Electro-Music Klee Sequencer Build and Applications Thread

Scott Stites · Posted: Fri Aug 01, 2008 7:34 pm Post subject:

Make sure the Klee shares signal ground with the 606 if you use that...

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

thanks scott

yeah i have a scope and DMM, they are ready and waiting
i'm just using this excuse to clean and tighten up all my panel wiring.

i noticed that i had V+ shorting to analogue ground, so that is fixed

ill work on the panel wiring for another hour or so and check out the analogue pcb, so far the digi one looks ok.

what was the chip that everyone was having problems with?

Scott Stites · Posted: Fri Aug 01, 2008 8:57 pm Post subject:

The CD4034. Are you getting any of the pattern LEDs to light at all?

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

im 9/10 through cleaning up my wiring scott
ill plug 'er all in soon and re-test once there is no more spagetti Smile

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

ok we have LIFE! woot

but still problems

B's LEDs are permanenty ON (except 2 which are always off)
i swapped 4034 and i get the same behaviour for A

the load doesnt seem to do anything, it never corresponds with my pattern switches. the only way i can alter the amount of bits on is hitting Inv B or 8/2:16/1 which seems to me to randomly switch the bit ONs and OFFs and possions of those that are ON.

but..!!

other than that seems awesome
all the voltages are working, havent tested triggers and gates as there is only so much room on my bench. getting some wild sounds, my housemates all came in and had a go Smile

Scott Stites · Posted: Fri Aug 01, 2008 11:28 pm Post subject:

Looks like the one CD4034 definitely is sour if it did the same thing in the A or B slot.

Here's something to try: put the good CD4034 back in the A slot. Go to 8X2 mode. Put the Random/Pattern Switch into Random position. Turn the ref level pot all the way over counter clockwise (normally, the Random Reference LED will light up even with no random input). If not, crank the reference pot the other way. If you still can't get the ref LED to light, put in an LFO or some other signal into the random signal input jack, and crank the random level all the way up and adjust the reference pot - the LED should light up in sync with the signal (flashing at the rate of an LFO for example). A DC voltage input will light the Random LED all the time, too, which would be preferable for this test.

The point is, we want to get the Random LED to light, which indicates that a random signal will feed the CD4034.

Get the Random LED to light and clock the Klee - as long as the random LED is lit, the pattern LEDs will light up one by one with each clock pulse. At the end of the eight, they'll stay lit. Adjust the random controls until the Random LED is off - now the LEDs should all turn off, step by step, with each clock pulse.

Clock the random input until only one LED is lit, then stop the clock. Flip the Random/Pattern switch to "Pattern" - this should cycle one bit through the eight LEDs of the 'A' section repeatedly.

This will ensure that the serial input of the CD4034 is good and that it will clock the bits correctly serially.

Next, stop the input clock. Punch the load switch and look to see if you get pulses on pins 13 and 14 of U6 (the 'A' CD4034). You should get some *really* fast pulses on those two pins each time you hit the switch.

If you don't see any pulses when you hit the load switch, connect an LFO to the external load input - with each clock of that LFO, you should see pulses on those pins - they're really fast, so look quick Very Happy

. If you still don't see anything, suspect the area around the CD4013 or the circuitry before it. Without those pulses, the CD4034 won't load the pattern bits.

Also, check pins 16 to 23 on the CD4034 for highs where you expect the pattern bit to be high (as set on your pattern switches).

I'd help you more, but I have to get to bed - got an early drive to Manhattan tomorrow. Should be back in the evening, which will be your morning. In the meantime, if you detect a problem with the load pulses, I think I cover that in the theory of operation thread.

Cheerios,
Scott

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

ok cool, well the ref light comes on without any dramas.

yup that worked

i use random ref, let one bit come up and then hit load, it locked in one bit

State Machine · Posted: Sat Aug 02, 2008 5:52 pm Post subject:

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

bill you are a champion, but in all seriousness, i dont want you to be out of pocket for this... shipping to australia is a expensive waste. i can buy some, no worries, you guys have done so much already.

i have to do a futurlec order soon anyway, and in all serousness, it is probably better if this thing isnt functioning for the next few weeks, only 3 months to go of uni and so much work to do Smile

don't need any more flashy light show distractions

State Machine · Posted: Sat Aug 02, 2008 7:22 pm Post subject:

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

ok cool will do

i had a look through the schematics last nite and it all makes sense, ill work off them and try probe for the load issues. if i have any dramas ill get back on here and voice my screams of pain Smile

Scott Stites · Posted: Fri Aug 08, 2008 11:41 am Post subject:

Here's the latest build doc with Luka's corrections and one from SDuck. Link will be in the docs thread for eternal reference...

