simple 556 osc's starve ?'s

[scriptstyle]

ok so i got this schematic off of thesquarewaveparade.com i built a module for it awhile back and didn't really use it. i also added a second 555 cuircuit as a clock out with led.

question is: i am running my lunetta off a simple +5v power supply, and i think the starve pot causes the power supply to bounce from 4.5- 5.5 or something close to that. Also the leds sometimes will dim. to the point where the regulator will get hot and i even got alittle smoke. now that i'm looking at this schematic again i think i might need to add some .1uf and 100uf caps around the power supply as well? should i do away with the starve pot?
any help would be appreciated! i really don't want to burn out the 4017 on my newly finished baby10!

[scriptstyle]

i forgot to mention that i am not using the touch contacts, and i also just noticed that the 2 cv inputs probally need 100k pulldown resistors.

[Inventor]

Well, I don't know what to say, something like that could be any kind of miswiring, but I presume that you are getting sound out of it, are you? The starve pot looks really suspicious to me. As a matter of fact, all those pots look like trouble in general.

This is just a guess, now, so take it with a grain of salt, but when you connect a pot as a rheostat like that you should put a series resistor with it to prevent from shorting the ends. By that I mean, your pot is 5k, but on its minimum setting it's like say 100 Ohms, or 2% of full scale say for example... Well, if that full 5V happened to appear across that little 100 Ohm tail-end of the pot's resistive material, you'd have V*V/R = 0.25 Watts of power applied to it. So the poor little segment of pot material can't take the heat and it lets the proverbial smoke out of itself! You should put enough resistance in series with the pot to prevent this.

I know of an awesome reference for pots, the Bourns corporation which makes all the world's great pots or most of 'em has a short little hand-drawn picture book that teaches engineers everthing they'd ever want to know about pots and then some. It's a quick, easy read, just cruise on over to the Bourns web site and contact them, ask for a free copy of "the pot book", if you are so inclined.

Also, did you put they electrolytic caps in the right way? They often give up their smoke or create shorts that cause other parts to get all smoky if they are in backward.

Aside from that, check all your wiring and tripple check everything really thoroughly. The voltage dip and LED dimming are both consistent with some kind of short somewhere. So, just some things to check. I have no doubt it's something other than what I've mentioned probably... who knows? Good luck...

[scriptstyle]

thanks for all the great info! i can say that it is definitely not the wiring because i was using it to test the sequencer when it was on the breadboard. now that you put it in perspective it dose seem that this schematic is short some resistors, and that totally makes sense. but it did seem like the smoke came from the regulator? now on the triple check i did notice that the back of my +5v and gnd jacks were shorting out to the frame when i mounted it (thats not good, when testing it wasnt mouted). with that fixed, 2 100k pull downs for the cv ins, some decoupling caps around the power, and a copy of that manual(that i'm going to try and obtain right away if possible) so i can add some resistors in series to the pots i should be golden? or should i do away with the starve all together? thanks for the quick reply!

[Inventor]

Yeah, I'd ditch the starve pot altogether. It's not wise to do that sort of thing. You see, I used to work as an electrical design engineer in digital, analog, and software. Now that I'm hobbying in DIY music electronics and software through this great forum, I'm starting to see some flagrant violations of sound design principles in the homebrew circuitry that people make. 99% of what I see is good, but some of the stuff is just plain silly. In fact one person, despite my warning, did something so silly that if he/she were to do that as a professional engineer it would foster laughter from coworkers and possibly cause his manager to question his value as an employee, haha. That starve pot is just plain wrong IMHO.

Although some small occasional design issues are popping up, I'm finding most of what I see to be good stuff and I'm also learning a lot myself! I've never studied music hardware and synthesizers before and the whole world of it is really kewl to me! So a few flagrant violations are certainly to be tolerated in the name of adventure and fun.

Oh, if the regulator gave up some smoke, I'd replace it for good measure.

I'm glad you found the short, that is most definitely the likely cause! Good debugging there. Hope you make lots of krazy sounds with your little mini-circuit, have fun!

[scriptstyle]

this guide to pots is a must have thanks amillion!!!!!!!!

[nicolas3141]

The starve pot goes against good EE practice, but that is because good EE practice is oriented towards getting good clean square waves out of a 555. The 555 is renowned for high instantaneous current draw when it switches so at a rough guess the starve pot will make those square waves somewhat less square on the rising and falling edges. That is not good EE, but might be musically interesting.

[Inventor]

Yes, good point nicolas, it probably is musically interesting!

[scriptstyle]

so maybe i will leave the starve pot and just add a resistor in series and just try and keep it like a fine tune, im not to sure yet.

[CJ Miller]

I recommend to make certain that you are using the CMOS version of the 555/556, as it uses far, far less current. As for the pot... I don't know. Try a different wiring? I don't want to cost you any burned ICs, but sounds like it could happen anyway. Put a diode between the regulator and the starve? Wire a pot with +V on one side, ground on the other, and pick the desired voltage from the middle (maybe with a bigger value pot)?

Another, maybe less risky trick, is to use signal output from one timer as the supply voltage for a second timer.

[loss1234]

i love the starve pot on mine! this circuit is meant to make pretty crazy noises and is really one of the most amazing uses of a 555 based chip i have ever had the pleasure of working with. nothing wrong with rules, but often when you are searching out crazy sounds, they are meant to be broken.

another cool thing with this circuit is the CV ins, they are not 1/v oct but they make a ton of racket! very lunetta like.

