary switch)
RX: Replaced with a B10k pot (lugs 1 and 2 wired to where RX would be soldered), based on this guy's build: http://www.youtube.com/watch?v=hoK76PDomeE

The result: My gain knob is backward (increases gain counter-clockwise), and only effective for a few degrees on the far CC end. If I turn it any further than a few degrees to the right, the signal drops to nothing, and within that effective range on the far left, it is far too distorted.

With some switch positions, I can definitely hear a tiny bit of sweep underneath all the distortion, but until I figure out why my gain is so high, I don't feel like it would be very productive to start troubleshooting anything else.

Any help would be appreciated, even if it's just to say "Dude, that's a mess, abandon hope."

[waltk]

Dude, you're almost there.  Don't give up.

[FarrisGoldstein]

I'm a wee bit stumped. Won't have time to play with it until later in the week, but I guess I'll just start checking voltages and component values. Any tips on the most likely cause of so much distortion?

[Mustachio]

I built this about a month or so ago. I used the GGG layout to etch my board and populate from.

It seemed to me getting the opto section of the circuit was touchy. I ended up using a much larger then suggested resistor. I ended up testing with a bunch of different photocell and led combinations. Ill open mine up later today and give you my exact values. I did a lot of reading on what other people used, didn't read much good about the H11F3's in this circuit. One guy used KE10720 photocells (thaishine*ebay-tayda) and a diffused LED, put them inside a fuse body to hold it all together and filled it with a clear epoxy resin. I believe the epoxy he used makes the light spread out more like a diffused led and maybe a more even distribution of light to the photocells. And much like the original units made. I made something similar with the same photocells because I had them on hand already. But I did test out a bunch of different ones and also Vtl5c7's I didn't have any 3's on hand which I believe would be the easiest option. Fine tuning this section is critical before you get it to sound correct.

I used Similar pot values as you. And I wired them backwards like you as well (haha) Maybe it was the way we read the layout. I didn't use a c150K pot because I couldn't find one I believe I settled on a 200K. For what its worth test with the switches set up so the range is low the direction of the sweep is up and its on low pass, gain low around first quarter and peak around half way or so. Now adjust that resistor or pot RX to about 120K see how it is there. I had similar outcome using different photocells and resistors.

I think your power side is fine and all you need to do is fine tune the optos. I don't know if this circuit will eat batteries fast , I'd imagine it would but I don't know. I would suggest adding the charge pump section of the circuit and running it off a normal 9v adapter. The trick is to bridge pins 1 and 8 on the 7660scpa and its silent!

If  you end up making your own opto parts like I did I suggest trying out some wide angle straw hat style leds or something thats diffused and has a large viewing angle, after testing out diff types with mine i get the feeling the more you can flood the area that the light is picked up in the better not necessarily brightness power but field of view.

good luck! 

[FarrisGoldstein]

Wow, that's encouraging, thanks.

Mustachio: What did you end up using for RX? A fixed resistor, or a pot? What value?

[FarrisGoldstein]

Oh, and one more question:

I ended up using a much larger then suggested resistor.

Are you referring to the photoresistors, or some other resistor.

If you do end up opening yours up and giving me the values for your optional components, we'll be BFFs.

[FarrisGoldstein]

While I was out running an errand, on a whim I stopped at RatShack and picked up two of their little candy bags full of 5 assorted photoresistors. Found two that seemed to be ~300k dark, and, goldang, what a difference. Still way too much breakup for my taste, but I'm getting closer.

After work I'll see if I can do some more rigorous testing of these LDRs and pick two better ones (and maybe some different LEDs, I'm using a cheap green one), then either tape 'em up or epoxy them. Then if I can get this gain down to an acceptable level, I just need to swap some lugs and put her in a box.

[Mustachio]

I believe I used a 120K , I socketed the resistor labeled RX that adjusted the led/ldr part of the circuit . I think it actually adjusts how bright the led side gets. I used a pot and after i got it where I wanted I took it out and measured with a dmm and popped in a resistor in that spot. Ill go open it now and report back in a few

[FarrisGoldstein]

You're right, RX adjusts the brightness of the LED, which is why a pot in its place acts as sort of a "sweep" knob by changing the peak light hitting the LDRs.

I think I've found a workable set of LED/LDRs, but there is still more grit/distortion on the signal than there should be, so I think I might have some problems elsewhere in the circuit. I can get rid of the distortion by turning down the gain to JUST above the point where I lose all audio, but then it seems there's not enough going through the circuit to drive the LEDs and get any effect to speak of.

More evidence that I have screwed something else up: If I flip the range switch to the on side that's actually wired up, then the circuit acts as kind of a gate: until the LED hits a certain brightness (variable, thanks to RX and the LDRs), the signal is silent, or vice versa if I flip the direction switch. So if I really bang the instrument, I'll get *some* noise, no matter how bright the LED starts at.

