#garfield Logs

Feb 27 2021

01:20 AM polprog: rue_mohr: ive got a bit more free time this weekend.. which converter was it? the one with the main choke?
01:31 AM polprog: ok, that
01:32 AM polprog: ill try
01:44 AM rue_mohr: you can use a ferrite ring, but choose carefully so you dont need 800 turns to get 2.5mH
01:44 AM rue_mohr: the good core I found was ~23 turns
07:09 AM polprog: ill take a look
09:30 AM polprog: hmm... core #8 does not seem to work. yellow, outer dia = 1.1"
09:32 AM polprog: attempt 2
09:32 AM polprog: im using a 2n2222
10:04 AM polprog: it started to oscillate, i put a C across the R
10:04 AM polprog: idk what my inductances are though, this is a small slab of ferrite from a broken large core
10:04 AM polprog: i have to get some Known Cores
10:05 AM polprog: ok, 270kHz. R=220, C=33n
10:09 AM polprog: 912kHz. R=1k, C=33n
10:09 AM polprog: *the fcc would like to know your location*
10:09 AM polprog: 5V @ 9mA
10:12 AM polprog: 12mA actually
10:13 AM polprog: i forgot to count how many windings...
10:13 AM polprog: i know the length of the wire tho
10:14 AM polprog: and the dimensions of the core
10:14 AM polprog: 800kHz, i think i can catch this on my sa
10:31 AM polprog: outer, 81 windings
10:47 AM polprog: output winding pk-pk 2.5V
10:48 AM rue_bed2: polprog, a 2222 is ok, you have to make sure your about 2.5mH
10:48 AM rue_bed2: it shouldn't fail to oscillate
10:48 AM rue_bed2: shouldnt need a cap
10:48 AM rue_bed2: maybe you have the feedback winding backswards?
10:48 AM polprog: it does oscillate, im not sure about the inductance
10:49 AM polprog: it is not backwards, i checked it
10:49 AM rue_bed2: oh I thought you got an L meter
10:49 AM polprog: nope
10:49 AM polprog: i could probably check an lc tank circuit with a known cap and the signal gen
10:49 AM rue_bed2: you need to watch the winding count, cause the pri:sec ratio totally determines the output voltage
10:50 AM polprog: yeah i got about 2x more windings on the sec
10:50 AM rue_bed2: you want 5:(12.7-5)
10:50 AM rue_bed2: it should run about 50% duty
10:51 AM rue_bed2: if not, the feedback thing isn't quite right, the diode should ground out the flyback
10:52 AM polprog: i think i dont have cores good enough
10:52 AM rue_bed2: if its running ~1khz your in the right area
10:52 AM rue_bed2: inductance might even be a bit high
10:53 AM polprog: ok, i have an inductor that says 0.8uH with 5 windings on it
10:53 AM rue_bed2: n^2
10:53 AM polprog: oh it's n^2?
10:53 AM rue_bed2: L~~ n^2
10:53 AM polprog: ok...
10:54 AM polprog: 25 ~ 0.8
10:54 AM polprog: so 0 ~ 0?
10:54 AM polprog: hmm
10:54 AM polprog: how to solve that
10:54 AM rue_bed2: I was going to make a program where you put in the turns and inductance, tell it the new incutace you want and it tells you the new turns
10:54 AM polprog: yeah that would be neat
10:54 AM rue_bed2: remember that website I used for calculating?
10:54 AM polprog: im tired :/
10:54 AM polprog: uhh. russian one?
10:54 AM rue_bed2: yea
10:55 AM rue_bed2: you dont need to put in the core dims and stuff
10:55 AM rue_bed2: just the inductance and turns
10:55 AM rue_bed2: then carry whatever u number is calcs to the second half
10:55 AM rue_bed2: as long as the dimentions are the same they cancal out
10:55 AM rue_bed2: I just woke up
10:56 AM rue_bed2: L = k*n^2
10:56 AM rue_bed2: k = L/n^2
10:56 AM rue_bed2: ---
10:56 AM polprog: http://zpostbox.ru/how_to_determine_permeability_of_unknown_cores_online_calculator.html
10:56 AM rue_bed2: n^2 = L/k
10:56 AM rue_bed2: yup
10:56 AM rue_bed2: just put in your turns and inudctance
10:56 AM polprog: thats for toroidal eh
10:56 AM rue_bed2: no
10:56 AM rue_bed2: doesn't matter
10:57 AM polprog: ok, let me try
10:57 AM rue_bed2: n = sqrt(L/k)
10:57 AM polprog: it says 9 turns
10:57 AM polprog: shall we try it
10:57 AM rue_bed2: no
10:57 AM rue_bed2: uH vs mH
10:57 AM polprog: oh shit, micro henries??
