Little mushroom.

[technix]

After struggling a lot about how to get LM399 onto a breadboard without burning it, here is my solution: mount the LM399 onto a SO-8 to DIP-8 adapter board.



Any project suggestions? What can I do with this LM399? Should I send it out to get calibrated along with my UT61D?

[metrologist]

why not just build up the 10V reference circuit on that board?

[technix]

why not just build up the 10V reference circuit on that board?

You sure, on that SO-8 to DIP-8 adapter board?

[metrologist]

It's only 3 resistors and LT1001? I would solder LT1001 to the pads, then super glue 3 SMD resistors to the top of the chip with either just solder bridges to the IC pins or small jumpers if needed. Then solder the LM399 back on top of that mess. Then maybe some hot snot to keep it stable.

[technix]

It's only 3 resistors and LT1001? I would solder LT1001 to the pads, then super glue 3 SMD resistors to the top of the chip with either just solder bridges to the IC pins or small jumpers if needed. Then solder the LM399 back on top of that mess. Then maybe some hot snot to keep it stable.

I wonder if LT1001 have SOT23-5 or SC70-5 versions? If so I may develop a board that have the LT1001, three resistors and LM399 properly soldered, and follows the REF01 pinout (and thus be a drop-in replacement of REF01)

[babysitter]

Consider using the waste heat of the LM399 to stabilize your resistors a bit. Remove Polysulfone cap, surround LM399+helper with something. Supress air currents.

[technix]

Consider using the waste heat of the LM399 to stabilize your resistors a bit. Remove Polysulfone cap, surround LM399+helper with something. Supress air currents.

Can I put the four resistors along with the de-cased LM399 on the top side of the board, the op amp on the bottom, bond the top components together using thermally-conductive silicone adhesive and seal the entire assembly then in epoxy? This can result in a package pin-to-pin compatible with REF01 and have a built-in self regulating heater.

[babysitter]

I wouldn't prevent you from trying. Document how it behaves at different stages of building, and use a temporary covering first before you go for final.

(Sent from mobile)

[Kleinstein]

There is no real need to remove the nice original cap from the LM399. It will still stabilize (though not as good) the temperature in the surrounding, especially if the is a cover around the whole circuit.

[David Hess]

It's only 3 resistors and LT1001? I would solder LT1001 to the pads, then super glue 3 SMD resistors to the top of the chip with either just solder bridges to the IC pins or small jumpers if needed. Then solder the LM399 back on top of that mess. Then maybe some hot snot to keep it stable.

I wonder if LT1001 have SOT23-5 or SC70-5 versions? If so I may develop a board that have the LT1001, three resistors and LM399 properly soldered, and follows the REF01 pinout (and thus be a drop-in replacement of REF01)

Unfortunately the LT1001, LT1012, and LT1097 are not available in smaller packages but the LT1880 is available in a SOT23-5 package and would be suitable.  For a dual I might consider the LT6011 in the MS-8 package.

[technix]

There is no real need to remove the nice original cap from the LM399. It will still stabilize (though not as good) the temperature in the surrounding, especially if the is a cover around the whole circuit.

I was thinking about decapsulating the LM399 and use thermal conductive material to bond the chip with the resistors and maybe the op amp, then pot the entire assembly to suppress airflow.

[Kleinstein]

The LM399 get quite hot inside - this is rather high for the OP and even resistors would see increased drift at that temperature (e.g. 90 C). So it is better to keep the OP cold and maybe have the resistors somewhere in between. If really needed for stability (high TC resistors) add an extra oven / control loop for the resistors and the ref chip with it's cap and maybe even extra insulation. 

Having to much heat conducted away from the LM399 will also cause trouble with the internal heater, as it might work to hard and cause internal gradients.

I would avoid potting, as stress can cause extra drift an the OP and resistors. More like a closed case with some glass-fibers inside if too large.