Problems Clocking Muxlicer w/ Tempi

Cwejman, Livewire, TipTop Audio, Doepfer etc... Get your euro on!

Moderators: Kent, luketeaford, Joe.

User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Problems Clocking Muxlicer w/ Tempi

Post by Hey_Steve86 » Sat Mar 02, 2019 9:05 am

I already asked this in the dedicated Muxlicer thread, but it got buried and never answered.

My question is regarding the clock input of the Muxlicer.

Previously I had been using it as my only sequencer. I was usually clocking it from my Mother 32 square lfo, or through the comparator mode of the disting.

I recently picked up a Tempi and Rene, as I wanted more in depth playable clocking and sequencing. My intention now is to use the Muxlicer more for its routing/switching and modulation abilities.

Now here is where I'm encountering my problems...the Mux seems to hate the Tempi direct clock outs (also the Rene gate outputs, go figure, lol).

I can only get it to work by sending a channel of Tempi through another module. Such as through the aforementioned disting, through the gate input of the Mother 32, and using its gate output, or hell, even just sending it through a channel of Maths seems the do the trick. Even heavily attenuated I might add! :doh:

Looking at the manual for the Mux it says it only needs a rising edge to step forward. What could possibly be going on that the Muxlicer will not respond to Tempi directly, but will through intermediaries?

I can only guess that the Muxlicer's is like a black market clock dealer who will never do business in person, or that it personally has some sort of grudge against Make Noise triggers... :razz:

All kidding aside; My system is rather small at the moment, so having to use other modules as intermediaries when they could be used for something else is slightly hampering.

I'm tempted to just get a dedicated clock divider to drive the Muxlicer as a more permanent partner.

As a side note, this unit was built by me and is only the second module I've build. I could have made a mistake putting it together, but it's seems to work fine in every other aspect.


Thanks for you time and any help you could possibly give me!

User avatar
luketeaford
Super Deluxe Wiggler
Posts: 2181
Joined: Wed Jan 08, 2014 9:54 pm
Location: Columbus, Ohio
Contact:

Post by luketeaford » Sat Mar 02, 2019 9:23 am

Have you tried setting the tempi output to triggers instead of gates?

User avatar
Basterdman
Learning to Wiggle
Posts: 47
Joined: Wed May 02, 2018 9:46 pm
Location: Oregon

Post by Basterdman » Sat Mar 02, 2019 9:55 am

This happens to me as well. If I send clock from my stillson mk2 no clock.
So I have to click my knit rider and send it clock from there. Drives me nuts

User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Sat Mar 02, 2019 10:47 am

luketeaford wrote:Have you tried setting the tempi output to triggers instead of gates?
I believe I have, but no harm in trying again.

*Edit* Changing the output to triggers didn't fix anything.

danishchairs
Wiggling with Experience
Posts: 338
Joined: Fri Dec 30, 2016 11:10 am

Post by danishchairs » Sat Mar 02, 2019 11:25 am

The TEMPI outputs +10 V when high. Is it possible that the Muxlicer doesn’t handle well such a high voltage? (I know that’s within the Eurorack “spec”, but it is on the high end.)

Probably not related, but here’s what the Muxlicer manual says about using the Address knob and CV input:
ADDRESS KNOB/CV IN
DIGITAL STEP CONTROLLER
Address CV IN and its associated manual control "Address Knob" sets the active step either via voltage or manually. At 0v or Run position, the module plays the entire sequence from step 1 to 8 but when they are in other value than 0 Muxlicer will jump to a step relative to the sum of both parameters.

It works on a synchronous way, so it only will works during playback, and will always wait for the next clock pulse to react.

With the Address knob top anticlockwise, CV ranges goes from 0 to 10v. With the pot at the middle CVs go from -5v to 5v, and with the knob top clockwise -10v to 0v will be accepted. Other voltages are just ignored and don't represent any danger for the module.

