#garfield Logs

May 28 2023

01:03 AM rue_shop5: trike 4, Rue 0
01:14 AM rue_shop5: The main drive sproket for the gang axle let go (slipping)
01:14 AM rue_shop5: so I re-shelved that and I'll work on one of the other 10000000 projects
02:03 AM rue_mohr: - order 6-32 taps
02:03 AM rue_mohr: - find the venting battery in the shop UPSs
02:03 AM rue_mohr: - look for radiator for house/shop cooling
02:03 AM rue_mohr: - design and print lots of L brackets
02:12 AM rue_mohr: ouch, $8 is typ for a 6-32 rap
02:12 AM rue_mohr: HSS
02:14 AM rue_mohr: - design and print lots of L brackets
02:24 AM rue_mohr: - write battery measure code for the dustbot
02:24 AM rue_mohr: - find the venting battery in the shop UPSs
02:24 AM rue_mohr: - look for radiator for house/shop cooling
11:35 AM rue_mohr: - write battery measure code for the dustbot
11:35 AM rue_mohr: ---
01:48 PM rue_mohr: ok, I watered the fish
01:48 PM rue_mohr: - write battery measure code for the dustbot
01:48 PM rue_mohr: - find the venting battery in the shop UPSs
01:48 PM rue_mohr: - look for radiator for house/shop cooling
01:51 PM rue_mohr: I wonder if I should cool the shop or the house..
01:53 PM rue_shop5: shop
04:30 PM rue_shop5: https://www.instructables.com/Wire-Hose-Clamp/
04:30 PM rue_shop5: it works
04:47 PM rue_mohr: - write battery measure code for the dustbot
04:47 PM rue_mohr: - find the venting battery in the shop UPSs
04:47 PM rue_mohr: I was able to get adapter hardware down to 1/4" pipe on the radiator and confirm it doesn't leak
05:15 PM aandrew: yeah there are a lot of standards, IEC 60601 and 62133 are kind of generic medial device standards, and UN 38.3 is a special certification for batteries for transport
05:16 PM aandrew: I spent most of the afternoon taking apart and trying to figur eout what's wrong with my mosquito magnet (propane fired mosquito attractor)
05:17 PM aandrew: it's down to one of two things: either the thermistor is kaput (it does change with temp, but maybe it doesn't completely work anymore) or the propane nozzle is iffy
05:17 PM aandrew: there is gas flow, it does ignite and heat up, but the 200k @ 25C NTC gets down to about 107k when it should be down closer to 14-30k
05:18 PM rue_mohr: huh
05:18 PM rue_mohr: is it over or udner firing?
05:18 PM aandrew: (when at temp) -- the controller is kicking out because after abotu 15m with the ingnitor on and watching the temp, it's not seeing enough temp
05:19 PM aandrew: it really is quite a simple design though. combustion chamber with 12v ignitor and 200k NTC and solenoid.
05:20 PM aandrew: when it powers up, it runs the fan for 3m to establish ambient, then opens the valve and runs the ignitor while watching temp
05:20 PM aandrew: when the temp is hot enough, it turns off the ignitor and runs
05:22 PM aandrew: it's all run off a stupid PIC16 so there's no reason why I could not roll my own as long as the combustion chamber's good
05:22 PM aandrew: I cleaned it all out, there was some corrosion but nothing bad
05:22 PM aandrew: and it clearly lights up I can feel the heat
06:47 PM rue_shop5: yep
06:47 PM rue_shop5: trying to find something good to do with the last few hours of the weekend
06:47 PM rue_shop5: I should try the VFD with a larger motor and see what it kicks that motor rail up to
08:07 PM aandrew: it seems unlikley that a thermistor would "half fail" so I guess I'm going to take the nozzle out and apart and try to clean it
08:07 PM aandrew: they're using oil burner nozzles
08:07 PM rue_mohr: for gas?
08:07 PM rue_mohr: propane
08:07 PM aandrew: for propane yes
08:08 PM rue_mohr: ?
08:08 PM aandrew: 1.5gph 45 degree, either A (hollow cone) or W (universal? used for humidity injection?) but can't seem to get a definitive answer
08:09 PM rue_mohr: I was able to pedal the trike up a hill!
08:09 PM rue_mohr: without catastrophic failure!
08:09 PM aandrew: nice
08:10 PM rue_mohr: the motor driver scares me
08:10 PM rue_mohr: I'm spooked that the larger motors might be able to kick the rails up to the 400VDC the caps are rated to
08:11 PM rue_mohr: tho, the 200V caps in the computer supply didn't blow
08:21 PM aandrew: yes - this was one of the two main critical monitoring points when I was in this biz
08:23 PM aandrew: bus voltage (and more importantly rate of change of bus voltage) and desat. shoot-through was not really a thing since the drivers were solid but if you saw the bus voltage heading for the moon we had a circuit (not software) which turned the braking resistor on to try to burn that energy off before it took out the caps
08:53 PM rue_mohr: Every weekend I get to walk around the yard and take all ma's empty planters back to her planter pile...
