#robotics Logs

May 26 2013

00:55 Guest8255 How's your cat?
01:09 Jak_o_Shadows H bridge is finished I think. Cleaned up some shorts, put the caps in. Need to test now
01:15 RifRaf beam me up Jak_o_Shadows, engage the h-bridge
01:16 Jak_o_Shadows haha, i'm just testing that I have no shorts atm
01:16 Jak_o_Shadows then I tentatively plug it in, and hope it works
01:19 Jak_o_Shadows On the other hand, can you recommend me a pre-built h-bridge that can handle 2.36A slight load, 15A stall, with 6A on startup? 16Vish
01:22 RifRaf sure have a few, will try find one
01:22 Jak_o_Shadows thanks
01:23 RifRaf IMS-1
01:23 RifRaf will try find a link
01:25 RifRaf http://www.aliexpress.com/item/IMS-1-The-smart-car-Coreless-H-bridge-motor-driver-module-Arduino30A-single-Reversible-brakes-PID/691967544.html
01:26 Jak_o_Shadows hmm, is there one that an australian reseller would have? so shipping time wouldnt be so long?
01:26 RifRaf they are faster than many aussie sellers
01:32 Jak_o_Shadows I think i should've gotten one a ew weeks ago'
01:33 RifRaf indeed, i told you about that one 2 months ago at least :)
01:33 GuShH Ard uin0
01:33 GuShH LOL
01:34 Jak_o_Shadows lol, why would I listen
01:34 RifRaf GuShH, thats just a word in the title, is not a plug into arduino thing, i used them how i wanted to
01:34 GuShH RifRaf: No, but look how it was written, ard uin0...
01:34 GuShH http://www.ebay.com/itm/IMS-1-30A-Single-channel-H-Bridge-Motor-Driver-Module-Arduino-Intelligent-Car-/300910084506?pt=LH_DefaultDomain_0&hash=item460fa3819a
01:35 GuShH they've got a 50A version too
01:35 GuShH http://www.ebay.com/itm/IMS-1-50A-Single-channel-H-Bridge-Motor-Driver-Module-Arduino-Intelligent-Car-/300910083403?pt=LH_DefaultDomain_0&hash=item460fa37d4b
01:35 RifRaf i have the 50A ones, cost me $20 i think each
01:35 GuShH seems well built
01:35 GuShH isn't it a copy of another module I've seen out there...
01:35 RifRaf must just be a different fet, the pcb is identicle
01:35 GuShH looks very similar
01:36 GuShH I've had good luck with chinese modules
01:36 GuShH good stuff for the mone
01:36 GuShH money
01:36 RifRaf made by this chinese dude that makes a few drivers it seems
01:37 GuShH when did chinese stuff actually become more or less usable or good?
01:37 GuShH what year
01:37 RifRaf must have a factory next door or something
01:37 GuShH I remember when their things were as bad as japanese electronics
01:37 GuShH very early on
01:37 GuShH well in China you've got entire cities devoted to electronics
01:37 GuShH I'd love to go on a tour, bring in a few tons of components with me.
01:37 RifRaf yes seems the last two years has been as good as anything
01:37 ShH g
01:37 GuShH they sell the stuff by the pound / kilogram
01:37 GuShH in paper bags
01:38 GuShH at least in many shops they do that.
01:38 GuShH similar to how HAM fests go
01:38 GuShH :( there ain't no real ham community here
01:38 GuShH just a few cats and no real meetings
01:39 RifRaf i get parcels from there every day or 2, alot of stuff is packaged in paper and tape, lots of tape
01:40 Jak_o_Shadows How should I test this h-bridge? see if anythi=g gets hot with no load attached?
01:40 RifRaf a minute
01:40 RifRaf with a load will test it properly
01:41 Jak_o_Shadows without a load would help prevent damage if I have any silly shorts
01:41 RifRaf if its gonna blow will do so in seconds
01:41 Jak_o_Shadows then screw it, and pray?
01:41 Jak_o_Shadows also, should I do the pwm straight away or not?
01:42 RifRaf you can use simple on and off logic to test
01:42 RifRaf then use pwm but if you know your pwm is good then just use it, starting with 0% duty cycle
01:48 GuShH no schematic supplied?
01:48 GuShH Jak_o_Shadows: got a scope?
01:50 Jak_o_Shadows nope :{
01:50 Jak_o_Shadows im fairly confidant of the schematic
01:55 Jak_o_Shadows BUGGER
01:55 Jak_o_Shadows HIP 4082 chip got hot.
01:56 Jak_o_Shadows as did the fets
01:58 Jak_o_Shadows well, the left hand side fets did at least
02:04 GuShH with a dummy load?
02:04 Jak_o_Shadows real load. Wheel didn't budge. I don't really have anything suitable for a dummy load
02:12 Jak_o_Shadows so what are the chances that the fets and hip4082 are gone?
02:13 RifRaf probly not too much unless you saw the magic smoke
02:13 Jak_o_Shadows that's good
02:16 RifRaf Jak_o_Shadows, are you using a fuse?
02:16 Jak_o_Shadows nope. Unless I can find a car fuse that would work (would it work?) I'm not
02:16 RifRaf well you should be, will save you alot if things are getting hot for no reason
02:17 Jak_o_Shadows might've found a short.
