LM399 heat loss within a vacuum

[robert.rozee]

To get the epoxy to stop outgassing later you would need several weeks of pumpdown before sealing. You will need to draw all the moisture and volatiles out of the pcb or the vacuum will quite quickly get worse [...]

why put the passive components within the vacuum chamber? they will not share the same temperature stability that is afforded the die of the LM399/LTZ1000. it makes far more sense to JUST place the reference (TO-46 can) within the vacuum chamber with leads brought out for attachment of any remaining components.


can ANYONE explain the results in the attached paper? not wanting to put too fine a point on it, but the paper discusses heat dissipation from a firetruck running TRAIN when placed in a mild (0.1atm) vacuum...

"Abstract: The vacuum tube transportation (VTT) system has been a promising direction of future
transportation. Within this system, a high-speed maglev travels in a low-vacuum environment
to reduce aerodynamic drag. However, the heat dissipation of on-board heating devices will be
compromised under low-vacuum conditions, and the device performance may thus be lowered. This
study investigates the low-vacuum conjugate heat transfer characteristic of a levitation electromagnet
module of a maglev using an experimentally verified numerical method. During the heating process,
the surface temperature distribution of the levitation electromagnet, and the temperature and velocity
characteristics of the flow field are examined. It is found that, as the vacuum level increases from
1.0 atm to 0.1 atm, the total heat dissipating from the levitation electromagnet module is decreased
by 49% at 60 min, the contribution of convection heat flux over the total heat flux is decreased from
49% to 17%, and the convection heat transfer coefficient of the levitation electromagnet is decreased
by 89%. This study can provide an efficient numerical model for low-vacuum heat transfer study
on a VTT system as well as help the evaluation and optimization of low-vacuum maglev thermal
management systems."
(my bolding added for emphasis, original from here:https://www.researchgate.net/publication/339111964_Influence_of_Vacuum_Level_on_Heat_Transfer_Characteristics_of_Maglev_Levitation_Electromagnet_Module, linked in reply #13 above)


cheers,
rob   :-)

[Kleinstein]

For many reference circuits the parts around the LM399 can also contribute quite a bit to the drift. Once you take the effort of a vaccum, secondary oven or similar it can make sense to use it for more than just the reference chip.

The VTT case is about getting rid of large amounts of heat and this is mainly by convection and forces air cooling (fan).  Convection and forced air cooling get less effective quite fast. So it is easy to get rid of the convection part, but this does not apply to the conductivity part.

The surprising but still true point is than thermal conductivity is not depending on the gas pressure. The reduced heat capacity per volume is compensated by the longer distance the molecules can travel. The concept of thermal conductiviy breaks down if the gas pressure gets so low that the distance of free travel for the gas molecules gets comparable to the dimensions. At a lower pressure the heat transfer no longer follows the same rules and the parameter thermal conductivty makes less sense, as the equation it is ued in no longer applies.

[DavidKo]

VTT is still the same case as what was mentioned before. You can block convection with isolation - mechanical or with lowering the pressure. Thus they have the issues with cooling, nothing new there, one of the standard things when you work with vacuum. It simply does not matter for your project the way you go. With modern materials there is no need to use vacuum in your application, since you do not have any issues with drag coefficient. Vacuum use is more or less obsolete, due to the challenges it comes with, except the areas where it is unavoidable or cheap to use.

Vacuum apparatuses - heavy, bulky, need to be constantly pumped, shape is more or less given due to the outer pressure
Insulation materials - depends on the depth of your wallet (cheap=bulky) and it will be mostly light even when if it is bulky, easy to work with, a lot of types ...

From what you have described, I expect that you want to have the insulation of the reference. You can go for sure with vacuum, but I expect that at the end you will not be able to achieve it (meaning vacuum long term stability), pay for it or it will not fit your application. Definitely it is interesting idea like VTT, someone simply checks the blind ways the hard way.

