Author |
Message |
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Sat Dec 13, 2008 9:57 am Post subject:
|
|
|
Ok, so the Klee is finished . I'm thinking in soldering the capacitor. Ok, 10n and 100n works. But they cut the high frequencies... maybe founding a capacitor value that solves the bounce but that respects the high frequencies in the clock in.
Where I can found it? locking for below 10n or over 100n?
Which will be the appropriated capacitor for this job? MLCC? metalized polyester film? ceramic?
Thanks! |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Dec 13, 2008 9:58 am Post subject:
|
|
|
Can you try putting a 1n cap between your switch lug and ground? I'm curious if it helps yours any. I suppose this could go between the signal lug and ground on the clock input as well. _________________ My Site |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Dec 13, 2008 10:01 am Post subject:
|
|
|
OK - just saw your message as I posted this. Try the 1N thing like I mentioned.
A work-around would be to replace the switch with an On-Off-On switch. Center position would disable the clock. One direction would be sub-audible rate - this one would switch in whatever cap value we want here. The other position would be without the cap - at audio rate, it doesn't matter where it goes when you flip the switch.
I'm going to go see if this 1n value restricts audio rate any..... _________________ My Site |
|
Back to top
|
|
|
sduck
Joined: Dec 16, 2007 Posts: 459 Location: Nashville
Audio files: 5
|
Posted: Sat Dec 13, 2008 10:02 am Post subject:
|
|
|
totally off topic -
50 pages! YAY!
is this the longest thread on electro-music? |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Dec 13, 2008 10:34 am Post subject:
|
|
|
I dunno - 50 pages....some kind of record maybe.
Sound - 1n doesn't seem to restrict my top frequency. Mine peters out at 10K pulse from a VCO whether the 1n is there or not.
Question is - does 1n make it any better for your clocking, using a DC level as input.
Again, I have to say, that is one cool Klee (I've been going over your pics trying to spot anything amiss). I wish I had the technology (read skill) to build a case like that.
BTW, I got sidetracked - have you tried putting in audio rate clock and audio rate external load, and then varied the two frequencies around while driving a VCO? Whoooaa.... _________________ My Site |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Dec 13, 2008 10:39 am Post subject:
|
|
|
No actually it stops clocking at around 12.5 kHz using the 1n. There's a slight bit of top end that it loses, but not that much. Have you had a chance to try 1n? _________________ My Site |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Sat Dec 13, 2008 11:04 am Post subject:
|
|
|
Scott Stites wrote: | Can you try putting a 1n cap between your switch lug and ground? I'm curious if it helps yours any. I suppose this could go between the signal lug and ground on the clock input as well. |
Let's see.
With 1n capacitor from ground leg jack IN, to signal lug jack I don't achieve positive results , sometimes the clock skip three or four steps.
With 1n capacitor in the switch (side of D2) leg to digital ground, the problem seems solved, it respects the audio range IN but some times when you turn on the switch it skips one step. Why?
As I observed, if when you turn ON the enable clock switch, the wave from the LFO which you are clocking your klee is in the positive period, I mean positive over the threshold of the clock input ( think is 2 volts?) the Klee skips one step. We think that because there is a bounce.
When you turn ON the switch and the LFO (clock) wave is under 2V, the Klee do not skip any step, just wait for the wave go over 2V.
Why it happens to me? Because I was clocking the Klee with the non AC output of a Bananalogue VCS, and this output of this module gives you a wave in the positive range. (Edit: It is an envelope that the end of the cicle trig the envelope again) So I had more possibilities of turn the switch when the clock in is receiving over 2V.
It have sense, no? Last edited by Sound on Sat Dec 13, 2008 11:17 am; edited 2 times in total |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Sat Dec 13, 2008 11:06 am Post subject:
|
|
|
Didnt see yor two last posts, I write English very slow! |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sat Dec 13, 2008 12:03 pm Post subject:
|
|
|
Quote: | Let's see.
With 1n capacitor from ground leg jack IN, to signal lug jack I don't achieve positive results , sometimes the clock skip three or four steps.
