Fine tuning the Keth raper circuit

[BeyondStupidity]

Ok, so i built the circuit, it looked all pretty, i plugged it in and it didnt work . I tried all the stuff that i've seen to try to fix it and it didnt work then i figured out that i had accidentally grabbed a mono cable rather than a stereo :Plol once i got my stereo cable in it worked .....ok, but it still has issues....

Problem
The problem i have now is that the aux/rim piezo is triggering over the bow piezo. when i hit lightly right over the bow piezo it triggers lightly but if i try to hit it hard the rim piezo just triggers over the top of it.

What i've tried
On my TD-10 exp i cant change the piezo sensitivity separate from the other piezo.
I tried 1, 2, and 3 foam discs under the rim piezo but it didnt really make a difference.
I built the circuit with a 100k pot that changes the sensitivity a little but not enough.

Is there anything else i can do to fix it?

[John F]

Hi,
Long time lurker, first time poster here. The problem that you're having is that the circuit is too sensitive. I've played around with the Keith Raper circuit quite a bit while building my 3 zone ride, and I must say it's a very clever circuit that works quite well when you get it adjusted right. I'm not sure what version of the circuit you're using, but the ones I've seen used a 100K series resistor (or 100K pot) and a 1M resistor from base to ground. To decrease sensitivity you need to turn the pot to maximum resistance. If that doesn't do it, you need to change the value of the 1M resistor and try readjusting the pot. I would start with changing the 1M resistor to 100K, then 47K, then 22K. Adjust the pot with each change and see if you get the sensitivity you want. I used a circuit analysis program to look at the Keith Raper circuit (I'm an EE) and found that it is sensitive and tricky to adjust, especially using non-adjustable parts. I redesigned it to make it easier to adjust and to handle a very wide range of piezo output voltages just by adjusting the pot. See the attached file. I'm using this in my 3 zone ride and it works well. Good luck!

[BeyondStupidity]

Ok, thanks John. I'll have to run to Radio Shack and get some smaller resistors to plug into the circuit. I'll let you know in a day or two how its works.

Thanks again!

[BeyondStupidity]

I found some smaller resistors and messed around with that a little and its getting better, but its still pretty unstable. it triggers most of the time and i'm still working with the sensitivity. I'm not an EE so i have some questions about your modified circuit.

1) For you does it seem to isolate the piezos better?

2) Does the wire from the transistor and diode connect above or below the 100K resistor(it looks like its in the middle and i'm not sure how what would work)? Oh, oh, or is that a 100k pot with the wire connected to the middle peg?

[SuperPuss]

John,
You do realise that circuit will never work don't you? ..... Your transistors don't have circles around them. As you'd be aware, all transistors (and diodes) contain a certain amount of smoke which enables them to work correctly and the circles are there to keep the smoke in. Without the circles, the smoke gets out and once the smoke gets out of a semiconductor, it usually stops working

Seriously though, your mod is similar to the mod I did to that circuit. I still haven't built one yet but I did the mod for another member in another post...... oh, and I just realised mine won't work either because, although my transistor has it's circle, my diode doesn't!!!


edit..... I also just realised my circuit contains an invisible 2.2M resistor!! Oooops, forgot to remove that label! Doh!

[John F]

Hey SuperPuss, I like your modified KR circuit. Looks a lot like mine. Great minds think alike! If you double check my schematic, I'm using a 2N3904 instead of a BC337. The 2N3904 is a special smokeless transistor. With all the anti-smoking laws here in the US, I had to use a smokeless transistor

For BeyondStupidity - there are 2 ways to get the isolation between the bow and edge piezo. One is physical separation. Try to move the edge piezo close to the edge (duh) if it isn't already. Then move your bow piezo as far away as you can consistent with where you would play on the cymbal. On my ride, the edge and bow piezos are about 6" apart. The other way is to get the isolation via the circuit. You need to adjust it so that it doesn't trigger on your normal playing of the bow piezo, but will respond to a good hit at the edge. I'm using a TD3 and it senses velocity to determine bow or bell sounds and I can hit the bell pretty good without triggering the edge. It takes a good whack to trigger my edge sound. Also, I would recommend trying either my circuit or SuperPuss' circuit as I think these are both capable of easier adjustment over a wide range of piezos than the original circuit.

