DivKid RND STEP

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

Moderators: luketeaford, Joe., lisa, Kent

User avatar
Richard deHove
Wiggling with Experience
Posts: 287
Joined: Wed Mar 20, 2019 4:32 pm
Location: Van Diemen's Land
Contact:

Re: DivKid RND STEP

Post by Richard deHove » Sat Jul 18, 2020 4:09 pm

So is there a bad batch or is it a trimmer issue?

closedLoop
Common Wiggler
Posts: 199
Joined: Sat Feb 02, 2019 9:01 am
Location: nyc

Re: DivKid RND STEP

Post by closedLoop » Sun Jul 19, 2020 12:03 pm

analogPedagog wrote:
Sat Jul 18, 2020 12:25 pm
Making it louder will keep the even-ness of the distribution but the range will be a bit wider and may include some levels beyond +/-5V at times.
I did a little playing with the adjustments of Rnd Step. I'm guessing from what I noticed is that when the potentiometers are fully clockwise, it's all the way up, and when it's fully counterclockwise, it's all the way down.

While I was carefully doing the adjustments, I noticed a strange thing. When I touched the potentiometers, the range seemed to double, and that effect seemed to linger for a little after adjusting, after which it returned to 'normal.'

Here's 2 quick shot of the 3 channels, one of the bipolar, one of the unipolar. This is with all the potentiometers fully clockwise. Channel 3 has substantially greater range.
rnd1.jpg
rnd2.jpg
You do not have the required permissions to view the files attached to this post.

User avatar
GryphonP3
Super Deluxe Wiggler
Posts: 1094
Joined: Tue Jan 29, 2013 3:20 pm
Location: Los Angeles

Re: DivKid RND STEP

Post by GryphonP3 » Sun Jul 19, 2020 3:31 pm

closedLoop wrote:
Sun Jul 19, 2020 12:03 pm
analogPedagog wrote:
Sat Jul 18, 2020 12:25 pm
Making it louder will keep the even-ness of the distribution but the range will be a bit wider and may include some levels beyond +/-5V at times.
I did a little playing with the adjustments of Rnd Step. I'm guessing from what I noticed is that when the potentiometers are fully clockwise, it's all the way up, and when it's fully counterclockwise, it's all the way down.

While I was carefully doing the adjustments, I noticed a strange thing. When I touched the potentiometers, the range seemed to double, and that effect seemed to linger for a little after adjusting, after which it returned to 'normal.'

Here's 2 quick shot of the 3 channels, one of the bipolar, one of the unipolar. This is with all the potentiometers fully clockwise. Channel 3 has substantially greater range.

rnd1.jpg
rnd2.jpg
Mine is acting exactly the same way. One channwl is stronger than the other two but nowhere near 5v, while the other two max at around 1.3v, and this is at fully clockwise settings.
My cat can eat a whole watermelon ...

BlinkyLights
Wiggling with Experience
Posts: 444
Joined: Wed Feb 13, 2019 12:16 am

Re: DivKid RND STEP

Post by BlinkyLights » Sun Jul 19, 2020 3:47 pm

Richard deHove wrote:
Sat Jul 18, 2020 4:09 pm
So is there a bad batch or is it a trimmer issue?
He wrote in this thread that some were simply calibrated conservatively and some may need an adjustment now.

(paraphrasing, but it's just a page or two back, go look...)

Somewhat similarly, I once had to recalibrate my Ultra Random Analog, also with a tiny trimmer on the back.

It required a scope, but I had the Mordax, so it was crazy easy. Only "hard" part was not over-rotating the trimmer.

User avatar
Richard deHove
Wiggling with Experience
Posts: 287
Joined: Wed Mar 20, 2019 4:32 pm
Location: Van Diemen's Land
Contact:

Re: DivKid RND STEP

Post by Richard deHove » Sun Jul 19, 2020 4:58 pm

BlinkyLights wrote:
Sun Jul 19, 2020 3:47 pm
Richard deHove wrote:
Sat Jul 18, 2020 4:09 pm
So is there a bad batch or is it a trimmer issue?
He wrote in this thread that some were simply calibrated conservatively and some may need an adjustment now.
(paraphrasing, but it's just a page or two back, go look...)
Somewhat similarly, I once had to recalibrate my Ultra Random Analog, also with a tiny trimmer on the back.
It required a scope, but I had the Mordax, so it was crazy easy. Only "hard" part was not over-rotating the trimmer.
Sure, I read that about calibration and am not above tweaking, but what about people like Gryphon (above)? Pots all the way up and still not right? Or is that 'within spec'? Not hating here, just wanting more specific info!

