[hpsdr] Cooling.

[Ben Hall]

Good morning all,

There was a question recently here about if a computer power supply fan 
would cool an HPSDR.  I've completed a rather quick, top-level only 
analysis.

Before I share results, let's talk about the limitations (flaws) of this 
analysis:

1)  It's fan-forced convection only.  IE:  all heat produced is removed 
by airflow through the case.  This is a conservative assumption, as the 
air inside the case will transfer heat to the case, and the case itself 
will cool by natural convection and some, albeit very small, radiation.

2)  It's top level only.  IE:  the components inside that actually 
generate the heat are not modeled.  It is possible that this analysis 
will show that a particular airflow is okay for the system as a whole, 
but some component inside may require more to stay under its maximum 
temperature.  Or, it needs a bigger heat sink.  It's likely that for 
Ozy/Janus, there are no components that will require further analysis. 
(Better said, they have a low power density.)

3)  We don't the heat generated by Atlas, Ozy, and Janus, now do we know 
what heat future boards may generate.  The spread sheet allows the user 
to input various parameters for heat load, air flow, and Inlet 
temperature.  I assume 100 watts, which may or may not make sense.  This 
is based on the fact that a www.mini-box.com picoITX supply can source 
120 watts.

4)  We don't know the maximum outlet temperature the components inside 
will tolerate.  I'd suggest no more than 110 to 120 deg F (43 to 49 deg 
C) for long component life.  Remember that if the outlet temp is 120 deg 
F, the components inside must be hotter than this to reject the heat 
they generate.

5)  We don't know the flow through the box.  IE:  if you put a 20 CFM 
fan onto any enclosure, the effective flow through the box will be less 
than 20 CFM.  Maybe much less - my experience is that most enclosures 
are pretty restrictive - a 20 CFM fan may only draw 7 or 10 CFM through 
the enclosure.  (Muffin fans like used in computers hate back pressure.)

The design of the yet to be proposed Pandora enclosure system will 
minimize flow restrictions as much as practical.

So, with all these limitations, how does this analysis help?  It gives 
us a rough idea of what's possible and what's not possible.

My thoughts will be to shoot for a 110 deg F output temp, and double the 
CFM required to maintain that outlet temp.  This is very conservative.

Some Results:  (I will post the simple spreadsheet somewhere, stay tuned.)

***Typical Climate Controlled Environment***

At 100 watts, an inlet temp of 75 deg F (23.9 deg C), with 10 CFM 
(assuming a computer fan of 20 CFM effectively drawing 10 CFM thru the 
case) fan, the outlet T is 106.6 deg F, 41.5 deg C.  That's okay.

***Typical Non-Climate Controlled Environment***

At 100 watts, an inlet temp of 95 deg F (35 deg C), with a 10 CFM flow, 
the outlet T is 126.6 deg F (52.6 deg C).  This is higher than I'd like.

At 100 watts, an inlet temp of 95 deg F (35 deg C), with a 20 CFM flow, 
the outlet T is 110.8 deg F (43.8 deg C).  This is okay.  This case 
requires more than a computer fan.  Maybe two of them.

Let's say the HPSDR only requires 50 watts, and use the same 20 CFM 
computer fan with a 10 CFM flow through the case.  Outlet T is 110.8 deg 
F (43.8 deg C).  This is good.

***Extremely Hot Environment***

In this case, I raise the maximum outlet temp target to 120 deg F.

At 100 watts, an inlet temp of 110 deg F (43.3 deg C), you've got to 
flow 40 CFM to keep the outlet temperature under the 120 deg F maximum I 
mentioned above.  At 40 CFM, that's 117.9 deg F (47.7 deg C) outlet T. 
That's probably an 80 CFM fan.

Even if we go down to 50 watts, this is still a tough environment:

At 50 watts, an inlet temp of 110 deg F (43.3 deg C), you've got to flow 
20 CFM to keep the outlet temperature at 117.9 deg F (47.7 deg C).

Bottom line - a computer fan may be able to cool the HPSDR depending on 
your inlet air temperature and thermal loading.

-- 
Thanks and 73,
Ben, KD5BYB