Quadrax Firmware 1.2.0

1.2.0 (2020-11-19)

  • NEW: Phase sync for clock-synced LFOs. See manual for details.
  • CHANGE: Various optimizations, reduced latency.
  • CHANGE: Sharper minimum attack and decay times.
  • CHANGE: Add 5% dead zone at end of pot travel to account for component tolerances.
  • CHANGE: Remove crossfade at the extreme ends of one-shot mode 1. Instead the waveform just switches from square to sine.
  • FIXED: Fixed initial phase of one-shot mode square waveforms.
  • FIXED: Improved trigger detection.

https://intellijel.com/downloads/firmware/Intellijel_Firmware_Updater_macOS_2020-11-19.zip
https://intellijel.com/downloads/firmware/Intellijel_Firmware_Updater_Windows_2020-11-19.zip

Updated Manual:

9 Likes

I was just testing out the new firmware, and I noticed that the problem with the Burst to lock up when turning the Rise knob clockwise is still present. During beta testing, I noticed it happened when turning Rise fully clockwise, but now it starts to happen when turned a little past 3 o’clock. This also causes the EOR gate on Qx to lock up as well.

Also, with Burst Retrigger option set to ON, I just noticed that it only retriggers during a falling portion of a burst. So if the SHAPE knob is turned fully clockwise, there is no retrigger happening at all.

Testing the Cycle (green) mode, turning both the Rise and Fall knob fully counterclockwise causes there to be a significant drop in range of the LFO. It is most noticable when the SHAPE knob is turned either fully clockwise or counterclockwise. The range drop causes an abrupt and very different change in timbre when modulating things like filter cutoff or any other timbre parameter destination.

Will check these out ASAP, thanks

1 Like

Hi,
I’m new here and want to thank you a lot for your work.

Quadrax is an awesome one. I’m really glad of this choice.

1 Like

Just hopped on this version. Seems successful except that channel 2 isn’t outputting voltage. The trigger chain works normal, but nothing comes out. Tips on how to fix it?

Sometimes firmware updates cause a bug where a channel goes dead, but this can be fixed with a Reset in System Mode. I’ll copy the FAQ entry below.

There are two possible conditions that can make it possible for a channel to become seemingly dead on Quadrax.

  1. If a CV input is assigned to the LEVEL of a channel and there is nothing present at that input the channel will not output any signal. You should either send a CV signal to control the level of that channel or clear the CV assignment.

Remember: You can quickly and easily clear all CV assignments by pressing and holding one of the left buttons to enter CV Assignment mode, then pressing and holding the bottom right button.

  1. There was a bug in a recent firmware update where under certain settings channels would become temporarily unresponsive. It is possible to correct this by performing a “Reset to Default” in the Quadrax System Mode. This requires the most recent firmware version and instructions on how to perform this reset are below as well as in the Quadrax manual.

To enter System Mode:

  1. To put Quadrax into System Mode, hold down the LEVEL button (Channel 4’s MODE/DESTINATION) button while you power on the module. Once in System Mode, you can perform a number of global functions, as discussed in System Mode Options, below.

Reset to Default

  1. In System Mode, long-press the LEVEL button to reset the module to its default settings.

To exit System Mode: 1. To exit System Mode and save your settings, press the flashing red LEVEL button.

1 Like

Clearing CV did it. Thanks!

1 Like

Hey there, I just got my Quadrax today and installed this firmware and have a few comments to make:

  1. The deadzone at the beginning and end feels extremely awkward and is a little annoying. It’d be nice to be able to turn this off in the menu, or even better add a calibration step so it can be perfect. Why is the dead zone even needed? Isn’t the pots job acting as a voltage divider feeding an adc?

