Need some help with an Hbridge chip.
rwgast_logicdesign
Posts: 1,464
Im trying to wire up two l6205 H Bridge chips. These chips have an over current protection scheme built in to them where they will rest them self after a certain amount of time if the chip senses to much current, this should never be an issue for me, as my motors draw 1.6 amps max, and the chips supports 3 to 5 amps. But the way the reset circuit works is causing a bit of confusion.
First of here is the data sheet
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00002345.pdf
Pages 12 shows the reference schematic I will be following to drive two motors, page 9 talks about ENa, ENb, cap and resistor selection. These two pages are the highlight of this post.
The first issue im having is the data sheet states
So basically the reference schematic shows that using a 100k resistor with a 5.5nF cap will give us a 2.2ms delay on over current, this is a basic RC Time Constant. Well I dont have any 5.6nF caps around, so I figure I will just use a 44nF film cap, as film is better for timers, and a 5k resistor. Does anyone see a problem with this, im wondering why STM would choose such a strange value like 5.6nF.
The next point of confusion lies in the wiring. So the two way I know of to commonly wire an H bridge are to use two PWM channels per motor one connected to inA and one connected to inB. So inA's channel would set the speed of the motor going forward, while inB would set the speed going backwards. I feel the better way to do this is use a single PWM channel per motor that is connected to the EN pin, then flip the INa, and INb pins high and low to get the correct direction. Now sense the current protection timer is set up on the enable pins, is sending the PWM to the EN pins of the l6205 even possible?? I would think that charging and discharging a cap would greatly interfere with PWM pulses by filtering them completely out.
First of here is the data sheet
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00002345.pdf
Pages 12 shows the reference schematic I will be following to drive two motors, page 9 talks about ENa, ENb, cap and resistor selection. These two pages are the highlight of this post.
The first issue im having is the data sheet states
CEN is also used for providing immunity to pin EN against fast transient noises. Therefore the value of CEN
should be chosen as big as possible according to the maximum tolerable Delay Time and the REN value should
be chosen according to the desired Disable Time.
The resistor REN should be chosen in the range from 2.2KW to 180KW. Recommended values for REN and CEN
are respectively 100KW and 5.6nF that allow obtaining 200ms Disable Time.
So basically the reference schematic shows that using a 100k resistor with a 5.5nF cap will give us a 2.2ms delay on over current, this is a basic RC Time Constant. Well I dont have any 5.6nF caps around, so I figure I will just use a 44nF film cap, as film is better for timers, and a 5k resistor. Does anyone see a problem with this, im wondering why STM would choose such a strange value like 5.6nF.
The next point of confusion lies in the wiring. So the two way I know of to commonly wire an H bridge are to use two PWM channels per motor one connected to inA and one connected to inB. So inA's channel would set the speed of the motor going forward, while inB would set the speed going backwards. I feel the better way to do this is use a single PWM channel per motor that is connected to the EN pin, then flip the INa, and INb pins high and low to get the correct direction. Now sense the current protection timer is set up on the enable pins, is sending the PWM to the EN pins of the l6205 even possible?? I would think that charging and discharging a cap would greatly interfere with PWM pulses by filtering them completely out.
Comments
So you cannot drive it push-pull directly - hence the two schemes (open-drain or RC filter) for
driving it without burning out the pin.
Look at the timing data to see if it can swtich fast enough for PWM for your purposes. Fault
current condition is rare and if you want to ignore the possibility just drive it via a 2k2 resistor
and omit the capacitor.