H-bridge to power drill motor and controlled by the Prop
eagletalontim
Posts: 1,399
And yet another project! The last was a bust, but this one I think I can fully accomplish with very little help 
This project is for a solar tracker to move my solar panels for optimum performance.
Since I have limited knowledge of the mathematical portion of electronics, I need some assistance with an H-Bridge to control a drill motor. This motor will turn on for a maximum of 1 to 2 seconds at a time in one direction, and the same for reverse direction, and will not turn back on for several minutes. If I could find a servo that was strong enough and was a full rotation, I would use it instead as it requires less amperage.
The design of the mechanical parts I am picturing as a frame that holds each panel that will eventually have an "X" and "Y" axis. For now, I plan to just go with one axis and add the other later. The panels will be connected to an all thread rod at the bottom of the frame. A motor will turn the all thread and adjust all the panels at the same time to face one direction. I have no idea how much torque it will take to move these panels and I am sure a hacked drill can handle it.
I have already built an H-Bridge with tip120's and tip125's but the start up current needed for the motor is just too much and any drag on the motor causes the transistors to heat up extremely fast. The drill motor just free spinning makes the transistors a little warm to the touch, but not bad. Using my meter in Amp mode, the free spinning amperage is around 2 amps and under a heavy load is > 10A @ 12V. Should I look for a different motor / gear box or use Mosfets or relays instead (which would make the circuit have a possibility of failing due to mechanical contacts)?
Edit : I just found this motor, but it does not list the load current
http://www.allelectronics.com/make-a-store/item/dcm-450/rt-angle-gear-head-motor-w/tachometer/1.html
This project is for a solar tracker to move my solar panels for optimum performance.Since I have limited knowledge of the mathematical portion of electronics, I need some assistance with an H-Bridge to control a drill motor. This motor will turn on for a maximum of 1 to 2 seconds at a time in one direction, and the same for reverse direction, and will not turn back on for several minutes. If I could find a servo that was strong enough and was a full rotation, I would use it instead as it requires less amperage.
The design of the mechanical parts I am picturing as a frame that holds each panel that will eventually have an "X" and "Y" axis. For now, I plan to just go with one axis and add the other later. The panels will be connected to an all thread rod at the bottom of the frame. A motor will turn the all thread and adjust all the panels at the same time to face one direction. I have no idea how much torque it will take to move these panels and I am sure a hacked drill can handle it.
I have already built an H-Bridge with tip120's and tip125's but the start up current needed for the motor is just too much and any drag on the motor causes the transistors to heat up extremely fast. The drill motor just free spinning makes the transistors a little warm to the touch, but not bad. Using my meter in Amp mode, the free spinning amperage is around 2 amps and under a heavy load is > 10A @ 12V. Should I look for a different motor / gear box or use Mosfets or relays instead (which would make the circuit have a possibility of failing due to mechanical contacts)?
Edit : I just found this motor, but it does not list the load current
http://www.allelectronics.com/make-a-store/item/dcm-450/rt-angle-gear-head-motor-w/tachometer/1.html
Comments
Here's a quick tip, don't use the TIPxxx darlingtons for heavy loads, though they are rated for the current and a lot of online circuits seem to use them they will always have a very high Vce(sat) of at least 1.2 to 1.8V drop at 2A and over 3V at 10A. Do your maths and you will know that even the running current means the transistor will have to dissipate over 3W running and more than 30W under load. Even 3W needs a good heatsink already. So those online circuits don't show the huge heatsink. Just use MOSFETs and very many different types have been mentioned in these forums suitable for direct drive. Did you have a particular H-bridge circuit you are using? I published a circuit years ago that was very easy to drive form the Propeller in that it followed the I/O pin and would disconnect the driver if the I/O pin floated. The MOSFETs I used were actually dual N+P in a small package, drive them right and you don't need anything big and fancy.