[Inventor]

[scriptstyle]

a scope is in my near future for me! a promised hand me down so i just hope it works and dosnt need to be calibrated. for the cap any tips on what would be good? if i use an electro cap the + goes towards the power? hey loss do you have this same cuircuit added to your lunneta? if so did you change anyhting add resistors? im quiet sure its a must to add the 100k pull downs to the inputs and the decoupling caps, but im just wondering cause mine is all soldered up already and once i put the add ons on im hoping to get it right.

and thanks to everyone for the good info

[Inventor]

If you use the same 5k pot, try a 0.01 uF decoupling cap. If your frequency is audio, say 1 kHz, that's a period of 1 ms, then divide that by R=5,000 Ohms to get 0.2 uF, then divide by say 10 to make the time constant be 1/10 times the period. Something in that ballpark would be good to try.

But I'd wait to get the scope so you can see what the starve pot will do. The reference divider of the 555 is driven by the power pin, so the starve pot might be doing more than just softening the edges.

Oh, and the cap goes to ground, like this:

[nicolas3141]

Okay, I got around to trying it. I took a std 555 (not the cmos kind) and built the simplest possible 50/50 square wave oscillator with it. Pins 4,8 to +V. Pin 3 to 100K pot and output. Pins 6,2 to other end of 100k pot and 100n cap. Pin 1 and other end of 100n cap to -V. Supply was 12V SLA with a 20K wirewound pot in series on the +ve side.

It oscillated nicely in the audio range sounding like a normal square wave through the amp and looking on the scope pretty clean. Might have been more like 55/45 rather than 50/50, but certainly pretty close to square. As the supply impedance was increased from zero up to 10k interesting things happened. None of them was similar to the effect of a simple filter as suggested above. Above 10k oscillation stopped.

As the supply impedance increased the output amplitude dropped and the frequency dropped. As the frequency dropped the pulse width also changed with the +ve half of the cycle getting up to about 70%. These effects weren't too hard to predict if I had thought about it. The +ve half of the cycle also developed something of a sawtooth look, getting to about 80% of the peak voltage immediately, then ramping up the rest of the way.

[nicolas3141]

... continues ...

At this point with the starve pot set at a few K the effect can be heard, but it isn't dramatic and with the addition of PWM control to the circuit you could probably get a similar effect and more of it.

But as you get close to 10K supply impedance things get weird on the scope and you can definitely hear something odd through the amp (although again I would have to say it is not a particularly dramatic effect).

The waveshape changes described above have impacted the +ve half of the cycle, now the changes start to impact the -ve half.

The frequency drops now due to the lengthening of the -ve half of the cycle and the pulse width returns to roughly 50/50, then goes beyond it to about 30/70 just before it flatlines. The zero volts part of the cycle becomes a mess of very high frequency oscillation (a blur on my scope) of about 1/4 the overall amplitude. Then just before it flatlines (ie at about 10k supply impedance) the very high frequency oscillations disappear and the -ve half of the cycle settles into a waveshape that is somewhat above zero volts and has a residual ringing (like an overshoot) that I can see is about 2 octaves above the fundamental.

So at high 'starve' settings you are actually getting more overtones, the complete opposite of the filter suggested above. Definitely interesting.

Now please note that I have not built the circuit at the start of this discussion, but a much simplified version of it. And when using chips outside their design parameters results can vary a lot from manufacturer to manufacturer and even from batch to batch. So your mileage might vary ...

[nicolas3141]

And one more comment in favour of starve pots is that they are quite a good idea when messing about with something you have either just built or have just opened up and don't yet understand. A bit like the old circuit-bending advice of starting your investigations with the machine running on 3/4 flat batteries. Can really limit the damage if you have soldered something in with the wrong polarity, have accidentally used a 5R where you should have a 5M or just done the classic trick of shorting out two pins when trying to probe one of them with a test lead.

What I do is keep some old mostly flat 9V batteries around for the initial fire up after wielding the soldering iron. I would only connect a high current supply like the 12V SLA battery to a circuit I was confident had no shorts or other goofs in it.

[scriptstyle]

ok the starve pot is staying with a 220ohm resistor in sereies, there is probally a better value but 220ohm seems to do the trick i think a trim pot is in order but im just doing the fine tune thing for now. added some decoupling caps, pull downs on the inputs, and everything seems to be all good. this thing does make some great noises! soon im going to make a post of my blocks so far: baby 10/decade counter, 4094 8 stage shift register, 556 blast disast, power supply, and my 40106 is still on the breadboard. my thinking is the more osc's in my lunetta the better? may be wrong due the fact that it is very hard to guess what these things will do when all hooked together[funn stuff].

[Inventor]

Thanks for all the info about starve pots. My engineering training tells me that they are bad practice, but I see the DIY community has come up with something rather interesting instead! Play that funky music, haha!

[gogmagog]

[widdly]

You could buffer the starve pot by first running the 5v-15v through a couple of gates in a CD4049. That would isolate the starved supply from other circuits sharing the same power source.

You could replace the starve pot with the cmos based voltage controlled resistor synthmonger (i think) has posted in another thread. Then you could voltage control the starve too.

[Inventor]