With 1n capacitor in the switch (side of D2) leg to digital ground, the problem seems solved, it respects the audio range IN but some times when you turn on the switch it skips one step. Why? |
OK - good. A quick question: I couldn't tell from your pictures, so does your clock input jack have a connection from its ground lug to digital ground?
Quote: | As I observed, if when you turn ON the enable clock switch, the wave from the LFO which you are clocking your klee is in the positive period, I mean positive over the threshold of the clock input ( think is 2 volts?) the Klee skips one step. We think that because there is a bounce. |
If I understand your question correctly, this is not due to bounce (it's normal). If you flip the switch on while the clock is high, this will naturally advance one step. The reason is, if there is a high on the input, when the switch is flipped, it sends a transition from low to high to the clock input, which is interpreted as a clock pulse. Normally, when I want to start at a certain point, I set the pattern so that it is on the step before the step I want to start on.
Quote: | When you turn ON the switch and the LFO (clock) wave is under 2V, the Klee do not skip any step, just wait for the wave go over 2V. |
Exactly.
Quote: | Why it happens to me? Because I was clocking the Klee with the non AC output of a Bananalogue VCS, and this output of this module gives you a wave in the positive range. (Edit: It is an envelope that the end of the cicle trig the envelope again) So I had more possibilities of turn the switch when the clock in is receiving over 2V.
It have sense, no? |
Yes, that makes sense.
Back when I was working on the Klee, I wanted to synchronize the manual/external load with the clock. In other words, right now when you press load, it loads instantly. You have to be careful to load right at a clock pulse in order for things to go smoothly (if that's what you're after). I didn't give much thought to the clock - the clock enable switch was something I saw Fonik do on an early design, and it made sense. Sometimes you just want to stop the clock while you set things up, or just to silence the Klee.
I had plans at one time of producing a synchronization module that would not only work well with the Klee, but would also work well with other tasks (so having the module without a Klee would be useful). That got sidetracked by this guy that wanted to jump in and start producing this kind of stuff on a "mini-Klee" or some such thing - that went away for one reason or another. I think perhaps it's time to design such a module - I don't recall seeing a sychronization module like what I have in mind.
What I'm thinking of is a module that (among other things) has an "arm" function, like on a scope. You arm the module either with a momentary switch or an external pulse. Then that module waits until it receives a clock pulse to execute the task. This would be very useful in syncing not only a Klee, but other modules to a master clock. It could perform other Klee helpful (but not Klee-centric) tasks as well.
Thanks for the feedback - let me know about that input clock jack. I imagine it's grounded, but thought I'd double check.
Take care,
Scott _________________ My Site |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Sat Dec 13, 2008 12:22 pm Post subject:
|
|
|
Scott Stites wrote: |
Sound - 1n doesn't seem to restrict my top frequency. Mine peters out at 10K pulse from a VCO whether the 1n is there or not. |
Mine, with the capacitor 1n in the switch leg stop to clock a 25,2Khz
Quote: | Question is - does 1n make it any better for your clocking, using a DC level as input. |
Yes with the 1n capacitor the klee skips only one step or sometimes 2. I will try other values under 10n.
Quote: | Again, I have to say, that is one cool Klee (I've been going over your pics trying to spot anything amiss). I wish I had the technology (read skill) to build a case like that. |
Thanks! lot of work designing the front panel!
I just have added the optional (not glided) voltage outputs. The front panel is less beauty but these otputs are necessary! It is more powerful.
I like very much that I put three switches instead a rotary switch. They give me 8 different positions and is a very easy mod, very easy to wire: pin 1 P209 to one leg of the three switches, units switch to pin2, tens switch to pin 3 and hundreds switch to pin 5.
