#garfield Logs

Mar 25 2013

00:14 rue_more hmm, Ipartly stripped it beofre I realized maybe I shouldn't
00:14 rue_more so maybe I'll put it back togethor.
00:24 rue_more kat, why should I replace the door pins with roller bearings?
00:24 rue_more wait, this dosn't add up, the force directions are wrong
00:25 any1121234 what door?
00:25 rue_more oh, the truck door
00:26 any1121234 a) i doubt you can, i think you will haveto replace the entire hinge
00:26 any1121234 b) bearings will last longer than you will and never squeal
00:27 rue_more after I said that it seems to be if anything it should need a thrust beraing
00:27 any1121234 if you replace the pin, replace the bushings, pins and bushings come in a kit
00:27 rue_more primary force is along the axis of the hinge
00:28 rue_more I was thinking about turning brass bushings
00:28 any1121234 no, the weight isn't severe, but the overhung weight pulls and pushes sideways very strongly
00:28 rue_more hmmm
00:28 any1121234 i'd get the kit, it's only like $5 per door and guaranteed to fit
00:29 rue_more any work on the car today?
00:29 any1121234 only moving stuff around, crowbar can't get the steering shaft universals off the shaft, i'll haveto torch it
00:30 any1121234 there's a pipe of some sort under the windshield, going to make cutting that difficult as heck (to remove the glass intact for resale)
00:30 rue_more did you ever say why your after the engine computer?
00:31 rue_more or did I misunderstand
00:31 any1121234 to run the engine, i think
00:31 rue_more oh, k
00:31 any1121234 it has a thottlebody injector
00:32 rue_more I udnerstand
00:32 e_more whispers "you have a microcontrol
00:32 y1121234 whispers back that she has the working theory of operation, but if i have the oem device, why not use
00:33 e_more whispers back "think of the
00:33 y1121234 whispers back that she wishes she had the
00:34 any1121234 after the engine is out, i still haveto put it in something with wheels,a nd make a way to get the existing wheel torque to the new wheels
00:35 rue_more k
00:35 any1121234 and i don't want to go 60mph downthe mountain thrut he woods, i do want to make 5mph under a heavy load, so some gear reduction is in order, as well as tying in the power brakes, etc
00:36 rue_more can you get a good ratio tranny from the wreckers?
00:36 any1121234 what i need is a $5000 forklift and a $5000 farm tractor (with attachments)
00:37 any1121234 don't need to, i will try to put a tiney chain sprocket where the Geo't tires are now, and a 18 inch sprocket on the thing i have to put the engine on
00:38 any1121234 i have seriously clogged up FF
00:41 any1121234 http://designerthinking.com/3wheeler.html
00:42 rue_more I remember
00:43 any1121234 k, so now i'll be putting the lil 3 cyl engine to drive the rear wheels thru a chain speed reduction, instead of 500lbs of batteries and dc gearmotors
06:05 jennie hello
09:03 rue_more jennie, your back
09:03 jennie yes
09:03 rue_more did you run that code?
09:04 jennie no
09:08 rue_more can you run that code?
09:08 rue_more the paste site cant include the math library
09:10 jennie ok
09:10 jennie hey,also what about the driver circuit ?
09:10 rue_more that code is the key on how to run it
09:11 rue_more I have to look carefully at the pwm current limiter
09:11 jennie ahh
09:11 jennie also I have talked to one of ##robotics user about PIC code and driver circuit
09:12 rue_more the current circuit is a bit flimsy on the control
09:12 rue_more you must not use a microcontroller to limit the current
09:12 rue_more they dont react fast enough
09:13 rue_more when I put my scope on a hard drive, they run the coils at 12V, with 50Khz pwm, at about 66 or 33% duty
09:14 jennie ok, if we do not use uC then what we will use ?
09:14 rue_more we can use the miro to sequence it
09:14 rue_more but not to do the pwm current limiting
09:15 rue_more ah
09:15 rue_more yes, changing to the fets may mean the comparitor will be ok
09:15 rue_more then I just have to tune the rc values on the filter
09:15 rue_more k
09:16 rue_more let me see if I ahve a circuit for you
09:17 rue_more hmm we can just use a pullup on the fet gate...
