LM3414 DIM input pin characteristics

can I see the schematic of your application?

what is the VIN and the level of the signal you are applying on DIM?

from where is coming (cable, other IC on the same pcb or ...)?

in the "failing" parts, can you check short circuit betwen DIM and VIN/VCC/GND or what?

Vincenzo

Sorry I can't release the schematic.

Anyhow I'm not sure how relevant that is to my specific questions... are you planning to provide that information (regarding DIM and VCC pins)?

The pin is driven via a 5V microprocessor (Freescale HCS08 series), driven through a 100Ω resistor.  The microprocessor is on a remote board, connected with anywhere from 0.1m and 1.5m of wire for the signal and ground.

We're analyzing a couple of the failed parts now, I don't have the information yet.

thanks,
Marc 

the pin is self-described in datasheet: a weak (internal) pull up keep the pin biased high when left floating; the pin is 6V (absolute maximum rating) so depending on how good was your remote signal (across a long cable, in case of mismatched line you could have ringing, causing overshoot above (you have a 5V digital signal) and/or undershoot (below -0.3V) that can damage the part... for this purpose, again, look for short circuits betwenn DIM Vs VCC and/or DIM Vs GND)

the more I can see, the better I can help, sorry

Vincenzo

Vincezo, the pin is described in the data sheet but it is not characterized.  Big difference.  

- in fact, it's described as a constant current source on the DIM pin, not a weak pullup.  What is the min/max current over operating? 

- again - what is the equivalent input circuit?  This is normal information for a data sheet, it is missing here.

Regards, 

Marc

Hi Marc,

  Dave Johnson asked me to look at this, and it seems the main issue you are concerned about is protecting these pins.  As such, I would like to recommend adding a TPD2E001 Protection Array:  http://www.ti.com/product/tpd2e001 .  Even if we fully understand the internal design of the circuitry, the real issue is to prevent damage, and the TPD2E is a practical way to do so. 

When using the TPD2E, be sure to connect heavy etches to the GND and VCC pins, direct to the respective planes would be best.

-Leonard

Thanks for the reply with some practical information, Leonard!

I took a look at TPD2E001 and if you are proposing that the VCC of that chip connects to the VCC of the LM3414 then I wonder, what will happen to the LM3414's internal regulator if you push a large voltage into the protection chip?  It will get shunted there.  If you leave the TPD's VCC unconnected, then it looks from the data sheet like it will allow up to 10V through.  

I was looking at a different chip, the tpd2s017.  It appears from the clamping waveforms that it should clamp the voltage to -0.8 (typ) to 5.5V, which is the best I could find.  And I wouldn't connect it to VCC.  I would have a series 1.5K resistor to the outside world, then the TPD then the DIM pin.  Then a 24V signal in would cause just under 250mW power dissipation in the TPD which is just acceptable....   thoughts?

well, we passed the one year anniversary of this question, and still no reply...

Well, I don't know what to say, Marc, except to apologize for the non-response.  A year ago, not sure what happened, and hopefully this will not taint your experience on E2E.  Let me look into this and get back to you ASAP. 

Basically you want to know what happens to the energy coupled through the TPD device, especially the Vcc rail, as it is also connected to other devices that have a Vmax rating, correct?  In most cases, the power supply impedance is comparatively very low and the spike is dissipated in the capacitance of the Vcc rail.  Wondering what your Vcc rail is supplied from?  A power supply or a battery? 

-Leonard 

p.s. do you have a local distribution FAE that you work with?  If so, who would that be?   Also, this App Note provides some guidance on the ESD protection devices: http://www.ti.com/litv/pdf/slla305 . 

Hi Leonard,

If you look back through this thread, I believe you'll find answers to your quetions:

- VCC is the LM3414HV VCC output (nominally 5-5.5V).  That is the only 5V rail in the device, the other rail is 48V.

- Dave Johnson is my local FAE.

So, from my previous post, I have 2 questions.  One is a quantification of how much abuse the VCC pin of the LM3414 can take.  (followup - most regulators go out of regulation when asked to sink current - thus the rail may get pushed well above 6V - how will that affect the internals of the LM3414?)

The second is whether the tpd2s017 would be up to the task, keeping it's VCC pin tied only to a 100nF cap as recommended (to it's GND), and not to the LM3414's VCC.  The latter has an odd 'max 5V' Vio rating, and the LM3414 DIM pin sources 30uA and extrapolating a bit from the data sheet, could have an open circuit voltage of up to 6V, so there's a question whether 30uA continuously into the Vio pin of the tpd2s017 would damage that part.

Thanks,

Marc

another month goes by....

Hi Marc and Leonard,

Did anything ever come of this?  I am looking to protect the DIM pin and am researching options.

Both the TPD2E001 and TPD2S017 appear to have a breakdown voltage (start clamping) at 11V, which is far higher than the max voltage rating on the DIM pin.  The LM3414 datasheet states the absolute max rating of DIM to GND is 6V.

I was thinking of using a simple 3.3V TVS diode (such as the LittleFuse SP4020-01FTG) which has a breakdown voltage of 3.5V.  This would keep clamp transients up to 1A under the 6V max rating of the DIM pin.  Thoughts?

Thank you for your help.

Never heard anything from TI.  I ended up using the  tpd2s017 in the product along with a 1.5KΩ series resistor, and we haven't had problems thus far....  Not the way I like to leave things, though.  Figures 4 & 5 in the data sheet looked the best for the application.  My driver was 5V so a 3.3V TVS wouldn't work for me.

Marc, thank you for noticing.  You're right the 3.3V TVS I mentioned would not do a very good job of keeping transients under the DIM pin's 6V limit.

I did a bit more research and chose the ON Semi ESD5Z2.5T1G TVS Diode.  The part has a 2.5VWM and lower breakdown of 4V.  It will keep a 5A transient under the pin's 6V max.