About the case... is the first time, Actually I only folded the two flaps! And now I have problems to put the screws. It would be better to fold the flaps to the opposite side. Next case that I will do will be better, it is relativity easy with the woods and the clams, with 1mm thick aluminum. It will be better with iron rulers instead of wood.
http://farm4.static.flickr.com/3243/3104602193_3859f74928_b.jpg
http://farm4.static.flickr.com/3248/3104604809_ec7529f76d_b.jpg
http://farm4.static.flickr.com/3108/3105439118_4a6834e58e_b.jpg
Quote: | BTW, I got sidetracked - have you tried putting in audio rate clock and audio rate external load, and then varied the two frequencies around while driving a VCO? Whoooaa.... |
I will try it when i finish to mount the klee. |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Sat Dec 13, 2008 12:33 pm Post subject:
|
|
|
Scott Stites wrote: |
OK - good. A quick question: I couldn't tell from your pictures, so does your clock input jack have a connection from its ground lug to digital ground?
|
I post only for answer this question, Yes the jack is grounded in the digital ground, I revised all the strap front panel ground visual and with the multimeter and is OK.
I keep reading your last post. Thanks. |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Sun Dec 14, 2008 5:45 am Post subject:
|
|
|
Scott Stites wrote: |
Sound wrote: | As I observed, if when you turn ON the enable clock switch, the wave from the LFO which you are clocking your klee is in the positive period, I mean positive over the threshold of the clock input ( think is 2 volts?) the Klee skips one step. We think that because there is a bounce. |
If I understand your question correctly, this is not due to bounce (it's normal). If you flip the switch on while the clock is high, this will naturally advance one step. The reason is, if there is a high on the input, when the switch is flipped, it sends a transition from low to high to the clock input, which is interpreted as a clock pulse. Normally, when I want to start at a certain point, I set the pattern so that it is on the step before the step I want to start on. |
I mean that when I turn ON the clock enable switch and the Wave is in the positive range as well inputting positive DC, the Klee advance two or more steps. And when the wave is in the negative range the Klee just wait for the next pulse and advance correctly only one step.
It happens to your Klee? |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sun Dec 14, 2008 11:04 am Post subject:
|
|
|
On mine, with 1n, it actually does pretty good. If I crank it up and down fast enough with a DC input, it'll eventually skip a step now and then, but that's a fairly unnatural use for the switch.
It does seem like yours is really sensitive for some reason; for example, mine has never clocked when turning the switch off when using a DC input. I had an offline conversation with State Machine; could be increasing R19 would stretch the pulse out. Actually - it's only around 30 uS, so increasing it would widen the time the pulse stays open, and probably keep your audio range intact. Thing is, of course, that has to be desoldered and played with. Basically R19 determines how fast C17 charges and discharges. So, a bunch of bouncy pulses would blend in with one pulse, and the registers wouldn't advance.
Not sure if we've gone there, but it might be worth it to check that C17 is 1n (1000 pF), and that R19 is at least 22K. And, pin 13 of U2 should be high 99.9999% of the time (it will only pulse low when U1 transistions from low to high, either from a clock pulse or when flipping the clock enable switch on with a DC input applied).
If you have the time, you might see how wide your pulses are coming from pin 12 of U2 - it should pulse high with each clock pulse.
Take care,
Scott _________________ My Site |
|
Back to top
|
|
|
EdisonRex
Site Admin
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
|
Posted: Sun Dec 14, 2008 5:54 pm Post subject:
|
|
|
Right. well I made the final push today, built the power supply this morning (Thomas Henry's 21st century one minus the 5V part) and finished the wiring at about 11pm UK time after having spent 6 evenings and the entire weekend wiring this m*****f****r.
Along the way, I discovered that I neglected to actually design the random reference LED, and the rand/patt switch onto the panel. I have space, but it's gonna be a ghetto mod, and in the meantime I have a random influence to the Klee which isn't bad, but it was mightily confusing at first (think all pots to the left).
Also despite all advice in the build manual, and after carefully checking all of the pattern switch orientations, they are exactly backwards, but I guess that just means down means engaged like a load switch should do. The merge switches, orientated the same way, work as expected.
I missed where I was supposed to wire up the ext/int range switch, so it's undone too.
And to top it all off, I grabbed a green LED from my drawer to use in the power supply and what do you know, it's a blinking one
I had some of the pattern LEDs wired out of sequence. I was seeing 2-1-3-4-5-6-7-8-12-11-10-9-13-14-15-16. Note to self: when splitting ribbon cables, make sure you really know the damn sides you're splitting. It was obvious when checking the load switches that the LEDs, and not anything else, was incorrectly wired.