Sound precedes music | Instruments for Neb2 module


User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Sat Mar 02, 2019 11:40 am

danishchairs wrote:The TEMPI outputs +10 V when high. Is it possible that the Muxlicer doesn’t handle well such a high voltage? (I know that’s within the Eurorack “spec”, but it is on the high end.)
I tried running a channel of Tempi through one channel of my Mixmode and it did work, but then it also was still working when I turned it up to unity gain...I should read up on how to use my multimeter to measure the voltages coming out. Actually, I should just learn how to use my multimeter period, lol.

*Edit* Interesting side note: I tried running it through my passive Pico Manual Switch, and while it didn't work unfortunately (I was really hoping it would since I don't really use it for anything right now) it did step through when I manually switched from input 1, which had the clock signal, to input 2, which had nothing plugged in.

Doesn't really solve my problem but it pointed out an interesting way to use the switch, lol.

danishchairs
Wiggling with Experience
Posts: 338
Joined: Fri Dec 30, 2016 11:10 am

Post by danishchairs » Sat Mar 02, 2019 9:12 pm

Making some measurements with Mordax Data:

TEMPI generally outputs 0V on low, and +10V on high. However, if the output of one of TEMPI’s clocks is multed to both one channel of the Data and the CLOCK IN of the Muxlicer, the voltage output is +3.5V on the low part of the pulse, and +8.65V on the high part of the pulse.

(For reference, I multed the same output from TEMPI to both the Data and one GATE input of a MI Stages. The voltage went from 0V to +10V.)

After disconnecting the TEMPI’s clock from the Muxlicer, I took the output of that same TEMPI clock and multed it into both the Data and the A1 input of a Befaco A*B+C module. TEMPI’s clock pulse was 0V on low, and +10V on high (just like it was when multed with Stages).

I set the offset on the A*B+C so that the low pulse on the A*B+C would go to 0V.

Another channel of Data was connected to OUT 1 of the A*B+C multed with the CLOCK IN of the Muxlicer. I slowly adjusted the output voltage (using the B1 LEVEL knob) until the Muxlicer’s clock would start.

The low voltage of a pulse/gate going into the Muxlicer’s CLOCK IN seems to need to be at least as low as +2.1V. (Any higher than that and the CLOCK IN won’t recognize it as a valid signal.)

The high voltage of a pulse/gate going into the Muxlicer’s CLOCK IN seems to need to be at least as high as +2.5V. (Any lower than that and the CLOCK IN won’t recognize it as a valid signal.)

No matter the settings, the output of A*B+C would never get higher than +7.48V when multed with Data and the Muxlicer’s CLOCK IN. (When disconnected from the Muxlicer, the pulse rises to +10V.)

The upshot is this: the clock pulses from TEMPI are not being recognized by Muxlicer because they never go low enough when they’re connected to the Muxlicer’s CLOCK IN.

So, why aren’t the TEMPI’s clock pulses able to hold their voltage levels in this (and perhaps other) situation(s)?

I don’t know. I’d guess that the TEMPI’s outputs aren’t able to source much current (if I’m using that phrase correctly). If whatever is connected to one of TEMPI’s clocks has high resistance/impedance, then everything works well (like with Stages). Conversely, if it has low resistance/impedance, then the voltage levels can’t swing between 0V and +10V.

I’m imagining that the Muxlicer has low resistance/impedance at its CLOCK IN port. Modules that can supply enough current to stay within that +2.1V to +2.5V range are going to be able to clock the Muxlicer fine. Modules that can’t supply enough current to provide those voltage levels are going to have trouble.

PS - in the course of working this out, I’ve discovered that I have a version 1.1 Muxlicer. There is evidently a 1.2 version. There is also a newer version of the firmware (from mid-Dec 2018). Perhaps newer versions of the hardware and/or firmware may not have these problems with the CLOCK IN port.

