Noise Ensemble of meekness

[moid]

Hello everyone

Imagine it's very late December, not quite New Year's Eve, and you've just survived four solid days of close encounters with members of your family who you normally try to only meet once a year because you have to (you know those types of people - they don't like guitars, they are tedious as hell, and are also rather stupid, but think they are clever). You're in the mood for making an unholy noise to somehow make up for four fairly awful days of irritation and having to be nice to people who you're related to... and then you find this schematic for a Noise Ensemble
http://www.diystompboxes.com/smfforum/index.php?topic=87493.0




It seems like just the thing to make some bizarre blasts of mayhem, so you breadboard it and it sounds good (or evil, which is closer to what I wanted) so I then decided to make a veroboard version, which is usually the point in any plan for me that goes wrong and of course it doesn't work (no sound goes through it)... However my daytime job then goes crazy and for the next three months the broken veroboard lies around gathering dust until today when I decided, hey, maybe testing the circuit with a DMM for voltage readings might be a good idea (don't say I don't learn anything!) and from that I discover some interesting (and depressing) information. Firstly that I'd put the 78L05 in the socket the wrong way round, so after I'd fixed that issue I discovered my second problem which was that due to me not grounding things properly at one point the PT2399 chip had 9 volts going into pin 1. Am I right in thinking that I should consider that chip to be fried now? I don't have any definite working spares to swap out, because at one point in the middle of fixing the 78L05 and before using the DMM to check the PT2399, I swapped my spare PT2399 into the circuit thinking that the circuit would roar into life and that spare PT2399 also got 9V to pin 1. Laugh? I almost punched myself.

So am I the proud owner of two dead PT2399's? I need to order some parts so I think I'll add a couple more PT2399 to the list, but I wondered if there was an easy way of testing whether a PT2399 still worked without building a circuit correctly and placing it in the circuit?

Something tells me that there'll be more wrong with the circuit anyway, so I'll add more hilarious anecdotes as I get the time to poke it with small metal objects.

[bluebunny]

I would vote for "fried".    You might want to check voltages in the empty socket before committing a fresh PT2399 to its new home.  (BTW, for future reference: 78Lnn is OGI, 78nn is IGO reading left-to-right from the front.)

[anotherjim]

As you say you didn't ground things properly, there may have been 9v on pin1, but was there 0v on pins 3 & 4? Even so, it might  only have locked the chip up by injecting static charge into it's inner workings. MOS devices are capable of this as the chips contain highly insulated areas that can hold static voltage for a long time. Like a vintage car that won't start with a wet distributor, it might if you can somehow dry it out.

Put the PT2399's on something conductive that all the pins are touching - aluminium foil, anti-static foam etc. Leave it for a while. If it's static locked, this can bleed off the charge and give you a working chip. Not promising it'll work every time, but it can.

[moid]

Thanks bluebunny and anotherjim Well some sucesses and more questions (kind of curiouser and curiouser as Alice would've put it). First of all anotherjim's trick with the PT2399 and aluminium foil worked! (well on one of the PT2399 chips anyway). So one of them makes the circuit emit an unearthly howling oscillating din, the other does nada, so I assume that one is fried and the other is OK? I say this because I've now got some more questions about the circuit (or my build of it anyway). At present the pot marked Din seems to be a blend between wet and dry signal, and the pot marked Shambles modifies the length of the repeats / the craziness of the results. Does that look likely? The reason I'm asking is that I'm getting constant feedback through the circuit if I play guitar or if I try line level inputs via a buddha machine. They both have a different result though.

Guitar - constant oscillating feedback that doesn't really seem to be modified much by playing - I can blend it in via the Din knob (I had to use a 20K pot instead of a 22K pot because I didn't have any 22K pots, could this be the issue?) but after only a short turn of the pot it overwhelms the guitar signal (which stays clean) and so I can't hear what I'm playing at all - regardless of what notes I play the noise is the same. Adjusting the Shambles knob causes the noise to alter pitch, tone, length of repeat etc but strumming the guitar doesn't appear to modify it much or at all. Also it never fades away which is what I need it to do, even if it takes 30 seconds to do so In the demos of this pedal the guitar notes sound like a weird fuzz with other sounds happening around it and they do fade away eventually which is what I'm after.

Buddha Machine (line level) - playing back a sample from this into the pedal results in semi clean sample playing through, and as the sample fades down in the volume the feedback rises to replace it, but this feedback does seem to be modulated by the audio from the Buddha Machine - it also never goes away when the audio input stays silent for a long time or is switched off - if there is power in the circuit then it will oscillate constantly.