How's it going, Luka?

em_klee_build_issue5.pdf

Description:

Issue 5 EM Klee Build Doc

Download (listen)

Filename:

em_klee_build_issue5.pdf

Filesize:

1.81 MB

Downloaded:

1758 Time(s)

sduck · Posted: Sat Aug 09, 2008 8:13 pm Post subject:

IT'S ALIVE!!!! Head for the hills!

After several months of careful assembly, I finally powered up my klee this evening. And nothing happened - in fact it browned out the rest of the synth, causing all sorts of indignation on my part. After about 15 minutes of staring at the thing dumbly, poking it with various test apparatus, finally found a chip I had put in backwards. Fixed that, powered it up again, and it seems to work perfectly! Only minor, cosmetic problems that I've discovered so far - the pattern switches work backwards than I thought they were going to, and one step led isn't working, but I can live with these minor points for the time being. The led will be easy to fix, and the switches aren't labeled with which way they're supposed to go, so having them backwards is no big deal.

Now it's off to bed - I've been working today, I'm tired, and don't want to get into the calibration thing yet, but it's very gratifying to have the thing working after all these months.

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

good work sduck.

scott im still at the same point. had a quick look over the schems but that is about it, between uni and big party we organised i had no time. cant wait till the end of semester. Smile

sduck · Posted: Sun Aug 10, 2008 7:21 pm Post subject:

Ok, here's my only debugging question so far. It may just be a matter of me not understanding how this is supposed to work - it may be operating correctly.

Operating the klee in 16x1 mode, invert b off, pattern mode, with one bit on, the leds cycle through all 16 steps correctly, and output A puts out voltages according to the knobs on stages 1-8, but nothing on stages 9-16. Output B is similar, except the knobby output is steps 9-16, but nothing on stages 1-8. Output A+B is all 16 steps.

Is this how it's supposed to work? The "know the klee" doc indicates that in this mode both A and B are supposed to output 16 step unique patterns, not 8 and then 8, each half empty - or have I misread this?

8x2 mode and all the others seem to be working as I understand they're supposed to. Because everything else is working, I'm strongly in favor of the option I presented in sentence 2 of paragraph 1.

Thanks!

bugfight · Joined: Aug 02, 2007 Posts: 188 Location: Arlington, TX USA

sounds like it is working to me.
they are 2 unique 16 step patterns, just most of the steps are 0v...
add more active bits and you will see what's going on i think.

sduck · Posted: Sun Aug 10, 2008 7:48 pm Post subject:

I may have spoke too soon, I have other wierd stuff happening. When I switch invert B on (16x1 mode), as it clocks through the steps, I get all the stage leds coming on and staying on (and the pitch rising accordingly) until it cycles through the second time, where they turn off one by one. Very odd, and I've probably described it badly, but it doesn't seem right. Maybe it is?

Scott Stites · Posted: Sun Aug 10, 2008 7:49 pm Post subject:

Scott Stites · Posted: Sun Aug 10, 2008 9:11 pm Post subject:

sduck · Posted: Mon Aug 11, 2008 7:20 am Post subject:

OH!

That's just great. Here I am, finished with this huge build process, ready and willing to do the usual debugging I normally have to do, and the only thing that needs fixing is my expectations? Man. I guess it's peanuts to the syndicate, eh?

So I fixed the led, the thing works as it should - I'm happy! Now it's calibrate the ranges, which are seriously wacked currently, put the knobs on, and we're in business. (oh, and build that new cabinet for all the displaced modules) Yay!

And yes, I'll post samples when they occur. Thanks Scott and all!

Scott Stites · Posted: Mon Aug 11, 2008 8:08 am Post subject:

Cool!

And Bugfight, I saw the flicker of the DFW show - that is one frikkin' beautiful Klee. The offset is a very nice feature addition. Lots of times I wish I could just shift the whole Kahuna up or down easily. I have to use a mixer and a DC offset to do that.

Cheers,
Scott

bugfight · Joined: Aug 02, 2007 Posts: 188 Location: Arlington, TX USA

fonik · Posted: Mon Aug 11, 2008 1:39 pm Post subject:

fonik · Posted: Mon Aug 11, 2008 1:42 pm Post subject:

Scott Stites · Posted: Mon Aug 11, 2008 5:55 pm Post subject:

I feel for y'all - it's a lot of work just explaining the damn thing. Laughing

I saw the video of the DFW show - I think I caught a snippet of Bugfight explaining the gate outputs. Fonik - didn't know you'd showed up at a meet with the Klee!

And the idea of mixing with the Baby 10 into a quantizer - that would be fun...