Controlling the OOPic from a GSM Phone

Hi Jonathan, I have links to several h-bridge devices and schematics here:

The main thing you want to watch out for when setting up any h-bridge is that your logic doesn't accidentally turn on both transistors on the same side of the bridge - since this will short the power supply to ground.

- dan michaels =======================

Dan,

Thanks for the links. Is there any way to make a more fool proof system that having both inputs low or high at the same time would not cause over current draw? The MCU will be programmed by students, and so it would be nice to make it as safe as possible.

Thanks,

Jonathan

Ha, well there's the $64 question. Foolproof for students? Sorry, that's an impossibility! [just let them fry the stuff, and learn - just kidding. ;-)]

When using an mcu tp drive the inputs, it's too easy to get things wrong - over and over. A fuse won't really help much with low-current transistors. OTOH, you can jigger the circuits by adding additional components for safety purposes, to prevent direct shorts. Maybe some of those links show this. In this case, sometimes people use a logic chip like a MUX/etc to drive the inputs. If you're gonna go to this much trouble, however, you might as well use an integrated h-brige chip, like L293D, SN754410, or driver like a TC4424. All relatively cheap.

- dan michaels =======================

If you have access to a scope, and depending on the type of driver circuit you are designing, you may be able to use an old trick. You can use two resistor-diode combinations to control the charge/discharge current going into a small capacitor. The waveform out of this cap would be used to drive the motor amplifier stage. Don't think digital, think analog. The charge and discharge sides are ramps. The on/off threshold of the receiving circuit is different for the "on" and "off" sides. The slope of the ramp causes one side to turn off an instant before the other side has a chance to turn on. You have to be careful to characterize this sort of a design for temperature/voltage/current/tolerance variations, but it works.

DRIVE PIN ----- CHARGE-RESISTOR ---(A) DIODE (K)--\ \-- DISCHARGE-RESISTOR --(K) DIODE (A)--|--- TO MDA | CAP TO GND

The same basic concept can be applied/modified to control timing of pulses in various types of circuits.

Another more "digital" approach is to use inverter gates and take advantage of the inherent gate delay (you may have to cascade a number of them to get the right delay).

All of these solutions need to be tested and characterized for the desired operating range and component tolerances (voltage, current, temperature, etc.). Spice is great for this.

Also, you may want to put one of those PTC resettable fuses in the line just in case. The transistors will probably do a good job of protecting the fuse, but it won't hurt to try.

Use a 2 channel Opto-Isolator with the input LED's wired in parallel and reversed.

-Beau Schwabe

That's a good one.

-Martin

See for a discrete circuit with current limiting. It shows the outputs using power darlingtons which have a parasitic diode enhanced for back EMF protection (at least, that's what was explained to me here :-). Use external Shottky diodes if you don't have such transisters.

Clifford Heath.

Use XOR (eXclusive OR) logic gates so that it is not possible to turn on more than one FET at a time.

Truth table for 2 input XOR gate:

AB Out

00 0 01 1 10 1 11 0

Scott

Great replies, all! Thanks. I think I will look into the opto isolator idea, I have some here. This is to run a latching valve that needs a brief small current draw. The integrated H bridge chip might be the way to go though, easy for students to use.

Thanks,

Jonathan