Is this what the circuit is supposed to do? Or is my PT2399 chip only semi working perhaps? I have two new ones hopefully arriving tomorrow so I will swap those in to see if it is the chip, but does anyone have any ideas / suggestions?

Maybe I need a louder signal from the guitar to go into the pedal? I'd still like the effect to fade away over time though, is there any way to do this? Should there be a knob to control feedback length perhaps, or do one of the resistors do this? I'm guessing maybe R3 in the vero layout because it seems to connect the feedback / repeats with the duration of the repeats (I may have this wrong, I'm guessing by looking at more 'normal' schematics for the PT2399 chip)

I noticed that if I touched any of the other power jacks on the daisy chain I'm powering the circuit with I could modulate the pitch of the feedback... so possibly because the circuit isn't wired into a box (crocodile clips are holding the power and ground cables together at present) it will behave weirdly?

Anyone's thoughts / dreams / curses are welcome and encouraged at this point, thanks!

[anotherjim]

Another PT2399 rescued! My work is done! ... err, no it isn't...


Blend is fixed - those 10k resistors around pins 13 - 15. Din pot is just a volume control from output pin14.
Suggest you look at the data sheet for the Pt2399 to see the block diagram of where things go in and out.

Shambles pot is the delay time setting. Increasing resistance of the pot means longer -  and lower quality delayed sound. That also gets wobbled by the signal from the input via a 47k resistor.

Input depends on a passive guitar really and you should be able to control things a bit using it's volume and tone knobs. You might get better results from non-guitar sources with a 100k variable series resistance to the input.

[moid]

Good news! (for women) as Mauro Pawlowski and the Evil Superstars used to sing. The constant feedback turned out to be poor grounding on my part (all the grounds were held together by crocodile clips) - I soldered them all to the power jack properly and the pedal behaves happily (or splutteringly well). All is well. Or is it?

For my next weird question (small drum roll, sense of tension in the audience is required) I would like to stick both the Mr Black Tiger Boost I recently built and the Noise Ensemble in the same stompbox so that I can put a blend knob between them to blend between a loud clean signal and a loud noisy as hell chaos signal. I'm fine with how to do this (I think). but what I wanted to know is can I make both circuits share the same power jack (and therefore also the same ground) - is this a good idea or a bad one? It saves me only having to use one power jack (yay) but i have no idea if is likely to violate some ancient law of electrical engineering and be the effects pedal equivalent of crossing the streams in Ghostbusters. Any guesses? Or do I just try it and hope nothing explodes?

[EBK]

is this a good idea or a bad one? It saves me only having to use one power jack (yay) but i have no idea if is likely to violate some ancient law of electrical engineering and be the effects pedal equivalent of crossing the streams in Ghostbusters. Any guesses? Or do I just try it and hope nothing explodes?

I personally thought stacking op amps was the crossing the streams equivalent: http://www.diystompboxes.com/smfforum/index.php?topic=116379.msg1077244#msg1077244 

I'll try to give a serious reply if I can find time later....

Here's a semi-serious answer in the meantime: it depends.  Electrically, different pedals can generally share the same supply, but it depends on how "chatty" each pedal is and how "chat resistant" it's opposite is.  Op amps are generally chat resistant, while discrete transistors are super social and have ADHD sometimes.

[moid]

Thanks for the useful advice. Well I couldn't wait, could I? It was like having one of those big red buttons with do not touch written on it... so I crossed the streams The universe didn't reassemble itself, and the pedal didn't explode, but when I combined this brilliant idea with a couple of other cables and some switches I found I could add some hideously high pitched feedback, or low grumbling feedback, both of which removed any signal. Whoops. They were cool for short periods of time, but they need to fade away, and I guess the issue with feedback loops is they don't ever fade away?

So those will be coming out tomorrow, because I have one mod that did work and that makes a nice effect (subtle more chorus like distortion) and I can now blend between clean Tiger Boost and Noisy Noise Ensemble. I'll draw up a new vero layout when I've got this finished. Thanks again everyone, this pedal sounds pretty good to me (maybe not as chaotic as I expected, but rather harmoniously musical in places... or maybe that's my sense of hearing being somewhat nuked!).

[slashandburn]

Cool. I built one of these a while back and loved it, even though I couldn't overcome the  note decay issue.  Always meant to revisit it and try to iron that out.

Sounds like the gates of hell. Low parts count. Cool circuit!

