MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

lessonquencer evolution
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author lessonquencer evolution
falafelbiels
OK so I've posted here before on my lessonquencer. I never opened the box up after my last post and I am not going to. I'm just going to build the lessonquencer mk2. -insert drama-
Mk1 is built up far too cramped to do anything useful tn there and is fun enough as it is.

So on Christmas eve I finally got around to doing some breadboarding and built me a clock (first time I got a 40106 oscillator to run!) and built up the counter and multiplexer again with 8 leds showing me the output. The aim was to get proper syncing and proper resetting to clock pulses and gates and getting rid of some glitchy stuff that sometimes goes on.

Now I took allyou guys's advice in the previous thread to heart and got it running smooth so far. w00t
it syncs, it resets, it isn't glitchy whatsoever. So the next day I stuck in bypass capacitors which completely destroyed the glitchness it had before I did that. Then I let myself be inspired by the CGS pulse divider for that "trigger conditioner" for the clock input and reset and after that all was fine.

GREAT

So then I went to my bed (had to visit my parents the next day off course, I wouldn't have my christmas be all lonely with the breadboard as my only companion) but of course I couldn't fall asleep because you know, I should have done something relaxing after all that anxiety and such but I hadn't. Just couldn't stop thinking about the stuff you know?

So OK now I started thinking about some other stuff the Mk1 ain't so good at. The gate output is somewhat wonky and I believe it is about current problems (actually just guessing here, no science behind there at all.

Anyway I think about 2 ideas for the Mk2 here:
- sticking some potential from the 12v bus through the cd4051 multiplexer, through the switch array, have 1 output from those drive a transistor followed bij a voltage divider (panel mounted pot maybe?), then out the machine to some gate awaiting module.
- Take potential from a voltage regulator and follow that same route, without the voltage divider.

I wonder if the regulator (say a 10V one) would have any benefit over the 12v bus +voltage divider because I'v never used one.

Any ideas or advice?

I will take more advice to heart and draw schematics as I go, I promise...
DGTom
I'm pretty sure your idea for the gates will work.

Give the MUX input +12, a voltage regulator would be overkill IMO, if you were using the V. Reg to feed pots which then get de-multiplexed into a sequence it'd be worth it, to get a good, clean consistant voltage supply at a precise point.

Anyway, you'll need buffers on the outputs. Not sure what you mean by 'wonky' hmmm.....

This is what I use on all my CMOS gate / clock outputs;



This X8 with the 4051 will give you 8 solid gate (stage) outs with LEDs for each.

Leave that seperate to the actual gate sequencing side tho.

I really like the vid of the previous sequencer, its an interesting idea. I've been working with the 4051s lil bro alot lately, binary patterns are fun applause
falafelbiels
I was kind of trying to get away with a bit less transistors. have some (regulated or not) voltage going through the multiplexers to the switches and the step led, then the output from the switches to a single, gate-bus feeding transitor buffer. Your design seems specifically nice for that last one. I got away with it on Mk1, but it might be bringing up the noise on ground I believe. Once I installed some bypass caps in Mk1 the wonkyness may well be over anyway. I actually opened it up today to check if stuff can still be installed, it sure can!

Yeah the binary pattern thing is real nice and on Mk2 I want the bit-inverting and more stuff the be gate-controllable. That may be great fun since I also have 2 CGS sequencer PCB sets lying around w00t Guinness ftw! .

sequencer overload yes please......
DGTom
ah cool, I can dig trying to minimize part counts... but tbh there is a balance to be struck between minimizing the cct. & getting acceptable stability, especially where sequencer type devices are concerned.

I don't think the problem is just the voltage itself, the output needs to be buffered in some way, CMOS are really only designed to interface with one another, not the outside world.

That trannie buffer is prob the smallest, most effcient way to do it, but, I agree with 8 steps it can get abit unruly.

Maybe take a look at using the 4050 Hex Buffer IC.
meridic
Using something like the 4050 may seem easier then transistors but remember, the 4051 has its i/o pins spread over both sides of the chips, so layout is more complex, you need to run jumpers to get everything to the 4050. Transistors clean this up abit, you can put them right where you need them. You can also simplify the transistor circuit down to just a transistor and a resistor if you wanted to, If you make the V+ feeding the transistor the same as the output voltage you want you can dump the voltage divider, I do not see much purpose for the 100k base resistor in this instance, you can certainly get away without it here. The LED can just go if you do not want it. But the LED is essentially free in this case, the same resistor limits current for it and the transistor.

