col

I have the general idea about "voltage inverters" but it will draw what is required to power it, no power is free and the current is the real draw that flattens the batteries. I am amazed at the little power that you are apparently using and would love to see some detail (either here or email me). I can see that the lm386 can draw very little power if run at a low gain (pins 1&8 open) but I really do need that power to get the full effect. My preamp is giving plenty of juice to the lm386 and is based on a tested design that was built to interface with it. That said, battery consumption isn't that bad, but for on board electronics, something that lasts a few days or weeks of intermitent use, will still require easy access and replacement. Assuming that I've not made some critical mistake in my measurements etc. (big assumption), you should be able to get a similar battery life using a similar circuit. The trick is in conditioning the signal prior to feeding it into the LM386. Unless you have some hardware on your axe that dampens the strings, or some other impediment, or your drive is very different, it should work. Of course, its possible that you are going for a different end result, and that you would prefer the sound of my setup 'juiced' to the max My reference to a novelty is related to the practical implications of battery life, not the quality of the effect. e.g. IMO a sustainer with a 3 hour battery life is a novelty no matter how good the effect - no-one who is serious about guitar playing is going to incorporate something that expensive and hasslemungous into their playing style - it would just be used as an occasional trick. On the other hand, a 20 hour battery life would move it into the realms of 'useful tool' - it starts looking more like a candidate for permanent installation. If you can have the thing switched on all the time you're playing without having to change batteries every few hours, then you can really start exploring its capabilities, and you have a better chance of developing a unique sustainer dependent playing style - of course, if you have it switched on all the time, it must not have a detrimental impact on your normal sound/style... when the instrument is silent, it won't amplify - the Dynamic Range Inverter amplification tails off below about 20mV. Its still an issue though - low level EMI feedback that would not have caused problems in a simpler system can still be amplified. So the EMI badness threshold is more sensitive. As far as complexity, right now its about 35 components chips and all - not a huge circuit . I have one half of a dual op amp spare in there, so I'm hoping to use that for some kind of squelch without making much inpact on the component count or board size - fingers crossed. There is no runaway effect - as soon as the EMI feedback gets up to sustainer level, the amplification is reduced, and the strings start to take over with normal sustain - you can still hear it squeal a little though which is not acceptable I did, although only briefly. The core is 60mm x 6mm x 2mm iron bar. the bobbin top and bottom are black cd case plastic. there is about 1mm of core at the top and bottom. The coil is 2mm thick - about 140 turns of .23mm guage wire, potted in epoxy. It is somewhat microphonic - it was my second attempt, and I made a few mistakes that I should be able to rectify when I get around do making another. The way I see it, if I can get the circuit to work well with this driver, I can then make a fab and spiffy one for my final installation. I hope to get around to pics and sound samples soon. sounds very interesting - I look forward to hearing more about that. cheers Col

Yeah, I've been aware of sustainers since the Edge started using the 'Infinite Guitar' As I've explained, my motivation for building and helping to develop this app is to try and get the same response from an amp sim/headphones/pc recorder setup as I can get from a LOUD amp in a room. So from my point of view, the more natural it sounds the better. In addition, I know that if the batteries only last 2 or 3 hours, it will be impractical. (Personally, I'm willing to sacrifice having a neck pickup in order to achieve the other goals.) Its kind of like a compressor on pcp... with a compressor, even with extreme settings, it doesn't invert the dynamic range, it just reduces it. I tried that first, and although it does help, it's never going to be a total solution. The problem is that some of the strings and notes on the guitar will vibrate more readily with the same driver input - while others (high E string) need much stronger input to come through... so even if you compress the signal so all input levels give an equal signal level to the driver, some strings will sing loud and rattle the frets while others will struggle to sustain at all. A limiter is of even less use - it just reduces the volume of the loud bits... Using a 'Dynamic Range Inverter' (DRI) goes some way towards improving things. A simple threshold noise gate (squelch) may be necessary to bring my DRI into line and kill the EMI feedback. It is set up so that the amplification drops back off below the amplitude that my high E string produces, however it does still make the thing more prone to EMI feedback Sound clips soon, scheme when it's finished, tested, debugged etc. cheers, Col