09:18 rue_more hmm
09:18 jennie also have a look at this chat , http://pastebin.com/GYnPdnjq
09:19 jennie http://codepad.org/cOAL58Py is PIC code
09:19 rue_more hmm
09:23 jennie also he said that if motor is spinning with current MOSFET driver circuit then it is ok
09:25 rue_more the mosfets on the bottom and bipolar on the top
09:26 rue_more hmm, the feedback isn't too hard to add, not seen it done like that before, cool
09:29 rue_more http://www.microchip.com/forums/download.axd?file=0;581305&where=msg
09:29 rue_more but becuase your pic isn't dsp, we need to use the op-amp for current limiting
09:30 jennie ok
09:30 rue_more and were using bipolar on the top, unless you want to get some P channel fets
09:30 rue_more hmm
09:30 rue_more I have to go to work
09:31 rue_more !time
09:31 tobbor My watch says its 07:17AM Mon Mar 25 2013
09:31 rue_more I'll be back at 17:00 or so
09:31 jennie I said that I have STB55NF06
09:31 jennie ok
09:31 rue_more 10 hours...
09:31 rue_more yes, the pwm is on the bottom, the N channel fets
09:31 jennie I need to buy anything ?>
09:32 rue_more the top ones dont have to switch as fast, so bipolar will be ok
09:32 jennie new parts for circuit
09:32 jennie ?
09:32 rue_more I dont think so
09:32 jennie so I have to make the circuit you posted above ?>
09:32 rue_more nope
09:32 jennie ok
09:32 rue_more its an example of what he was talking about
09:32 jennie ahh
09:32 jennie got it
09:32 rue_more we need a slightly different circuit
09:33 jennie what about the one you made ?
09:39 rue_more its close
09:39 rue_more we have to change the npn out for the fets, but its not as easy as that
09:41 jennie ok
16:46 jennie !time
18:43 any68455593 in the olden daze, i just applied power, and just let the current trip oscillate,, the only thing new since the 1980's is syncing the current cutoff to latch and then reset to on when main pwm cut off
18:43 any68455593 it's just wierd to see this being a design problem in here
19:29 rue_more I just dont want to get too facny with it
19:31 jennie hi
19:32 jennie hi rue_more
19:41 rue_more hello
19:43 rue_more jennie, I'll work out your new circuit
19:43 rue_more and I have to take apart the other two printers and the server
19:43 rue_more and test the other 6 monitors
19:43 rue_more and I have 2 hours
19:43 rue_more to do it
19:43 rue_more all
19:44 rue_more kat, the circuit we made the transistors do get REALLY hot
19:45 jennie and you want them to be cool , but how you want them to be cool ?
19:45 rue_more jennie, I dont know when you monitor the phase voltage to know when to do the next switch
19:45 rue_more they should be no more than 30c in a room 25c with no heatsink
19:46 jennie Phase voltage ?
19:46 rue_more remember the comparitor thing you said?
19:46 rue_more thats what thats about
19:46 jennie so should I start making that circuit ?
19:47 rue_more you dont have the circuit to make yet
19:47 jennie then how I will measure phase voltage ?
19:47 rue_more not sure, I'm inclined to say to just do it open loop first
19:52 M0TRN_th1 void main(){for(;;){printf("p3n1s\t");}}
19:52 rue_more who the hell are you?
19:52 jennie he is my friend
19:52 M0TRN_th1 just followed topic ;)
19:53 rue_mohr be nice or your out :)
19:53 M0TRN_th1 ok :)
19:53 jennie he wants to participate in motor driver circuit
19:53 rue_mohr ah
19:53 M0TRN_th1 correction: jennie wants me to participate in motor driver circuit ;)
19:54 M0TRN_th1 I was helping her with PIC code
19:54 rue_mohr need to convert the pwm circuit to fets
19:54 M0TRN_th1 that's what I suggested first
19:54 rue_mohr ah, help jennie run this C code...
19:54 M0TRN_th1 but is the PWM done by PIC in software?