[Conrad Hoffman]

A bit OT, but I just received an ADR1399KEZ in the LCC package. My mistake as I really wanted the leaded device. Datasheet says they're similar but the LCC has potentially more hysterisis. I'm thinking of dead-bugging it, which will change the thermal conditions substantially compared to a PCB layout. Not vacuum, but probably less heat loss than the leaded package, especially if I use very fine leads. I might even use fine manganin, as that package has force and sense leads for the diode, making lead resistance a (hopefully) non-issue. Manganin is 22 W/m-k vs 398 for copper. Thoughts?

[Andreas]

Datasheet says they're similar

They need more heater current due to lacking thermal isolation (the hat is missing).
I would use additional thermal isolation around the device.

with best regards

Andreas

[iMo]

It has got sense leads, but with rather large current (like 2.2mA)..

[DavidKo]

A bit OT, but I just received an ADR1399KEZ in the LCC package. My mistake as I really wanted the leaded device. Datasheet says they're similar but the LCC has potentially more hysterisis. I'm thinking of dead-bugging it, which will change the thermal conditions substantially compared to a PCB layout. Not vacuum, but probably less heat loss than the leaded package, especially if I use very fine leads. I might even use fine manganin, as that package has force and sense leads for the diode, making lead resistance a (hopefully) non-issue. Manganin is 22 W/m-k vs 398 for copper. Thoughts?

What about putting it on flex board and create a sandwich between 2 insulators. I'm still thinking about how to not over insulate the reference. You still need some heat loss for correct operation. Probably measurement will be necessary - checking the heater current at maximum expected ambient temperature and tuning the insulation thickness or adding some artificial loss (screw or metal sheet which can move and influence the heat loss).

[Kleinstein]

The LM399 / ADR1399 runs quite hot at some 90-95C and the reference part only needs some 1 or 2.5 mA and thus not much minimum power. So it would be hard to over-insulate. The thermal resistance no longer goes up linear with insulation thickness once the thickness is larger than the source size - it turn to more like a log law (and even slower for a point source) and thus deminisching returns of thicker insulation. The wires alone give plenty of heat loss unless really thin.

A small flex board for the LCC version could make sense , though one still needs to somehow hold the reference in place.

[DavidKo]

...
A small flex board for the LCC version could make sense , though one still needs to somehow hold the reference in place.

Fixing should be done by the insulation material itself (thick foam insulation from packaging material that is quite rigid and easily available), maybe glass wool in small cut out for direct contact. Put it between two flat parts and done. I suggest that EPS will not be suitable due to the high temperature of reference itself.
Such an approach can be used for whole family (LMx99, ADR1399 both). For classic version the flex board is not necessary since the legs can be easily insulated or 4 holes can be "drilled" directly to the insulation material. Maybe even the in the case of death bug approach mentioned by Conrad, the sandwich can help to fix the legs.

[iMo]

ADR1399 LCC in 1cm thick expanded polystyrene with 6x 0.1mm dia copper leads.. 

[DavidKo]

Are you able to calculate that the isolation of the wires will be longer? It will increase both resistances, but maybe it make sense to make the thermally insulation of wires longer.

Sorry for the dummy question, but I'm not to much familiar with thermal simulations. In case of 1 cm insulation the heat transfer of wires already dominates and the total loss will be cca 0.23W cca 15mA@15V.

[iMo]

The wires in that example go straight through the polystyrene wall (here 1cm thick), thus the length of the wires is 1cm. If the walls of the box will be thicker, the wires will be longer. With 2cm wall the Q_wires will be half. There are two shape factors for edges and corners, which depend a bit on the ratio of the wall thickness and the size of the box.
Below a table with the results, with params as above, you may see for 20mm wall one needs a better shape factors.. Mind the inner box size is the size of the LCC package (5x5x1.5mm).
Also note the Q depends on the temperature difference between LCC surface and outer box temperature (I set to 70C), not on the ADR's Vcc*Icc..