Newbie- L297/298 Problems

[brentonw2004]

Hello. I recently built a bipolar driver using a L297/298 combo. I basically followed the datasheets given for the 297/298 and added some screw terminals and adjustments for control. I just finished assembling everything today, but I cannot get it to work. When powered on, the motor just vibrates and hums. I am using two separate computer power supplies to power the driver. One for the logic and one for the motor supply with the grounds tied together. I have the connection for SYNC left open, CLOCK going to a 250Hz 5V square wave, and CW/CCW grounded. I am certain that the motor is wired in the correct order. I have tried the driver with CONTROL enabled and disabled with very little change. I have not observed any smoke, excessively hot components, or anything else to suggest that the circuit is fried. I tested and made sure all the chips were getting power. I have attached a copy of my schematic that I made. Any advice would be greatly appreciated as to the validity of my schematic or suggestions for troubleshooting my issue. Thanks!

-Brent

[armstrong360]

Try a much slower clock, I suspect the motor can't accelerate to that speed instantly...

[pminmo]

Nothing jumped out as wrong with the schematic at a glance. Could be the physical implementation of the circuit, as how the high current paths are handled is important. Take a look at the pcb for my implementation of a DIY L297-8 here: http://www.pminmo.com/l297-8/l297-8.htm The fact that the L298 isn't warming up is a little troubling, as they run hot. Make sure you have adjusted vref to provide the current the motor needs. You can start without a motor by putting two high value resistors (4.7k for example) in place of the motor coils. At that value your circuit shouldn't chop, so in full step mode, you should be able to see the voltage across each resistor switch between +24V and -24V (approximately).

[brentonw2004]

Hello all. I got a chance to mess with my circuit some more today. I am still having very little luck figuring it out. I lowered the step clock to 2 Hz with no progress. The motor still simply twitches, just much slower. The motor will hold stationary at its holding torque when no step clock is applied. The motor that I am using is a pretty small 1A, 7.5ohm motor. One very peculiar thing I observed is that when I vary the amplitude of the square wave down to 1V the motor actually turns some, very very erratically but it did make a couple very jerky revolutions. However, if I speed the clock up really any at all it goes back to twitching. Correct me if I'm wrong, but I don't think the L297 is even supposed to respond to such a low clock signal. I built circuit very similar to this one, but I used the L293D. I used it with the same motors and it it worked perfectly. The only real difference in the circuit is that I added the capacitors mentioned in the L298 datasheet, removed the diodes, and didn't have to mess with chopping. I hooked the new circuits motor coils to 2k resistors as pminmo suggested and hooked up a scope to them. One side appeared to functioning somewhat normal, but the other side was extremely noisy/random, so I am think I may have a fried L298. Any suggestions for actually testing this chip other than just pulling it out of the board and replacing it? Also, what type of capacitors should c2, c3 in my schematic be? I used normal electrolytic can capacitors for both. As for the physical implementation of the circuit, any particular error in mind to avoid? I laid the components out very similar to the one in the datasheet, motor terminals close to the diodes, which are close to L298, minimal wire jumpers/top traces. I used 20mil traces for all of my board. Finally, I left the L298 on for a few minutes and it actually is getting hot. I guess I must just not have gave it long enough when I was testing it last. Any additional ideas? Thanks for your help so far!

-Brent

[vger]

Brent,
What is pin 1 (sync) connected to? With the internal osc disabled this functions as an input for syncing several L297's. You also might try tying pin 20 high with a 10K just in case noise is triggering low reset.
Steve

[pminmo]

C2 and C3 aluminum electolytics. The motor supply voltage side would benifit from a low esr type, but isn't absolutely necessary. You say 20 mil traces, that would be a problem on the high current ground's, motor supply and motor outputs. Another thing I just noticed is no filter on the current sense path's. The L297 is kind of known for needing a small rc filter on it's sense inputs. If you look at the app note pcb, you will see that the gnd area on the perimeter of the board on three sides is quite large, that aids in keeping a good 0V at all points where there is heavy current flow and the analog portions of the circuit need an establised 0V such as the L297 sensing the motor current from the sense resistors. If you look closely the low side of the sense resistors take an independent path to GND at one point at the L298 before coupling to gnd. Also note the use of wider traces on from the L298 to the sense resistors in the app note, those are wide for a reason.
With a high value resistor, you should see steady voltages on the phase outputs with a DC meter.