#garfield Logs

May 07 2020

04:36 PM polprog: https://polprog.net/rozne1/ircjunk/computers/beetle/
05:32 PM rue_mohr: wow thats fast
05:33 PM rue_mohr: polprog, Point Of Sale?
05:33 PM polprog: yeah
05:33 PM rue_mohr: 24VUSB
05:34 PM polprog: what's fast? done a lot of stuff today
05:34 PM polprog: yeah. these ports are why i got it
05:34 PM polprog: :P
05:34 PM polprog: powdered usb :P
05:34 PM rue_mohr: 1.4Mhz on your converter
05:35 PM polprog: it was slower under load/temperature
05:35 PM rue_mohr: generally you run a converter between 40khz and 500kHz
05:35 PM polprog: i think it was mostly load related cause it would jitter a LOT
05:35 PM polprog: that would be about 700kHz at the end
05:35 PM polprog: pretty fast
05:35 PM rue_mohr: try bringing it down t about 300kHz
05:35 PM polprog: hmm, good idea
05:35 PM rue_mohr: with some small caps on the middle stages
05:36 PM rue_mohr: it'll convert better
05:36 PM rue_mohr: *should
05:36 PM polprog: i guess it should
05:36 PM rue_mohr: less transient time on the transistors
05:38 PM polprog: its kinda interesting to make that thing oscillate
05:53 PM polprog: ive put some pictures into the computers directory, feel free to surf
05:53 PM rue_mohr: they were all the computer images
05:54 PM rue_mohr: did you have one of cores?
05:54 PM rue_mohr: hmmm whats this guy up to with his power supply...
05:55 PM polprog: i mean the directory above
05:55 PM polprog: one of waht cores?
05:55 PM polprog: what*
05:56 PM rue_mohr: did you say you tested multipel cores?
05:56 PM polprog: yeah, i tested some of the ones from the box.. i dont have many
05:56 PM rue_mohr: k
05:56 PM rue_mohr: I find most of them suck
05:56 PM polprog: the one labeled "1" was the best, then there was one i wound myself, same as in https://polprog.net/rozne1/ircjunk/dcdc/handwound1_10mhz.png
05:57 PM rue_mohr: but those FRI beads, are darn good
05:57 PM polprog: yeah i unwound some of the toroidal cores and wound my own wire on them
05:57 PM rue_mohr: the EMI coupled inductors are generally good
05:57 PM polprog: common mode chokes?
05:57 PM rue_mohr: yup
05:57 PM rue_mohr: high uH/turn
05:58 PM polprog: i think the one i tried first was a choke, but later i tried a different choke and got only 19V, same load
05:58 PM polprog: maybe the freq was too fast
05:58 PM polprog: what is concerning is that the bottom trace on that pic doesnt look like an inductor trace at all
05:59 PM polprog: it's def too fast
05:59 PM rue_mohr: 42...
05:59 PM polprog: ?
05:59 PM rue_mohr: I was looking for a part number
05:59 PM rue_mohr: uc8343
05:59 PM rue_mohr: but lower voltage start
06:00 PM rue_mohr: 3842 wont start till its up to 16V
06:00 PM rue_mohr: I was looking for soemthing that would start with a 12V supply
06:01 PM polprog: i think that picture was something vaguely similar to 3842, also didnt start off under 16V
06:01 PM polprog: yeah i played with a 3842
06:01 PM polprog: https://polprog.net/rozne1/ircjunk/dcdc/uc3842_1.png
06:02 PM polprog: https://polprog.net/rozne1/ircjunk/dcdc/uc3842_2_sch.png
06:02 PM rue_mohr: there is a series of them
06:02 PM rue_mohr: the undervotlage is one of the things that different
06:02 PM polprog: now that i think, these chokes have two windings, so in could in theory make an isolated design
06:03 PM rue_mohr: you got it working tho, good show
06:03 PM rue_mohr: you also can make a non-isolated one
06:03 PM rue_mohr: you used the second winding there..
06:03 PM rue_mohr: https://polprog.net/rozne1/ircjunk/dcdc/uc3842_2_sch.png <-- nice
06:04 PM polprog: what's on breadboard on the other desk has both windings connected by a clip
06:04 PM polprog: thanks, thats partly from the datasheet :P
06:04 PM rue_mohr: remember to say what fet your using
06:04 PM rue_mohr: TheUC3842 and UC3844
06:04 PM rue_mohr: have UVLO thresholds of 16V (on) and 10V (off) The
06:04 PM rue_mohr: UC3843 and UC3845 are 8.5V (on) and 7.9V (off) The
06:04 PM rue_mohr: UC3842 and UC3843 can operate within 100% duty cycle.
06:04 PM rue_mohr: The UC3844and UC3845 can operate with 50% duty cycle.
06:04 PM polprog: hmm, cut cornerson the to-220.. thats probably P55NF06 or P16NF06, whatever fell into the tray
06:05 PM polprog: beforehand
06:05 PM rue_mohr: so look for UC3843 for 12V applications
06:07 PM polprog: i like all these UC384* are fixed frequency
06:08 PM polprog: id like to see the face of the FCC lab testers if they would get a device with variable frequency DCDC :P
06:08 PM polprog: RF jammer :P
06:09 PM rue_mohr: er, no....
06:09 PM rue_mohr: pin 4 is the osc control
06:10 PM polprog: ah
06:10 PM polprog: hey, it looks like it uses a relaxation osc inside
06:10 PM rue_mohr: yup
06:11 PM polprog: that probably means i could use one of my dual diodes as a varicap.. mayhem ensues
06:18 PM polprog: goodnight
06:18 PM rue_mohr: polprog,
06:18 PM polprog: huh?
06:18 PM rue_mohr: so, tiny13 as a power converter
06:18 PM polprog: ill see if i have some
06:18 PM rue_mohr: obviously with some kinda regulator for its power
06:19 PM polprog: should be doable with a big enough cap
06:19 PM rue_mohr: but, legal solution do ya think?
06:19 PM polprog: what do you mean by legal?
06:19 PM rue_mohr: obviously it cant do the driving, but...
06:19 PM rue_mohr: I think its a bad design choice
06:19 PM rue_mohr: if it locks up..
06:19 PM polprog: well, its got an adc, and could drive a fet
06:19 PM polprog: hmm yeah, im always considering that
06:19 PM polprog: what would happen if X locks up
06:20 PM polprog: i think Jartza used an AVR to generate a sine for his retrofit commodore pSU
06:20 PM polprog: but that was just for clocking the computer, it took a time signal from the zero crossing of mains
06:23 PM rue_mohr: hmm
11:07 PM rue_shop1: ok, test my diode logic circuit