The random being permanently on confused me even though I knew I didn't have the switch wired (it's on if you don't wire the switch, I can see now). Turning the reference pot up has calmed down the effect, and the random pot off. It still seems to have an effect.
More tomorrow. It does sequence, it does load, the gates all work, it needs a calibration, and I am going to regret using bright blue LEDs since the place it's going in the studio will be around eye level. Although, because of the way I mounted them, they are just nice and bright from any angle except straight on, where it's like getting hit with 16 blue lasers.
Anyway more later. I better read up the parts of the manual I skimmed, about the calibration and testing. And drill two more holes. And find the wire for the external range switch.
Anyway, yeah, it's up.
I'll do a picture or movie or something tomorrow. I'm too tired to do it now. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated.
Home,My Studio,and another view |
|
Back to top
|
|
|
EdisonRex
Site Admin
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
|
Posted: Mon Dec 15, 2008 12:11 pm Post subject:
|
|
|
My Klee
Couple more wiring problems worked out, had to cut almost all of the wire ties, but I have more. The silliest thing I did was put a 7805 in instead of a 7815, it ran but had some odd quirks and would not calibrate (doh!). All better now.
Oh yeah, fixed the lack of a rand/patt switch by changing the random signal to a normalled connection. Just the random reference level LED to find a home for. Won't be a big deal. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated.
Home,My Studio,and another view |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Mon Dec 15, 2008 12:39 pm Post subject:
|
|
|
Cool, E-Rex!! Looks like you slid through the obstacle course and came out in one piece.
Great panel! I think that's the most compact Klee I've seen. Nice save on the rand/pat switch solution. I doubt you'll miss that LED.
Bright LEDs - What did you use for limiting resistor value?
Oh, yeah - IMO, all power supply LEDs should blink.
Take care,
Scott _________________ My Site |
|
Back to top
|
|
|
EdisonRex
Site Admin
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
|
Posted: Mon Dec 15, 2008 1:17 pm Post subject:
|
|
|
Bright LEDs - 6K8, but these are really bright LEDs. They were pretty cheap so if they bother me I'll swap em but to tell you the truth it's pretty cool.
I think I still have a problem with gate bus 2. 1 and 3 are fine, so I can't see offhand what's going on with not getting gates all the time and I get what seems like switching noise when I use gate bus 2 exclusively.
As to compactness, it seems to be quite usable. Remember that I had the mockup around for almost a year. I started designing this panel back in November of 2007... I should have the rest of the case done after Christmas, but at least the Klee is ready for NYE. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated.
Home,My Studio,and another view |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Mon Dec 15, 2008 5:26 pm Post subject:
|
|
|
Very nice Edison! How did you paint the front panel? .
It is very pleasant to finish a job after long time designing it, actually the Klee value go above its work!
Scott, I will update when I have tested the last things.
Regards! |
|
Back to top
|
|
|
EdisonRex
Site Admin
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
|
Posted: Tue Dec 16, 2008 4:30 am Post subject:
|
|
|
Hi Sound, yes, thank you - I think we both spent almost the same amount of time designing and building. Even though they are very different, you panel is excellent too.
The panel decoration is photographic inkjet paper adhered to the panel with contact cement and then layered with clear coat spray. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated.
Home,My Studio,and another view |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Tue Dec 16, 2008 10:04 am Post subject:
|
|
|
Scott Stites wrote: |
It does seem like yours is really sensitive for some reason; for example, mine has never clocked when turning the switch off when using a DC input. I had an offline conversation with State Machine; could be increasing R19 would stretch the pulse out. Actually - it's only around 30 uS, so increasing it would widen the time the pulse stays open, and probably keep your audio range intact. Thing is, of course, that has to be desoldered and played with. Basically R19 determines how fast C17 charges and discharges. So, a bunch of bouncy pulses would blend in with one pulse, and the registers wouldn't advance. |
Although, that the solution of the 100n capacitor from clock enable switch could be enough (The audio range input is cool but maybe not a very useful characteristic in the klee) I tested increase the value of R19, (without capacitor from clock enable switch to ground) and don't solve the inconvenience. When the clock signal is more than +2V if I turn the ON the enable clock switch, the klee skips some steps. I have tested increasing some values until 1,5MΩ.