User avatar
analogPedagog
Steady State Fate
Posts: 966
Joined: Wed Apr 14, 2010 10:56 pm
Location: New York
Contact:

Re: DivKid RND STEP

Post by analogPedagog » Sun Jul 19, 2020 7:35 pm

Hey all,

So I performed some tests using my Rigol O'Scope and compared them to the Mordax data.

First, I highly recommend not using the voltage monitor to measure the output. The noise transients are very quick and the voltage monitor does not have that kind of bandwidth and will alias out everything nearing or going above 2kHz.

Also compared to the Scope mode on Data and the output is also a bit aliased and not showing the same peak levels as my 70Mhz bench top unit. However, I highly recommend using the scope mode vs Voltage Monitor as it will give you a much better picture of the actual output levels.

The faster your trigger speed, the more detailed the output level will be, as you will be allowing more samples of the noise to be seen on your Scope within a given time division.

That being said, I'm sure that because I was being so conservative that there will be modules with lower output - that may or may not work for you. I will say again that all you need to do is try and turn up the levels. If that does not work for you and/or you are having any other issues - please just contact me via the SSF website contact form and I will take care of you.

Apologies again for any inconvenience caused. Happy to make it right.

-Andrew
SSF

User avatar
GryphonP3
Super Deluxe Wiggler
Posts: 1094
Joined: Tue Jan 29, 2013 3:20 pm
Location: Los Angeles

Re: DivKid RND STEP

Post by GryphonP3 » Mon Jul 20, 2020 12:44 am

analogPedagog wrote:
Sun Jul 19, 2020 7:35 pm
Hey all,

So I performed some tests using my Rigol O'Scope and compared them to the Mordax data.

First, I highly recommend not using the voltage monitor to measure the output. The noise transients are very quick and the voltage monitor does not have that kind of bandwidth and will alias out everything nearing or going above 2kHz.

Also compared to the Scope mode on Data and the output is also a bit aliased and not showing the same peak levels as my 70Mhz bench top unit. However, I highly recommend using the scope mode vs Voltage Monitor as it will give you a much better picture of the actual output levels.

The faster your trigger speed, the more detailed the output level will be, as you will be allowing more samples of the noise to be seen on your Scope within a given time division.

That being said, I'm sure that because I was being so conservative that there will be modules with lower output - that may or may not work for you. I will say again that all you need to do is try and turn up the levels. If that does not work for you and/or you are having any other issues - please just contact me via the SSF website contact form and I will take care of you.

Apologies again for any inconvenience caused. Happy to make it right.

-Andrew
SSF
Thank you for the good support andrew, i'll contact you over there.
My cat can eat a whole watermelon ...

User avatar
Glitzerstrahl
Common Wiggler
Posts: 91
Joined: Tue Jan 22, 2013 8:10 pm
Location: Tokyo
Contact:

Re: DivKid RND STEP

Post by Glitzerstrahl » Sun Jul 26, 2020 10:36 am

I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....

malnourish
Learning to Wiggle
Posts: 43
Joined: Mon May 18, 2020 3:43 pm

Re: DivKid RND STEP

Post by malnourish » Sun Jul 26, 2020 10:48 am

Glitzerstrahl wrote:
Sun Jul 26, 2020 10:36 am
I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....
Anyone else feeling the same way? I was hoping to pick one of these up in the near future, and was specifically hoping to do pitch oriented things so I don't want to bite if it's unreliable.

closedLoop
Common Wiggler
Posts: 199
Joined: Sat Feb 02, 2019 9:01 am
Location: nyc

Re: DivKid RND STEP

Post by closedLoop » Sun Jul 26, 2020 11:02 am

Glitzerstrahl wrote:
Sun Jul 26, 2020 10:36 am
I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....
Just for clarity - you're feeding a v/oct signal to the input of Rnd Step, and then triggering it randomly?

I've never thought to work that way. I always feed the output of S&Hs into a quantizer.

malnourish
Learning to Wiggle
Posts: 43
Joined: Mon May 18, 2020 3:43 pm

Re: DivKid RND STEP

Post by malnourish » Sun Jul 26, 2020 11:22 am

closedLoop wrote:
Sun Jul 26, 2020 11:02 am
Glitzerstrahl wrote:
Sun Jul 26, 2020 10:36 am
I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....
Just for clarity - you're feeding a v/oct signal to the input of Rnd Step, and then triggering it randomly?

I've never thought to work that way. I always feed the output of S&Hs into a quantizer.
Typically that would be the expected norm, although this was a significant portion of the marketing pitch, that it would be stable enough for extracting v/oct without drift (or minor enough drift it would be highly useable for this use case).

