Can an LM199 heater survive reverse polarity power?

[intabits]

While testing my new eBay LM199, using 15v and a 7k5 resistor on the zener, only 0.7v came out. I soon realized that I had the leads plugged in backwards on my power supply.
The current limit was as low as I could get it on my crappy supply, somewhere in the range 20-40mA.
After swapping the leads I get 6.849v as can be expected. But the current is only around 1mA, obviously from the zener. The heater is drawing nothing (should be more like 8mA).

So now I don't know if I've burnt out the heater, or if it was never working anyway, due to the part being a fake.

Has anyone else here made this mistake and had the (genuine) part survive?

[mimmus78]

Also one of mine died like this.

Inviato dal mio Nexus 6P utilizzando Tapatalk

[rob77]

the "eastblock" tesla clone MAB399 is definitely toast after reverse polarity applied to the heater - it's experimentally verified

[TiN]

If you open up LMx99 datasheet, you can clearly see heater is not just a passive resistor but rather complex circuit involving diodes, transistors, etc. So result of large negative bias is magic smoke, trapped in hermetic can  .


National Semi LM399 die photo. Courtesy branadic (Dipl.-Ing. A. Bülau) from the EEVBlog

[Andreas]

Hello,

to be more specifically:
The datasheet only allows a maximum negative voltage of -0.1V

So the LM399 seems to be a bit picky.
Other devices often allow -0.3V reverse polarity voltage.

With best regards

Andreas

[intabits]

Thanks for the responses, they are definitive that the heater is now cactus.

I've ordered a couple more (LM399s actually, I thought I had a 199), and put them on small plugable PCBs, and WITH reverse polarity protection diodes!
I'll probably include a buffer amp as well.

Does anyone here have any recommendations for a better or more modern buffer amp than the LM108 or LT1001 used in the data sheet examples?

...you can clearly see heater is not just a passive resistor but rather complex circuit...

Yes, The datasheet has the schematic for that also. A quick glance told me that it was unlikely it could cope with reverse power, but wanted to be sure.

[David Hess]

Based on the heater schematic, it looks to me like reverse voltage on the heater at 20mA or higher would burn out the JFET used for startup.  If it shorted from drain to source, then the zener diode used for startup would look like a normal diode.

The maximum negative voltage specification of 0.1 volt refers to the substrate diode between the negative terminals of the heat and zener reference.  It has nothing to do with connecting the whole device up backwards where pins 2 and 4 are shorted together.

[David Hess]

Does anyone here have any recommendations for a better or more modern buffer amp than the LM108 or LT1001 used in the data sheet examples?

There is nothing wrong with the LT1001 and its descendants which are still produced.  The LM108 is no longer produced but lives on with super beta input parts like the LT1008, LT1012, and LT1097.

My first choice in a low cost application would be the OP-07 which is inexpensive because it is widely sourced.  The lower input bias current of super beta input parts would be better if an LC filter is used to reduce reference noise.