replicating the Eardrill Pulse Balloon

Post your patches here! Put the format(s) in the subject.

Moderators: Kent, luketeaford, Joe.

Post Reply
steevolution
Learning to Wiggle
Posts: 22
Joined: Tue Feb 23, 2021 1:41 pm

replicating the Eardrill Pulse Balloon

Post by steevolution »

Has anyone used this module? Or does anyone have any tips on building out similar functionality with other modules?

As I understand it, pulses in raise the output voltage, with subsequent voltages raising the output higher. In the absence of any pulses, the out slowly decreases with this rate controlled by the gravity knob. The idea is to replicate the motion of batting a balloon up in the air and watching it slowly fall.

So far, my plan to recreate this has been to send the pulses into a slew limiter (function generator), setting the up slew to be quick and down slew to be slow. However pulse length has a huge impact on how quickly the "balloon" rises and the output isn't as subtle/smooth as I'd like. I'm thinking of either:

A) using the pulse to generate a decay envelope and seeing that, then shaping the envelope curve / length for different effects. From my reading of the Eardrill pulse balloon, this doesn't seem to be how it's implemented, but I think with some wiggling it could be functionally similar?

Or

B) sending the output of the slew limiter to a sample and hold that's triggered on each incoming pulse. Then attenuting the incoming pulse and mixing it with the output of the sample and hold. The idea being that each pulse could only make the output incrementally higher. The only issue is this could take too many pulses to get to a reasonable level. I guess I could first send the sample and hold throw some Max logic and use that to set a minimum level the first pulse can drive it to. Or I could use a vca to attenuate the incoming pulses and attenuate it more if the output is already high.
User avatar
KSS
Super Deluxe Wiggler
Posts: 7606
Joined: Mon Jan 25, 2016 7:28 am

Re: replicating the Eardrill Pulse Balloon

Post by KSS »

It's a little like the *old* video game lunar lander. Code for that can be found online. Along with at least a few different versions.

And you're correct to use an integrator aka slew to replicate it. This handles the Up part easily. Fed with constant voltage you get the expected slew. But fed with pulses -whose width and rate do matter as you've pointed out- you get the 'batting' of the 'balloon'. Time allows the discharge patch to drop the voltage as the balloon falls.

Change the rate and width of your pulses to get the results you expect. Because there are so many contributing factors to the result, this is a sensitive arrangement and can be tricky to set up and use. But that's also why it doesn't have to end up boring a few minute or hours later.

The Roland Jupiter4 EGs work like this too.

I've added a similar function to my 1011 50SQ clone of the Triadex Muse. Euterpe in the full version.
I've implemented both 'digital' and analog versions.
The analog is simply charging and draining a capacitor with CV and controls to dial it in, and is akin to using a slew.
In the 'digital' version the variable pulse output of the Muse is used to inject voltages into a Serge style SSM counter, and the spaces between pulses are used to either reset or decrement per clock. Since my 'SSG' has only 64 max steps, you can scale this across any voltage range -including 1V-Oct to use the output as note CVs
User avatar
Snufflepuff
Wiggling with Experience
Posts: 278
Joined: Sat Dec 19, 2020 11:53 pm

Re: replicating the Eardrill Pulse Balloon

Post by Snufflepuff »

What do you end up doing with the output voltage?
User avatar
KSS
Super Deluxe Wiggler
Posts: 7606
Joined: Mon Jan 25, 2016 7:28 am

Re: replicating the Eardrill Pulse Balloon

Post by KSS »

The accumulator outputs are just another couple CV sources. To be used whenever-where ever. The top is aimed more towards pitch CV and Gate-Trig, while the bottom is more aimed at being a modulation voltage. That's in line with how the original ARP2500 matrix was set up, which is what the APT 50SQ modules and format are based on.

The nice thing about them -the accumulators- in the 1011 is that they're both being fed from the same pulse train. But as one is analog -aka smooth- and the other is 'digital' -aka stepped- they are not quite the same. And can be made wildly different by controlling the input parameters for each.

I left off that both sections also have a comparator for Gate-trigger outputs.

Here's a few drawings which may help, since it's kind of abstract without context.

I've basically broken out the MUSE's LPT internals to a banana or stackcable patch bay in the narrow 1011X module. So there's a lot of recursive patching possible. It also provides two knobs for use in setting the parameters of one or both accumulators. I've got the top one labeled SCALE here, but it might end up being unlabeled so it can more easily be chosen as scale or set point-threshold for the G-T comparator. The lower knob is similarly assigned at installation for some specific use. There's no microprocessor or memory on the module pair and Xpander so multi-use knobs aren't really possible.

