<kristianpaul>
I will not do that with my M1 wpwrak ;)
<kristianpaul>
I remenber when adam did some powercycles interruptions to some rc2..
<kristianpaul>
I was told NOT to try that at home :-)
<wpwrak>
kristianpaul: i was thinking of some 1000 cycles at around 0.5 Hz :)
<kristianpaul>
:-|
<wpwrak>
the problem is that labsw sometimes gets confused by interference. this seems to come from the relay contact closing (or maybe the current that flows through the wires). it's hard to reproduce with anything but M1. M1 gives me a "hit ratio" of about 10%. with my other lab instruments, it's well < 1% if it happens at all
<kristianpaul>
M1 in quarantine at wpwrak's apartment :-)
<kristianpaul>
Your scope should be scared of that interference field ;)
<kristianpaul>
(just kidding)
<wpwrak>
the scope is another issue. seems that i can't quite trigger on the damn behaviour. part of it are design limitations of the scope. part of it are bugs.
<wpwrak>
the interference is actually just M1 powering up. not sure where exactly it comes from. i suspect it's a current surge from the power supply (e.g., to fill all the input caps)
<wpwrak>
i ran labsw switching 7 V / 3 A from a lab supply and nothing untoward happened. tried another at 5 V / 2 A and only very few anomalies appeared (~ 2% or less). i get the "best" results with one that makes a lot of noise (beeps) when i short its outputs. so that one has to wait until daytime.
<wpwrak>
of course, this was still well below the M1 score, which is in the 10-20% range :)
<GitHub142>
[autotest-m1] xiangfu pushed 1 new commit to master: http://git.io/2p6UgQ
<GitHub142>
[autotest-m1/master] update clean to new boot.crc name - Xiangfu Liu
<GitHub177>
[scripts] xiangfu pushed 3 new commits to master: http://git.io/gwiqbA
<GitHub177>
[scripts/master] compile-flickernoise: add build boot.crc.bin - Xiangfu Liu
<GitHub177>
[scripts/master] compile-milkymist-firmware.sh update - Xiangfu Liu
<GitHub177>
[scripts/master] add compile-openwrt-milkymist.sh - Xiangfu Liu
<kristianpaul>
haha, lekernel should changes his nick to vjlekernel ;-)
<wpwrak>
verifying that power cycling in standby is indeed harmless. so far, it seems to be like that. (1280 cycles and counting)
<wpwrak>
will have to do run where i expect corruption afterwards. had too many that didn't do much lately. wouldn't do if the problem just went away due to some other effect.
<kristianpaul>
lekernel: midi allow me to control patches as i were with keyboard right? and also i can read midi from patches and modify running render parameter?
<kristianpaul>
btw a friend of mine told me that 4 hours may be a small amount of time for the arduino/mm1 workshop, mainlly because too much information at the same time, anyway just droUping the comment