19:54 M0TRN_th1 so far we've just been toggling on/off
19:55 M0TRN_th1 this PIC doesn't have hardware PWM
19:55 rue_mohr http://codepad.org/i9arFt4j
19:55 rue_mohr a microcontroller MAY NOT do this pwm
19:55 M0TRN_th1 ok
19:55 rue_mohr it has to be done by a currect detector citcuit
19:55 rue_mohr thats pc code
19:55 M0TRN_th1 I know
19:55 M0TRN_th1 :P
19:56 M0TRN_th1 so you're making 3 sine waves
19:56 rue_mohr it builts (overbuilds) the table of on/off sequences for hte motor
19:56 rue_mohr yup, 3 phase
19:56 M0TRN_th1 but if the first is < -0.5 it becomes -1, and if the other two are above 0.5 they become 1.0
19:56 M0TRN_th1 why?
19:56 rue_mohr trig fn's top out at 1
19:57 rue_mohr I'm turning them into "equiv" digital pulses
19:57 rue_mohr trinary anyhow
19:57 M0TRN_th1 ok
19:57 M0TRN_th1 then you can just use constants
19:57 rue_mohr when the phase is 1 the driver outputs high, when 0 the driver is off, when -1, the driver goes low
19:57 M0TRN_th1 like if < 0.5*pi then 0 otherwise 1
19:58 M0TRN_th1 basically you are turning the sine waves into square waves out of phase by 120 degrees
19:58 rue_mohr it was just to make the officiel table and to help jennie understand it more
19:58 rue_mohr yup
19:58 M0TRN_th1 ok
19:58 rue_mohr its written like that to be clearer about what its doing :)
19:58 M0TRN_th1 but I don't think jennie understands the code
19:58 M0TRN_th1 in any case it makes sense
19:59 M0TRN_th1 but if you used the sine values as pwm outputs it would be better for the motor?
19:59 rue_mohr he was making it, but got stuck converting an integer degree angle to a radian angle
19:59 M0TRN_th1 I don't know much about motors only about code
19:59 rue_mohr the motors coils require current limiting
19:59 rue_mohr because of the motors magnetics the current rises 'slowly'
20:00 rue_mohr the circuit needs to cut off the current when it gets too high and wait for it to come down again
20:00 M0TRN_th1 so it needs some kind of feedback into the PIC?
20:00 rue_mohr no
20:00 rue_mohr the pic may not be involved
20:00 rue_mohr thats what the comparitor is for
20:00 rue_mohr pic is just there to sequence
20:01 M0TRN_th1 ok
20:01 M0TRN_th1 but still it will need an input?
20:01 M0TRN_th1 external trigger
20:01 rue_mohr jennie, found out something I'd not known of tho, using a comparitor on the phases to detect when the motor is ready to go to the next step
20:01 rue_mohr I suggest now it just be open loop and use fixed step timing
20:02 M0TRN_th1 this was what we were trying to do
20:02 M0TRN_th1 but the motor was running slowly and without much power
20:02 rue_mohr yup
20:02 rue_mohr because there wasn't enough current in its coils
20:02 M0TRN_th1 with 2 of that DC motor driver IC
20:02 rue_mohr because of the driver
20:02 rue_mohr we were redoing with more powerfull drivers
20:03 M0TRN_th1 ok so using FET's
20:03 rue_mohr some fets
20:03 rue_mohr the pnp transistors still stay in
20:03 M0TRN_th1 this is what I suggested a week ago but the problem was the voltage for the FET's was too high for that motor?