08:53 PM rue_mohr: aandrew, OOooo cool
08:54 PM rue_mohr: I got 400V caps, and 170V busses (+-170V) I wonder if just using a votlage trigger for a brake resistor would do?
08:55 PM rue_mohr: - write battery measure code for the dustbot
08:55 PM rue_mohr: - find the venting battery in the shop UPSs
08:55 PM rue_mohr: do I feel like sniffing UPS' right now?
08:55 PM rue_mohr: hmm
08:57 PM Tom_L: is that like sniffing glue?
09:03 PM rue_mohr: no, more sulphor
09:04 PM rue_mohr: ok dustbot
09:04 PM rue_mohr: wow, the editor is burried in like 20 programs since
09:06 PM rue_mohr: aandrew, well, want to give me pointers on the VFD conversion as I go?
09:08 PM rue_mohr: I need to check my shoot through, I'm using 2 different fet drivers, opto isolated ones, they have the input diodes back to back so they cant be triggered at the same time
09:08 PM rue_mohr: I think I got some IR2110 tho
09:08 PM rue_mohr: in general, I didn't intend it for this...
09:09 PM rue_mohr: hah no, my lo side is at -170V
09:16 PM aandrew: using a pair of comparators would probably be good... level 1 = IRQ/NMI "oh fuck" level, latching, and level 2 = hardware fet driver, non-latching
09:17 PM rue_shop5: I dont mind doing just hardware
09:17 PM rue_shop5: er, oh I think I see what you mean
09:17 PM aandrew: the desat one was interesting - we would turn off both halves of the h-bridge if one started coming out of saturation (I mean the other was already off of course, but we wouldn't allow it back on, it's a fault)
09:17 PM rue_shop5: not as in software control first, but like system fault badness
09:17 PM aandrew: exactly
09:18 PM aandrew: the software needs to know shit is going supercritical so it can do the needful
09:18 PM rue_shop5: this is just 2 switch driving
09:18 PM rue_shop5: I'd not expect a transistor to lock on...
09:18 PM rue_shop5: hmmm
09:18 PM aandrew: so I mean ideally if the bus voltage is spiking and software could control it, yay, but you had the hardware failsafe to try to protect the system
09:18 PM rue_shop5: I wonder if these IGBT could even take a hit to short out that cap bank
09:18 PM aandrew: I had my head in a unit intentionally fucking with it (2ph -> 3ph) to try to get some more power out of it and I heard this little squeak and the unit faulted out
09:19 PM aandrew: that was the hardware failsafe. It saved my noggin
09:19 PM rue_shop5: you know what I'm up to with the 3 DC motors eh?
09:20 PM rue_shop5: the incomming 120V is doubled to +-170VDC
09:20 PM rue_shop5: the motors are driven from the middle of the doubler (neutral) to either the + or - rail by the IGBT
09:20 PM rue_shop5: forward/reverse
09:21 PM rue_shop5: due to only needing 2 transistors per motor, I can drive 3 motors with the hardware from 1 vfd
09:22 PM rue_shop5: so my problem is that the inductive kickback of the motors from the pwm charges the rail opposite to the one that was driving it during the pulse
09:23 PM rue_shop5: in a test with a 18VAC transformer, I get +-22V, a small motor is kicking the opposite rail up to 85V
09:23 PM rue_shop5: there is a sanity resistor, it used to balance the charge across the 400V caps (they used two banks of 400V in series)
09:24 PM rue_shop5: now its a discharge/calmdown set of resistors
09:24 PM rue_shop5: ~11w steady loss
09:24 PM rue_shop5: (170V, 5.1k)
09:25 PM rue_shop5: I think I actually get more like 150V on the rails
09:25 PM rue_shop5: the IGBT are 1200V 150A, I suspect the only thing that could damage them is shunting out the cap bank
09:26 PM rue_shop5: its 6600uF each side
09:26 PM rue_shop5: the incomming AC is only "15A"
09:33 PM aandrew: right so you've basically got three indepenent half bridges with a common bus
09:33 PM rue_mohr: yup
09:34 PM aandrew: that does seem like a very high kickback though, is there really any actual energy there or is it something a diode+zener could easily handle?
09:34 PM rue_mohr: well, I kinda wonder if its just cause the votlage is so low
09:34 PM rue_mohr: its a tiny motor I'm testing with
09:35 PM rue_mohr: I think at 170V those resistors would bite in harder
09:35 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1280612.jpg
10:37 PM rue_mohr: I'm going to need something to convert stepper control signals to DC motor drive
10:37 PM rue_mohr: Am I ready to try this with EGMK?
11:32 PM rue_mohr: -- Rue this week ---
11:32 PM rue_mohr: - fix the timer box
11:32 PM rue_mohr: - add brakes and or shifters to the trike