Do a search for "transistor switch" if you want to know more. Any of the popular search engines will give you more information then you could ever want on using transistors as switches.
falafelbiels
yes well I thought since there is always just 1 of the stepleds burning I wouldn't need buffering, but do these buffers also reduce the noise on the power supply? Soldering transistors gets me confused with the weird pinout and all, datasheets aren't always so clear to me on this so the 4050 I find quite nice actually. This wiring and bridging will always remain a problem when working on stripboard anyway.
I don't know if I get the "transistor switch" really. Is there an advantage to using transistors over say cd4066 switches? I really like CMOS...
falafelbiels
DGTom wrote:
I don't think the problem is just the voltage itself, the output needs to be buffered in some way, CMOS are really only designed to interface with one another, not the outside world.


Yes they do that fine, but I since the 4051 is an analog multiplexer, kind of like the 4066 but switching among several in/outs, it's not really interfacing with the pots and transistors and all, rather having that stuff interfacing with each other through it, right/wrong? hmmm.....
meridic
Quote:
yes well I thought since there is always just 1 of the stepleds burning I wouldn't need buffering,


It depends on your goal, if you want individual gate outs, and LEDs you probably want something that would increase the drive abit. If you are not doing individual gate outs and just going from the 4051 to the summer then you good, each CMOS output can drive an LED and summer no problem. The one thing to keep in mind is LED current, the CMOS limits the amount of current the LED can see, so it limits brigtness. You must chose your current limiting resistor properly, to big and you can not see it glow, to small and you start draggin down the output.

Quote:
but do these buffers also reduce the noise on the power supply?


Nah, you do not need to worry about noise in CMOS much, at least in not the synth world. Atleast I have never worried about it, nor encounted a problem from it. At high speeds a little noise can cause alot of confusion though.

Quote:
Soldering transistors gets me confused... ...say cd4066 switches? I really like CMOS...


The two types of switches can do some of the same things but they are abit different. To make the synth analogy of transistors, take a VCA and put DCV on the audio input and a VCO on the CV input, run the output to an amp. What happens? Well the VCO modulates the dc, if that VCO is a 100hz sine it modules that DC into a 100hz sine so the audio output is now a 100hz sine. Lets go abit further and say that there is a 1:10 ratio between the CV input and the audio input, so a 1 volt change on the CV makes a 10 volt change on the input. So throw on a 1 volt sine on the CV input and you get a 10 Volt sine on the audio output. This is what people call an amplifier, which is abit misleading, it is more of a magnifying mirror. The small AC signal is reflected back on the larger DC signal. So as long as the DC is bigger then the AC times the 1:10 ration you get gain. If the DC is 10Volts and the AC is 1 volt you get 10 Volts out, but if the AC goes up to 2 volts and the DC stays at 10, you do not get 20 Volts, you get 10Volts and alot of distortion.

OK! now lets relabel things, that CV input becomes the base, the audio input/output become the collector/emitter, the DCV is the +Power supply and the 1:10 ratio is the beta of the transistor. This applies to all those three terminal gain devices, with a triode the CV input becomes the grid, the audio input/output become the cathode/plate the 1:10 ratio becomes the mu, with FETs it is the gate source and drain.

So how does that become a switch? replace that AC 100hz sine wave with a gate or trigger, the output now is just a pulse with the same duration as the input pulse, but it can be a greater voltage or current, or of less if you wanted. So you are actually just replacing the CMOS pulse with a pulse of the +V supply of the same duration. You can use a CMOS switch in this fashion if you ran +V into the input and ran your pulse into the control input. You are 'switching' on the transistor with you input pulse so that +V can flow, and it will flow as long as that pulse is present.

Dont let tansistors confuse you, they are not nearly as hard to use as they seem. You did name this "lessonquencer" after all, is it much of a lesson to use the things you are already comfortable with?
falafelbiels
Oh I get the difference, thanx anyway, but I easily get confused over the pinout of transistors as I read them from the datasheet. I find it real easy to solder em in the wrong way around. ICs don't give me that head-achy problem.
Having said that, I really don't get the PNP/NPN thing.

Quote:
is it much of a lesson to use the things you are already comfortable with?


you certainly have a point there.



Still, today I soldered a 100n bypass cap and the CGS style clock input and reset in the Mk1. All wonkyness is over and it syncs tightly, EXCEPT...
It keeps skipping the first step... It just runs 7 step patterns and I really don't like it. The first step CAN fire up when using reset and the jam inputs, also by inverting the bits that drive the multiplexers so the multiplexers are not at fault here. I have done something wrong rewiring the counter as I had breadboarded. Anybody had this before, just 7 step patterns from a cd4029?

Too bad as I just today got my Tyme Sefari in, I thought let's deglitch that lessonquencer and get rocking! not so...
meridic
Quote:
I really don't get the PNP/NPN thing.


