Heating TC4422 during tests / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Heating TC4422 during tests

Move Thread

LAN_403

Tonic

Mon Jun 23 2008, 10:12AM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

I have no idea why it's heating. I did put it on breadboard, connected together both GND pins (it's a TO-222 version), connected them to PSU's ground, connected Vdd to +15V. I did place a 100nF decoupling ceramic capacitor closely to Vdd and GND pins. No signal is connected and when I switch PSU on, it heats pretty bad. What am I doing wrong? To be honest, at first test, I forgot to connect both GND pins together (nah, read datasheet before doing something..), could it be a reason why it heated dramatically and destroyed driver?

GeordieBoy

Mon Jun 23 2008, 10:19AM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Things to check:

1. Make sure you haven't left the input pin floating. This can lead to operation in the linear region and cross-conduction or cause high-frequency oscillations. Both cause excessive heating.

2. Check that you have interpreted the datasheet correctly and correctly identified the pin numbers counting from the correct end of the device etc.

3. Scope the device output and find out what it is actually doing.

4. Measure the supply current and find out what it is actually drawing from the 15V supply.

Marko

Mon Jun 23 2008, 10:23AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

It looks like it's already dead. And for package, I guess you meant TO220?

People are probably going to address that you left the input floating but I don't think it's a problem, I always test my drivers like that fine. To be sure, fix the input to Ucc or GND and scope the output.

Tonic

Mon Jun 23 2008, 10:56AM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Yes, TO-220, my typo. Then, Marko, are you suggesting that not connecting together GND pins is a reason?

To GeordieBoy:

1. You mean connecting floating pin to ground? In that case, I have feeling that after connecting to ground it heats much lesser, but it still heats. I guess it's caused by destroyed structure or something.

2. I've checked few times and it seems everything is connected properly.

3). Nothing happens, only flat line, which jumps to up when I power up, then slowly returns to first place on oscilloscope monitor.

4). Now I've found out that amper meter in my digital multimeter is somewhat destroyed, and shows 0 whole time. I've checked with another working circuit and it's still same. I have no idea what's current driver is drawing.

Probably it's just died. Perhaps it's time to check with another sample, but I would like to know about testing, before I kill another one by stupidity :)

I've put a radiator, so driver heats much longer and I have time to figure out what's going on and switch off.

...

Mon Jun 23 2008, 02:01PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Are you 100% sure the gate on your fet isn't shorted? Try just connecting the tc4422 with only v+/gnd and the input grounded and see if it still heats.

Tonic

Mon Jun 23 2008, 03:05PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Nothing was connected to driver's output.

Now I've tried with another sample and it luckily works. But it still bothers me, why previous driver died in horrible agony? The floating input sounds very plausible to me, since on breadboard I'd a NE555 oscillator and it was working also and wasn't connected with driver, so I guess the output generated sorta of magnetic field and driver catched it.

What about not connecting together GND pins? The datasheet states that duplicate pins must be connected together, but nothing more. Why is that?

By the way, it's a pretty nice driver, it charges/discharges 10nF capacitor through GDT with very little affect on rise time :)

EDIT:

Forgot to ask. Can I use TC4421 (inverted one) like as TC4422? Since I'm building a small half-bridge, I would like to use only one driver and 1:2:2 GDT.

Marko

Mon Jun 23 2008, 03:24PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Well, overheating is really the only thing that could kill the driver.

It can be result of output overload, latchup, or inadequate dissipation.

The Ic could have latched up if it received reverse voltage on the output, without having freewheeling diodes installed. You *need* freewheeling schottky diodes if you are using a GDT.

From other side, it may have been enough if you just accidentally shorted the output to something, or miswired the pins for a fraction of second to kill the driver. They are quite unforgiving as output impedance is very low.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob