Author |
Message |
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Mon Mar 09, 2009 11:58 pm Post subject:
40106 clock bleed Subject description: arrrhgg, why?!?! |
|
|
I'm having so many troubles with my lunetta.
I dont know why, but my 40106 clock signals are bleeding into and effecting each other.
Any one else have this problem?
Very standard wiring except that i have an led indicating with 1k resistors to indicate speed.
Is it a grounding problem? I've triple checked my grounds and it seems fine! If my leds are blinking, then i cant see how it could a grounding issue...
Would diodes on the outputs help? that doesnt make sense either since im not trying to modulate the clock outs with each other.
i cant figure it out! |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Tue Mar 10, 2009 6:39 am Post subject:
|
|
|
I've just finished yesterday, adding a CD4051 functioning as a:
- 1 input
- 8 output
multiplexer. Which of the 8 outputs that is active, is chosen by the 3 control bits of the 4051.
Each of the outputs goes to 2 output sections:
- a 4K7 resistor that is feeding super-bright LED, which goes to the common Ground
- 1k resistor that goes to the output banana jack
I have a 0.1uF ceramic cap. as close to the power input and ground pins of the 4051 as possible, as well as a 10uF cap. for the circuit board, that the 4051 is on.
I'm not having any bleeding problems at all.
Also in my machine, 1 of my 2, 4 bit shift registers has an LED in a slightly different fashion: the output goes through a 47k resistor into the Base of an NPN transistor. The transistor's Collector is connected +V and it's emitter outputs through a 4k7 resistor, to that output's LED. No bleeding.
Anyone else please correct me if I'm wrong in this, but I suspect that:
- the resistors for the LEDs from your 40106s, may be too small, allowing too much current to be pulled from the 40106s, at the same time - especially if there is up to 6 of the LEDs being run from the 40106.
I would use the 47k to transistor to 4k7 to LED method, for your 40106, to see if that helps.
Also - do you have a 0.1uF cap., as close to the 40106's power input as possible, for buffering the power, for that chip?
I hope this helps! |
|
Back to top
|
|
|
slacker
Joined: Nov 18, 2007 Posts: 301 Location: England
Audio files: 11
G2 patch files: 1
|
Posted: Tue Mar 10, 2009 12:02 pm Post subject:
|
|
|
I'd agree with Rych the 1k current limiting resistor for the LEDs is probably too small. Depending on your supply voltage and the LEDs you could easily be pulling 5-10mA per LED, multiply that by 6 if you're using one per CD40106 stage and that's quite a lot. You might even be stressing the chip, I can't remember how much current the outputs are happy with.
First thing I'd try is disconnect the LEDs and see if the bleed through goes away, if it does that's your problem.
On my Lunetta I use Ultrabright LEDs with 50k current limiting resistors which keeps the current draw down to about 0.05mA each. |
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Tue Mar 10, 2009 2:00 pm Post subject:
|
|
|
i'll try increasing the resistance of the rezzys in front of the leds and see how that helps. If not, i'll go with the NPN suggestion.
I'll also try adding a .1 uf cap to ground connected to the V+
If it is a problem with the leds, i might just take them out and replace them with a simple on/off or modulate switch to fill up the holes in my panel.
But, i did build up a second set of clocks with another 40106, this time excluding the leds, which i connected to common V+ and ground. I plugged the out put of the one oscillator into the amp and got a nice strong squarewave. But when i twist the knobs of my first set of clocks which is only interacting with 2nd via the power and ground commons, there still some noticeable interaction going on almost like some faint pwm.
Thanks for the suggestions. Every one here is always so helpfull.
Once i get my 2 sets of clocks worked, i'm going to finish trouble shooting my 4040 divider and then get onto that multiplexer which im very excited about. |
|
Back to top
|
|
|
Adam-V
Joined: Jan 29, 2007 Posts: 300 Location: Australia
Audio files: 1
|
Posted: Tue Mar 10, 2009 2:21 pm Post subject:
|
|
|
Sounds like a decoupling problem. Do you have the 40106 supply decoupled? If not stick a 100n cap from Vcc to Gnd as close to the chip as possible. That might solve your interaction issues.
Cheers,
Adam-V _________________ Digitalis Effect | Fractured Symmetry (www.spiralsect.com) |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Tue Mar 10, 2009 4:19 pm Post subject:
|
|
|
I just finished wiring up 6 more clocks (40106); my 4 Stage LFO and my Liquid Voice. (Somewhat modified Liquid HiHat).
I used proper decoupling (0.1uF / 100nF as closely as possible to each of the chips) as well as 10uF on the power input to each of the 2 boards.
I'm having absolutely no bleeding / cross-talk, with any of the new modules.
So it's a very good chance, that's the problem you're having. (Seeing above, that you had added a 0.1uF after.)
Here's to hoping for ya. |
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Tue Mar 10, 2009 8:46 pm Post subject:
|
|
|
would you guys expand a bit by 'as close as possible to the chips'?