[moid]

I may have come up with a solution of sorts for the note decay issue (sputtering out too fast I assume you mean?). I added an SPST with a cable from the audio in of the footswitch to Lug 3 of the Shambles knob. When the switch is off you have the normal Noise Ensemble. With the switch on it seems to extend the length of the notes and smooths the noise effects out a bit so they are more like a harmonic chorus fuzz sound! Sounds good to me anyway. I don't know why this works I just tried randomly connecting points on the circuit to see if anything interesting happened and it did

Here's my layout - I really prefer it with about 30% of the clean boost mixed in

[EBK]

I don't know why this works I just tried randomly connecting points on the circuit to see if anything interesting happened and it did

You are insane... and lucky.   
Happy to hear that it works and that you took notes.

[moid]

Ahhh more craziness for you all After playing the pedal through a friend's amps, he suggested sticking a booster pedal in between the guitar and the pedal I've just made - this sounded great, the effect was even stronger and wilder. OK so I thought I'll take this pedal home and so some surgery and put in an SPDT switch with audio from the Tiger Boost to the central lug, then one of outer lugs connected back to my blend pot, and the other outer lug connected to the input of the noise ensemble. When the switch is in one direction for the original pedal sound Boost and Noise blended by a pot, everything is fine. When the switch is in the direction of Boost into Noise pedal, I get nothing by motorboating rumbling as if there was a bad ground connection somewhere and no sound. Both circuits work happily independently of each other and will mix together via a pot, but I can't put the boost into the noise...any guesses why?

I still have an audio signal from the footswitch going to the Noise Ensemble, with a cable into the same hole that connects via the new switch to the Boost. Could this be the issue? If so I don't think I can get around it. If anyone has any bright ideas please let me know or this new switch will be coming out tomorrow morning, and I'll just have to leave it as it is (still a good pedal though). Thanks.

[moid]

OK after some research I thought that perhaps my problem might be fixed by using a buffer, because presumably the Tiger Boost puts out high impedance signal, and the Noise Ensemble (being a digital chip) wants a low impedance signal. So I built one of the buffers from the AMZ website onto breadboard. I have tried a signal through the buffer to the amplifier to check that sound went through it (that worked, although I just realised I used a low impedance signal, bugger, let me guess, I should've used a guitar?). When I connected the breadboard to the spaghetti zone inside my pedal so that the signal from the Boost went into the buffer and then into the Noise Ensemble I got some strange results. If I used a low impedance input (my buddha machine) I got fantastically crazy glitchy, warped results! It's awesome OK sometimes it does silence or bursts of pure noise, but it's fun. If I plug a guitar in, all I get is motorboating noise and no guitar at all. I tried increasing the resistor value that goes to VR in the below circuit as recommended by the designer from 1M to 2M2 and finally 5M5 and there's no discernible difference The designer says to go to 10M or higher for hot humbucker pickups (I was testing single coil, I don't think they are that fierce). I have tried using separate power and ground to the breadboard for the buffer circuit and also tried using the power socket of the pedal itself for power and ground - there's no difference I could hear.

Is 5M5 still to little to hear any signal? I haven't got anything larger and I made that by running a 2M2 and a 3M3 in series. I'll need to buy larger resistors to get any bigger, but I'm wondering if maybe that will never work with this circuit? Is there a way to use a DMM or something else to measure impedance so that I could see if the signal going in and out of the buffer is behaving as expected (high going in, and low coming out?) I looked at some websites but they recommended using an oscilloscope and I don't have one. If that's the only way to do it, I guess I'll buy some 10M resistors and hope something happens.

I used the following two circuits (first is the power, second is the buffer) I used a 2n5457 as the JFET.


Any thoughts would be nice to hear, thanks!

[anotherjim]

The input of the noise ensemble is wired as a comparator. It produces a square wave from the input signal into the delay chip. The way the input is biased means any little bit of noise going in will trigger it. A buffer added to the input is only making it more sensitive to noise. That might well be your motorboating effect. The noise could be radiated pickup from the output wiring.
Is this thing in a screened box?
Does every piece of circuit have a capacitor on it's power supply input?

[moid]

Yes, I built it in a hammond enclosure (aluminium). Both circuits have a capacitor on their power input (Noise Ensemble has 100uF, the Tiger Boost has 47uF).  I had the motorboating issue whether I had a buffer or not. OK so it seems that it's not possible to stick the Tiger Boost output into the Ensemble's input? The weird thing is if I plug my guitar into a different effects pedal and then plug that into the Noise Ensemble it is fine with that; it just refuses to allow the Tiger Boost to work in series. Weird. Thanks for replying though - should I give up at this point? The rest of the circuit works, I was just hoping for more (I didn't breadboard this idea before I built the circuit - the idea came to me after building).

