L298 Heatsink Question

[jasonm]

Hey,

I have a question regarding the L298 dual H bridge driver. I am using one in conjunction with an L297 stepper controller and a dsPic, to drive a 1.2A/phase bipolar stepper motor. The supply voltage to the motor is 15vDC, but this is obviously chopped to keep the current at 1.2A/phase.

My question concerns the L298, and more specifically the method of how to calculate its power dissipation. For the other higher current h bridges I am using, I know the resistance of the MOSFETs so I can work out power by P=I^2 * R. For the L298, how should I work out what the power dissipation is? I know the current but I don't know the voltage drop between the supply and the motor - or am I able to work this out from the motor specifications (coil resistance, inductance etc)?

Any help you are able to provide will be most appreciated

Jason

 

[Russlk]

You can't compute the power without knowing the voltage, so unless you have a datasheet or are able to measure the voltage drop, you are out of luck.

 

[jasonm]

Thanks for the reply.

I realise I need to know the voltage, but how would I work this out? Particularly as the L298 is being run by the L297 and hence at a 20kHz chopper frequency. Should I measure the average voltage across one of the coils of the motor? And then using (15-motorV) * motor current per coil * number of coils, this should give me the power dissipation?

Many thanks

Jason

 

[Nigel Goodwin]

I'd be inclined to simply run it and see how hot it gets! - as crudely as a damp finger if that's all that's available.

You're using a 15V supply, is the motor actually rated at 15V?.

 

[jasonm]

The motor datasheet says and I quote:

"Power supply voltage can be any value as long as the driver output current is controlled at the rated current."

The motor specs are:
Bipolar stepper motor
1.2A/phase
3.0 ohm coil resistance/phase
2.1 mH coil inductance/phase

Unfortunately I can't run it and see how hot it gets as I am at the PCB design stage (the stepper drive is part of a larger system) and I need to allow space on the PCB for the heatsink if necessary, or combine it with the other two H bridge (discrete) heatsinks. Is there any way of calculating the power dissipation from these specs? All I have to go on is in place of the actual unit I am testing using a late 1980's robot arm stepper controller (not using an L298 but it looks very similar) and the heatsink is the size of a small housebrick.... it barely gets warm however so I assume I need something smaller!

Many thanks

Jason

P.s., no idea why ohm is a link to this page, there is no link in the text here!

 

[Nigel Goodwin]

The motor datasheet says and I quote:

"Power supply voltage can be any value as long as the driver output current is controlled at the rated current."

The reason I asked is because it's common to run steppers off a higher voltage with current limiting via wirewound resistors - this gives increased stepping speed and better torque.

If you google for "stepper motor tutorial" you will find plenty of information about it!.

As for trying it, you could breadboard the design and test it like that - which is probably a good idea anyway!.

 

[Russlk]

The data sheet (what's that?) lists the voltage drop as 3.2 volts at 1 amp. If you plan on 4 watts dissapation at 1.2 amps, you should be OK. I don't know anything about stepper motors, but assume only one coil is driven at a time.

 

[Nigel Goodwin]

I don't know anything about stepper motors, but assume only one coil is driven at a time.

Depends on the mode you drive them in, but for the easiest mode only one coil at a time is driven.

 

[philba]

are you planning to half step? it's good to reduce resonances and get higher speed but does increase the concurrent coil current. 3.2V drop is suprising to me - there are 2 bipolar transistors in each energized circuit so I would have guessed somthing on the order of 1.4V but I wouldn't bet against the datasheet Note that with half stepping, you will need to add 50% more current due to the overlap. Definitely use a beefy heatsink. I'd also add a header to power a fan in case your heat sink isn't enough.

the 297 is a current sensing chopper so no resistors are needed.

Man, I sure don't like the L298 - a real heat generator, needs a stepper controller and talk about a prototype unfriendly package. Any reason you didn't use the A3977? Uses DMOS output transistors for much less heat generation. other than a hard to solder plcc package, it's pretty nice.

 

[jasonm]

The data sheet (what's that?) lists the voltage drop as 3.2 volts at 1 amp. If you plan on 4 watts dissapation at 1.2 amps, you should be OK. I don't know anything about stepper motors, but assume only one coil is driven at a time.

I have spent a lot of time looking over the datasheet but I was unaware that was the meaning of that particular specification - now you point it out it is obvious! Thanks!

Philba - Yes, half stepping will be employed intermittently (when the error in the control loop gets below a certain threshold). Yeah, the L298 is a pain isnt it! To be honest that is the first I have heard of the A3977.. but it looks really good! Unfortunately I am pretty far along with the current design and as I have about a week to route the PCB from tomorrow it is a bit late to change.. but definitely on the list for v2!


Thanks for all your replies, much appreciated.

 

[philba]

have you looked at the open source designs on https://pminmo.com/ ?

there is at least one allegro 3977 based design. several 297/298 ones as well. schematics and artwork included.

Phil

 

[jasonm]

have you looked at the open source designs on https://pminmo.com/ ?

there is at least one allegro 3977 based design. several 297/298 ones as well. schematics and artwork included.

Phil

I haven't come across that site before, thanks for the link!