hi im trying out some uln2003an's that i got for free from my local hackerspace whom recieved it as a donation. they came in these tubes to be assembled with a machine. so they sat around my house a bit in their tubes for the machine, and ive finally gotten around to using them. i hooked it up in line with what im seeing for any schematic to drive a motor to them, but every last chip is burning up. im trying to drive a 6 wire nema 17. This is the situation:
-i got the center taps tied to 19v(laptop power supply cuz it does 5amps)
-i got the vcc going to the same 19v
-ground going to ground for 19v, while also tied into arduino ground
-logic pins plugged into an arduino
-coil ends going to outputs on uln2003an

when i run it with no delay, i get a musical hum from the electricity.
when i run it with a 10 ms delay, i get 1/4 second of motor movement before the pop sizzle smoke
when i hooked it up to NO uccontroller, and only simulated logic with a 5v regulator, i would zap the pin for a quick tap and the motor would move just that one step, then start sizzling regardless of the now logic low state, and would continue to squirt out smoke like a fog machine unless i pulled vcc or ground. kinda like it was locked on. theres no resistors in the circuit aside from internals.

it shorts out my power supply and i have to reset it.
when i had my arduino hooked to the laptop via usb cable, for both power and txrx, and it turned the motor that 1/4 second, it also killed my laptop which is itself, plugged in. so i believe it got zapped with 19v all the way thru the duino, thru the usb port, and messed some stuff up and shorted out. no immediate devastating dmg yet, but i just wanna turn the motor with my 5amp supply, because ill need to drive not one, but 10. and with prices, 10 uln2003s are much cheaper than 10 l293ds. any help would be a godsend. ive fried circuitry all the time but this is new since i cant find a single schematic with external resistors and i usually fry chips cuz im forgetting those.

 

You offered no details at all of the motor? the ULN 2003 is not really suitable for direct motor control except maybe the very smallest.
500ma Peak!
Max.

 

theyre nema 17- 76 oz-in motors. they should be pretty low amperage, at least in testing since i have 0 load on it*(i should stop saying them, im only trying to get just one working then ill scale up). but yea this is the link where i bought the motors, he has a partial of the datasheet in the description and below that is the link to the full datasheet from minebea motors

https://www.ebay.com/itm/112553267528
https://www.eminebea.com/en/product...__icsFiles/afieldfile/2017/09/29/17pm-k_1.pdf
my exact model number isnt on there, but from reverse calculating, a 76oz in motor spits out 535 mnms so its likely closer to one of the ones with a holding torque of 500.

im glad you mention that though, because i did notice that all the videos appeared to show floppy drive motors not any nema class motors. would anyone recommend a better way to drive 10 steppers in relative sync? i know theres mechanical differences but im just trying to make the normal 4 logic signals drive 10 motors instead of just 1. so they all stop at the same time and start at the same time, regardless of how far the shaft actually turned

 

Hello,

It is no wonder that the ULN burns out.
Even the smallest 500 nM motor from the shown datasheet takes 0.9 Amp as the ULN is rated only 0.5 Amp.

Bertus

 

It appears that these are stepper motors used in the bi-polar mode, (4 connections).
Measure one coil resistance and that and the voltage you are using will tell you the current, a stepper motor draws the same current at 0rpm as the operating mode.
Max.

 

am i mistaken in thinking that the uln is 500ma per output? as in when you divvy up 2 amps across 4 coils, that produces 500ma per coil? the reason i was under that impression is i found discussions on the uln vs the l293d, and people pointed out that the l293d can push out 600ma total, whereas the uln can handle 500ma per channel. this is what has me all bent up, because i ran 3 of these nema17s off a single l293d and they turned, it only stopped when i added a 4th. im not sure what im missing here, if an l293d can turn 3 of these off just 600ma, how can a uln not handle a single one with a whole 3500ma total capacity?

 

Hello,

The L293 is likely to small to.
An L298 has a total current of 4 Amp, wich also is likely to short.
I do not know more powerfull chips by head.

Bertus

 

theyre hybrids, they have 6 leads, 2 center taps. so when i hook them up to the l293d, i hook them up bipolar, but i was also under the impression that since ulns only sink current, they couldnt drive a bipolar stepper anyways. ive been tying my 2 center taps to 19v as shown in the original schematic and using it as unipolar to accomodate. or so i thought it should

 

Hello,

Perhaps you could have a look here:
http://www.ti.com/product/DRV8825
or here:
https://www.st.com/en/motor-drivers/stepper-motor-drivers.html?querycriteria=productId=SC2132

Bertus

 

The rated current of the motor is Per Coil.
Stepper motors are traditionally high current devices, a fairly small one would be 600ma rated. Typical, 1a - 3.5a.
At 19v that is Way too high.
Unipolar mode uses the 6 leads, bi-polar the two and have the higher current rating of the two.
Max

 

maybe im unclear.. the l293d WORKS GREAT. drives them bipolar with no issues. i could likely hook up 10 of them to a hefty enough power source and they should all work perfect. but for 3$ a chip that gets a bit much when theres a .33c option. but no, the l293d is not too small, like i said, i was able to run 3 of them off a 12v 2a power supply. thts 3 coil a1 wires put into a single a1 output, 3 a2 wires to a single a2 output, b1 and b2 also. my chips could just be bad too. like i said i got them for free from a local nonprofit. i was just seeing if anyone knew something about them i was overlooking

 

those drivers look ok, hadnt considered using pwm but ill give it a look

 

am i mistaken in thinking that the uln is 500ma per output?

