LM199/399 application

[pmmunro]

I am working on a Standard Cell replacement based on a Linear Technology data sheet circuit which I have also found to be very similar to TI's data sheet.

This all seems fine, but during some research, I found a reference in "The Art of Electronics" 2nd Edition, (Horowitz and Hill). The second last paragraph on page 339 states "One caution when using the LM399: The chip can be damaged if the heater supply hovers below 7.5 volts for any length of time". Apparently this does not appear in the third edition of H & H, although I have not been able to check myself and I haven't found anything to support this caution on the internet.

Could anyone who has detailed experience of the LM399 offer any clarification please? I am not at all sure what is meant by hovering nor what constitutes "any length of time" or why a low heater voltage should damage the IC.

PMM

[Radio Wrangler]

I've not used it, but I have been involved with precision, ovened quartz oscillators.

Some early generation ones used a nichrome resistive heater with a transistor circuit switching on and off, as directed by a thermostat. Small variations in the oscillator frequency could be traced to the switching.

The replacement was a linear temperature controller. The heat from the power transistor was not wasted, its heatsink was the oven mass. I wondered why they retained a resistive heater section. They could have used just the power transistor alone, and had faster heat-up. The answer I got was that if the heater supply was low, there needed to be something to limit the current to a non-damaging level.

I think that is what may be going on in the LM399, it the heater is all-transistor. It could be at risk of over-current.

David