So I came back to the idea of the capacitor from enable switch to ground and starting from 1n i increased the value putting 1n capacitors en parallel. When the capacitor value is 5n, my Klee only moves one step, each time that I turn the enable switch ON. And the klee responses to the audio clock input until 19,5Khz, much more than enough for the Klee so I'm satisfied with the results.
Quote: | Not sure if we've gone there, but it might be worth it to check that C17 is 1n (1000 pF), and that R19 is at least 22K. And, pin 13 of U2 should be high 99.9999% of the time (it will only pulse low when U1 transistions from low to high, either from a clock pulse or when flipping the clock enable switch on with a DC input applied). |
Yes, R19 and C17 are the right value.
Pin13 is always high and only down with the clock pulses.
Quote: | If you have the time, you might see how wide your pulses are coming from pin 12 of U2 - it should pulse high with each clock pulse. |
I expend a lot of time with this point and that was great because I knew my oscilloscope a little more. I have not achieved the measure because is very thin, the oscilloscope only detect it in the function "peak detection", This is an OWON PDS 5022 and I thing that don't have high resolution or more probable is that I still don't know well how it works. It gives you the length of the cycle but not of a pulse.
Well, the Klee is opened I will change the push buttons for other ones better, Scott if you want that I measure or test anything just tell me it, I will do it with pleasure.
With this experience now, I know a little more my klee, my oscilloscope and my VCS bananalogue.
Sincerely thank you,
Oscar. |
|
Back to top
|
|
|
Sound
Joined: Jun 06, 2006 Posts: 842
Audio files: 1
|
Posted: Tue Dec 16, 2008 10:15 am Post subject:
|
|
|
EdisonRex wrote: | Hi Sound, yes, thank you - I think we both spent almost the same amount of time designing and building. Even though they are very different, you panel is excellent too.
The panel decoration is photographic inkjet paper adhered to the panel with contact cement and then layered with clear coat spray. |
Yes I did the cardboard front panel prototype Because I saw that you did it. . In the beginning I was obsessed with the professional look of a front panel, but now I did my dual Mankato just drilling a piece of aluminum and labeling by hand with a permanent marker. I thing that will be a great idea to put color in a next front panel.
Regards. |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Tue Dec 16, 2008 11:43 am Post subject:
|
|
|
Thanks for the info, Sound. You've gone above and beyond the call of duty, indeed!
I think the fourth run of Klee boards will include pads for the capacitor, in case anybody runs into the same situation.
Take care,
Scott _________________ My Site |
|
Back to top
|
|
|
EdisonRex
Site Admin
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
|
Posted: Tue Dec 16, 2008 1:34 pm Post subject:
|
|
|
Glad I took Bill up on that offer of new 4034s. The Klee was acting weird, in that things I was debugging would change state and start working. This was with the old SCL 4034s. All of a sudden, the 9-16 bank stopped working altogether, and then a BANG and some magic smoke. The 4034 'sploded.
Also the 1-8 bank was misbehaving too, dropping bits and stuff. Replacing both seem to have fixed a LOT of small problems I was having which I could not pin down to wiring.
Moral of the story is, if you got those SCL 4034s, get rid of them. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated.
Home,My Studio,and another view |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Thu Dec 18, 2008 5:25 am Post subject:
|
|
|
my klee2 run like a charm, almost from the first powering up!
however, unfrequently the A&B output behaves not as it should: the CV glides although i turned the glide pot fully ccw. it happens only sometimes, and when it happens the potentiometer has no effect at all. i did not note any rule here, yet. i rewired this section two times and replaced the potentiometer.
could it be capacitance/cross talk? the wires to the potentiometer are twisted and are tied to the wires going to the jack socket. if, why would it happen for the A&B output only? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
|
Back to top
|
|
|
|