Sound precedes music | Instruments for Neb2 module


User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Sun Mar 03, 2019 5:46 am

Thank you so much for doing all that testing! Amazing! :yay:

The issue was really baffling me. Seems like I'll just have to keep doing what I'm doing, at least for now.

I will have the check what version my Muxlicer is.

User avatar
Nagasaki45
Common Wiggler
Posts: 229
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Sun Mar 03, 2019 2:56 pm

I think about getting the muxlicer, and if I do it will be clocked by tempi, so I'm interested in the discussion. After some online research here are my findings:

- From @danishchairs checks it sounds like an hardware issue. According to the schematic there's a 1k resistor at the input (R7) which might be too low. My limited understanding says that the input resistor should be 100k.

- Having said that, on December 12 2018 befaco released a new firmware, version 1.2, saying something about fixing a clock bug. Maybe it is, maybe it isn't. I tried to read the code and couldn't understand exactly what was wrong and what's the fix.

Did anyone try to contact befaco about that? From my experience they are quite responsive to such questions. I don't mind sending them an email and pointing them here. Hope this info is more useful than confusing.

Cheers,
Tom

User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Sun Mar 03, 2019 3:26 pm

I just looked at the back of my module, and I have version 1.2.

Unfortunately I cannot test out the updated firmware because I do not have an ISCP programmer.

I would be interested to know if changing the resistor on r7 would fix the problem. Seems like an easy enough thing to do.

I will confirm that simply plugging the tempi into a buffered multiple first seems to work as well. At least with the Mother's 32's multiple, I have no others to test at the moment...

User avatar
Nagasaki45
Common Wiggler
Posts: 229
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Sun Mar 03, 2019 4:12 pm

@Hey_Steve86, the version on the back of the module is the hardware version. It doesn't necessarily mean that the software version is 1.2. I couldn't find any information about when they updated the hardware and whether they match hardware and software versions. Anyway, it sounds more like an hardware issue to me. Will contact befaco and ask them.

User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Sun Mar 03, 2019 4:24 pm

@Nagasaki I understand that it's just the hardware version, was just confirming that it seems the problem exists for both hardware versions. I have no idea what firmware mine currently has, but it's definitely not the newest one since I built mine in October.

I would be interested to see what insights Befaco can impart!

User avatar
sherwood
Learning to Wiggle
Posts: 20
Joined: Mon Aug 31, 2015 10:48 am
Location: NC

Post by sherwood » Mon Mar 04, 2019 10:31 am

Nagasaki45 wrote:I think about getting the muxlicer, and if I do it will be clocked by tempi, so I'm interested in the discussion. After some online research here are my findings:

- From @danishchairs checks it sounds like an hardware issue. According to the schematic there's a 1k resistor at the input (R7) which might be too low. My limited understanding says that the input resistor should be 100k.
YOu're right, To conform to the Doepfer standard it should have a 100k resistor to ground, in parallel with the clamping diode D2. The 1k resistor is there to limit current.
- Having said that, on December 12 2018 befaco released a new firmware, version 1.2, saying something about fixing a clock bug. Maybe it is, maybe it isn't. I tried to read the code and couldn't understand exactly what was wrong and what's the fix.
if they changed the code so that the clock pin didn't drive the input it would likely fix the problem. Either the internal pullup is too low in value or the pin is used as an output to try and send signal to the clock detect pin.
Last edited by sherwood on Mon Mar 04, 2019 10:49 am, edited 1 time in total.

User avatar
Nagasaki45
Common Wiggler
Posts: 229
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Mon Mar 04, 2019 10:46 am

sherwood wrote:the atmega may have an internal resistor that can be configured to load PD4... many micros do. That would be easier than hacking in a physical resistor.
Checking again, the clock input is pulled up by the atmega internal resistor.

User avatar
Befacosynth
Common Wiggler
Posts: 245
Joined: Thu Jul 01, 2010 10:17 am
Location: Barcelona
Contact:

Post by Befacosynth » Mon Mar 04, 2019 1:48 pm

Hey peepz!