User avatar
Glitzerstrahl
Common Wiggler
Posts: 91
Joined: Tue Jan 22, 2013 8:10 pm
Location: Tokyo
Contact:

Re: DivKid RND STEP

Post by Glitzerstrahl » Sun Jul 26, 2020 7:46 pm

closedLoop wrote:
Sun Jul 26, 2020 11:02 am
Glitzerstrahl wrote:
Sun Jul 26, 2020 10:36 am
I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....
Just for clarity - you're feeding a v/oct signal to the input of Rnd Step, and then triggering it randomly?

I've never thought to work that way. I always feed the output of S&Hs into a quantizer.
Correct, and to be clear the signal is already quantized as its coming from a quantized sequencer.
Also the triggers are always on the beat.

I've tested it with a sequence from Rene (quantized), the SQ-1 and the new quantized pitches from FW 2.06 of Pam's. Regardless of the source the output does not match the pitches of the sequence 100%, so randomly I get a note that is not in the sequence and therefor throwing the melody out of tune.

With my other Sample & Holds (Intellijel Noise Tools 1U or Doepfer A-148) I'm not seeing this behavior. No matter how frequently or infrequently I trigger them the pitches are always one from the sequence they are being fed.
malnourish wrote:
Sun Jul 26, 2020 11:22 am
Typically that would be the expected norm, although this was a significant portion of the marketing pitch, that it would be stable enough for extracting v/oct without drift (or minor enough drift it would be highly useable for this use case).
Exactly. That's what's so surprising to me. :eek:
Maybe I just got a bad unit.... :despair:

And just to avoid confusion, I love my other SSF modules.
Design is always spot on and never had any issues in the past, so no shadow on the brand and their other stuff.
Last edited by Glitzerstrahl on Sun Jul 26, 2020 11:44 pm, edited 1 time in total.

ima_jrk
Common Wiggler
Posts: 53
Joined: Sat Nov 09, 2019 5:51 am

Re: DivKid RND STEP

Post by ima_jrk » Sun Jul 26, 2020 9:55 pm

Glitzerstrahl wrote:
Sun Jul 26, 2020 7:46 pm
And just to avoid confusion, I love my other Instruo modules.
Design is always spot on and never had any issues in the past, so no shadow on the brand and their other stuff.
Ochd is Instruo. RND STEP is SSF.

User avatar
Glitzerstrahl
Common Wiggler
Posts: 91
Joined: Tue Jan 22, 2013 8:10 pm
Location: Tokyo
Contact:

Re: DivKid RND STEP

Post by Glitzerstrahl » Mon Jul 27, 2020 12:07 am

ima_jrk wrote:
Sun Jul 26, 2020 9:55 pm
Glitzerstrahl wrote:
Sun Jul 26, 2020 7:46 pm
And just to avoid confusion, I love my other Instruo modules.
Design is always spot on and never had any issues in the past, so no shadow on the brand and their other stuff.
Ochd is Instruo. RND STEP is SSF.
LOL yep. Sorry mixed it up when I wrote the post. Edited to reflect.

closedLoop
Common Wiggler
Posts: 199
Joined: Sat Feb 02, 2019 9:01 am
Location: nyc

Re: DivKid RND STEP

Post by closedLoop » Mon Jul 27, 2020 9:30 am

Glitzerstrahl wrote:
Sun Jul 26, 2020 7:46 pm

Exactly. That's what's so surprising to me. :eek:
Maybe I just got a bad unit.... :despair:

And just to avoid confusion, I love my other SSF modules.
Design is always spot on and never had any issues in the past, so no shadow on the brand and their other stuff.
Are you grabbing from the Unipolar or Bipolar output?

As I note above, I never work the way you're trying to work, but I did a quick test patch. From the Bipolar output, I'm getting pretty good pitch tracking feeding USTA's CV out into Rnd Step, and triggering with PNW. I'm checking the tune on the Mordax Data tuner.

Things do get a little weird from the unipolar.

User avatar
Glitzerstrahl
Common Wiggler
Posts: 91
Joined: Tue Jan 22, 2013 8:10 pm
Location: Tokyo
Contact:

Re: DivKid RND STEP

Post by Glitzerstrahl » Mon Jul 27, 2020 11:49 am

closedLoop wrote:
Mon Jul 27, 2020 9:30 am
Glitzerstrahl wrote:
Sun Jul 26, 2020 7:46 pm

Exactly. That's what's so surprising to me. :eek:
Maybe I just got a bad unit.... :despair:

And just to avoid confusion, I love my other SSF modules.
Design is always spot on and never had any issues in the past, so no shadow on the brand and their other stuff.
Are you grabbing from the Unipolar or Bipolar output?

As I note above, I never work the way you're trying to work, but I did a quick test patch. From the Bipolar output, I'm getting pretty good pitch tracking feeding USTA's CV out into Rnd Step, and triggering with PNW. I'm checking the tune on the Mordax Data tuner.

