PWM Driver not working

Hi,
i just finished setting up the PWM driver and I hooked it up to the arduino and to the power.
the power LED comes on.

I uploaded the following firmware onto the arduino:
[www.reprap.org]

but after it finished uploading the board does not respond.
symptoms:
-the solarbotics motor I hooked up does not move.
-the LEDs do not light up (except the power LED})
-I can tell the Arduino is running code successfully since it does print to the terminal window.

I checked the diodes, the LED's the power and input connections and the screw terminals.

any idea how to troubleshoot this?
-Leav

Also, I added a heatsink to the transistors, but it's just one heatsink for all of them... it's not shorting stuff out is it?

well since the backplate of the transistor ist connected to the ground pin, and all transistors are equal and equaly grounded that shouldn't be a problem
(as long as you don't hit another trace with the cooler)

To make sure you can remove the cooler for a testrun, but I think the problem is elsewhere.
maybe you should try with other pins
int pwm_a = 9;
int pwm_b = 10;
int pwm_c = 11;
or check the wiring

'sid

yeah... I kindoff found the problem... turns out transistors don't work when soldered on backwards...

4am + soldering = flipped transistors

-Leav

p.s.
why are they impossible to get out??!??! *head/desk*

Edited 1 time(s). Last edit at 11/06/2008 10:08PM by Leav.

well.. in that case

relax, heat your soldering iron, take a deep breath... try again. one pin after another and make use of the desoldering pump you kept handy

'sid

The tab of the transistor is connected to the middle pin which is the collector, so if you use one big heatsink it will short all the outputs together.

You can get heatsink mounting kits which consist of a mica or plastic washer to insulate the tab and a plastic tube with a flange to insulate the screw. You could get by with two, leaving the heatsink connected to one transistor.

To remove the transistors without damaging them, the easiest way is to use a solder sucker to remove almost all the solder. You can then waggle each pin individually to break it free from the pad and then the transistor will fall out.

If you don't have a solder sucker you can try to place the soldering iron across all three pads and pull the transistor with pliers.

Edited 1 time(s). Last edit at 11/07/2008 03:53AM by nophead.

---

[www.hydraraptor.blogspot.com]

Quote

The tab of the transistor is connected to the middle pin which is the collector

Uuups that's true unfortunately
And now I'm confused.. is that for all to220 , because I ripped apart some old thermoprinter where all TOcased transistors where cooled by ONE large aluminum part which holds the mechanics.

Anyway.. sorry leav, my fault.

'sid

On all the TO220s I have come across the tab is connected to the middle pin, which is the collector on bipolar transistors, the drain on MOSFETs and the ground on most 3 pin regulators.

However there are some devices, including Darlingtons, that do have electrically isolated tabs. I expect it compromises the thermal performance, but in the case of a print head driver, with a lot of transistors that need heat sinking, then you would get a big cost saving sharing a heatsink without the need for insulation washers and the associated assembly time. Also the lack of external isolation would more than compensate for the extra thermal resistance of the internal isolation.

---

[www.hydraraptor.blogspot.com]

Thanks for the help guys

I went out and bought a solder sucker like you suggested and managed to get them out!

Since it was so hard getting them out, and I have reason to believe I fried them (connecting backwards and also shorting the tabs), I got new transistors to solder in instead.

just a quick question: the store only had TIP122 (instead of the current TIP120), they seem to be a drop in replacement with better performance, but I am not sure.

Spec sheet: [www.fairchildsemi.com]

could someone confirm this?
(alternatively, is there a simple way to test a transistor?)

Thanks!
-Leav

Edited 1 time(s). Last edit at 11/07/2008 03:22PM by Leav.

Through plated PCBs are tricky when removing multi-pin things. I have been known to chop the component off first so the pins can be removed individually. Overheating the board and damaging tracks and vias is tooooo easy and often the components are cheaper then rebuilding or bodging the board.

Leav,
Yes TIP122 just seems to be a higher voltage version of TIP120, everything else looks the same.

The easiest way to test a transistor is with a multimeter that has a transistor test function. A simple transistor can be tested with an ohmmeter but a Darlington is actually two transistors, a diode and two resistors, so it is a bit more complicated to test. I suggest you try one by soldering it in and if it works solder the other two.

Swapping the base and emitter in that circuit will not have damaged them. As long as you did not get them too hot removing them they should still work.

---

[www.hydraraptor.blogspot.com]

Thanks for the reply


Higher voltage means that they can handle it higher, not that they need it higher, right?

Since I already have the TIP122s I think i'll solder them in, and save the TIP120s for backup, especially since they got very hot when I tried to remove them. I touched them and didn't get burnt but my finger was sensitive for a couple of hours.

Thanks again!

-Leav

Yes the voltage is the maximum breakdown voltage between collector and emitter. When the transistors is off only a very small leakage current flows between the collector and emitter until you exceed the breakdown voltage, at which point it will conduct. If the current isn't limited by the load then the transistor could be destroyed. Not a problem in this 12V system with back EMF clamping diode.

By hot, I mean you would have to get the inside of the transistor more than 150C to damage it.

---

[www.hydraraptor.blogspot.com]