[ORDERING] An all new QLFO (Quad [Syncable] LFO): JME b101

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, Joe.

BARE BONES
Wiggling with Experience
Posts: 289
Joined: Sun Aug 10, 2014 5:15 pm
Location: Bath

Post by BARE BONES » Fri Aug 07, 2015 2:37 am

Jack Gull wrote:
BARE BONES wrote:is the sync input normalled to all inputs if you just patch in one master clock?
Each is normaled to the previous, so syncing A only means that you also sync B C and D
Syncing B means that you also sync C and D
Syncing C means that you also sync D
More if the feature set gets laid out for us soon
what did i miss :)?
perfect !

do the expander boards have additional circuitry or are they literally a panel of jacks/pots with an easy connection to the main board?

User avatar
falafelbiels
still learning to wiggle
Posts: 4356
Joined: Tue Aug 04, 2009 7:15 am
Location: Rotterdamm son

Post by falafelbiels » Fri Aug 07, 2015 2:45 am

Nice stuff!

Can I be in for a 101 and a 101-2 board and ic set?

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Fri Aug 07, 2015 4:22 am

BARE BONES wrote: do the expander boards have additional circuitry or are they literally a panel of jacks/pots with an easy connection to the main board?
they have additional circuitry

@falafelbiels
(to be sure) no panels?

User avatar
mbartkow
Veteran Wiggler
Posts: 726
Joined: Wed Jul 31, 2013 7:31 am
Location: Moonbase Alpha

Post by mbartkow » Fri Aug 07, 2015 5:29 am

Hi!
This seems to be a nice project, but are you aware of the aliasing problems related to digital generation of wideband waveforms? The manifestation of aliasing in an LFO output is that the waveform is irregular. I can clearly hear it in your videos when the frequency is pushed over 50Hz - the modulated tone is all jerky and unstable.

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Fri Aug 07, 2015 5:50 am

Of course i am aware of aliasing. But are you sure? I spent quite some time testing the output and i didn't hear aliasing.
Keep in mind that sound came out from macbook's speaker and was recorded by a phone so there might be something else going on.
In the meantime i was monitoring through headphones and didn't hear anything strange

I'll have a look at it again. In case i notice something strange i'll just add a digital filter, no big deal

User avatar
mbartkow
Veteran Wiggler
Posts: 726
Joined: Wed Jul 31, 2013 7:31 am
Location: Moonbase Alpha

Post by mbartkow » Fri Aug 07, 2015 6:23 am

I can clearly hear in your examples that the modulated tone from the VCO is jerky and unstable in pitch, unlike a proper modulation from a decent analog LFO. I am listening to it through studio monitors. But the quality of reproduction is irrelevant as it does not affect pitch.

Adding a filter _after_ you produced the waveform with aliasing is too late. No filter will remove aliasing that is already there. Do not confuse aliasing with the presence of sampling spectral images (that will be inaudible at 50kHz). Producing an alias-free waveform is not a trivial thing. Assuming you employ a classic approach with phase accumulator and wave table - incrementing the phase means dropping samples. In e.g. a sawtooth waveform this means you sometimes hit the peak of the waveform and sometimes you miss it, which yields an irregular waveform with a nasty interference from the sampling pattern. When such a waveform is used to modulate a pitch of a VCO, it can be quite disturbing, once you notice that the modulation is irregular. Compare it with an analog LFO and you will hear the difference.

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Fri Aug 07, 2015 7:21 am

Well, thank you for making me notice this.
There is indeed some aliasing but, i should be able to get rid of it by filtering the samples.
At the highest generated frequency the alising start being noticeable at about 3.5kHz. And i can increase the sampling frequency so i can push it even further away.
Unfortunately i don't think i have enough memory to store all the bandlimited waveshapes**, but i could make something smart like reading the bandlimited transitions from memory. I'll try and see



**which is why i didn't use them in the first place

User avatar
Fubard
Common Wiggler
Posts: 59
Joined: Tue Aug 26, 2014 6:10 pm
Location: Austin, TX

Post by Fubard » Sun Aug 09, 2015 3:08 pm

Count me in for one as well! :yay:
Thanks!

User avatar
falafelbiels
still learning to wiggle
Posts: 4356
Joined: Tue Aug 04, 2009 7:15 am
Location: Rotterdamm son

Post by falafelbiels » Mon Aug 10, 2015 7:06 pm

Jack Gull wrote:
@falafelbiels
(to be sure) no panels?
Myeh well they look good but I am pretty broke, so I am going to cut my own...

User avatar
scozbor
100th Member!
Posts: 1746
Joined: Wed Apr 30, 2008 4:17 pm
Location: Maleny

Post by scozbor » Mon Aug 10, 2015 7:48 pm

lazy people need full kits for this! looks amaze.

cmon synthcube/thonk

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Tue Aug 11, 2015 5:59 am

:lol:

That would be really nice. I'll try and reach steve after this first batch
Anyway, added all who contacted me.

Some improvements:
- A better (faster) Tap tempo code
- Polished code so better performances. i had enough space to add 2x oversampling, which was enough to remove most if not all artifacts. I couldn't hear differences while modulating from this or an analog LFO. Will post some demos when i have time, probably tomorrow.

Also, adam contacted me and pointed out a website that makes metal eurorack panels for a good price, according to him. i'll contact them and see what can be done.