20:04 rue_mohr it turns out the normal driver uses 12V pulses that are 1/3 of 50Khz (cant remember how many ms)
20:04 M0TRN_th1 ok
20:04 rue_mohr (for the pwm)
20:04 M0TRN_th1 so just shorter pulses with higher voltage
20:04 rue_mohr pwm current limiting
20:05 M0TRN_th1 well
20:05 rue_mohr it could probably be done as fixed duty, but I'm not gonna go there
20:05 M0TRN_th1 it's a lot simpler in the PIC code
20:06 rue_mohr that program is just to build the step table
20:06 M0TRN_th1 yeah
20:06 M0TRN_th1 but we already have the step table
20:06 rue_mohr the io needs to be redone for the new driver circuit
20:06 M0TRN_th1 ok
20:06 rue_mohr so the step table needs to be redone
20:06 rue_mohr the new driver uses on/off signals for the high and low side
20:06 M0TRN_th1 but the step table ought to be essentially the same
20:06 M0TRN_th1 ah
20:06 rue_mohr if you turn them both on, you short the power supply (with current limiting)
20:06 M0TRN_th1 ok so it's just an extra output for each phase
20:07 rue_mohr still 6 outputs
20:07 M0TRN_th1 and all derived from the existing table
20:07 M0TRN_th1 just if enable is on then either one of the other two will be on
20:07 M0TRN_th1 and if enable is off then none of the others will be on
20:07 M0TRN_th1 that translates to 3 outputs per phase
20:07 rue_mohr but its not high/low and enable/disable its high on/off low on/off
20:07 M0TRN_th1 or so no enable
20:07 rue_mohr yea
20:07 M0TRN_th1 that makes sense
20:08 M0TRN_th1 so the table transform is simple
20:08 rue_mohr so that program builds the information to make the new table on the controller
20:08 M0TRN_th1 if the first is enable and the second is disable then it becomes
20:08 M0TRN_th1 sorry
20:08 rue_mohr I'm gonna change, go to the shop and make up the new circuit
20:08 M0TRN_th1 first enable and second high
20:08 M0TRN_th1 then it's
20:09 rue_mohr you have H L inputs, H on L off makes the motor phase high
20:09 M0TRN_th1 yeah
20:09 M0TRN_th1 ok
20:09 rue_mohr H off L on makes the phase low
20:09 rue_mohr H off L off disconnects it
20:09 M0TRN_th1 and both off is all cases where enable was low
20:09 rue_mohr H on L on shorts the power supply
20:10 rue_mohr so 1, -1, 0, and self-destruct respectivly
20:10 M0TRN_th1 so basically any time where enable was 0 it should be 0 0, and any time where enable was on it should be 1 0 if output was high or 0 1 if it was low
20:11 rue_mohr kinda, some of the inputs are inverted, I just have to go work that out now
20:11 any68455593 rue, scope them, that's really the only way to be sure why they are hot
20:11 M0TRN_th1 she doesn't have a scope
20:11 rue_mohr you two are pretty sharp, where are you from?
20:11 M0TRN_th1 ...
20:12 M0TRN_th1 me? I'm from Denmark but I live in England
20:12 rue_mohr ah, that expalins a few things
20:12 jennie me ?
20:12 M0TRN_th1 what about you?
20:12 rue_mohr canada
20:12 rue_mohr it expains the time diff
20:12 M0TRN_th1 yeah
20:12 jennie it is 6:29AM here
20:12 M0TRN_th1 it's almost 1am here
20:13 any68455593 what explains {printf("p3n1s\t");} ?
20:13 M0TRN_th1 lol
20:13 jennie lol
20:13 rue_mohr 6pm here
20:13 M0TRN_th1 any68455593, just the topic line :)
20:13 M0TRN_th1 joined the chan and thought I'd pay in advance XD
20:13 any68455593 oh, then {printf("cunt\t");}
20:14 M0TRN_th1 well there's rude and then there's rude
20:14 any68455593 equally accepable, right?
20:14 M0TRN_th1 don't ask me, ask rue it's his chan
20:14 M0TRN_th1 and yours doesn't compile mine does
20:15 y68455593 writes an interpreter that doesn't need to compil
20:15 M0TRN_th1 try tcc
20:15 M0TRN_th1 ish
20:16 M0TRN_th1 so anyway
20:17 M0TRN_th1 rue_mohr, I think we got the PIC code nailed down, but I don't know about that circuit
20:17 M0TRN_th1 where do we put the FET's?
20:18 y68455593 copy/pastes C code on where to put
20:19 M0TRN_th1 any68455593, what an annoying nick
20:19 M0TRN_th1 is it some "phone code"?
20:20 Tom_itx yours is rather annoying as well
20:20 M0TRN_th1 so is yours
20:20 Tom_itx good we got all that established then
20:20 M0TRN_th1 aye
20:21 Tom_itx why you botherin with pic anyway?
