Thanks Duane, The LT1158 looks like a really good choice. You can see the data sheet for it at http://www.linear.com/pdf/lt1158.pdf . There's a Full H Bridge example circuit diagram given on Page 19 with examples of how to implement current limiting protection - very important - and there's also info in there about driving more than 1 MOSFET in parallel. Apparently the LT1158 can drive up to 4 MOSFETs from each gate driver. I also considered the IR2112 device from IR but their datasheet is far less clear; the thing basically does the same as the LT1158 although the LT1158 avoids the thorny issue of potential shoot-through (where both MOSFETs are on together effectively tying Vcc to ground - goodbye MOSFETs!) by some smart logic timing. What I also really like is the ability of the LT1158 to take input directly from a sensing MOSFET to provide overload current protection. So, my project starts to take form. I will use 4 LT1158s driving 4 x IRCZ44s sensing MOSFETs on the high side and 4 x IRFZ44 non-sense MOSFETs on the low side. This should be good for a 20-30A controller and I can double or triple up on the MOSFETs once I've proved the design. I'm now also clear that the best way to generate the drive PWM signals will be to have L+R PWM drive signals for speed and L+R FWD/REV logic signals for each motor. Now the PIC bit is another question. Can anyone point me to a clock timing diagram for a typical RC servo output? I know that each servo gets a pulse between 1 and 2 msecs but Chan 1 pulse comes before Chan 2 pulse etc. This leads me to think that I can use interrupts to call the output PWM changes and use regular code for the input detection. Once I've detected a pulse in on Chan 1, I know that it will only go low somewhere between 1 and 2 msec later and THEN I can start to worry about Chan 2.... or maybe I should consider just straight line timing code the whole damn thing? Lost yet? Let me recap. I'm trying to control a 2-motor robot such as you see on Battlebots. What I want is for a PIC to take 2 servo channel inputs that correspond to F-R and L-R and generate PWM + FWD/REV logic which will feed the LT1158 via protection logic control (see bottom of data sheet page 19) which will feed the two motors vi MOSFETs. Hmmmmm, I wonder what should happen when say I'm going full speed ahead and I let go the stick? Should the PWM just stop? or how about it brakes the motor(s), or what if one selected reverse while going full ahead? Maybe a brake pedal would be cool on a 3rd channel? Now - anyone know where I can actually buy LT1158s? If they're an arm and a leg, maybe I may have to go back to the IR2112 and brew my own shoot-through avoidance circuitry - not that big a deal really. Thanks to y'all for advice so far. Ted Mawson PICman@portfoliopm.com ----- Original Message ----- From: "Duane B" To: Sent: Friday, February 08, 2002 8:52 PM Subject: Re: [PIC]: Robot Motor Controller > I don't know if this will help you'll out or not, but you might want to take > a look at eh Linear Tech part LT1158 - a Half Bridge N-Channel Power MOSFET > Driver. I'm using it in my design, well 4 of the actually, in conjunction > with IRCZ44 - SenseFET's. The are rated at 60V, 50A, Rds(on)= 0.028. > > I controlling the LT1158 with a National Semiconductor LM629 Precision Motor > Controller, which, in turn, will be ran be a PIC. > > Duane > > > ----- Original Message ----- > From: "Ted Mawson" > To: > Sent: Friday, February 08, 2002 10:40 AM > Subject: Re: [PIC]: Robot Motor Controller > > > > David Harris wrote... > > > > "What kind of current do you want?" > > > > Answer... > > I'm thinking in the 100-150 Amp range. Basic outline is for a PIC that > > senses 2 standard R/C servo channel pulses on 2 inputs and converts this > to > > signals for 2 H Bridge drives, 1 per motor. > > > > Questions that I have at the moment are... > > 1. Whether to poll the inputs or trigger the pulse width count using an > > interrupt? > > 2. Whether