That's all for now.
:nana:

User avatar
synaptech
Wiggling with Experience
Posts: 347
Joined: Wed Aug 10, 2011 12:12 pm
Location: San Jose, CA

Post by synaptech » Mon Aug 17, 2015 7:42 pm

Count me in for one full set of the following:
B101, B101-2, B101-3, B101-4!

euromorcego
Super Deluxe Wiggler
Posts: 2762
Joined: Sat Jan 18, 2014 6:22 am

Post by euromorcego » Tue Aug 18, 2015 3:17 pm

also definitely interestedt, for now the B101 QLFO (PCB set + Panel + Programmed PIC).

If you can make metal panels happen it would be awesome.

User avatar
Crashlander42
Common Wiggler
Posts: 207
Joined: Sun Jun 24, 2012 12:20 pm
Location: Winter Park, FL
Contact:

Post by Crashlander42 » Wed Aug 19, 2015 7:27 am

Interested in a B101 FS & B101-2 FS.

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Wed Aug 19, 2015 7:47 am

Great! I added all other who contacted me.

:)

I am requesting and getting quotes from some businesses who makes aluminum panels. Trying as hard as i can but i'll have to raise prices in case i'll do metal panels (about 5 euros) as they really are expensive

User avatar
gbiz
Super Deluxe Wiggler
Posts: 1076
Joined: Tue May 07, 2013 11:20 am
Location: Sussex, UK

Post by gbiz » Wed Aug 19, 2015 8:06 am

I'm in for a 101 pcb/panel/mcu & a 101-3 pcb/panel.

I'd prefer metal panels too. I realise they'd cost extra.

User avatar
yan6
Ultra Wiggler
Posts: 981
Joined: Sat Jun 15, 2013 7:14 pm
Location: Ontario, Canada

Post by yan6 » Wed Aug 19, 2015 8:28 am

Dude can i get in on a pcb, panel and pic or FS of

B101, B101-2, B101-3, B101-4

also interested in metal panels

User avatar
mbartkow
Veteran Wiggler
Posts: 726
Joined: Wed Jul 31, 2013 7:31 am
Location: Moonbase Alpha

Post by mbartkow » Tue Aug 25, 2015 1:45 pm

Jack Gull wrote: - Polished code so better performances. i had enough space to add 2x oversampling, which was enough to remove most if not all artifacts. I couldn't hear differences while modulating from this or an analog LFO. Will post some demos when i have time, probably tomorrow.
I would be very interested to see/hear some examples. From what I know, two times oversampling does not solve the problem of heavy aliasing in direct synthesis of wideband waveforms, like a sawtooth. In fact, two times oversampling makes almost no difference. The spectrum is so rich of overtones that it would require rather a 50 times oversampling, which would be a waste of cpu. The topic is well studied and there are several methods for digital generation of classic waveforms that don't break a bank.

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Fri Sep 11, 2015 12:03 pm

Sorry for the absence.
Between work and stuff i had to rewrite most of the code to address the issues pointed by mbartkow.

now, there you go, sawtooth and 50% square wave.

Image

Image

i also added some other neat stuff: if selected waveform is noise you get random impulses on the gate output. probability is set by the frequency pot/cv

as for the orders i'm still waiting for a couple more people to decide if they're in or not.

User avatar
mecanikill
Electrical Demon
Posts: 1029
Joined: Thu Apr 23, 2009 9:45 pm
Location: Raleigh N.C.

Post by mecanikill » Fri Sep 11, 2015 12:37 pm

I can't wait to get this looks like the end all lfo!

User avatar
mbartkow
Veteran Wiggler
Posts: 726
Joined: Wed Jul 31, 2013 7:31 am
Location: Moonbase Alpha

Post by mbartkow » Fri Sep 11, 2015 12:42 pm

This looks like a properly bandlimited saw and square. The spectrum shows no trace of aliasing. I hope it's kept so in higher pitches as well. Anyways, good job! :tu:

User avatar
Kroaton
Common Wiggler
Posts: 90
Joined: Sun May 24, 2015 8:11 pm
Location: Bucharest, Romania

Post by Kroaton » Fri Sep 11, 2015 1:33 pm

Will there be any way of changing the exponential-logarithmic response of the CV inputs like in the Frequency Central - Ultra Wave (video of that in action here) when using the CV expander?

Also, any chance of there being a PCB+PIC option without the panel?

sammy123
Super Deluxe Wiggler
Posts: 1223
Joined: Sun Mar 24, 2013 12:36 pm

Post by sammy123 » Fri Sep 11, 2015 2:10 pm

I can't wait. Nice work.

User avatar
falafelbiels
still learning to wiggle
Posts: 4356
Joined: Tue Aug 04, 2009 7:15 am
Location: Rotterdamm son

Post by falafelbiels » Fri Sep 11, 2015 4:56 pm

Yeah still in

Jack Gull
Wiggling with Experience
Posts: 338
Joined: Sun Jan 26, 2014 11:46 am
Location: Italy

Post by Jack Gull » Sat Sep 12, 2015 5:14 am

Kroaton wrote:Will there be any way of changing the exponential-logarithmic response of the CV inputs like in the Frequency Central - Ultra Wave (video of that in action here) when using the CV expander?
not at the moment, but i've been meaning to do that for the saw -> pulse transition only. I just have to find out a clever way to do it
Also, any chance of there being a PCB+PIC option without the panel?
of course, just tell me what you want and i'll put you on the list ;)

Post Reply

Return to “Music Tech DIY”