Im a bit confused, what im gathering is your suggesting a cap to ground right next to pin 14 of the chips? |
|
Back to top
|
|
|
RF
Joined: Mar 23, 2007 Posts: 1502 Location: Northern Minnesota, USA
Audio files: 28
|
Posted: Tue Mar 10, 2009 9:15 pm Post subject:
|
|
|
Yup _________________ www.sdiy.org/rfeng
"I want to make these sounds that go wooo-wooo-ah-woo-woo.”
(Herb Deutsch to Bob Moog ~1963) |
|
Back to top
|
|
|
scriptstyle
Joined: Jan 22, 2008 Posts: 250 Location: nj
|
Posted: Tue Mar 10, 2009 10:53 pm Post subject:
|
|
|
i love this photo it should help you out a little good trick for old gear with bad engineering.
^photo stolen from the old crow site. |
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Tue Mar 10, 2009 11:02 pm Post subject:
|
|
|
so i just unhooked all the leds, nothing changed except the tonality of the 40106 osc, they still bleed together.
Then i added a 100 nf... which causes some pitch changes but thats it.
i dont get it, each one should be perfectedly isolated. Could it be my battery supply? Im planning on going off wallwart power once i get the whole unit built up, but for prototyping its just the 9volt. Last edited by stolenfat on Tue Mar 10, 2009 11:11 pm; edited 1 time in total |
|
Back to top
|
|
|
scriptstyle
Joined: Jan 22, 2008 Posts: 250 Location: nj
|
Posted: Tue Mar 10, 2009 11:08 pm Post subject:
|
|
|
i was just curious to see what my leds are pulling(ma). so i did quick test. im using small general purpose leds, on 5v, and a 4k7 limiting resistor. the result is .652ma i was shotting for less... |
|
Back to top
|
|
|
scriptstyle
Joined: Jan 22, 2008 Posts: 250 Location: nj
|
Posted: Tue Mar 10, 2009 11:17 pm Post subject:
|
|
|
ok you say tonality changed? this may be from the battery, but i dont think so. stupid question but do you have the outputs after the resistor for the leds? |
|
Back to top
|
|
|
synthmonger
Joined: Nov 16, 2006 Posts: 578 Location: flada
Audio files: 3
|
Posted: Wed Mar 11, 2009 3:00 am Post subject:
|
|
|
How bad is the bleed through? And they are effecting each other how exactly?
You're going to get some audio bleed through no matter what. It shouldn't be terrible though. |
|
Back to top
|
|
|
RF
Joined: Mar 23, 2007 Posts: 1502 Location: Northern Minnesota, USA
Audio files: 28
|
Posted: Wed Mar 11, 2009 5:24 am Post subject:
|
|
|
Do you have the outputs and/or LED wires bundled together on the way to the output jacks? You might be introducing some crosstalk that way. _________________ www.sdiy.org/rfeng
"I want to make these sounds that go wooo-wooo-ah-woo-woo.”
(Herb Deutsch to Bob Moog ~1963) |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Wed Mar 11, 2009 6:41 am Post subject:
|
|
|
Now, this is becoming a puzzle. Could you give us a close up photo of your circuit board; both sides? A long with a posting of the schematic that you're working from?? |
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Wed Mar 11, 2009 1:03 pm Post subject:
|
|
|
well... im going to be embarrassed- it's a really dirty circuit on protoboard, but i've checked everything, all my solder work, all the connections, its all groovy. I mean, the thing works and does what it supposed to but i hear every oscillator through each output.
(one knob even functions like a lowpass filter, which makes NO sense to me)
The schematic- there IS NO schematic. I'm just kinda going with the simple resistor and capacitor network. Each out put goes right to the jack now, no leds.
I did have one thought that maybe since my panel is made out of metal, maybe the pot bodys are causing the interaction through the panel's natural conductivity.... maybe i should try grounding them next time i get into it.
but heres some pictures: I dont know how these pictures are going to help, any one wanna catch a plane to SFO?
Description: |
|
Filesize: |
112.41 KB |
Viewed: |
20794 Time(s) |
|
Description: |
|
Filesize: |
62.34 KB |
Viewed: |
20794 Time(s) |
|
Description: |
|
Filesize: |
101.72 KB |
Viewed: |
20794 Time(s) |
|
Description: |
|
Filesize: |
122.46 KB |
Viewed: |
275 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
synthmonger
Joined: Nov 16, 2006 Posts: 578 Location: flada
Audio files: 3
|
Posted: Wed Mar 11, 2009 1:25 pm Post subject:
|
|
|
Organize your wires and make your solder joints better. That should help a lot.
How bad is the bleed-through? Like I said, you will definately hear each oscillator in every output, but it should be minimal. |
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Wed Mar 11, 2009 1:38 pm Post subject:
|
|
|
i like the chaos, its artsy.
but better solder work? If i had a pcb there would be no problem, but this protoboard stuff gets messy and i dont know if theres any way around it.