We have a Tempi and a muxlicer open in our desk, poking around and looking into this (again).

We are trying to avoid changing hardware AND software (wich would solve the incompatibility) to just change hardware, that will maximize retro-compatibility.

Will keep you posted!!
www.befaco.org
support((at))befaco.org

User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Mon Mar 04, 2019 8:04 pm

Befacosynth wrote:Hey peepz!

We have a Tempi and a muxlicer open in our desk, poking around and looking into this (again).

We are trying to avoid changing hardware AND software (wich would solve the incompatibility) to just change hardware, that will maximize retro-compatibility.

Will keep you posted!!
Thanks so much! Look forward to hearing from you.

User avatar
Nagasaki45
Common Wiggler
Posts: 229
Joined: Thu May 11, 2017 1:21 pm
Location: London

Post by Nagasaki45 » Tue Mar 05, 2019 4:03 am

Befacosynth wrote:Hey peepz!

We have a Tempi and a muxlicer open in our desk, poking around and looking into this (again).

We are trying to avoid changing hardware AND software (wich would solve the incompatibility) to just change hardware, that will maximize retro-compatibility.

Will keep you posted!!
Yeah! Thanks for looking into this.

danishchairs
Wiggling with Experience
Posts: 338
Joined: Fri Dec 30, 2016 11:10 am

Post by danishchairs » Tue Mar 26, 2019 9:49 am

Befacosynth wrote: ... We have a Tempi and a muxlicer open in our desk, poking around and looking into this (again).

We are trying to avoid changing hardware AND software (wich would solve the incompatibility) to just change hardware, that will maximize retro-compatibility. ...
I heard back from Befaco this morning, both about compatibility with TEMPI, and about the differences between v1.1 and v1.2 hardware. This post is about compatibility with TEMPI.
Befaco from email wrote: About the Tempi Incompatibillity...

Tempi schematics are not available so we borrowed one with a friend and extract the output scheme from it, and test them together.

After this we found, It is difficult to solve without modifying either Muxlicer or Tempi.

We are working on finding the easiest solution, but in the mean time you can just put and active buffered or Mult in between.
So, it’s still an open question for the time being. I’m tempted to hook up a potentiometer in replacement of R7, to see if a different value would solve the problem (without messing up the “clock detect” circuit with R8).

Sound precedes music | Instruments for Neb2 module


User avatar
Hey_Steve86
Common Wiggler
Posts: 109
Joined: Thu Mar 02, 2017 8:19 am
Location: Philadelphia

Post by Hey_Steve86 » Tue Mar 26, 2019 1:46 pm

danishchairs wrote:
Befacosynth wrote: ... We have a Tempi and a muxlicer open in our desk, poking around and looking into this (again).

We are trying to avoid changing hardware AND software (wich would solve the incompatibility) to just change hardware, that will maximize retro-compatibility. ...
I heard back from Befaco this morning, both about compatibility with TEMPI, and about the differences between v1.1 and v1.2 hardware. This post is about compatibility with TEMPI.
Befaco from email wrote: About the Tempi Incompatibillity...

Tempi schematics are not available so we borrowed one with a friend and extract the output scheme from it, and test them together.

After this we found, It is difficult to solve without modifying either Muxlicer or Tempi.

We are working on finding the easiest solution, but in the mean time you can just put and active buffered or Mult in between.
So, it’s still an open question for the time being. I’m tempted to hook up a potentiometer in replacement of R7, to see if a different value would solve the problem (without messing up the “clock detect” circuit with R8).
Thanks for the update! Too bad the problem is trickier than they hoped....

More reasons for me to finally pick up a standalone buffered mult.

