Help understanding Darlington CB leakage current and its ramifications

gerbermultit00l · 2005-07-21T15:22:32+00:00

Just when I think I'm advancing as an engineer I come across somethingthat is simple in theory but completely stumps me. I am going througha product design review. I am stuck on a comment a senior designengineer made.Sorry for the lack of ASCII artwork depicting my circuit but I don'tknow how to do that, nor can I read circuits others have posted intheir threads. They look garbled. I'll try to describe this verbally.I have a circuit that operates off a 13.5V supply. I am switching arelay coil on and off with a FMMT614 Darlington NPN transistor actingas a low side driver. One side of the relay coil connects to the 12Vsupply, the other side connects to the transistors collector. Theemitter is grounded. The base is pulled to ground through a 100K ohmresistor. Base current is supplied by a microcontroller output portthrough a 10K ohm resistor. I think that's about it for the timebeing. Simple I hope.Oh yeah, I have a back to back zener/rectifier diode clamping networkacross the relay coil to suppress transients to a safe level protectingthe transistor.This circuit must operate from -40C to 85C. It must surviveenvironments as high as 125C without damage.I thought this was a no-brainer. That's when my mentor suggested Icheck out the collector to base leakage current of the transistor. Hedid not elaborate any further.I have two problems:1.) Why did he say this?2.) This data is not furnished in the data sheet nor in anyapplication note I found so far.I assume he asked me to check this since the Darlington is a highcurrent gain device. A small amount of leakage C->B can turn into asubstantial current flow C->E. Could the device turn itself on andactivate the load!I've read a few threads posted here through the years that argue whereC-B leakage current goes. Into the base load (pulldown), into the B->Ejunction, both? That I'm not clear on. I don't mean to open that canof worms again but feel I must. Can any amount of C->B leakage currentflow into the...

J

John - KD5YI 20 years ago

drain on the micro and none of these other problems. It is >a match made in heaven.

From the Mouser catalog:

Fairchild BSS123, 100V, 6 Ohms, SOT-23, $.08/250. Supertex VN2110K1, 100V, 4 Ohms, SOT-23-3, $.24/1000. Siliconix SI2328DS-T1, 100V, 0.25 Ohms, SOT-23, $.55/500

I didn't look any further. I don't understand why you would need a 100V device to operate a 12V relay, but I'm sure you do. If you find that you could use a 60V device, you'll have a much wider selection. But, you knew that already.

Good luck.

John

Vote

J

John Larkin 20 years ago

Car?

John

Vote

J

John - KD5YI 20 years ago

In that case, the Rds of the Fairchild part may be too high for you.

device to operate a 12V relay, but I'm sure you >do.

Ah! As I recall from SAE J1211, you can have transients well over 100V at times. But, I don't know if that is applicable to your particular case.

Cheers, John

Vote

R

Rich Grise 20 years ago

formatting link

Pick about a 30V or 33V one of those and put it across your supply leads, maybe with a hash choke or some ferrite beads.

Good Luck! Rich

Vote

G

GerberMultit00l 20 years ago

UInfortunately a trans-zorb will not work for me. My assembly requires a direct battery feed (low resistance path to module). The device I need would have to be large to absorb transient spikes on an automobiles charging system without some means to limit current.

The relay coils add some current limiting but the transient suppressor would still be too big. I'm working in very tight confines. I have a hard enough time finding room for a 1210 package resistor. Don't want to go to a through hole part. We haven't had a lot of luck with that. The assembly cannot be wave soldered due to its physical layout. Hand soldering is not that reliable.

Thanks,

George

Vote

J

John - KD5YI 20 years ago

Hmmmm. He must have some way to protect his low-level (micro) from transients because he is only worried about the relay's switch. So, why not put a 40V TVS from the FET drain (or darlington collectors) to ground? Nothing would happen for the 24V case and 60V of the 100V, 400ms transient would be dropped by the relay. The relay provides an impedance for the TVS to work with. The relay might pull in temporarily. This way, he could use a

60V FET and a really low power TVS.

Just an wild thought.

John

Vote

G

Ge0rge Marutz 20 years ago

"Hmmmm. He must have some way to protect his low-level (micro) from transients because he is only worried about the relay's switch."

Correct.

" So, why not put a 40V TVS from the FET drain (or darlington collectors) to ground? Nothing would happen for the 24V case and 60V of the 100V, 400ms transient would be dropped by the relay. The relay provides an impedance for the TVS to work with. The relay might pull in temporarily. This way, he could use a

60V FET and a really low power TVS."

A few reasons I don't do this.

1.) I am switching high current loads (60A). I cannot allow the relay contacts to close during such a high transient. It could tack the relay contacts shut. I would no longer have control over the load. This would not be good.

2.) The cost of two components and machine use to place them on the board exceeds the small incremental cost to place a single higher priced 100V part. On top of that, I have very little board space available. I would be pressed to find room for two small SMA package TVS's, more or less the SMC size that would be needed. IMHO Sometimes its better to try and immunize yourself against transients by letting them crash against a barrier vs. trying to absorb them.

Ge0rge

Vote

Help understanding Darlington CB leakage current and its ramifications

Join the Discussion

Didn't find your answer?