The 1011D display 'may' end up with a uP, but for now it's more old school lightshow -like the original Triadex product- and at least initially all analog.

The B0 input is a 'jam' input to ignore-over-ride what the 1011 LPT is putting out. The little black tact switch is a manual way to do this and the orange jack allows external sources to be patched in. As on the original MUSE this row is used to provide an always off-low to the shift register via the long slide switches. That frees up its LED space -which is always dark in the original- for the yellow seen here. Yellow chosen because it's like merging into oncoming traffic so proceed with caution. Orange input jack to show that it's a combination of yellow caution and red=trigger in 50SQ.

The VC input on the lower accumulator is the decay time input. Higher voltages lengthen the decay. I think the rest of the labeling is self-explanatory. The 5 left I/O spots upper and 4 left lower are for the basic MUSE controls that will be missing if the 1011A module is not present.

I'm also adding a synth voice -Based on the ARP Solus- to the 1011A which fills the empty space between the speaker and controls of the original MUSE.
This provides built-in destinations for the two accumulator parts of the main generative portion of the set. And makes the Euterpe cabinet a potentially self-contained yet useful and fun standalone synth-SEQ combo.

1011- B MUSE D = Basic Muse Digital = Bemused.
1011A- A MUSE NG = Analog Muse NextGen = Amusing. NG alternatively stands for engine, relating to the controls of the LPT gen and the included synth voice.
1011D- Analog Muse Display = Amused.
1011X= I/O expander
While one only 'needs' the 1011 itself to create pulse trains like the original MUSE, the other modules expand on this and the two module Euterpe set include the full I/O of the Triadex MUSE. The Euterpe cabinet will be shaped like the original, and is approximately the same size. It's designed as a base for the main 50SQ cabinets, leaving he desktop more clear for other units while presenting a unique appearance. Covid and its remaining aftermath have impacted this but I'm hopeful it will come to pass this year.
Attachments
APT-50SQ-1011-1h-1011X-2a-CLR-TopDetail.jpg
APT-50SQ-1011-1h-1011X-2a-CLR-BotDetail.jpg
MuseSet-Euterpe-1a-CLN.jpg
Treat utility modules as stars instead of backup singers.
Treat power supplies like Rockstars instead of roadies!
Chase magic sound, not magic parts.
User avatar
cbm
Eardrill
Posts: 2476
Joined: Thu Dec 04, 2008 10:07 pm
Location: Alameda, California

Re: replicating the Eardrill Pulse Balloon

Post by cbm »

Image

Conceptually, a Pulse Balloon section can be recreated using an attenuator with offset (Force), a slew set for zero attack and a longer release (Gravity), and a VCA for Tap.

The output of the attenuator is sent to the input of the VCA.

The output of the VCA feeds the Balloon slew.

The offset control of the attenuator is the Force source.

In addition to being the Balloon output, the output of the slew is fed back to the attenuator to accumulate state.

A Tap pulse is the control voltage to open the VCA. The pulse needs to be about as long as the minimum time on the slew you're using. When the VCA is open, the current slew state, plus the Force offset is sent to the Balloon slew driving it higher.



I haven't looked at that code for over a decade, but I don't think there's anything magic lurking inside. I had to look at the manual to remember how this worked. I just patched it up using the slew in a Frap Tools Falistri, one section of a Frap Tools 321 attenuvertor, and a VCA, with reasonable results. Tuning of all the sections to work together is the slightly tricky part.
Chris Muir
Eardrill Archive - http://www.eardrill.com (on permanent vacation)
User avatar
KSS
Super Deluxe Wiggler
Posts: 7606
Joined: Mon Jan 25, 2016 7:28 am

Re: replicating the Eardrill Pulse Balloon

Post by KSS »

Thanks cbm! :tu:
User avatar
cbm
Eardrill
Posts: 2476
Joined: Thu Dec 04, 2008 10:07 pm
Location: Alameda, California

Re: replicating the Eardrill Pulse Balloon

Post by cbm »

After sleeping on this, I realized that because the Pulse Balloons was a digital module, there was a virtual sample/hold in the feedback path.

When a new tap comes in the current Balloon value is "sampled," and the new Force addition uses that value. There is never a time when the feedback loop can directly push the Balloon higher, as there almost certainly is with the attenuator/slew/VCA version I described above. It's only a trigger pulse wide, but still enough to exhibit a slightly different, more finicky, behavior from a Pulse Balloons module. I don't have an actual Pulse Balloons module to test the difference, though.
Chris Muir
Eardrill Archive - http://www.eardrill.com (on permanent vacation)
Post Reply

Return to “Synthesis Techniques”