True...the driver is crucial. There is talk that not all lm386's are equal (though their specs are the same) but mine have not been anything special and all have worked. Yes, there are a bunch of different specced lm386 versions. They are tuned to work best in different contexts - smaller and larger voltage supplies etc. They have different max ratings etc. Other than the min/max supply requirements, I'm not sure that the differences would matter much to us. Best to avoid the M and MM versions - those seem to be surface mount, and lower overall power rating. The N-1,N-3 & N-4 seem to be similar but do have different max and min power and dissipation specs. My chip is an LM386N-1 The "Ruby" uses a simple buffer but the "fetzer" is a low powered preamp. The preamp is there to raise the gain a little and to prevent loading. Without it, the "sensitivity control would act as a volume for the whole guitar if you think about it! The idea of the buffer/preamp stage is to help match impedances so as to allow the lm386 to work efficiently also. Yeah, I should have been clearer on that ... in my circuit, the attenuator is between the AGC and the input of the LM386 - there are buffers on the main circuit input The thing about the power consumption is that the LM386 - when its driving an 8ohm load - draws a LOT more current than normal stomp box circuitry. So adding in a few op-amps really doesn't make much of an impact... So if you can get them to 'condition' the signal to allow the LM386/driver combo to work only just as hard as it needs to and no harder, it's like buying a dollar for a cent - you spend a few milliamps to save a few tens of milliamps (or more as it turns out (I hope)). Nah, it's not that kind of thing - this is proper sustain - lovely stuff. All my testing has been with headphones - or just the guitar/sustainer with no powered output. My 'Volume Inverter' is causing some EMI feedback when there are no notes being played and the strings aren't muted - I suspected i might. This quickly excites the strings and turns into normal sustain, but it's still unacceptable. Shouldn't be too hard to fix though (famous last words ). I thought that spazzy might be getting sustainer feedback supported by traditional 'loud guitar in a small room' feedback, but hey, I havn't posted ANY evidence that my setup even exists so who am I to say. What I am truly concerned about is the idea of feeding a normal pickup with the output of a full-on guitar poweramp though (I wonder when this thread will run out of batteries) cheers col

Hehe, this is probably premature - but I'm excited so I'll post it anyhow I've been messing about trying to get the lower few frets and open note on my high E string to work, and I think I may have found a partial solution. The thing is that it also evens out the overall response more, and best of all, reduces the current draw on my setup to between 12 and 35 mA !!! thats getting really good for battery consumption. OK, so I'm excited, and I've probably missed some big gotcha, I will keep testing it, and fingers crossed it will work out just fine. The basic solution is the result of some circuit 'bending' I was doing. I have built what might be called a 'Volume Inverter'. Maybe someone here with a background in electronics may be able to shead some light on this? any 'correct' way of achieving similar results? What happens is, you feed it a really low signal, you get back a big one, you feed it a big signal, you get back a small one... e.g. 2V in => 250mV out 600mV in => 300mV out 40mV in => 600mV out (The response starts to drop back down again at 20mV, at 14mV output is back down to 300mV) for a sustainer, this is cool. It means the amp works hard to bring a small signal up to a level where it can get the strings moving, then it eases off using just enough juice to keep them going. Ideally, this effect could be even more exaggerated - if 2V in gave 5mV out and 20 mV in gave 800 mV out, we really would be cookin' on gas. The less the sustainer interferes with the natural vibration of the strings the better, so if the thing only worked when it was absolutely required it would be great - much more natural sound - much less battery usage == less of a novelty/effect, more a very useful tool. (of course, there could be a knob to tweak the response so we can have extreme sustainer if required ) Right now, the thing I'm most worried about is that I may be abusing the chip... I've checked the datasheet, and everything is well within the absolute max ratings, but I'm not sure if long term use might still cause problems - I guess I'll have to wait and see. Another likely issue is that amplification of small signals might make the system more sensitive to EMI. If this is a big problem, possibly some sort of 'squelch' may be necessary - don't want to have even more circuitry though Col.

Bummer. what does the white powder look like ? is it on the transistor, or on the solder joint ? do you have a multi-meter? If so, when you replace the blown bits (assuming they're blown), you should check the voltage on the transistor pins, and measure the current going through it to make sure they are within the limits on the datasheet for that transistor. If I remember correctly, you are using a supply greater than 9v ? Maybe some of the other component values need tweaking ? fwiw, I always use sockets for transistors - i get stuff called SIL socket strip, and break off 3 hole chunks It distorts because its designed to . It is a guitar amp after all. A hot passive humbucker can pump out as much as 2.5 volts when you play hard, it won't be that much when the note is decaying. With the sustainer constantly exciting the strings who knows... (just checked, mine is steady at about 100mV The Ruby is based on an LM386 which has a gain from 20 to 200, even at minimum setting, it will probably start clipping at about 400mV (guessing). As you can guess, you don't need much more than minimum gain to get clipping with this setup A few more measurements from my circuit... When the thing is sustaining, the voltage after the gain control is at about 120 - 140 mV. I reduce to 60 - 80mV with a pot before the lm386 input and still get a good sustain - (this is a good thing - the less you put into the lm386, the better the battery life). Of course, how much juice you need to feed the lm386 depends on how efficient your driver is, and possibly even on which kind of lm386 chip you are using !? (not sure - enyone?) So, I set my LM386 up with minimum gain (no cap or pot between pins 1 and 8) and put an attenuator on the input. (My bridge pickup is a duncan JB, it has loads of power, possibly a more subtle single coil pickup would need some power amp gain or a preamp instead of just a buffer - not sure.) As far as posting my circuit, I will do that after some more tweaking and testing. I'm no electronics Whiz, and I don't want to 'publish' some junk circuit that doesn't work properly and breaks down after a few weeks . Its also very complicated, and not finished..... It's not even soldered up yet - still on a bread board. cheers col.