10:57 AM polprog: 900...
10:58 AM polprog: no i odnt have the tools for that :v
10:58 AM rue_bed2: no no n^2
10:58 AM rue_bed2: 0.8uH 5 turns
10:58 AM rue_bed2: k = L/n^2
10:59 AM rue_bed2: k = 0.032
10:59 AM rue_bed2: --
10:59 AM rue_bed2: n = sqrt(L/k)
10:59 AM rue_bed2: l = 2000
10:59 AM rue_bed2: whats sqrt of 62500?
11:00 AM rue_bed2: my command line calc cant do fn's yet
11:00 AM polprog: looks like between 200 and 300 turns on this calc
11:00 AM polprog: 280
11:01 AM polprog: 2499uH
11:01 AM rue_bed2: ok, so not that core
11:01 AM rue_bed2: find one that was used as a cable filter
11:01 AM rue_bed2: like on a usb cord
11:01 AM rue_bed2: or a power adapter
11:02 AM polprog: not mains filter?
11:02 AM rue_bed2: the core from the linked inductors also have a really high u
11:02 AM rue_bed2: :) yea
11:02 AM polprog: linked inductors?
11:02 AM polprog: so mains filter is OK?
11:02 AM rue_bed2: should be
11:02 AM polprog: ive got one that went on a falt flex, long flat one
11:03 AM rue_bed2: yup, that shoudl be high u
11:03 AM rue_bed2: often hard to wind them :)
11:03 AM rue_bed2: its not even 9
11:03 AM rue_bed2: how did I wake up so early?
11:04 AM rue_bed2: oooh wait, it was supposed to be a nap, I didn't have supper
11:04 AM polprog: ill put 5 turns inm abd try to measure L with a scope
11:04 AM rue_bed2: I hope I didn't start supper
11:04 AM rue_bed2: polprog, OR
11:04 AM rue_bed2: the inductance is just to keep the freq of the osc reasonable
11:05 AM rue_bed2: if its running somewhere from 10khz - 20khz its good
11:05 AM rue_bed2: but you also want it to need enough turns so you can get the ratio close
11:06 AM rue_bed2: close is +- 0.5V
11:06 AM rue_bed2: the good news is, that the transfomer is the only catch to building this supply
11:21 AM polprog: ok, im measuring the inductance of the flat core
11:21 AM rue_mohr: with a slot, right?
11:21 AM polprog: dinner o_o
11:21 AM polprog: yes
11:29 AM polprog: this doesnt make sense
11:29 AM polprog: L=? C=1u, f_0 = 35khz
11:30 AM rue_mohr: freq kinda high
11:30 AM polprog: L = 1 / C * 4pi^2*f^2
11:30 AM rue_mohr: I dont suggest a cap
11:30 AM polprog: 64uH
11:30 AM rue_mohr: the slew will throw it
11:30 AM polprog: no, im measuring the inductance of the flat core using an LC tank
11:30 AM rue_mohr: ooh
11:30 AM polprog: i put 9 turns on it
11:31 AM rue_mohr: hmm, I wonder if I can work out another way of doing this
11:31 AM polprog: permeability 2849
11:31 AM polprog: thats better
11:31 AM rue_mohr: hmmmm
11:31 AM polprog: website says 56 turns on this one
11:31 AM polprog: thats better
11:32 AM polprog: lets measure a different core
11:35 AM rue_mohr: 56 is reasonable
11:35 AM rue_mohr: not tooo small that you cant get the secondary ratio
11:35 AM * Tom_L overslept
11:35 AM polprog: still 20 is nicer
11:35 AM Tom_L: but feels better
11:36 AM polprog: no, im 20 and i feel like shit
11:36 AM Tom_L: we abuse ourselves when we're young and pay for it later
11:40 AM rue_mohr: polprog, that design is quite forgiving
11:40 AM rue_mohr: you just want to get the secondary ratio right
11:41 AM polprog: rue_mohr: ok, i found a core that will do me 18 turns for 2500uH
11:41 AM polprog: note to future polprog, core numner 13, green, outer dia 13mm, inner dia 6, permeability 28324
11:42 AM polprog: number*
11:48 AM polprog: ok. 20 windings x2 + 40 windings
11:56 AM Tom_L: at least it's future polprog instead of other polprog
11:57 AM polprog: :)
11:58 AM Tom_L: get a couple rue's together and they talk all night
12:05 PM polprog: nope, they dont work either
12:07 PM polprog: wait, it does
12:07 PM polprog: 10khz!