I have never used a half bridge but I have heard of them. The only time I used an H bridge was on an old project that switched on the water to mist my outside AC unit when it kicked on. The water valve was a small DC motor that just needed to be pulsed in one direction or the other one time to turn the water on or off. It was a small DC motor and the H Bridge was well able to handle it.
I do have some IRF510 mosfets laying around but they are only 3A. I do have some of these as well, but they are P-Channel : http://www.digikey.com/product-detail/en/DMP3010LK3-13/DMP3010LK3-13DICT-ND/3076574 I don't think I have all the right parts to build a complete unit, so it looks like I may have to order parts
As stated in the first post, I really suck when it comes to the math part of electronics. Most of the time it is a hit or miss if I use the right resistor values. I am learning though!
Maths? What maths? We are only talking simple functions such as R = V/I therefore V = R*I and I = V/R etc The one applicable to your circuit is W=V*I so knowing V & I you know it's too hot. There is no need to hit and miss with anything so basic. You have an LED which drops 1.8V at 10ma and with a 5V supply that means the resistor must drop 3.2V at 10ma, therefore R = 3.2V/10ma = 320R or 330R to round it.
The IRF3708s have been mentioned in this forum but you also need to find a suitable more or less matching P channel. Forget Vgs(th) in the datasheet, this can be misleading, always head straight for the graphs and look at Rds(on) at Vgs at an elevated temperature and actual load current. Don't forget stall current too, but you can reduce this by sensing the voltage or current and cutting down on PWM.
you need 12V for the drivers, the inputs are 3 or 5V logic compatible, and dead-time handling and shoot-through prevention
can all be handled with the drivers (they vary in their support). The motor supply can be any voltage independent of the
driver supply BTW, 0.1V to 600V (with the right drivers)
You really need shoot-through prevention in hardware for an easy life, and for 10A handling go for 10 milliohm or less
MOSFETs
Read the datasheet of some of these driver chips like the HIP4081A, IRS2004, FAN7380...
Here is the circuit I am using for controlling the valve at my hydro power station, and that`s a windshield wiper motor.
It uses 2N3055 and MJ2955 transistors. (complementary pair) Can handle up to 15A.
Controls the motor connected to J601, via signal from Prop-pins at J602. One pin for each direction.
The "flyback" diodes over the transistors, is 1N5408. Without them you`ll burn the transistors momentary. (I`ve tried...
J603 and J604 are for control LED`s to see if there is a control signal present.
Hope it could be a tip. Works fine for me.
God luck with your interesting project.
You Norwegians are mad! Looks like a circuit from the 70's when the 3055 and it's buddy were power kings.
With this circuit you could easily produce magic smoke by putting a high on both inputs! Also if there is motor stall there is nothing to prevent excess current through the small drive transistors as the emitters of the power stage would effectively be connected to ground/supply. A valve is somewhat different though and doesn't stall
3055's and 2955's had their day, its a no-brainer to use MOSFETs for an H-bridge.
For instance I've got a 3-phase MOSFET bridge I've designed that handles upto 5A and 100V without
any heatsinking, and its surface mount taking a couple of sq in of PCB. Add heatsinks and it'll do
10A at a few watts dissipation. Using 3 2n3055's and 3 2n2955's plus driver transistors would take
up lots of space and dissipate more heat in the driver transistors than my entire bridge!
Since the circuit I am using can fail if both inputs are active, I have added a 2 second delay in my code that turns both pins off before it is allowed to turn on another one. This is a peace of mind for me, but there is always a chance it can still fail. I will be adding a fuse to the H-Bridge to prevent damage in case of a short.
For a future build, I will more than likely be going with a 4 channel mosfet h-bridge driver like the HIP4081A. Thanks again guys for the great information. I feel like I am starting to understand the Mosfet world a little better!
Using predictive method will work better on cloudy days when optical trackers start becoming inaccurate. There is benefit to having the panels pointed directly at the sun even on cloudy days.
Sandy