From novak Thu Feb 13 17:49:54 1997
Subject: heat load estimates
Content-Length: 2943
Status: RO


This message is a summary of some thinking and calculating that 
I've done regarding the heat loads on the He-4 that are expected 
from (1) an operating 1.5 K pumped pot, and (2) the JFETs.  I also 
compare these to the total maximum heat load that we can afford to 

(1) Pumped pot:

From the paper by DeLong et al. that we were studying last year, I 
infer that maximum heat load that can be dumped on a pumped pot 
(as measured by you last summer for the SPARO prototype pumped 
pot, for example) is related to the additional boiloff expected on the 
He-4 due to an operating pumped pot by just about a factor of two.  
In other words, for the two pumped pots that you ran last Summer, 
which had maximum heat loads of 20 mW and 60 mW, the power 
that we would have to absorb in our main He-4 reservoir to operate 
these pots would be about 40 mW and 120 mW, respectively.

I'm pretty sure that these results are valid for all pumped pots, not 
just the particular ones designed by DeLong et al.  I confirmed this 
with Harvey at the SOFIA workshop.  He pointed out, however, 
that the He-4 vapour in the pumped pot line can be used to COOL 
the various temerature stages.  If this can be done very efficiently, 
then the factor of two mentioned above can be reduced to values 
approaching unity.

In summary:  The 'hit' on the heat load to the main He-4 bath that is 
due to the operation of a pumped pot will be between 20 and 120 
mW, depending on length of capillary you choose for SPARO 
(assuming it is between the limits of the systems you ran last 
Summer) and the efficiency of vapor cooling via the pumped pot 

(2) JFETS:

The JFETS in Hertz dissipate 3 mW each when operational.   So if 
we go with the JFETs in Hertz (that are the same as in Penguin), 
then we will be able to operate them by dumping 3 mW times 18 = 
54 mW to the main He-4 bath.  I have carried out an estimate that 
shows that this amount of power is sufficient to keep the JFETs at 
the optimum operating temperature.  IE no additional heating will be 
required (except at startup, when a supplementary heater must be 
used to get them unfrozen).  

In fact, I bet we can run the 18 FETs on half of this amount of 
power (ie about 25 mW) with very little increase in SPARO's 
detector noise.  This is because (1) In Horowitz and Hill it is shown 
that the voltage noise in a JFET is proportional to the inverse one-
fourth power of the drain current.  (2) the power dissipated in the 
FET is proportional to the drain current, and (3) the FET noise is not 
the dominant noise source for Hertz, anyway.

In summary:  The 'hit on the heat load to the main He-4 bath that is 
due to running the JFETS should be about 25 mW.

(3) Maximum heat load permitted:

In order to put the above numbers in perspective:  We want to have a 
total heat load of 300 mW or lower.  This will give us our 2day+7 
hours hold time minimum for SPARO.