[SuperPuss]

Hey SuperPuss, I like your modified KR circuit. Looks a lot like mine. Great minds think alike! If you double check my schematic, I'm using a 2N3904 instead of a BC337. The 2N3904 is a special smokeless transistor. With all the anti-smoking laws here in the US, I had to use a smokeless transistor
.......

Doh!! Forgot about the anti smoking laws!! LOL
Same thing over here in Aus with anti smoking (which is a good thing IMHO). I might have to switch to the 2N series too.... after all, the BC series are ancient now (the BC stands for 'Before Christ')

A little useless trivia... the older Germanium transistors used much harsher chemicals in their operational smoke and this tended to eat away the casing and made them slowly lose their working properties. The old CK722 PNP's suffered a lot from this, especially the blue metal cased ones.
The AF series TO72 cases suffered smoke related cancer problems too. Their circles always developed cracks and these can be seen on old schematics along with coffee and solder flux stains

[John F]

A little useless trivia... the older Germanium transistors used much harsher chemicals in their operational smoke and this tended to eat away the casing and made them slowly lose their working properties. The old CK722 PNP's suffered a lot from this, especially the blue metal cased ones.

OMG, someone who's heard of a CK722! I actually bought a "grab bag" of these for about $1 when I was in college. This would have been in the mid 1970s.

[SuperPuss]

OMG, someone who's heard of a CK722! I actually bought a "grab bag" of these for about $1 when I was in college. This would have been in the mid 1970s.

Cool

My dealing with those were from my early days of experimenting with crystal set radios made from toilet rolls (for the aerial coil), germanuim diodes and those nice chunky metal tuning gangs that looked more like elaborate heatsinks!! I added a simple two stage amplifier to one of my radios using a pair of CK722's and thought it was wonderful.
I popped a few of these in my quest to design an amplifier capable of driving a small speaker though! all good fun though

[BeyondStupidity]

I just put one of your modified circuits together(its not working) and didnt realize that i have 2n4401 and you're circuit calls for 2n3904. Would that difference be enough to make the circuit not work? I checked the wiring a few times and i'm pretty sure i got it right. so that makes me think its the transistor.

[BeyondStupidity]

ok, i messed around with it more and got it working, but, its too quiet now.... lol

I'm guessing you use the same theory as last time and move up the value of the resistor that is connected below the 100k pot, from 10k to 21k to 47k to 100k and so on?

[SuperPuss]

ok, i messed around with it more and got it working, but, its too quiet now.... lol

I'm guessing you use the same theory as last time and move up the value of the resistor that is connected below the 100k pot, from 10k to 21k to 47k to 100k and so on?

Speaking on behalf of John here because his mod is similar to mine but yeah, changing that resistor should do the trick. You could also change the value of the pot to 250K or 500K and that will do the same thing.

[John F]

Agreed. You need to raise the total circuit resistance. Also, a 2N4401 should work OK. Most general purpose NPN transistors should work OK. I looked at a number of different ones with the circuit simulator software including 2N3904, BC337, 2N4401, 2N2222, and 2N2369. They all seemed to work about the same, only requiring a minor tweak of the pot.

[BeyondStupidity]

Ok! I think i have one that is working good! now i just have to make a few more of them and i'll be good! thanks for you're help and i'll let you know how it all works out.

[Sabre]

I couldn't find the specific parts outlined in the circuit, so I just used the NPN transistor and silicon switching diode that were available, and the circuit works very well for me, so as John points out, knocking yourself out to find the actual parts really isn't necessary.

I didn't use a trim pot because I couldn't get one in the right resistance range, but I'm going to build another board and try a variety of different resistances just to see if I can tweak a little more volume out of mine as well. I'll be interested to see what your testing reveals BS.