Hi
I am trying to design in a LM3414 to a lighting system I am building but having some problems.
I built 3 boards according to the schematic on the LM3414 spec sheet. Each board had multiple LM3414 circuits
1. All three boards seemed initially to work; I checked each circuit individually with no signal connected to the EN pin and my LEDs lit
2. I am assuming this means that the EN pin floats high
3. When I input to the EN a 3.3v 10KHz PWM signal the chip immediately literally caught on fire.
4. The PWM signal was from a micro, so low current but there was a little ringing on the square wave edges.
5. the third board stopped working before I had a chance to test it with a filtered PWM and the LEDs only glowed slightly.
6. I notices in this post http://nbitwonder.com/forum/viewtopic.php?f=6&t=47 that someone had the same experiences with me and was hoping that there is someone else who may be familiar with this and can lend guidance.
Thanks
Jason
Jason,
Can you be more specific about which schematic you used and any changes which were made compared to what is published?
It would also be helpful if you could detail the operating conditions:
Hi John,
I don't see any obvious issues with those component values, especially given that they work in steady-state (no PWM dimming).
Do you have any scope captures of the ringing at the DIM pin? From what you have said so far, connecting the dimming signal seems to be at least in part the mechanism which causes the failure.
Regards,
John
I didn't capture the traces. I can reproduce them if necessary. First I will try the again with a filtered signal, once I reconstitute my board.
Do you have any reason to believe that negative ringing may have been responsible.
thanks
Unfortunately that is never a simple answer. All TI datasheets include an absolute maximum ratings table. Exceeding the values specified may cause irreparable damage to the IC. In the case of the LM3414 DIM pin, the absolute maximum ratings are (-0.3V) {min} to (6V) {max}. We cannot make any guarantees about the product if these ratings are exceeded.
Based upon what you have stated about the LM3414 working properly with no dimming signal applied and becoming damaged when the PWM is connected to the DIM pin, it is a very logical place to start checking for problems.
I just ran into a related problem with LM3414, where we're getting some dim-pin related failures (no fires, but apparent failure of the pullup current or input circuitry possibly due to over voltage)
The data sheet is lacking some needed information about the dim pin. can you please provide:
1. equivalent input circuit
2. Iih and Iil (current at Vih and Vil), and leakage current
Also I'm trying to figure out how to protect this pin, as I want to drive it from quite a long way away.
Looking at the block diagram there's a 30uA pullup current source on there - can this be easily damaged via overvoltage? Would current limiting (resistor) solve this?
Is it reasonable to use protection diodes to GND and VCC (the VREG output) to protect this pin, or should VCC not be used for this (in which case, maybe a zener or Transil diode)?
Thanks!
Did you scope the DIM input? - Mine had spikes, although the PWM was from a micro. So the things you say about internal damage are probably true.
For on/off static use, I tied the DIM to + 5 (needed to generate a 5V supply), or to GND.
for PWM operation, had to filter: 100 ohm series resistor at the PWM output, then just before the LM3414 a filter: cap and 100ohm resistor
The cap and resistor are in series but go from the DIM input to gnd. Use an appropriate cap for your PWM frequency - probably 0.0x uf.
Look at the signal at the Dim input before and after putting these parts in to make sure you're OK.
I didn't 'scope the unit with problems it happened in the field. It has only happened in 2 units out of hundreds we've built already.
Filtering that line with R-C might not be a good idea, as it will create a slow rise/fall time which can't be good. In any case, the proper filter or protection can't be designed without knowing more about the pin characteristics.
TI - still waiting for your response. I'd like the missing information concerning the pin.
Thanks.