| Author |
Message |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Fri Oct 24, 2008 12:24 pm Post subject:
|
 |
|
Well I finished building my Klee today. Unfortunately its not working.
Upon applying power, pattern LED #16 alone lights up (in both pattern switch positions). However, flipping the invert B switch shuts it off. Pattern switch #16 is working according to my multimeter.
None of the other pattern LEDs will light after throwing pattern switches and pressing Manual Load.
any ideas?
Pete |
|
|
Back to top
|
|
 |
State Machine
Janitor


Joined: Apr 17, 2006 Posts: 2810 Location: New York
Audio files: 24
|
Posted: Fri Oct 24, 2008 1:31 pm Post subject:
|
 |
|
| Quote: | | Well I finished building my Klee today. Unfortunately its not working. Sad |
Well, it's a big build Pete and it's no wonder. mine did not work either upon first power up. We will get it up and running
| Quote: | | Upon applying power, pattern LED #16 alone lights up (in both pattern switch positions). However, flipping the invert B switch shuts it off. Pattern switch #16 is working according to my multimeter. |
I suspect something is going wrong in the area of U5. When you power up and then press LOAD, the following pins should be at nearly 0V. U5 pins 1 and 13. If not, then do further investigation in that area like bent under pins, wrong components, etc ....
| Quote: | | None of the other pattern LEDs will light after throwing pattern switches and pressing Manual Load. |
Yes, the LOAD circuitry is whats giving you the trouble.
The CD4034 truth table is a bit tricky as there are several data transfer modes associated with this chip. The load circuitry keeps it all straight when you hit the manual load and if anything is wrong with it, things go sour.
Also, Keep in mind, however, that there have been several BAD chips of this type floating around. Please tell me the marking on yours so I can tell if you have the bad lot. Troubles with the CD4034 IC's can drive one nutty !!! They can exhibit all manner of weird behavior when they fail.
Anyhow, Lets start here with my first suggestion and work our way through things from there
OK, I will check back later.
I am working on Scott's "Appendage" Ribbon Controller PC board layout for a bit tonight then I will check back.
Bill |
|
|
Back to top
|
|
 |
Luka

Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
Posted: Fri Oct 24, 2008 1:48 pm Post subject:
|
 |
|
that is similar behavour to mine photon
except one side runs (except still no pattern switches)
perhaps you got 2 bad 4034s
im going to swap my dodgy 4034 in the next couple of weeks (finally) and will report back _________________ problemchild
melbourne australia
http://cycleofproblems.blogspot.com/
http://www.last.fm/user/prblmchild |
|
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
|
|
Back to top
|
|
 |
State Machine
Janitor


Joined: Apr 17, 2006 Posts: 2810 Location: New York
Audio files: 24
|
Posted: Fri Oct 24, 2008 6:14 pm Post subject:
|
 |
|
| Quote: | | I'll run through the things you suggested. I don't think I'll be able to tonight, but hopefully before the weekend is over. |
OK you can look at those thingsbut I can see from that image that you have the bad lot of chips .... PM me your address Pete and I will mail out a set of (2) TI chips right away free of any charges ...
| Quote: | | Excited to hear you're working on the Appendage! That thing has me foaming at the mouth. |
It is an amazing ribbon controller and very capable S/H. So many options!
OK, I will await your PM with that address
Also, if anyone else has this same IC as shown in the photo Peter took with those same markings, please let me know and I will send you a set of CD4034 IC's that WILL work.
Bill |
|
|
Back to top
|
|
 |
Luka

Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
Posted: Fri Oct 24, 2008 7:38 pm Post subject:
|
 |
|
no pete not yet
im looking into buying a laser printer at the moment
im think i may have found an ex-office a3 laserjet (for $5 )
and then ill gather some paper for testing and some aluminium scraps
do i have to make an aluminium bracket to hold the plate? it says somewhere that the bracket should be made from the metal you are trying to etch and i noticed you had a brass bracket. im not too sure how to make something from alu tbh. is there alu solder or something? _________________ problemchild
melbourne australia
http://cycleofproblems.blogspot.com/
http://www.last.fm/user/prblmchild |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Oct 24, 2008 10:01 pm Post subject:
|
 |
|
| Quote: | | However, flipping the invert B switch shuts it off. Pattern switch #16 is working according to my multimeter. |
Sorry for being out of the loop on this - yep, Bill is right - it won't hurt to get a couple of non-suspect CD4034s.
However - I'm not entirely certain a bad CD4034 would cause pattern LED to switch off by flipping Invert B (if the pattern isn't being clocked). Pin 1 of U7 (CD4034) supplies that signal. The output of Pin 1 feeds the pattern LED, voltage stage, and also goes to pins 12 and 13 of U4, which is a CD4093 NAND gate. It also goes straight to the Invert B switch.
The Invert B function acts by inverting the output of this pin when it's being fed back to the input of either U6 (in 16X1 mode) or the input of U7 (8X2 mode) - this happens downstream of the bit 16 LED. In actuality, U4 always inverts this signal, but the Invert B switch decides whether the input of the shift register the bit is feeding is inverted or not - the switch should only pass the signal that is present on pins 12 and 13 of U14 (Invert B off) or the signal that is present at pin 11 of U14 (Invert B On). It doesn't actually invert it at Pin 1 of U7, so flipping the Invert B switch shouldn't affect whether step 16 is lit or not, if the registers are not being clocked.
So, in the meantime, you might want to make sure the Invert B pin is wired up correctly - the "center post" (pole) of the switch should have continuity to Pins 1 and 5 of U9 (CD4053). One "outside" pin (throw) of the switch should have continuity to pin 11 of U4 (CD4093) and the other "outside" (throw) of the switch should have continuity to pins 13 and 14 of U4 (CD4093). This is all on the Digital Board, BTW.
Cheerios,
Scott _________________ My Site |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Oct 24, 2008 10:39 pm Post subject:
|
 |
|
It's quite possible that if the invert B switch is wired up incorrectly, the ouput of U4 pin 11 (CD4093) could be fed directly back to pin 1 of U7 (CD4034) and that would illuminate the LED when the switch was in a particular position. Pin 11 of U4 will be high if the CD4034 is putting out 0V - its input is held low by R89 in that case, so its output would be high, and that high could be what is lighting the LED instead of the CD4034 (IE, the CD4034 may not be doing anything at all).
Once you've made sure that all is OK, then let me know if your clock LED is flashing with your clock input. On my Klee, on power-up I don't get a pattern load on power up until I've clocked the Klee at least once.
If your clock LED is flashing with the clock input, then check to be sure you are getting a clock signal at pin 15 of U6 and pin 15 of U7 (the two CD4034s). If you are, and you have another LFO, or any voltage source that will transition above two V and below, try sending an external load signal into the the external load jack. It could be, in this case, we might want to take a closer look at the load circuitry. The CD4013 (U5) handles this job, though it does require the circuitry before it to feed it a signal it can recognize. Sending an external load signal eliminates the manual circuitry for the time being and just concentrates on the main portion responsible for loading the patterns.
I think I may cover this portion in the rather truncated "Theory of Operation" thread. _________________ My Site |
|
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Sat Oct 25, 2008 8:37 am Post subject:
|
 |
|
Hey Scott
Thanks for the help.
Yup...the invert B switch was wired wrong. Good call.
I tried externally clocking. It seems to work. The Rate LED flashes in time with the LFO (as do the gate bus 2 and master bus gate LEDs, (should they? Maybe I should double check all the wiring)). I haven't tried seeing if I have a clock signal at pin 15 of U6 and pin 15 of U7 yet. I'll get back to this when I can spend some real time.
Thanks again
pete |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Oct 25, 2008 10:06 am Post subject:
|
 |
|
Cool! Yes - I'd say both your master gate bus and gate bus 2 flashing at the same time is normal right now. If none of your pattern LEDs are lit, that likely means the CD4034s aren't putting out any bits. With no bits, that means no signal is reaching gate bus 1 or gate bus 3 regardless of how the gate bus switches are set. Gate Bus 2 is logical NOR, so if no signal is reaching gate bus 1 or gate bus 3, then gate bus 2 will put out a signal with each clock (neither 1 NOR 3 are high).
Master Gate Bus will always flash with the clock.
Take care,
Scott _________________ My Site |
|
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Sat Oct 25, 2008 2:03 pm Post subject:
|
 |
|
While looking over my wiring, I noticed that only one wire was connected to the Clock Enable switch (its wired to Clock input signal lug). There were no signs of a broken wire or even that the other lug ever had anything soldered to it. I wasn't sure how I had overlooked this, so looking through the build doc's I think I found a small error:
on page 35, it shows the above mentioned Enable Switch lug to Clock Input signal lug connection.
then on page 60, wire 2 of P102 cable is shown going to the Clock Input signal lug (not the other lug of the Clock Enable Switch).
I don't think this is the cause of my Klee's troubles at the moment, but it you may want to note it for a future rev. of the Build Docs.
apologies if this was already mentioned somewhere in this (or another) thread. |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Oct 25, 2008 3:02 pm Post subject:
|
 |
|
| Quote: | on page 35, it shows the above mentioned Enable Switch lug to Clock Input signal lug connection.
then on page 60, wire 2 of P102 cable is shown going to the Clock Input signal lug (not the other lug of the Clock Enable Switch). |
Yeah, in issue 5 of the build doc, a note was added on page 60 about that connection. Reason being, at the time the build doc was drafted, the clock enable switch was treated as an optional function that could be added, so the clock enable switch wasn't on Figure 6-11.
Cheerios,
Scott _________________ My Site |
|
|
Back to top
|
|
 |
Luka

Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
Posted: Thu Nov 06, 2008 2:20 am Post subject:
|
 |
|
damn!
i finally got a new 4034 and it burnt out as soon as i put it in
it kinda worked for a minute then the LEDs flickered and it went back to the behaviour of how it used to work with a busted 4034.
why would it overload like this? any ideas
the circuitry seems find on either side of my 4034s as i can swap the living 4034 from socket to socket and the sequencer works accordingly. strange
what an anti-climax! _________________ problemchild
melbourne australia
http://cycleofproblems.blogspot.com/
http://www.last.fm/user/prblmchild |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Thu Nov 06, 2008 7:38 am Post subject:
|
 |
|
| Quote: | | i can swap the living 4034 from socket to socket and the sequencer works accordingly |
Does this mean your remaining CD4034 works as expected in either socket? IE, in 8X2, it'll work OK for register A while it's in the U6 socket, and then work OK (in 8X2) for register B while it's in the U7 socket?
What brand was the newly-defunct CD4034 - TI or...?
Lemme know!
Take care,
Scott _________________ My Site |
|
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Thu Nov 06, 2008 1:31 pm Post subject:
|
 |
|
Ok-
I replaced the U6 & U7 (thanks Bill!) and rewired the invert B switch,
but the Klee still does not load a pattern.
So I've been going over my Klee with a fine tooth comb. I re-checked all wiring and cables and everything seems fine.
However...When I put my DDM on pins 1 and 2 of J1 (+V & digital GND; digital board) I'm reading about 60K (!) between those pins.
Then I checked the power and ground pins on all chips against their respective pins on J1 (on the digital board).
all read zero ohms except the U14 & U15 (the 2 LM324's)
a) pin 14 (U14 & U15) and pin 1 (of J1) reads 62K on both chips
b) pin 7 (U14 & U15) and pin2 (of J1) reads ~5k on both chips.
visual inspection reveals no shorts that I can see.
any suggestions?
thanks!!
Peter |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Thu Nov 06, 2008 1:58 pm Post subject:
|
 |
|
So far you found two of the three sections that were corrected in build issue 5
| Quote: | Changes From Issue 4:
Corrects errata of issue 4 (Page 52 mis-label of Pin1, P101-P211).
Page 60 - Inserted note about connection to Clock Enable Switch
Pages 92 & 93 - Corrected LM324 power pin labeling |
The power pins on the LM324s are pin 4 (positive) and pin 11 (ground). So, I bet you're fine there. Also, you may be fine with the resistance between the power and ground pins of J1 - they'll read something other than infinite if you've got your ICs in the sockets, etc.
Okee doke. IIRC, your clock LED was flashing with each pulse, right? If that's the case, we should check some voltages:
1. Each time your clock LED flashes, you should see the voltage on pin 15 of each CD4034 go from ground to close to the positive rail. Are you getting that?
2. If so, make sure your voltage on the IC power pins is at the positive voltage. The ICs that really matter here are:
U1 Pin 8
U2 Pin 14
U3 Pin 14
U5 Pin 14
U6 Pin 24
U7 Pin 24
Do you have an oscilloscope or a logic probe?
Lemme know!
Scott _________________ My Site |
|
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Thu Nov 06, 2008 2:08 pm Post subject:
|
 |
|
| Scott Stites wrote: | So far you found two of the three sections that were corrected in build issue 5
Scott |
HA!
Yes, clock LED flashes in time with input pulses. I'll check those voltages and let you know. Thanks again.
Yes I have a scope.
Peter |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Thu Nov 06, 2008 2:20 pm Post subject:
|
 |
|
| Quote: | | Yes I have a scope. |
You can check those clock pulses on pin 15 of the CD4034s with that!
The key player in loading patterns is U5, the CD4013. So (just to be sure) check to see that it is indeed a CD4013 and doesn't have any folded pins or anything else weird about it.
If U5 looks OK, check pins 13 and 14 on each CD4034 to see if they pulse each time a pattern is told to load. These are very, very fast pulses. The easiest way to check them is to send a signal into your external load jack, preferably an LFO. You should see your oscilloscope twitch, or display a pulse (depending on how nice and spiffy your oscilloscope is) with each repetition of the LFO that's pulsing your external load input. In fact, you don't even need a clock hooked up for this test - we just need to be sure that your CD4013 is sending out a pair of pulses each time the external load goes high. Your LFO can be cooking along pretty quick, too - it's a fast circuit, so even audio rate should give you pulses.
Cheerios,
Scott _________________ My Site |
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Thu Nov 06, 2008 2:24 pm Post subject:
|
 |
|
BTW, all of the pulses we're looking for here should pulse low to high. The pins should be at 0V when there is no load signal. _________________ My Site |
|
|
Back to top
|
|
 |
Luka

Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
Posted: Thu Nov 06, 2008 4:23 pm Post subject:
|
 |
|
| Scott Stites wrote: | | Quote: | | i can swap the living 4034 from socket to socket and the sequencer works accordingly |
Does this mean your remaining CD4034 works as expected in either socket? IE, in 8X2, it'll work OK for register A while it's in the U6 socket, and then work OK (in 8X2) for register B while it's in the U7 socket?
What brand was the newly-defunct CD4034 - TI or...?
Lemme know!
Take care,
Scott |
yes the remaining 4034 works in either socket
in 8x2 mode i have a properly functioning 8step sequencer (except my pattern switches and load switches are not working) and this applies to the 4034 being in either socket, there is no difference
my chip is
MC14034B
these chips are expensive too, shame for it to burn so quick
when it was in and working, prior to it burning out, my pattern switches were not working as expected but i cant really talk about it too much because as soon as i started working out their behaviour the chip burnt
all i know is they were not completely influencing the pattern LEDs as i expected them to _________________ problemchild
melbourne australia
http://cycleofproblems.blogspot.com/
http://www.last.fm/user/prblmchild |
|
|
Back to top
|
|
 |
Luka

Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
|
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Thu Nov 06, 2008 5:56 pm Post subject:
|
 |
|
Hmmm....OK. I know Bill may not have a chance to be around his computer here for a little while, so don't get alarmed if you don't hear from him for a little bit. _________________ My Site |
|
|
Back to top
|
|
 |
Luka

Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
|
|
Back to top
|
|
 |
State Machine
Janitor


Joined: Apr 17, 2006 Posts: 2810 Location: New York
Audio files: 24
|
Posted: Fri Nov 07, 2008 1:23 pm Post subject:
|
 |
|
| Quote: | Bill perhaps ill take you up on your offer to send one out
i'll PM you my details
|
OK, back faster than anyone had though .... Sure, send me a PM with your shipping information and I will send you two "TI" chips that seem to always work. The MC part number you have is an RCA which have not given anyone trouble as of yet and I sold a bunch of those.
Before you place the new chips into your circuit though, please make sure NO inputs are floating. Older CMOS devices, when their inputs are left floating, can tend to oscillate and literally BURN themselves out due to excessive current draw. Why is this? It's because the CMOS transistors are in transit more often and draw larger average currents that could be damaging. this is why CMOS chip current consumption is non-linearly and directly proportional to input frequency.
Bill |
|
|
Back to top
|
|
 |
|