12:07 PM rue_mohr: more turns
12:07 PM rue_mohr: oh wait
12:07 PM polprog: wait, no
12:07 PM rue_mohr: I went lower, but ok, 10k - 20k
12:07 PM polprog: 100khz
12:08 PM rue_mohr: more turns
12:08 PM polprog: 180..
12:08 PM polprog: i have a 1k pot to adjust it
12:08 PM rue_mohr: hu?
12:08 PM polprog: instead of the 22k resistor.
12:08 PM polprog: also im running at 5V
12:08 PM rue_mohr: oh I dont reccomend adjusting that
12:08 PM rue_mohr: there is a sweet spot
12:08 PM polprog: why
12:09 PM polprog: well, im trying to find it this way
12:09 PM rue_mohr: a tradeoff between the drive current and the timing
12:09 PM rue_mohr: iirc the 22k was close to ideal
12:11 PM rue_mohr: hmm
12:11 PM rue_mohr: how many turns is your pri/feedback?
12:11 PM polprog: i only have the first two now, they are both 20 turns
12:11 PM rue_mohr: hmm
12:12 PM polprog: ill put a 40 turn third one
12:12 PM polprog: 13mA idle is a safe current...
12:12 PM rue_mohr: ...
12:12 PM rue_mohr: 20*(5/(12.7-5))...
12:13 PM polprog: the capacitor help, the swing on the base is bigger and i can go lower in frequency
12:14 PM rue_mohr: you have that diode, right?
12:14 PM rue_mohr: on the feedback?
12:14 PM polprog: yes,
12:14 PM polprog: i also just moved it to the base and its more stable now
12:14 PM rue_mohr: something sounds odd
12:14 PM polprog: i have the transformer right.
12:14 PM polprog: ill put the third winding now
12:15 PM rue_mohr: so the secondary needs to gen 7.7V
12:15 PM rue_mohr: turns ratio 1.54
12:15 PM rue_mohr: secondary should be 31 turns
12:15 PM rue_mohr: to get 12.0 out
12:16 PM polprog: the voltage swing on the winding is 1Vpp
12:16 PM polprog: only
12:16 PM rue_mohr: thats not right
12:16 PM rue_mohr: should be 5V
12:16 PM polprog: yeah, its odd
12:16 PM rue_mohr: are you sure you have the feedback polarity right?
12:16 PM rue_mohr: if its working right, the duty (without capacitors) is 50%
12:17 PM rue_mohr: or close
12:17 PM polprog: yes, ive been paying attention to that while soldering to the header
12:17 PM rue_mohr: hmm
12:18 PM polprog: the duty is 50%, its just that without the diode the voltage swing on the base is like 200mV
12:18 PM rue_mohr: the diode has to be there
12:18 PM rue_mohr: its taking the flyback and resetting the core
12:18 PM polprog: it is there
12:19 PM rue_mohr: it should be full 5V swing
12:19 PM polprog: im seeing it start very nicely, for a moment the transistor is in a state where it switches on/off completely
12:19 PM rue_mohr: like a wonderfull square wave
12:19 PM rue_mohr: hey, if your transistor is slightly leaking,
12:19 PM rue_mohr: try a 470k B-E resistor on the 2222
12:19 PM rue_mohr: ?