How would cleaner solder joints help isolate the sound? I've checked each one, they dont contact. (except now for the mess i got my self into last night, gotta get me some more braided copper)
anyways, i started building up a second set of clocks, i only have 3 of the oscillators completed and they already affect each other. _________________ home made noise and electronic ill-logic |
|
Back to top
|
|
|
scriptstyle
Joined: Jan 22, 2008 Posts: 250 Location: nj
|
Posted: Wed Mar 11, 2009 2:00 pm Post subject:
|
|
|
Ok so I counted 6 electrolitic caps for 6 osc, corect? 1 100n decoupling cap. Where is your 100uf supply decoupling cap? |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Wed Mar 11, 2009 3:01 pm Post subject:
|
|
|
I cant really see in the pics, but the unused CMOS inputs, are they tied up to anything? _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Wed Mar 11, 2009 5:01 pm Post subject:
|
|
|
whoa, no supply decoupling cap, i might have a 100uf, but i know i have a 220 uf. I'll try adding that from Vcc to ground and see what happens.
And also, all the inputs are used, all the outputs are used. So no need to tie empty pins to the ground.... right?
I started a better more organized set of clocks last night, which will probably take the place of this mess.
will continue my work later today.
thanks for the help and suggestions so far guys! _________________ home made noise and electronic ill-logic |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Wed Mar 11, 2009 5:15 pm Post subject:
|
|
|
stolenfat wrote: | And also, all the inputs are used, all the outputs are used. So no need to tie empty pins to the ground.... right?
|
Yup, right, used is used, it's just the unused ... CMOS can do really odd things with unconnected inputs. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Wed Mar 11, 2009 7:37 pm Post subject:
|
|
|
Looking at the solder side, this may not improve anything, but it does help in bug searching:
- when I built circuits on perfboard (like yours, with solder pads), I made sure that every single thing that came through the holes, was soldered. Every single pin on the socket ...... for the chip.
I would do the socket first and the start working around it for parts that are being added.
Sometimes, I would place the caps and / or resistors in to measure out the space that they take up, before soldering them down.
Once 1, 2 or more things were in place and looked comfortable, I would then fold the first leg down so that it touched the point that it were to be soldered to.
I would then FIRST solder the hole / pad, that that leg came through; cool it by blowing on it lightly, and THEN solder it to it's destination point.
(If the destination point were already soldered, I would apply just a LITTLE bit more, for it's internal flux to give a little bit of reaction to the solder destination.) And then again, blow on that and cool it.
(With this method that I learned from my dad in the early 70's, I've rarely had cold solder joints.)
With your mention of hearing every clock FROM every output, I'm wondering if you're having a Ground problem. A lack there of ....... a good example of that, is when you're working with a multiplexer (CD4051 or CD4052 as examples.)
Twice, when I've not had the Ground to the IC itself soldered properly, if at all and I power the chip up - ALL of the LEDs would come on. (Each of them searching, for a proper route to Ground.)
That sounds mildly similar to the trouble that you're having.
You mentioned this problem happening again, with your new chip with only 3 clocks hooked up.
I can't see the same mistake being made twice, but ..... it's such a simple circuit that I'm having a minor difficulty figuring out what else it could be.
Finally - as to cross talk - in the tune that I recorded today, in setting up for it - yes I CAN hear VERY, VERY FAINTLY, cross talk from the 6xosc. chip, coming through. Once everything is up and running though, it's completely drowned out and doesn't affect anything in any way that I'm able to hear.
Ooops forgot - for a larger decoupling cap for your chip, you could with anything from 1uF to 100uF. If you were going to go with a 220uF (overkill for a single chip), I'd then see if I could add something a lot smaller as well, as the 0.1uF. 0.47uF or 1uF? (For a single chip, I THINK, a 1uF with a 0.1uF should be fine.)
I've got 10uF on each circuit board of mine, with 0.1uF for every chip's power input. (1 for + and another for -, if the chip is bipolar.) |
|
Back to top
|
|
|
RF
Joined: Mar 23, 2007 Posts: 1502 Location: Northern Minnesota, USA
Audio files: 28
|
Posted: Mon Mar 16, 2009 7:06 pm Post subject:
|
|
|
So how are you coming on narrowing down that clock-bleed, stolenfat? _________________ www.sdiy.org/rfeng
"I want to make these sounds that go wooo-wooo-ah-woo-woo.”
(Herb Deutsch to Bob Moog ~1963) |
|
Back to top
|
|
|
stolenfat
Joined: Apr 17, 2008 Posts: 476 Location: Sunny Oakland California
Audio files: 1
|
Posted: Mon Mar 16, 2009 10:43 pm Post subject:
|
|
|
well, i kinda got frustrated with it and decided to take a break from the circuitry. I cant figure it out for the life of me.
I never got the bleed to dissipate.
I promise to come back with more of an update on the situation soon, and thanks again every one for the help. _________________ home made noise and electronic ill-logic |
|
Back to top
|
|
|
|