User avatar
robduarte
Learning to Wiggle
Posts: 30
Joined: Tue Jan 08, 2019 6:06 pm
Location: Tallahassee

Post by robduarte » Sun Jun 02, 2019 4:02 pm

I'm interested in hearing what the solution is since I plan to build one from a kit soon. If it's a matter of replacing a resistor or shoehorning an extra resistor in parallel somewhere, I'm happy to do it. Thanks for looking into it @Befacosynth

Michelob
Learning to Wiggle
Posts: 3
Joined: Mon Dec 28, 2020 2:28 pm

Re: Problems Clocking Muxlicer w/ Tempi

Post by Michelob » Mon Dec 28, 2020 2:38 pm

Hi everyone,

First time posting, digging out this old thread.

Trying to feed Muxlicer with Beatstep Pro clock, but I got the same issue.

Is there any 2020 update about this ?

Some details :

- my Mux is brand new
- Malekko Quad LFO perfectly receives the BSP clock
- if I take some triangle or square LFO from Malekko as a relaying clock to Muxlicer, well things work great.

Thnaks anyone, and if there's some better place in this forum to talk about this, please let me know.

Michel

User avatar
deftinwulf
Veteran Wiggler
Posts: 659
Joined: Sun Nov 03, 2013 10:34 am

Re:

Post by deftinwulf » Mon Dec 28, 2020 3:31 pm

Befaco from email wrote: About the Tempi Incompatibillity...

Tempi schematics are not available so we borrowed one with a friend and extract the output scheme from it, and test them together.

After this we found, It is difficult to solve without modifying either Muxlicer or Tempi.

We are working on finding the easiest solution, but in the mean time you can just put and active buffered or Mult in between.
(above emphasis added by me)

Man, I would be SO interested if anyone can explain why this sort of behavior between modules exists. I went back and forth with 4MS support about my QCD because for whatever reason, it was incapable of triggering my Qu-Bit EON without a buffered multiple inserted in the chain. I demonstrated it in numerous videos - QCD->EON, no trigger; QCD->buff mult->EON, perfect trigger. In the end, 4MS support said that they tested QCD with EON on their end and it worked fine, so it must be a problem with my EON. I've since moved on, but I never found that to be a satisfactory answer. I believe that 4MS was telling the truth, but there *must* be more to it, right?

To hear Befaco suggest it as a useful workaround leads me to believe my hunch was correct. Sorry for derailing the thread, this is just really interesting to me and the first I've seen it discussed on any other modules.

User avatar
rockwoofstone
Veteran Wiggler
Posts: 564
Joined: Tue Oct 02, 2012 5:32 am
Location: UK

Re: Problems Clocking Muxlicer w/ Tempi

Post by rockwoofstone » Mon Apr 26, 2021 7:54 am

Wondering if anyone every found a permanent resolution to this that didn't involve needing to use a buffered-mult between the clock and Muxlicer...

owsfaratu
1-Post Wiggler
Posts: 1
Joined: Thu Apr 29, 2021 1:02 pm

Re: Problems Clocking Muxlicer w/ Tempi

Post by owsfaratu » Thu Apr 29, 2021 1:08 pm

I've had issues with this as well... trying to clock from a keystep 37. I'm gonna be getting a buffered must anyway but would be nice to know whats going on. Clocks fine if i use the rise out of Contour 1.

User avatar
rockwoofstone
Veteran Wiggler
Posts: 564
Joined: Tue Oct 02, 2012 5:32 am
Location: UK

Re: Problems Clocking Muxlicer w/ Tempi

Post by rockwoofstone » Fri Apr 30, 2021 4:10 am

I contacted Befaco about this, and they sent me instructions to make a small modification to my v1.1 Muxlicer which fixed the issue. The solution is (unsurprisingly) to buffer that input on the module, but it only needed a few resistor changes, removal of a couple of diodes, and addition of a transistor. All fixed in just a few minutes, and works fine now, so if you're looking to do the same, give them a shout, and they'll send you the instructions.

Post Reply

Return to “Eurorack Modules”