12:20 PM rue_mohr: it shouldn't need it, the core completely reverses polarity
12:20 PM rue_mohr: :/
12:22 PM polprog: i know that in theory it should work, its just i cant get it right
12:22 PM polprog: now i have a 100k across B-E and it still needs <1k on the base to start ringing
12:22 PM rue_mohr: can we blame the core?
12:22 PM polprog: maybe
12:22 PM polprog: the inductance seems to be right
12:22 PM polprog: ill measure it, sec
12:23 PM rue_mohr: it should auto start, cause the feedback is tied to +5
12:23 PM polprog: it does auto start, but for a short while
12:23 PM rue_mohr: it works by turning itself off
12:26 PM polprog: hey wtf
12:27 PM rue_mohr: ?
12:27 PM polprog: the core has nearly no inductance
12:28 PM rue_mohr: ah
12:28 PM rue_mohr: I used a small toroid
12:28 PM rue_mohr: but it was a special one for sure
12:28 PM rue_mohr: I think it was from the video line of a monitor
12:28 PM polprog: it as if it wasnt there at all (im doing my LC tank circuit measurement again
12:28 PM polprog: )
12:29 PM polprog: i wanted to see the actual L of the core
12:29 PM polprog: its none
12:29 PM rue_mohr: I wonder if I can come up with a reference osc based on 10 turns
12:31 PM rue_mohr: better have breakfast so I'm ready to go test toms cap!
12:32 PM Tom_L: better take a big hammer along
12:32 PM rue_mohr: ?
12:32 PM rue_mohr: I suspect it'll fit fine
12:37 PM polprog: the coil's either open circuit or too small inductance
12:37 PM polprog: i cant get it to oscilalte in an LC tank even
12:37 PM polprog: >_>
12:38 PM polprog: crap
12:38 PM polprog: i have to take a break
12:38 PM polprog: it kinda works though
12:38 PM polprog: think about the cap
12:40 PM rue_mohr: hmm
12:40 PM polprog: bbl
02:32 PM rue_mohr: >:| drat!
02:32 PM rue_mohr: Tom_L, too small
02:33 PM rue_mohr: and he wasn't there, so I didn't have anything to know which way its too small
02:37 PM rue_mohr: I'll see if I can adjust it when I work it out!
02:39 PM rue_mohr: its much deeper than needed, there isn't much room from the tank to the bottom of the thread, which isn't an issue
03:00 PM rue_shop2: rods are 8.5mm
04:03 PM rue_shop2: well I'm sad nothings working
04:12 PM Tom_L: how much too small?
04:13 PM rue_mohr: he wasn't there, I didn't bring any tools
04:13 PM rue_mohr: so there will be another round yet
04:13 PM rue_mohr: it started
04:14 PM rue_mohr: didn't make it a full turn tho
04:14 PM Tom_L: i can try another if you want if we can figure out how to fix it
04:14 PM rue_mohr: I want to know more
04:14 PM Tom_L: the tpi looked ok though right?
04:14 PM rue_mohr: its possable that the threads are 100%, and its too deep for the root on the tank
04:15 PM rue_mohr: yes, again, didn't have anything to check with
04:17 PM Tom_L: spray something on the thread and see what rubs off
04:17 PM Tom_L: if that's possible
04:17 PM Tom_L: need a gage
04:19 PM rue_mohr: no worries, he was supposed to be there, I'm going to try again later
04:19 PM Tom_L: not much you can do to fix that one
04:20 PM rue_mohr: if the thread tips are just too deep, its easy
04:20 PM Tom_L: and if it didn't make it 1 thread it may need to be more than .005 - .010" bigger
04:21 PM rue_mohr: otherwise I can TRY centering it up in the lathe and using a modified tap and...
04:21 PM rue_mohr: let me find out
04:21 PM Tom_L: can't hurt it any, it's scrap right now
04:21 PM rue_mohr: its pretty tho!
04:21 PM Tom_L: :)
04:21 PM Tom_L: bak in a bit gotta run an errand
04:22 PM Tom_L: small thread was about right though wasn't it?