Last, but not least, with the rotary switch in the middle (Bandpass) position, none of the pots/switches (except for gain) appear to have any effect on the signal at all, and there is no filter effect, just gain.

[Mustachio]

Ok down in the studio now with it. I used a 120K resistor for RX I remember having similar issues as you untill i upped that resistor , i used a 10K pot at first and it was problems. so I moved up to a 500K pot and settled around 120K . I see your using a 10K pot try just grabbing a 120K resistor or what ever you have on hand or a pot and up that till you get it sounding good. also do the testing in the dark if your photocells are exposed. ill try to post a pic of the ldr set up I made.

[Mustachio]

Any luck with a larger resistor or pot ?

here's how I made my ldr set up

[FarrisGoldstein]

I haven't had a chance to try it yet. I'm a tad skeptical, as I don't see how changing that resistor would have an effect on the gain/breakup of the circuit.

Thanks for the photo, that's pretty. I still have the 14-pin socket in my opto portion (since I was previously using the H11F3s). It would probably be a whole lot easier if I replaced them with sockets like yours.

[Mustachio]

Im pretty sure mine acted like yours till I made that resistor higher then the 10K pot I used at first just like you did. And since its socketed its easy to test. I used a 500K pot then measured it when i found the sweet spot around 120K I think it was sweeter around 140K but I didnt have a resistor on hand at that range.

I used the inline cuttable sip sockets from thaishine (ebay USA)/ tayda .

I remember reading a lot about the ldr set ups from other people and I think RG talked about how the old leds had different forward voltage and current much lower then todays. I used a high bright amber led. It works and it works well but truthfully I think it could be better. From all the diff lil tests I was trying I have a feeling that a diffused LED with lower voltage and current and a wider field of view would do best. I seen one post on here where a guy made his inside a fuse body and it worked out great. He filled the inside with a fast setting clear epoxy and I have a feeling that spreads the light out better and floods the photocells with even light compared to more of a directional LED. I'm gonna experiment more since that section is socketed in my build even tho I really like how mine sounds now.

But really its worth trying a larger resistor there, Or if you wanna drop the money on vtl5c3's every one seems to have good luck with them and I think the under 10K resistor would work with them. Smallbear has some new vtl5c3 equvilant for cheaper.

Also my build had the same problems with the BP and HP and switches they seemed to do nothing almost except give a higher squeal untill I fixed up the LDR side of the circuit . the high pitch sound was still there untill I bridged pins 1 and 8 on the charge pump IC (7660SCPA). The way mine is now Its loud when I increase the gain about 30% is where I get the best womp and Its a bit over unity level. All my controls work now as well.

So really please just try to change that RX higher, now I'm dying to know if it will fix you! haha good luck man . (Test it in the dark so the ldr doesnt get light bleed)

[FarrisGoldstein]

I put a 500k pot on RX. While I can get slightly "better" results now, even with zero light getting to the LDRs, the sound is broken up too much at any gain level sufficient to drive a dynamic filter.

I might have to play more with the LDRs (or hunt down some VTL5c3s next payday) and different LEDs. I have a ton of assorted LEDs, but only ~12 LDRs to play with.

[FarrisGoldstein]

A couple of other things for me to think about next time I have time to play with this:

1) What's up with my gain pot? Do I really have it wired wrong, or is there something wrong with my input resistors/caps? Even if I reverse it, it still doesn't make sense to me that, beyond the first (or last) ~20%, the signal goes completely dead.
2) The ever-present distortion sounds suspiciously clipped/fuzzy, like there are some diodes getting in the way.

[Mustachio]

Yeah the signal dropping totally after you turn the pot up to high is weird mine doesn't do that . Just thinking about it do you have the non/bi polar 15uf cap and resistor at the end i think its coming off the rotary switch and to the output. Just going off memory I think RG said it was sort of a protection on the output, and I'm wondering if the signal is overloading at the end.

and check out this thread

http://www.diystompboxes.com/smfforum/index.php?topic=95269.0

and this

http://www.diystompboxes.com/smfforum/index.php?topic=14219.0

[FarrisGoldstein]

Yep, I have the 560ohm resistor and 15uF cap between the output lug and the selector lug of the rotary switch.

I'll update later on my results with swapping the gain pot, LEDs, and LDRs.

[FarrisGoldstein]

After yanking and replacing my LED a bunch of times, it now appears that my LED does not sweep/change when I play, it just stays illuminated at whatever brightness RX currently (pun intended) allows. I must have shorted or knocked something loose.

I have found, however, what I think are two decent LDRs. With an external light source, the wah/quack sounds awesome if I manually lighten & darken the LDRs with a hand or a piece of paper.