Positive Negative Positive/Negative Positive Negative. One is just backwards. They do the same thing on their own, but when paired they can do other things. Complimentry pairs. So current flows towards the negative, draw it out, in PNP you have one P at +V, N at -V and the other P at groung, in NPN you have an N at +V, the P at -V and the other N at ground. What does the current do?

Quote:
All wonkyness is over and it syncs tightly, EXCEPT...It keeps skipping the first step... ...s from a cd4029?


Go dig out the data sheets for the 4029/4051. Study the timing diagrams and then draw up your circuit from exactly how you have it on the breadboard and figure out the timing diagram for your circuit, the answer is there! You are close, very very close. Just make sure to sit down and actually trace out the circuit as you have it on the breadboard, do not follow any schematic you drew up before you put it on the breadboard. Staring at a schematic you built from trying to find a wiring error is a sure way to never see the error. TRACE it out, one of the best ways to troubleshoot a circuit. I have stared at circuits for hours trying to figure out why it did not work right, only to find the problem right away when i drew up a new schematic tracing from the circuit. A mistake you make yourself is the hardest one to find, you thought it was right when you made it after all.

By the way, never poke a glazed ham right after it comes out of the oven, you will regret it.

Why did I make a ham? Why am I still up, I just keep cooking instead of sleeping.
DGTom
meridic wrote:
I do not see much purpose for the 100k base resistor in this instance, you can certainly get away without it here.


Before I built it - that was a specific buffer for these 4051/2 gate multiplexing thingys - I would have aggreed with you... that 100K was just something I came across whilst experimenting, perhaps its not ideal, but, it really did make the outputs behave alot better in practice.

I think its something to do with strobing the A/B inputs non-sequentially & at high freq.s hmmm..... I'm not sure but it seemed to fix the problem I was having.

meridic wrote:
Just make sure to sit down and actually trace out the circuit as you have it on the breadboard, do not follow any schematic you drew up before you put it on the breadboard.


damn! I wish someone had given me this advice 9 months ago!! So many swears could have been avoided Mr. Green

falafelbiels, I keep a little notebook, where I copy out pinouts for ICs, transistors etc. makes a good, quick referance I can double check before soldering stuff in, much better than the dozens of .pdfs on a hard disk.
meridic
Quote:
I think its something to do with strobing the A/B inputs non-sequentially & at high freq.s hmmm..... I'm not sure but it seemed to fix the problem I was having.


perhaps it is creating some slew limiting when combined with the in circuit capacitance, smoothes out the strobe. I am going to have to think on that after I get some sleep and have abit better focus.

Quote:
damn! I wish someone had given me this advice 9 months ago!! So many swears could have been avoided


Nah, it would not have saved you any swaring, probably caused more. You still would have spent hours of fighting with it before tracing it out, then you would trace it out and see how much time you wasted on a simple obvious wiring error, more swears and some disgust with yourself, a little closer to just giving up. It might have been better if I did not share that hint.

Quote:
I keep a little notebook, where I copy out pinouts for ICs, transistors etc. makes a good, quick referance


I just tend to draw them were ever is convienent, that way when I need them I can spend hours looking before finally giving up and asking the internet. once again swears, abit of disgust with myself, and a little closer to just giving up.

Electronics is a great hobby
Luka
smile

for transistors i know draw the pin-out on the plastic bag and put a dot of a colour enamel paint on each one to know they are from that batch (also so i could write down beta if i needed to match them)

but yeh i tend to draw opamp pinouts on anything i can find and always loose them :(
meridic
you should get some tattoos
falafelbiels
And by tracing out one means drawing it out? I'll surely do that once I feel somewhat more sober (some time next year probably). My dabbling in electronics is about as organized as the rest of my life, so now that I have some spare time, I best also spend that on the very necessary tidying up around the house. After all of that my head should be a bit more clear and I'll have at it again.

Meanwhile last night I have been fooling around with the "half-functioning" Mk1 and the TS anyway, great fun, it DOES reset to 1 so no real problem actually. I may just really not open it up anymore this time and make sure Mk2 is perfect, as that was my goal all the time...

I'll take all this good advice to heart in the next year.
thumbs up
falafelbiels
SlayerBadger!

Right so almost four months later I stuck a diode from Q4 to preset enable and it works like a charm now!

Next I'll have to think of some logic to have it NOT reset to the step selected by the jam inputs after step 8... I have plenty logic chips around to put in there so it can be done whenever inspiration and initiative strike simultaneously. Another switch in the panel to (de-)activate the logic circuit, yessssss.

But for now this is a really nice sequencer!

w00t
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group