04:22 PM rue_mohr: hah, didn't check
04:23 PM rue_mohr: I can hit it with my tap if I need it bigger :)
04:23 PM rue_mohr: I'm having big problems trying to stay awake
04:31 PM aandrew: rue_mohr: yeah I could *not* sleep last night. fought the bed the whole night, and at 4am finally just got up
04:31 PM aandrew: I hope that doesn't happen tonight
04:32 PM Tom_L: i'm game to try again if you want but i'd rather have a thread here to test against
04:32 PM Tom_L: that way i can leave it in the fixture until it fits
04:34 PM Tom_L: see what you can figure out on that one i guess
04:36 PM rue_mohr: I got lots of sleep
04:36 PM rue_mohr: but I'm exhausted
04:36 PM Tom_L: my allergies are kickin me but i'm better than i wass
04:36 PM rue_mohr: Tom_L, let me figure it out
04:36 PM rue_mohr: must be spring
04:37 PM rue_mohr: <rue_shop2> rods are 8.5mm
04:37 PM Tom_L: problem you will have it following the exact path of the thread if you tried to make it bigger
04:37 PM rue_mohr: unless its just the tips
04:37 PM Tom_L: and if it didn't make it 1 turn it's quite a bit too small
04:38 PM Tom_L: how many fewer threads you think it will need?
04:38 PM Tom_L: i think looking at the model you should have 10 good threads
04:39 PM rue_mohr: there is also gasket overhead
04:39 PM Tom_L: should have measured the male thread depth
04:39 PM rue_mohr: I wanted to talk to the other guy, if he'd been there I'd have probably compared it up with him and maybe brought the tank back
04:39 PM rue_mohr: I had no tools
04:40 PM rue_mohr: I was expecting him to be where
04:40 PM rue_mohr: there
04:40 PM Tom_L: dunno where i'd look for that thread here as a test
04:40 PM rue_mohr: and I'm literally nodding off every 10 mins
04:40 PM rue_mohr: I want to get things done damnit
04:41 PM rue_mohr: I cant when my eyes keep snapping shut and it takes everying just to lift my arm
04:41 PM Tom_L: you remember what brand that bottle was?
04:42 PM rue_mohr: General
04:42 PM Tom_L: no
04:42 PM rue_mohr: but 6 manufacturers use the same
04:42 PM rue_mohr: also Ansul, and kidde
04:42 PM rue_mohr: and a few others
04:42 PM rue_mohr: General is out of business now
04:42 PM Tom_L: are they ever put out of service?
04:42 PM Tom_L: wondering if the local place may have a bad one i could get
04:42 PM rue_mohr: they have to be serviced every 5 years
04:43 PM Tom_L: only way would be if they fail the pressure test
04:44 PM rue_mohr: the connector would still be ok
04:44 PM rue_mohr: they fail out the bottom or side
04:44 PM Tom_L: right
04:44 PM Tom_L: i figured that
04:44 PM rue_mohr: a tank may also be dispossed of if the head is damaged
04:44 PM rue_mohr: or just tank damage
04:44 PM rue_mohr: if its had the * kicked out of it, you take it out
04:45 PM Tom_L: are there any that have threads close to that? so i can avoid them if i decide to look
04:45 PM rue_mohr: no
04:45 PM rue_mohr: not that I'm aware
04:45 PM Tom_L: much smaller or bigger but not close
04:45 PM rue_mohr: its either going to be an internal thread (one of a few) or that 1 external
04:45 PM rue_mohr: iirc thats the only one threaded that polarity
04:46 PM rue_mohr: every other design is an external head thread
04:46 PM Tom_L: male on the tank
04:46 PM Tom_L: err no, female on the tank
04:46 PM rue_mohr: no, this is the only male tank
04:46 PM rue_mohr: all others are female
04:46 PM Tom_L: i'm a bit fogged over today
04:47 PM rue_mohr: I'm really messed up
04:47 PM rue_mohr: I'll take another asa and Bcomplex
04:50 PM Tom_L: looks like there are 4 test companies here
04:51 PM Tom_L: the one i was thinking of is still there
04:57 PM rue_mohr: <rue_mohr> <rue_shop2> rods are 8.5mm
05:11 PM Tom_L: my bud is going to make me a test plug
05:11 PM Tom_L: he has OD thread mics and thinks he can do it easy
05:12 PM Tom_L: going off the data from the machinery handbook for that thread
05:12 PM Tom_L: then i can rerun the threadmill more than once if need be until i get a nice fit
06:46 PM rue_shop2: !! :S
06:47 PM Tom_L: !?!!??
