[Micro_CNC]

I have built the L297/298 as per the reference design by ST and it is fully functional in all but one respect, the resistors that go from ground to pins 1 and 15 on the L298 and pins 13 and 14 on the L297 have blown up quite spectacularily a few times now (but have also worked fine on other times) and I can't seem to figure out why. The datasheet for both the 298 and the 297 doesn't seem to offer any explainations as to what the function of these resistors is and what kinds of loads they are expected to handle, nor does it explain why they get hot enough to pop sometimes but not others.

My setup on this is enable tied to high, reset tied high, half/full tied to high, Vref going to center on a 500k voltage divider, control tied to ground and I'm using the reference values for the RC network that runs the chopper. I'm running EMC2. Anything I haven't listed is set to whatever the reference design calls for.

The reference design I am refering to is the very first result when you search "L297"
on google.

Any explaination as to what these resistors do or how to determine what the peak amps they use would be appreciated. Electronics and computer jargon doesn't phase me either, so don't worry about it. I just can't seem to figure out this chopper thing.

Regards, Matt

[jeffs555]

Those are the current sense resistors, and carry the full motor current. The resistors are probably burning up because they do not have a high enough power rating. To calculate the power in a resistor, you square the current and then multiply by the resistance. For example, if you were using 2 amps and a 0.5 ohm resistor, the power dissipated in the resistor would be 2 watts(4x0.5). For a safety factor, you usually double this so you would want a resistor rated 4 watts or higher.

[Micro_CNC]

Thanks, I'm going to put in an order for new resistors sometime this weekend so i should hopefully be able to make a thread titled "its alive" by Wednesday so in my circumstances (5 amps peak, doubled up 298's) i would need about 15 watt resistors correct? thats not doubled but I can always shut it off and adjust it if the temps get too high and I'm not willing to pay $30 per resistor.

Also what would changing the resistance do?

Do you know of any circumstances that my 1/8W ones wouldn't fry under - they ran fine for at least 3 trials (and they weren't hot at all) after that it was intermittent and they often blew up. EDIT - I think I might have figured it out, I can recall not always measuring the resistor value just pluging it in - is it possible that they were low enough to take advantage of the A^2 * R with the A changing with the R so I was experiencing about 3 A average so with an 1/8W resistor that would be 0.375 ohms thus allowing them to run maybe slightly warm but not hot enough to pop, so they didn't pop because of manufacturing variances.

Could you see any problems with just connecting those lines straight to ground? I tried that and while things do indeed get hot they are still within the safe temps outlined in the datasheet (130 max - 75 as a "normal" temp). It pulls about 3-4.5A as compared to 2-3.5A.

What does changing Vref actually do? I know it varies how many amps are pulled but what is it actually adjusting? is it like PWM or something?

thanks,

Matt

[pminmo]

Try to keep the vref under a volt. At 5A that would mean a sense resistor of .2 ohms or less. Avoid wire wound resistors. You won't be running 100% duty cycle so the resitor wattage can be less.

[jeffs555]

Could you see any problems with just connecting those lines straight to ground? I tried that and while things do indeed get hot they are still within the safe temps outlined in the datasheet (130 max - 75 as a "normal" temp). It pulls about 3-4.5A as compared to 2-3.5A.

What does changing Vref actually do? I know it varies how many amps are pulled but what is it actually adjusting? is it like PWM or something?

Connecting them directly to ground would be a very bad thing to do.

The current limiting counts on the voltage drop across the resistors. It works like this:
The motor current goes through the L298 out pin 1&15 through the resistors to ground. From Ohms law, the voltage across a resistor is equal to the current times the resistance(ie the resistor converts the current into a voltage). For example, with a 0.5 ohm resistor, 1 amp would give 0.5 volts, 2 amps would give 1 volt, etc. Inside the L297 is a voltage comparator that compares the voltage across the resistor with the reference voltage. When the voltage across the resistor goes higher than the reference voltage, the L297 shuts off the L298 for a fixed amount of time. With a direct connection to ground instead of a resistor, the only resistance would be that of the wiring and connections, and it would take a very high current to generate a voltage equal to the reference voltage.

[Micro_CNC]

That is exactly what I was looking for, an explaination of how the device is actually working. Thank you very much. It is on a breadboard though and the motors are pretty high resistance so I'm not pulling too many amps but yeah it gets hot.

So if my motors are at 0.8 ohms per coil and I'm running 12V then would I make those resistors about 0.8 ohms each and turn the Vref till I get around 1.5A per coil / 3A per motor. Therefore staying under the 4 amps limit for each pair of 298's so I shouldn't have a problem with temp that way either. Does that sound right? The only reason I'm not saying 0.2 ohms like in pminmo's post is because I forgot to mention I'm running 12V rather than 36V and I do not know how that would fit in, and I am sort of thinking of the resistor and motor as a voltage divider now which the equal resistance of the resistor and motor would allow a greater range of adjustment via the Vref.

In short is 0.8 ohms and 5 watts about right?

If one of you could reply in the near future then it would be much appreciated as then I could order those resistors tonight.


Matt

[pminmo]

The vref on the L297 has a 3V max. If you use .8 ohms, and set for 1.5A, your vref would be set to 1.2V. If you use a .2 ohm resistor, at 1.5A your vref would be set to .3V. You said you doubled up the L298's for the current. At 1.5A there was no need to double the L298's. They will do 2A/phase. Your L298's will need to be heatsinked regardless, they get hot.

Steppers loose rpm performance/power at lower supply voltages. The L298 will drop around 3V itself, if you have a .8 ohm resistor at 1.5A you will drop 1.2V across the sense resistor. So 12V (supply) minus 3V (L298) minus 1.2V sense resistor will give you a little less than 8V to the motor.

[Micro_CNC]

A heatsink isn't an option for me until the machine is running, once it is then I will use it to make 8 heatsinks and a nicer enclosure for the L298's. They do have a 90mm tube fan blowing on them at the moment and its keeping them at around 30C at 3A per pair. The datasheet says that they have a max power dissipation of 25W which at 12V figures out to be about 2A per L298 without a heatsink so with that in mind I doubled them, I will keep it at that until I make a more suitable cooling solution. That voltage will have to suffice because I do not have a higher rated power supply and I don't have the parts to make one. For my current uses it has enough power as is, In the future I plan on buying gecko drives and running something in the range of 50V.

I do understand the reasoning behind keeping Vref under a volt now though. Thank you.

So, from the help I have received from both of you I understand that ~0.2ohm and 5W (to allow the L298 pair to hit 4 amps per phase if I decided to go that high) is about right.

My only remaining question seems to be why not go with wire-wound resistors? Is it something to do with current in the coil (wound thus coils I'm assuming)? If not wire-wound then what type should I use?

Matt.

[pminmo]

.2 ohm 2w is plenty. You don't want to use wire wound because they have some inductance and that can affect your sense circuit performance negatively.

[Micro_CNC]

Thank you PMinMO and Jeff, I ordered them from digikey this morning and hopefully they will be here for wednesday.

Regards
Matt