Things do get a little weird from the unipolar.
Thanks for checking! I’m using the bipolar since the unipolar always samples from the internal random source. Maybe the
problem is just with my unit... will do some more testing and report back...

kpreid
Common Wiggler
Posts: 129
Joined: Wed Aug 29, 2018 10:18 am
Location: California
Contact:

Re: DivKid RND STEP

Post by kpreid » Mon Jul 27, 2020 12:57 pm

Perhaps the problem is timing? Any sample-and-hold, no matter how good at sampling it is, placed after a sequencer or other stepped CV source will produce "wrong" results if the sampling is performed at the same time the CV source is in the middle of shifting to a new value.

The test/workaround for this would be to add a small delay between the trigger controlling (or produced by) the sequencer and the trigger input of RND STEP, using a trigger delay module, delayed envelope, or similar. If that eliminates the unquantized results, one can then experiment with making the delay as short as possible.

closedLoop
Common Wiggler
Posts: 199
Joined: Sat Feb 02, 2019 9:01 am
Location: nyc

Re: DivKid RND STEP

Post by closedLoop » Mon Jul 27, 2020 1:39 pm

Glitzerstrahl wrote:
Mon Jul 27, 2020 11:49 am


Thanks for checking! I’m using the bipolar since the unipolar always samples from the internal random source. Maybe the
problem is just with my unit... will do some more testing and report back...
I had completely forgotten the unipolar is always to the pink noise source - thanks for the reminder.

User avatar
analogPedagog
Steady State Fate
Posts: 966
Joined: Wed Apr 14, 2010 10:56 pm
Location: New York
Contact:

Re: DivKid RND STEP

Post by analogPedagog » Mon Jul 27, 2020 5:39 pm

Glitzerstrahl wrote:
Sun Jul 26, 2020 10:36 am
I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....
It is the right idea that the trigger can/should be delayed in some cases.

Being such a fast trigger on RS of 50us, it could be sampling faster than the seq responds - during the transition between notes. And then you have to account that it isn’t always going to be perfect every time.

A longer sample duration that some s&h may utilize would close the sample switch later, once the transition to the next note has completed, but at the expense of high freq performance (ie sample rate reduction) and would allow more of the sample source to creep onto the output.

User avatar
Glitzerstrahl
Common Wiggler
Posts: 91
Joined: Tue Jan 22, 2013 8:10 pm
Location: Tokyo
Contact:

Re: DivKid RND STEP

Post by Glitzerstrahl » Tue Jul 28, 2020 12:20 am

analogPedagog wrote:
Mon Jul 27, 2020 5:39 pm
Glitzerstrahl wrote:
Sun Jul 26, 2020 10:36 am
I’m finding the s&h performance to be pretty poor honestly. Trying to feed a 16 step sequence of pitches and hit it with random but in sync (e.i. Not Euclidean for example) gates to pick up a melody. I find the resulting pitches to sometimes be out of tune, whereas when I perform the same patch with for example the Intellijel Noise Tools S&H it’s rock solid... I don’t have a scope to verify or screenshot but but it’s obvious to the ear. Thus I’ve resorted to only using it for random modulation instead....
It is the right idea that the trigger can/should be delayed in some cases.

Being such a fast trigger on RS of 50us, it could be sampling faster than the seq responds - during the transition between notes. And then you have to account that it isn’t always going to be perfect every time.

A longer sample duration that some s&h may utilize would close the sample switch later, once the transition to the next note has completed, but at the expense of high freq performance (ie sample rate reduction) and would allow more of the sample source to creep onto the output.
Yep the timing could absolutely be the reason, but since my other sample and hold modules do not produce these artifacts I'll just use those instead for pitch and the RND STEP for other tasks. :love:

User avatar
Dick Champion
Learning to Wiggle
Posts: 29
Joined: Sun Oct 03, 2010 6:52 am
Location: Copenhagen

Re: DivKid RND STEP

Post by Dick Champion » Sat Aug 01, 2020 8:02 pm

I'm having the similar issues with dropping voltage and wobbly sampling on the RND STEP.
When using o&C to monitor the pitch of an oscillator held by RND STEP, I can see it dropping ever so slightly.
Dont know if that is within reason.
But the annoying thing is, I wanted to do the "quantize-before-s&h"-trick... but the pitch is to darn off.
Even when I just feed the RND STEP a steady voltage from the uScale, it deviates just enough to be useless on each sample.
Kinks from Mutable is actually better at sampling incoming quantized voltages.
I also have a Disting MK3 but it's acting weird.
Oh, and I probably blew my other Disting by some kind of short. This is not my day.

Post Reply

Return to “Eurorack Modules”