06:48 PM Tom_L: i just double checked the numbers on the model and it shows the major diameter as 1.505
06:48 PM Tom_L: whether it cut that or not is anybody's guess
06:55 PM rue_shop1: so it should be an easy fit, right?
06:55 PM Tom_L: i would hope
06:55 PM Tom_L: i figure if i rerun the same program and use that as a baseline i can adjust it from there
06:56 PM Tom_L: unless you think you can salvage that one
06:57 PM rue_shop1: I'd like to try
06:57 PM Tom_L: gonna knock the points off first?
06:57 PM Tom_L: did you take the tank home with you?
06:57 PM rue_shop1: I'm kinda dizzy, exhausted, and am having to breathe kinda heavy to do things
06:57 PM rue_shop1: no I did not
06:58 PM Tom_L: maybe you should have
06:58 PM rue_shop1: <rue_mohr> I had no tools
06:58 PM rue_shop1: <rue_mohr> I was expecting him to be where
06:58 PM Tom_L: yeah
06:58 PM Tom_L: you said he was too busy though
06:58 PM rue_shop1: <rue_mohr> he wasn't there, I didn't bring any tools
06:59 PM Tom_L: right now my bud is not doing much since aircraft is slow so he practially volunteered to make the plug
06:59 PM rue_shop1: no he wasn't there
06:59 PM rue_shop1: he was supposed to be there
06:59 PM rue_shop1: wtf is wrong with me!
06:59 PM rue_shop1: arg
06:59 PM rue_shop1: waste of one of my 104 days a year to get my thigns done
07:00 PM rue_shop1: if I know the hospital wouldn't just hold me for 4 hours and say they cant find anything I'd freaking go in
07:00 PM rue_shop1: damnit
07:00 PM Tom_L: i like these handbook pdf's .. i can print the page and actually see the print now
07:00 PM rue_shop1: :)
07:01 PM Tom_L: i'd have to be spewing blood before i'd go there and then someone would have to drag me
07:02 PM Tom_L: so you think you only need about 6 - 8 threads?
07:02 PM rue_shop2: I have to do t all over again
07:03 PM rue_shop2: I have to call him and organize again
07:03 PM rue_shop2: I have to go there and talk to him and measure and stuff
07:03 PM rue_shop2: maybe just pretend I wasn't here yet
07:03 PM rue_shop2: I'm having problems
07:03 PM Tom_L: but you were able to get in?
07:03 PM Tom_L: even though he was gone
07:03 PM rue_shop2: he has it sitting outsize
07:03 PM rue_shop2: but I have to do again
07:03 PM Tom_L: oh
07:04 PM rue_shop2: ( I cant drive right now)
07:08 PM Tom_L: guess i'll go take a nap
07:09 PM rue_shop2: maybe my iron is out
07:11 PM Tom_L: naw, the plug just came loose
07:22 PM rue_mohr: I think I'm low on iron
07:28 PM Tom_L: wonder how i got your number.... was looking thru my phone for my bud's number and saw yours
07:29 PM Tom_L: that or when i was on the missing person's list
07:35 PM rue_mohr: :)
07:36 PM rue_mohr: I think its really low iron
07:36 PM rue_mohr: I think this has happened before
07:37 PM rue_mohr: latley all my claws have been snapping
07:37 PM rue_mohr: and apparently thats a symptom
07:37 PM rue_mohr: might also explain the excesivly cold fingers and toes
07:42 PM rue_mohr: Sorry, there are currently no appointments to be booked online, there may still be some available, so please call the office
07:42 PM rue_mohr: ^^ so much for the new system