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

"Or" circuit question
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author "Or" circuit question
Leisure Cove
Gang,

I just wired up what I think is a proper "Or" wired circuit (from the diagram on Doepfer's DIY page). Looks easy enough, right?

Here's my question:

I patched the output of the circuit to the Gate In on my A-140, and the three inputs on the Or circuit came from three of the outputs of an A-161 Clock Sequencer. It works just fine, in that each input in the Or circuit is triggering the A-140 via the Or circuit's output. Great!

But here's what's made me nervous: when any of the outputs of the 161 are triggering the 140, the LED next to that step is lit. Not too strange, I know, but what IS strange is that the LEDs next to the other steps in the 161 that are patched into the Or circuit light as well.

I am terrified of Magic Smoke, and unattached the patch cables as soon as I saw this.

So, my questions are:

1. Should I be worried about this?
2. What would you recommend I do to troubleshoot?
3. Does your Or circuit cause this behavior as well?

Thanks!
DGTom
This is my understanding of whats going on here, if I'm wrong I'd love to be corrected.

The problem with that cct. (as shown on the doepfer DIY page) is its totally unbuffered, you can OR stuff together like that inside a cct. no problem but hooking it up to modules is problematic.

Take a look at the Logic schem. for the CGS Pulse Divider / Boolean Logic;

http://www.cgs.synth.net/modules/cgs36_pulse_divider.html

or;

http://musicfromouterspace.com/analogsynth/mmlogic.html

bumps the part count up a little, but the results will be much better - chuck an AND, NOR & NAND in for good measure grin
Leisure Cove
Good call, DGTom!

Are we in consensus-- I should NOT use the Doepfer circuit (with the result I mention above) with my modular?
wetterberg
I would also say that unbuffered circuits are a bit more sketchy than one would like in a modular - not because of safety, I wouldn't worry about that, but more wrt the added dependencies, which add complexity in a massive way.

I wholeheartedly recommend these circuits, which add just a few more parts but seem to work much better (at least judging from my humble protoboard testing)

DGTom
it will work, as you'd observed, but, it assumes that it won't ever get two (or more) high states at once. A proper OR gate will output high when either one or more inputs are high, as it stands this cct. will actually 'mix' the inputs so if 2 inputs go high at once the output is 'more than high' feed it some audio & have a listen thumbs up

read thru the Ray Wilson mm2logic page & it will explain why the basic cct. will work, but adding some schmitt triggers makes it work better.

The cool thing is its a very simple, very efficent little build. The 40106 has 6 gates. You need 2 to make either an AND/OR - but you get NAND & NOR as well (just tap off inbetween the gates)

*Edit; def. echo wetterberg here. Grab a 40106 (or 10) & chuck this together on perfboard. The pulse stretcher is also VERY usefull.
wetterberg
yep, the mml circuits are all based around logic chips - the 40106 should work, too.
DGTom
Ray Wilson wrote:
MML lets you keep one IC on hand, the simple hex inverter (74HC14, 74C14 or CD40106). These are hex (because there are six of them in the package) inverting Schmidt triggers which when combined with resistors and some 1N914 diodes (which are very cheap and obtainable) can give you whatever CMOS compatible gate you need.


He uses the schmitt triger to basically do what the transistors do in the CGS Boolean logic cct. thumbs up
Leisure Cove
DGTom wrote:

*Edit; def. echo wetterberg here. Grab a 40106 (or 10) & chuck this together on perfboard.


Will do! thumbs up

And so begins my CMOS obsession...
Scott Stites
Ray Wilson's M2L write-up is the best I've ever seen on the subject.

The beauty of M2L is that it doesn't limit you to the number of inputs for a particular logic gate - you don't have to resort to finding a three input AND gate IC, for example. This allows one to consolidate a lot of functionality to just a few ICs rather than specifying a battery of other logic ICs.

The Klee Gate Bus is an orgy of M2L.

daverj
Leisure Cove wrote:

But here's what's made me nervous: when any of the outputs of the 161 are triggering the 140, the LED next to that step is lit. Not too strange, I know, but what IS strange is that the LEDs next to the other steps in the 161 that are patched into the Or circuit light as well.


Are you sure that you wired the diodes in the correct direction and that the diodes are good?

Even though that circuit is unbuffered, no current should flow backwards through the diodes from one input to another.

What type of diodes did you use?
Leisure Cove
daverj wrote:


Are you sure that you wired the diodes in the correct direction and that the diodes are good?

Even though that circuit is unbuffered, no current should flow backwards through the diodes from one input to another.

What type of diodes did you use?


1n4148s, bought from Radio Shack. I very well may have oriented them backwards! d'oh! I'll double-check it when I get home/post a photo of the circuit.

Good call, Daverj! I hope this turns out to be the case. I'm going to build some 40106 logic circuits for sure, but am into the idea of passive utility circuits and want to add these simple diode + resistor OR ones to my collection. Don't ask me why, but I think passive is sexy.
daverj
Your biggest problem would be if they are not all facing the same way, or if any of them are shorted (which happens easily with 4148s). I've also had some be marked backwards, though not often. They can also be damaged easily with too much heat from a soldering iron.