Hi again... A quick explanation of how i discovered this gargantuan thread: I've been playing guitar on and off for about 20 years, however for the last 5 years, after starting a family, it's been mostly acoustic - babies and loud noise don't get on I have tried numerous ways of getting my Rock'n'roll fix without the extreme volume, but nothing seems to work - without the volume, the guitar isn't alive. My most recent bout of stomp box building was a combo of Runoffgroove thunderchief & condor cabsim. These boxes are superb - they sound very convincing, however i was still left with the same feeling of something missing... then a thought came into my head. Back in the 80s I remember reading about the sustainer system U2s the Edge was using. I followed the releases of the sustaniac and fernandes systems and even had sustainer lust for a short time but empty pockets and a move into a more hendrixy style put it to the back of my mind.... Anyhow all this sustainer stuff came back to my mind - I started to wonder if an electro magnetic sustainer could replace that missing something that you get with loud volume. So I googled "DIY guitar sustainer" and here I am hehe. After one abortive attempt, I have constructed a working driver - its not perfect, but good enough to experiment with. I used some 6mmx2mm iron bar that i got from B&Q. The coil is 2mm deep with about 135 turns of .23 wire for about 7.5ohms. I made the driver core too long, it overlaps the strings a bit too far... also the bobbin wasn't quite wide enough so even getting it up to 7.5 ohm was a struggle - epoxy all over the place :-| fortunately i bought a pack of disposable latex gloves for protection. My next one will be better The main thing I've been working on is some circuit ideas. After breadboarding it, I figured that the Fetzer/Ruby isn't quite right - I couldn't get the fet to bias correctly. Not sure if this is because its a cludge of two circuits that were not designed to go together... or if I just messed it up. Anyhow, after some messing about with different input stages, i've settled on an op-amp based pair of unity gain buffers - one feeds the sustainer amp, the other is the guitar output - I Hope this setup will prevent any 'loading' that often happens when a guitar signal is divided between 2 amps without buffering. I'm having the usual issues with the top E string - it struggles on the first 6 or so frets, but improves higher up the neck. To get the high E going I had to turn up the gain so much that the A and G strings were much too 'lively'. Tweaking the output cap can help fix this, but it messes up the response of the low E and A... So I figured some sort of compression/Auto gain control(AGC) might help. After lots of web research, I have tried two approaches, a preamp with loads of gain that clips all input to about the same level, and a non-clipping AGC based on an LM13700 chip. So far, the Auto gain control approach seems much better. The guitar responds in a more natural way - more the effect I was hoping for, similar to being in a room with an amp turned up loud . The balance accross the strings is much better, so the LM386 doesn't have to be dimed ( i disconnected pins 1 and 8 and use an attenuator on the input to set the level ! What this means is that the thing draws much less current so the battery should last longer. My initial measurements are between 30 and 70 mA, so a 9v alkaline should last in the region of 10 hours (I'd be happy with that, so I hope its correct ). The downside of the LM13700 is that its fairly chunky, and with that, the buffer op-amp and the LM386, along with ~30 smaller bits and pieces, the circuit won't be so compact. Theres enough room in my axe, but maybe not some others. I still want to mess about with some combinations of filters to tweak the Phase and frequency response to control the Harmonic effects - I find the 180º phase trick to be a bit extreme, but some level of harmonic bloom is very nice - it would be good to have control over that with a knob on the front of the guitar - maybe a push pull job that doubles as an on/off switch. Fwiw I'll post a schematic when i'm happy with it. Hopefully, someone with better electronics skills than me can then take it and iron out all the noobie errors I've noticed some potential alternatives out there - lots of different multi-function ICs exist, the most promising I have found so far is the LA4160 which is a combined pre-amp, AGC, poweramp that has phase compensation and on/off click protection... it seems to be available in a traditional DIP package (i've not tried tackling surface mount components yet so DIP is good ). and it's "cheap as chips". If someone can hack together a circuit with it, it could be very promising. here's the datasheet for it Anyway thats enough from me... cheers, Col

noooooooooooo 22nF = 22000pF mili -> micro -> nano -> pico... each one is 1000 times smaller than the last btw, hi all