Note on the data sheet that there's also a maximum rating of 2.5 amps total current, not just the 500 mA/output rating. See the spec for maximum total emitter terminal current below:



Depending on your motor(s), you're probably exceeding that rating, as well.

 

ok i see that now. so the most current i can sink with this chip is 2.5 amps? if thats correct then im puzzled and back to square 1.
in this datasheet for the l293d https://www.mouser.com/datasheet/2/389/l293d-954810.pdf it says that its peak current ability is 1.2 amps. thats 600 per coil, in alternating directions. if the l293d is less capable than the uln at 2.5 amps. why is the l293d driving them perfectly, and the ulns are frying?

 

in this datasheet for the l293d https://www.mouser.com/datasheet/2/389/l293d-954810.pdf it says that its peak current ability is 1.2 amps. thats 600 per coil, in alternating directions. if the l293d is less capable than the uln at 2.5 amps. why is the l293d driving them perfectly, and the ulns are frying?

Dunno. One way or another, though, you are exceeding one or more of the part's maximum ratings-- either the maximum allowed collector current on one or more outputs, or the total output current, or the package power dissipation. Or their could be a fault (e.g., a short) somewhere in your wiring.

 

I generally have done it with logic gate mosfets.
Max.

 

Am I getting this right? With the ULN driver the motors are used as unipolar with the centre of each winding connected to 19 V and with the L293 the motors are being driven with push-pull drive putting 12 volts across the entirety of each winding? That would mean the ULN driver would have to handle roughly three times the current that the L driver has to.

This must return to MaxHeadRoom's earlier instructions to measure winding resistance.

 

The ULN driver has a rating of 2.5 A maximum out of the ground pin. The L driver has a peak current rating of 1.2 A per driver. These spec's are for entirely different things and cannot be directly compared. The L driver also has a much higher total power rating. Total power for the ULN is not spec'd numerically but must be determined using allowable temperature and thermal resistance.

[EDIT - added:]
"... says that its peak current ability is 1.2 amps. thats 600 per coil, in alternating directions"

No, that is not what it means. The steady state current per channel is 600 mA as stated in the bullet points at the start of the (dismal) datasheet. The 1.2 A figure is per channel as a non-repetitive 100 µs pulse. The badly done datasheet fails to add any detail to "non-repetitive" - it certainly doesn't mean just once, and there is insufficient detail to make any reasonable calculations, making it a largely useless spec.

 

Am I getting this right? With the ULN driver the motors are used as unipolar with the centre of each winding connected to 19 V and with the L293 the motors are being driven with push-pull drive putting 12 volts across the entirety of each winding? That would mean the ULN driver would have to handle roughly three times the current that the L driver has to.

This must return to MaxHeadRoom's earlier instructions to measure winding resistance.

you are correct. i was wiring them differently to match the chips characteristics.

and indeed i did. im still having trouble understanding 100% EXACTLY whats going on, but i got the end result. i did as max said and measured resistance, and that seems to be the root cause. i was always under the impression that a current must be drawn, and that giving it full juice expecting it to only pull what it needs would work.

it turns out with my 12 ohm resistance across the full winding, and my 6 ohm at the center tap, with 19v i was pumping 3 and some change amps into it. so i run it backward with a target amperage of under 500 and landed with a 47ohm resistor at each output to land me somewhere in the mid 400s. hooked it up and worked like a charm.

sorry for bothering you guys, and thanks for helping. i wish there would be a resistor symbol on the datasheet where they demonstrate inductive loads, thats mainly what i was looking at, but pretty much every other drawing or video on it didnt show resistors either, but i had a hunch that was it. like i said at the first post, thats usually what fries my chips

 

Normally when you buy a stepper motor you either match the motor to the drive voltage and circuit you want to use or match the drive voltage and circuit to the motor you want to use. If you do that, then you don't have issues of excessive current.

Using a higher voltage than the motor was intended for and adding series resistors actually can improve the performance of a stepper because it allows faster rise of current. The winding inductance normally limits the rate of rise and that rate is proportional to the applied voltage divided by the inductance. The resistors do waste power, but the alternative is a more elaborate drive circuit.
This looks like it explains it fairly well, though I just skimmed it. MaxHeadRoom and others may have some other recommended reading.
gmpwebsite.com/wp.../LR-Stepper-Drives-Choosing-Current-Limiting-Resistors.pdf

"sorry for bothering you guys" Not to worry. It often takes some back and forth to get some critical detail from someone having problems.