  2. Not sure if that’s the same issue as @mtamayo mentioned, but using the CYCLE mode at audiorate with FALL all the way down turning RISE all the way down makes for a weird jump in pitch and feels also very unnatural. It would be nice having finer control over the short times than the long ones - so kind of a exp rather than log response over time if that makes sense? MAYBE this is really just me, but I feel like half of the turn is just about going from slow to even slower times…

  3. there is some serious pitch variation in the LFO mode without anything patched into CV. This was caused by CV A mapped to the RISE parameter. But I also noted that the frequency control in this mode shows a similar - but reverse - behaviour to issue #2 of the frequency jumping high in the last 30° of turn (probably caused by the dead zone…)

Generally I feel that the “tactility” could be immensely improved by being able to dial in finer adjustments. Being a former MATHS user this is something which I never even had to think about.

regarding issue #2:
this is the shortest rise time I am able to dial in VERY carefully, after this it jumps to almost 0ms

Experimenting a bit more I just found a few other things that appear to be bugs:

  • RISE and FALL does not respond to CV assignment inputs when turned fully counterclockwise for AD (unlit), AHR (red), and CYCLE (green) modes. You need to turn the RISE or FALL knob to approximately 9 o’clock to start getting a response from CV input. Also when it actually starts to respond to CV inputs it appears to have a non-linear response - e.g. having RISE or FALL set to 9 o’clock, it takes 5V input to move the Fall and Rise parameter from the 9 o’clock value to get to 12 o’clock position value. Applying 10V does not get to the max knob position value.

  • When in Burst mode (yellow) and LFO mode (magenta), application of CV assignments to the RISE parameter also appears to maybe have a non-linear response - e.g, if RISE is set to fully counterclockwise, applying 10V to the RISE parameter does not get the full range of values for pulse rate (Burst mode) and LFO rate (LFO mode).

  • There is a bit of a drop (about 2.5V) in amplitude/level for LFO mode (magenta) when Rise is fully clockwise and the SHAPE knob is fully clockwise/counterclockwise. Most noticeable when FALL knob is also fully clockwise/counterclockwise.

@mtamayo thanks for the thorough feedback, we’ll definitely be looking into these and do a patch release to fix

One thing about the rise and fall modulation, the CV inputs track 1V / octave, while the knob curve is not linear. So the first part of the knob is more for fine-tuning shorter times while later it becomes a lot more broad. If you’re in the fine tune range, then no amount of CV will move you to the max position because it’s several orders of magnitude larger.

2 Likes

This is because of the “snap to zero” behaviour that was added in the new firmware. See also the previous comment about the knob curve.

You may want to compare to the previous firmware and see how it behaved before, I’d be interested in hearing what you think.

So the first part of the knob is more for fine-tuning shorter times while later it becomes a lot more broad.

To be honest I can’t really retrace this behaviour or it would need to be more extreme to really achieve your intention…

How can I get the older firmware for comparison? I only have the possibility to choose 1.2.0.0 from the dropdown list

I’m not sure what you mean you can’t retrace it? in the positions before ~1 o’clock the times are under 2 seconds and near the start of the travel of the knob they are in the ms. range, after the 1 o’clock position they get much longer to the multiple seconds range.

Thank you for the reply Kamil, and for clarifying about the tracking. I had missed in the manual that it tracks at 1v/oct. I think that would only be applicable to LFO (magenta) and Burst (yellow) modes however because it doesn’t appear that the other modes should track at 1v/oct? Is there a non-linear response to those other modes as well even though they’re not supposed to track at 1v/oct? Also you mention the fall modulation, but 1v/oct tracking would only be applicable to the Rise parameter, or am I missing something?

Not sure if I’m doing something wrong, but it appears that when a channel is in AHR mode (red) and has the Trigger Link option (red) on the same channel, the Trigger Link does not work if the prior channel has no Link option (unlit) selected. For example, Channel 1 is set to any mode with no Link option (unlit), and Channel 2 is set to Trigger Link (red) and AHR mode (red) . . . when Channel 1 receives a trigger or gate, Channel 2 is non responsive. The only way to get Channel 2 to do anything is to set Rise to fully counterclockwise in which it then only acts as a decay envelope. I made sure all my CV assignments were cleared on all channels. I didn’t see this behavior noted in the manual, so I’m not sure if this is a bug or if I am missing something.

The trigger is only momentary so in AHR mode the envelope does not have time to rise unless the rise time is basically zero. Maybe it would make sense to have it link the gate in this situation.

1 Like

Ah ok, yeah it would make sense to keep trigger on all the other modes but a gate link for AHR. Otherwise it would just turn AHR mode into single stage decay.