20:22 M0TRN_th1 well because jennie had found this circuit that drove one of those motors with a PIC and that DC motor driver
20:22 M0TRN_th1 so I wrote the code to help
20:22 M0TRN_th1 but the driver doesn't seem to be strong enough
20:22 Tom_itx 754410 type?
20:22 jennie Old driver http://www5.picturepush.com/photo/a/12511203/img/Anonymous/driver2.jpg
20:22 Tom_itx 1A driver
20:22 M0TRN_th1 yeah
20:22 jennie new rue_ designed driver http://www5.picturepush.com/photo/a/12511168/img/Anonymous/complete1.jpg
20:23 M0TRN_th1 yes L293D was the first one, Tom_itx
20:23 M0TRN_th1 I don't know anything about motors and their drivers
20:24 M0TRN_th1 I just know how to code and basic physics
20:24 Tom_itx c?
20:24 M0TRN_th1 yes
20:24 Tom_itx ccs c?
20:24 M0TRN_th1 no
20:24 Tom_itx i used that on pic
20:25 Tom_itx and picbasic pro for a while
20:25 M0TRN_th1 I just know C well enough that it didn't take me long to figure out how to code in that mikroC PIC dialect that jennie had found
20:26 M0TRN_th1 I've not done much PIC programming before
20:26 M0TRN_th1 just some really basic stuff in assembler
20:26 any68455593 i hope rue is oscoping that TIP41 to see why it's hot
20:26 M0TRN_th1 I wish jennie had a scope
20:26 M0TRN_th1 it would make debugging her circuits remotely a lot easier
20:27 Tom_itx generally it's the crossover that makes em hot
20:27 any68455593 could be
20:28 any68455593 generally speaking it's a lot faster to deliver base drive than 10v of gate drive at really high freqs, but you can easily top out the driver or bipolar/fet base/gate performance
20:29 M0TRN_th1 ok
20:29 M0TRN_th1 so you want to drive FET's from the bipolars?
20:29 M0TRN_th1 and do you want to drive them in high frequency PWM from PIC or just on/off as a sequencer?
20:30 M0TRN_th1 we can do both, they run at 4MHz so it should be feasible to make softPWM
20:30 any68455593 i wasn't in favor of the fets at all, but it's equally feasable to drive bipolars with fets (IGBT)
20:30 M0TRN_th1 well if we can avoid the FET's so much the better
20:30 M0TRN_th1 just drive them directly from the PIC?
20:30 any68455593 which is why i hope rue can scope the TIP41's to see why they are hot
20:31 M0TRN_th1 ok
20:31 any68455593 so damn much deja vu going on
20:31 M0TRN_th1 jennie, you need to get a scope already
20:31 jennie ok
20:31 any68455593 someone sold a loaded working 7904 on ebay today for $87,, i nearly barfed
20:32 M0TRN_th1 heh
20:32 M0TRN_th1 I have a 2236
20:32 M0TRN_th1 does me well
20:32 M0TRN_th1 bought it for �75
20:33 Tom_itx mine's an old phillips analog
20:33 y68455593 has 2235, 7603s, 7603Rs, 7903s, and 7
20:33 Tom_itx i like my saleae analizer though
20:33 M0TRN_th1 what can I say, if it turns on, and it shows waveforms, it's all good :)
20:34 M0TRN_th1 mine came with new probes and everything and works like a charm
20:34 M0TRN_th1 bought it at a ham rally, even persuaded the guy to take a cheque as I'd spent my cash
20:34 rue_shop2 uh, i presume the idea was not to just buy a chip that did thsi all right?
20:35 M0TRN_th1 the 2236 has a nice multimeter and frequency counter as well
20:35 any68455593 that doesn't work right
20:35 M0TRN_th1 all of it works last I checked
20:35 any68455593 put on colorburst and try to get the freq to display, i dare you
20:35 M0TRN_th1 well
20:36 M0TRN_th1 since most of what I do is related to HF signals it works quite well
20:36 any68455593 for 30 minutes i could not get the Tek sales person and a teh AT A TEK SHOWROOM to make it work, so i didn't buy it
20:36 M0TRN_th1 heh
20:36 Tom_itx rue_shop2, chips simplify life
20:36 any68455593 colorburst is only 3.58Mhz, that's HF
20:36 M0TRN_th1 yeah
20:36 Tom_itx but we know you're not about that
20:36 rue_shop2 M0TRN_th1, jennie I presume the idea was not to just buy a chip that operates the motor for you?