If you have an ohm meter, you can test them, even in this circuit, since one end of each is not connected to anything when nothing is plugged in.

With nothing plugged into any of the jacks, take the ohm meter and touch the leads to either side of a diode. Now do it again and reverse the leads. You should only see a low resistance with the leads in one direction and a high reading with the leads reversed.

Repeat both measurements with each diode. Be sure you do get a low reading one way and high the other. And be sure that the low reading happens with the same color lead on the same end of each diode (ie: with a given color lead at the stripe of the diode).
Tim Stinchcombe
Coming a bit late to the thread, but the behaviour you describe is almost as if the diodes are not there at all. The A-161 outputs are very basic: the individual outputs of the 4017 counter chip (doing the 'sequencing') are attached to the base of an NPN transistor; collector of this to +12V; emitter to a) 10k to ground, b) LED + 2k2 to ground, c) 1k to output jack.

It would almost work well enough if you wire-or'ed the outputs together without the diodes (but then one would need to start looking at the danger of exceeding the base-emitter reverse voltage so as not to pop the transistor): the LEDs of all connected outputs would quite likely light when just one output was actually active. If the diodes were reversed then you shouldn't be getting the A-140 to gate at all. So perhaps the diodes are shorted, or wired incorrectly in some way...? Don't think it was mentioned before, but all the black-band ends should go to the same point. Failing that, some basic checks with a DVM as already mentioned would be a good idea.

Tim
DGTom
My preferance for M2L is only based on what I've observed / heard on my breadboard. If I have two sub-audio clocks I'll use diodes no problem, but, results get a little iffy IME with more than 2 inputs & faster speeds - I think changes / differances in pulse width start to effect the cct. more.

passive is sexy, but not when it starts making strangeness (unless you specifically want strangeness)

I'd keep the passive OR cct. you've got, after checking for those shorted diodes cos that'd make alot of sense, srsly - run some VCOs thru it! diode mixing sounds kewl!
Leisure Cove
In the interests of providing too much information, here is a photo of my OR circuit and a video of the problem in action:





The red wires are connected to the jacks' tip terminals. The black wires are the ground (the bottom-most black wire is connected to the one a row above it via a jumper on the back).

As you can see in the video, the diodes are "meter healthy" but are somehow allowing the signal to flow in both directions through them, which is puzzling. Looking at the photo, can you tell-- are they installed backwards? I'm not sure what else could be going on.

Thanks again for your help so far, everyone!
Scott Stites
These "simple" problems always seem the worst for me grin.

The diodes sure look right to me. This may be a long shot, but perhaps you should check to see if you've correctly identified which lug is the tip lug and which lug is the sleeve lug on the jack itself. My distorted mind tells me that if you accidentally mixed up the two, all three of the signals would be shorted together and would be passing through the resistor on to the output jack, as well as back to each other, since they'd be shorted together. The ground lugs on the three inputs would be a diode gap above ground, which would be provided by direct ground connection from the output jack.

Just as a sanity check, make sure that you have selected the proper lugs on your jacks, if you haven't done it a zillion times already.

If at all possible, grab an unused jack (if you have a spare) and plug a cable into the spare jack. Use your DMM and put one lead on the tip of the plug on the other end of the cable. Then connect the other lead of the DMM to the jack lug that you think is the tip connection (IE, the same lug you wired your red wires to). If it's the right lug, you will have zero ohms resistance. If it's the wrong lug, you will have infinite resistance. We're just checking to see if you have continuity from the tip of the cable you have plugged in all the way to the tip lug of the jack - if so, then you've identified and soldered to the correct lug. If not, that's your problem.

If you don't have a spare jack, you could use one you have wired up, but a spare jack would get rid of any in-circuit obfuscation that may arise.

It's an easy mistake to make, BTW. That's why I always check mine, because I hate re-wiring 3.5mm jacks. That's when I experience my worst banana envy.

daverj
The diode markings look right, and the meter readings are what they should be. But when connected it's acting like the inputs are shorted together.

Scott's suggestion is a good bet. Wiring up the jacks backwards would give the result you're seeing. Also double check the other side of the board to make sure the input red wires aren't getting shorted together on the board.

I think we've all soldered to the wrong pins on mini-jacks more than once, so that's a good thing to check.
Leisure Cove
Smart! I'll give it a shake when I get home and will report back with the results. Fingers crossed...

Thanks thumbs up
Leisure Cove
I am happy to report that my OR circuit problem has been fixed!

Big thanks to everyone in this thread for their suggestions and help-- especially daverj and Scott Stites, who rightfully suggested I had the tip and ground lugs reversed on my jacks. Once they were corrected, the circuit began to function as intended.

In this video, the Step 1, 5 and 7 outputs from an A-161 Clock Sequencer are feeding the OR circuit's inputs. Its output is set to the Gate input on an A-140 Envelope Generator, the output of which is opening and closing an A-132-3 VCA.

The "bass" tone is a self-oscillating Frostwave Resonator, sent into the VCA. The "tin drum" sound is a Boss Dr. Pad (triggered by the 161's Step 4 output):



It's peanut butter jelly time! It's peanut butter jelly time! It's peanut butter jelly time!
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group