to drive 4 PWM outputs or just 2 with a 3rd direction control > > (per motor)? The latter has the advantage that it is less likely to get > > into a drive for all 4 FETs in the bridge. > > 3. How to handle the current in the output stages? MOSFETs like the > BUZ11 > > can take 30A, 50 Volts, and have an on resistance of 0.03 Ohms and there > are > > many more out there that can handle 80V and 50A. I'm thinking of > > paralleling up 2 or 3 MOSFETs per channel with the resistance of a > carefully > > designed PCB providing some kind of current sharing resistor capability. > It > > seems like a no brainer to use opto-isolators to separate the PIC output > > from the MOSFET gate drivers. > > > > Enough detail? > > > > Ted PICman@Portfoliopm.com > > www.portfoliopm.com > > > > ----- Original Message ----- > > From: "David P. Harris" > > To: > > Sent: Friday, February 08, 2002 11:27 AM > > Subject: Re: [PIC]: Robot Motor Controller > > > > > > > What kind of current do you want? > > > David > > > > > > Ted Mawson wrote: > > > > > > > Hi Bob, > > > > > > > > Your description is exactly what I was thinking of. I'd be very > > interested > > > > in seeing what you've got so far. > > > > > > > > Have you got the H bridge FET drive stuff fully worked out? I have > some > > > > specific ideas in this area, seems to me that it should be possible to > > make > > > > a very high current H bridge by using MOSFETs that can have on > > resistances > > > > of 0.03 Ohms! > > > > > > > > Ted > > > > ----- Original Message ----- > > > > From: "Bob Blick" > > > > To: > > > > Sent: Thursday, February 07, 2002 10:16 PM > > > > Subject: Re: [PIC]: Robot Motor Controller > > > > > > > > > I've done something like that for a fighting robot. Takes two R/C > > > > > receiver channels and generates pwm for two motors. Treats one > > > > > stick as fwd/rev and the other stick as left/right. One motor goes > on > > > > > each side of the bot. Also monitors the current on the h-bridges and > > > > > limits current. Has extensive amounts of code designed to reject > > > > > spurious signals coming from the receiver. > > > > > > > > > > I was planning to do a web page about it. If you can wait, then it > > > > > would come with some explanation, but I can send it to you if you > > > > > want to figure it out yourself. > > > > > > > > > > Uses a 16F876, coded in HiTech C. > > > > > > > > > > Cheers, > > > > > > > > > > Bob > > > > > > > > > > On 7 Feb 2002 at 18:10, Ted Mawson wrote: > > > > > > > > > > > To save me re-inventing the wheel, can anyone point me to a PIC > > project > > > > that takes 2 Radio Control servo pulse inputs (from an R/C receiver) > and > > > > generates the PWM drive for 2 DC motors running in H bridge mode via > > FETs? > > > > > > > > > > > > Thanks, > > > > > > > > > > > > Ted > > > > > > > > > > > > -- > > > > > > http://www.piclist.com hint: The list server can filter out > > subtopics > > > > > > (like ads or off topics) for you. See > http://www.piclist.com/#topics > > > > > > > > > > > > > > > > > > > > > > -- > > > > > http://www.piclist.com hint: The list server can filter out > subtopics > > > > > (like ads or off topics) for you. See http://www.piclist.com/#topics > > > > > > > > > > > > > > > > > > -- > > > > http://www.piclist.com hint: The PICList is archived three different > > > > ways. See http://www.piclist.com/#archives for details. > > > > > > -- > > > http://www.piclist.com hint: The PICList is archived three different > > > ways. See http://www.piclist.com/#archives for details. > > > > > > > > > > -- > > http://www.piclist.com hint: The PICList is archived three different > > ways. See http://www.piclist.com/#archives for details. > > > > > > -- > http://www.piclist.com hint: The PICList is archived three different > ways. See http://www.piclist.com/#archives for details. > > -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listserv@mitvma.mit.edu with SET PICList DIGEST in the body