02:17 RifRaf oh really :)
02:20 Jak_o_Shadows between gate and drain of one of the lower fets.
02:20 Jak_o_Shadows Would that cause the hip 4082 to heat up?
02:27 GuShH shorts on an unprotected circuit are an easy way to destroy fets :p
02:27 GuShH Jak_o_Shadows: you can use a series resistor to limit current also
02:27 GuShH fuses are a preventive measure only
02:27 GuShH if it's slow-blow you can still damage your fets
02:27 GuShH it only takes an instant
02:28 GuShH I should know, I blew many fets in my time!
02:33 Jak_o_Shadows well, that's fixed now.
02:33 Jak_o_Shadows Couldn't spot any shorts in the hip 4082, so i'm just hoping that's fine
02:38 Jak_o_Shadows well, that didn't fix it
02:39 GuShH hope? hope isn't good enough!
02:41 Jak_o_Shadows yeah
02:42 GuShH hope is for religious people only
02:42 ShH runs
03:55 Jak_o_Shadows well, the didoes arn't in the wrong spot either
04:26 Jak_o_Shadows I guess i'll just go and walk and buy one tommorow
05:03 GuShH Jak_o_Shadows: buy what tomorrow?
05:57 Jak_o_Shadows Motor driver.
05:58 Jak_o_Shadows I realise buying it from a store isn't as cost effective, but I want it soon.
05:58 Jak_o_Shadows Unless somebody knows about one with good shipping?
06:10 Jak_o_Shadows I can't figure out what's wrong with what i'm doing, as I think the circuit's fine. Be easier to just buy one.
06:14 GuShH Jak_o_Shadows: double check your wiring
06:14 GuShH post schematic
06:15 GuShH you can just as easily have problems with a store-bought module...
06:16 Jak_o_Shadows http://images.elektroda.net/6_1183479474.jpg is pretty similar, excluding the overcurrent latch, and $2. Als R3/R4/R5/R6 are 20R, the fets are irf640, R2 is just plain missing, and i'm using the same diode for D2 and D1. Also VDD is 17ishV, and 3-4 (BLI/ALI) are tied to +5
06:19 Jak_o_Shadows What would I measure with a scope?
06:23 GuShH Jak_o_Shadows: you often probe input and output waveforms, start from where it's supposed to begin and move your way toward the outputs
06:24 GuShH that way you can tell at which point the failure occurs
06:24 GuShH so then you can hunt down the offending part/s
06:25 Jak_o_Shadows Start with checking the gate voltage for the fets?
06:25 GuShH a high voltage drop on the gate would suggest an issue on that particular fet
06:27 GuShH Jak_o_Shadows: if you are sure your drive signals are correct, it can only be an output problem so measure your four fets to see if they're shorted or open
06:27 Triffid_Hunter Jak_o_Shadows: umm you need R2, that's the dead time selector which prevents the high and low side fets from shorting out
06:27 Triffid_Hunter also make sure all your fets are around the right way
06:28 Triffid_Hunter Jak_o_Shadows: pro tip: put a car tail-light in series with the supply. if it lights up, kill the power
06:28 Triffid_Hunter that'll get your initial functionality tests out of the way with a couple of back to back LEDs (plus resistor) as a load
06:31 Jak_o_Shadows I think the fets are around the right way.
06:31 Jak_o_Shadows 47nF ceramic capacitors are right for the two caps?
06:32 GuShH sure
06:32 GuShH pin 8 dis -- DISable input. Logic level input that when taken high sets all four outputs low. DIS high overrides all other inputs.
06:32 GuShH When DIS is taken low the outputs are controlled by the other inputs. The pin can be driven by signal levels of 0V to
06:32 GuShH 15V (no greater than VDD).
06:32 GuShH there's a truth table on the datasheet also.
06:35 GuShH Jak_o_Shadows: LEDs and buzzers (if the frequency is low enough) are a way to troubleshoot if you don't have a scope, if you can make the clock low enough you can use a logic probe also).
06:36 GuShH Are you sure RL1 is closing as well?
06:36 Jak_o_Shadows I don't have that hole overcurrent sense bit. That's connected straight to ground
06:36 GuShH (all grounds in place, none missing?)
06:38 GuShH if your PWM is low frequency enough use a soundcard to probe it, protect it with a series resistor it's AC coupled already.
06:39 Jak_o_Shadows I may have access to a scope tommorow anyway.
06:39 GuShH of course anything higher than 10KHz and you won't even see anything resembling the actual waveform
06:40 GuShH a good logic probe can be a useful tool when scopes are not available :p
06:41 Jak_o_Shadows First few tests I did with a duty cycle of about 0.78, then I got sick of that and just put BLI to +5, and ALI to gnd
06:42 GuShH also in cases where you need a yes-no (but you don't care about waveforms) a good frequency counter can be useful.
06:42 GuShH of course you need to know it's input characteristics as to not blindly trust it
06:43 GuShH for instance a cheap DMM with a frequency mode may have a low input impedance and require a relatively high level to count...
14:25 Lola8088
22:01 Curious drops a ne
23:26 MrCurious http://wumo.com/2013/05/27
23:56 Rif ouch, who left a needle laying on the bloody floor :/
23:57 MrCurious i stepped on a tack last night