20:37 M0TRN_th1 well I've used it to check a DDS VFO made with a Si570 and it worked fine up to 50MHz or so
20:37 M0TRN_th1 rue_shop2, yes that was the idea
20:37 jennie chips are not available here
20:37 M0TRN_th1 we want to be able to tweak it and play with it
20:37 M0TRN_th1 and what jennie said, she's in India
20:37 Tom_itx dick smith has chips
20:37 any68455593 rue, i think in the first time slot of this adventure, no one had a clue anywhere about what it took to drive the motor, cause no one had scoped a working motor
20:37 Tom_itx i thought she was an aussie
20:38 any68455593 i thought she was down the street from rue
20:38 M0TRN_th1 lol
20:38 jennie lol
20:38 rue_shop2 I thought I was a pickle
20:38 M0TRN_th1 she thought you were aussies
20:38 any68455593 i'm useless
20:38 jennie x2^
20:38 rue_shop2 I'm still not sure
20:38 Tom_itx i'm never sure
20:38 rue_shop2 GDS
20:38 rue_shop2 irfz34
20:39 rue_shop2 ok
20:39 any68455593 i said IRFZ44 days ago, it's cheap on ebay
20:39 M0TRN_th1 anyway rue_shop2
20:39 M0TRN_th1 if you know a chip that will do it I can get one and scope it out
20:39 M0TRN_th1 in fact I could probably just smash open an old HDD and scope it out from that
20:40 M0TRN_th1 but that would be cheating :)
20:40 any68455593 rue did that yesterday
20:40 M0TRN_th1 ok
20:40 M0TRN_th1 I only joined today after incessant requests from jennie
20:40 any68455593 that's why today he is just saying : go buuy a driver chip
20:40 Tom_itx IR2136
20:40 M0TRN_th1 btw my normal nick is just th1
20:40 any68455593 my normal nick is really useless
20:40 Tom_itx or dumbass
20:40 any68455593 yeas
20:41 Tom_itx or kat
20:41 M0TRN_th1 as long as it's not sillykitty
20:41 any68455593 or fcking_stupid or NotWorthCr4p
20:56 any68455593 ...
20:57 rue_shop2 so the fet HAS to be driven by a totem pole, a pullup is way too slow on the pullup time
20:57 any68455593 charging gate capacitance sucks
20:57 any68455593 current
20:58 any68455593 tis why i don't generally use them
20:58 any68455593 and that current hasto be delivered a a voltage way above what a bipolar needs
20:58 e_shop2 takes a large portrait of a person smiling and places it infront of
20:58 rue_shop2 kat, youwant to design a circuit for us?
20:58 y68455593 studies the "sm
20:58 rue_shop2 then I can go take apart printers
20:59 any68455593 not until you oscope the TIP41 getting hot, cause then you should have the problem licked
20:59 M0TRN_th1 rue_shop2, the outputs of PIC are totem pole
21:00 rue_shop2 but their not 12V totem pole, thats why we have to use 3 transistors to drive the fet
21:00 M0TRN_th1 ok
21:03 y68455593 wonders if rue considered a
21:13 rue_shop2 pls design something kat, I have a lot of thsings I want to do today
21:14 jennie yes, please design kat
21:19 M0TRN_th1 nite guys, it's 2am here, thanks for the help
21:23 rue_shop2 bah, there is a oscillation in the gate thats causing even a .04 ohm fet to overheat
21:23 rue_shop2 jennie, why dont you buy a 3 phase motor driver pack with built in fet drivers
21:23 rue_shop2 you can still do the sequencing
21:24 rue_shop2 I didn't expect these issues, this is taking more time than I have
21:30 jennie where to buy ?
21:45 jennie which ones ?