What part of the Noise Ensemble is the comparator? If I changed some values of whatever part that is, would that make a difference? Does it matter that the Tiger Boost is an Op-Amp boost and not a transistor boost?

[anotherjim]

The 2399 input pin16. Actually an op-amp. The -in pin is 16, the +in is inside the chip, wired to the 2.5v REF at pin2. The op-amp output goes inside to the delay AND to pin15.
You have the input connected to pin2 via 100k. So both inputs of the op-amp have 2.5v. The op-amp has a sh!t load of gain and will hugely boost any difference between the 2 input pins. It's on a hair trigger.
You might calm things down with some negative feedback. Try 1M between pins 15 and 16. That will reduce the sensitivity and a guitar may have to be put through some other effect or boost first.

[moid]

Thanks for explaining that. I'll try that at the weekend. If a 1M resistor works, then I might be able to replace it with a pot and then vary the amount of craziness in the circuit

[moid]

It's Friday night and Friday night is only good for one thing? Yes, audio experimentation of course! (What else were you thinking?)

A 1m2 resistor (don't seem to have any 1m resistors here) connecting pin 15 and 16 created a constant clock noise sort of tick, rather loud, but did allow guitar audio through with a really interesting vibrato tremolo effect - sounds good, except for the a clock noise / loud clicking which is almost as loud as the signal. Changing the settings of any of the pots did not remove the clicking, but I did discover that when the Boost pot is set to almost nothing / completely off I appear to have a monophonic synthesiser (and a damn loud one that made me jump and chuck my headphone off pretty quick - some manly cursing may have been involved as well)

While playing with the resistor (which is sort of clamped in place with various crocodile clips, insulated ones) at one point one end of it bridged pin 15 and pin 14, the other end connected just to pin 16. This created an interesting freeze hold oscillation effect - whatever note was played on the guitar was held forever (well I got bored after 1 minute with no obvious drop in volume). As soon as a new note is played, the oscillation pad effect changes to something very similar to the new note. This could be potentially very cool, but does need to fade eventually I think - there's no way of stopping it - like a Freeze pedal or Superego if it's stuck on the same sample and you can play over it, but it never goes away.

Next I tried to fit a 2m2 resistor, but in the process of attempting to bend some hook shapes into it's wire, it pinged out of my hand and magically became invisible on the kitchen floor. Where it will no doubt remain hidden until tomorrow when I'm bound to walk over it barefoot and re discover it. Breakfast could be exciting tomorrow.

I had a 3m3 so I fitted that instead. Now the clock noise is much much faster, same volume though. The vibrato tremelo is almost gone and the guitar plays through the Tiger Boost, then into the Noise Ensamble and seems to be louder (of course) and have slightly longer decay than without it. I think what I need is something like a 2M pot if such a thing exists - 1M is really cool but very noisy, 3m3 is very similar to the original sound, so something inbetween might work.

Does anyone have any idea how to remove the clock noise / ticking, because that is rather annoying. I did try connecting the resistor to ground as well to see if that helped, but that cut all signal fullstop... maybe a capacitor to ground coming off the resistor? I'm guessing here (ok, flailing wildly about in my own special way!)

Thanks

[anotherjim]

Try a capacitor across pin 15 & 16 with or in place of a resistor. Might tame the oscillations  - Cap probably small, 1nF down to 33p.

[moid]

Thanks anotherjim - I tried a variety of capacitors today and all the small ones didn't change anything (47pF to 1nF), so I tried larger capacitors and somewhere around 10uF the entire signal disappeared... no noise, but no guitar either. I tried smaller capacitors in series with resistors in case that helped, but it didn't seem to make any obvious difference. I did read a thread on the tagboard forum about clock noise caused by LFOs and ways to insulate a circuit to avoid that, but I this circuit doesn't have an LFO so it shouldn't be affected in the same way?

http://guitar-fx-layouts.42897.x6.nabble.com/Lets-talk-about-ticking-in-modulation-and-modulated-delay-pedals-td37219.html

If you want to read it. I think I'll just close the box now, the rest of the circuit (the original bits, plus the blend between them) works really well and I'm happy with it, so I might now try to get on with that four way splitter I was asking questions about in another thread. At least this build didn't get consigned to my box of 'why on earth did this not work circuits!' Thanks again