[Amps] Coil cooling for tubular coils.

[Will Matney]

Traian wrote:

>Hi Will,
>
>If you mean circulating the air trough inside the coils copper tubing,
>this shall be not so efficient. Considering the diameter (small) of the
>copper tubing used for the coils and its length, the pressure drop
>shall be very high and the airflow very low. Using liquid instead
>of air shall be a better approach in this case. Other way, the use
>of low speed high volume airflow for cooling of the entire tank shall be ok.
>  
>
// I agree, but on the high freq coil (10 to 6 meters) your using a 
large diameter piece of tubing with a short length, maybe 6 turns at the 
most. After I posted this, I ran a small experiment with a 10 meter coil 
for a quad of 4CX250B's I had. It was 1/4" tubing (3/8" O.D. X 1/4" I.D. 
Type K). I put a high flow muffin fan in the bottom of a small cardboard 
box and put a small hole in the closed side for the tubing. The fan fit 
tight in the open side, and I held it in temporarily with 1-1/2" wide 
packing tape. I put the tubing in the other side through the hole and 
fired up the fan. Sure enough, you could fill a steady air flow coming 
out of the end of the tubing. I done this because of thinking about the 
same reason your mentioning, resistance to air flow in a coil. Now if a 
long coil is used like in some low freq. coils, it won't work so well. 
But, in the low freq coils, the coil generally heats on the end closest 
to the higher frequency so if that was the end getting the air flow, it 
may help a slight bit. How much, I couldn't say as I didn't test a long 
coil. A long coil has so many turns and each turn causes a resistance to 
the flow like your mentioning. The difference in the low and high freq 
coils can be compared to a low ohm and high ohm resistor. In the test, 
of course one hole was all that had a flow but it was a muffin fan which 
can't compress air like a blower. To do the coil cooling, even on one 
with a blower, the blower would need to be sized for the extra air to do 
this. I could try a test using a vacuum cleaners hose (in the outlet) 
supplying the blowing air in place of a blower. I don't have a blower 
right now to try it.

// As Rich mentioned, a series on holes, evenly placed, under the 
longitudinal axis of a coil, laying flat, not vertical, would provide an 
air flow through the coils windings.

// The thought I had about a perforated coil form may work but haven't 
tried that yet. It would be a form which would be mounted vertically. 
The top closed off with holes all about the circumference for most it 
length. The bottom where it mounts against the chassis would be open 
where air from under the chassis could enter. This would allow moving 
air to come up through the form and out the holes behind the coils. I 
would want the high freq. end to be at the bottom of the coil. Another 
thought would be to add the holes at one end of the form just because of 
the HF effect.

>When using the optional fan for some amps (Alpha, ACOM), this
>may contribute not only for better cooling of the power transformer,
>but for the output tank also, as the air from the PS is directed to the
>RF compartment though the holes located near the output tank circuit.
>  
>
// That it should and really should be included without being an option 
IMHO. The old Heath SB-220 blew air directly on the tubes first and then 
on to the tank circuit.

>I think that the location for the holes of the PS and RF compartments
>separating wall was not chosen by aesthetics, but by technical reasons.
>  
>
// You are correct!

>The Harris RF-110 output tank is placed above the pressurized cooling
>air duct and their separating wall have some holes, so a part of the
>cooling airflow is directed to the output tank. The coils are tunable,
>by using tuning cores (rods) inside the coils ceramic former.
>There are larger holes for circulating the air trough the entire tank,
>but some smaller holes are placed just below the ceramic formers
>of the lower freq bands coils, so a small airflow is directed inside the
>coils, probably for cooling the coils and their tuning cores also.
>  
>
// Even the small tunable coils supplied by several manufacturers had a 
hollow ferrite slug (The slugs hole was a hexagon). The nose of the form 
had a plated brass cap which was press fit through a hole in the 
chassis. This allowed air to flow through the center of the form and the 
slug.

>I saw a Philips 10 kW transmitter (4CX15000A) for which the output
>tank included a large variable inductance (variometer).
>The first turn of the variometer has copper fins attached, like a heatsink
>and the air exhausted by the final tube was directed to the variometer.
>My thought was that the variometer coil shall heat enough indeed, as to
>justify the use of the hot air exhausted by the power tube for cooling it!
>  
>
// Correct again. In engineering, that probably had a heat problem at 
the first design. It was caught and a change order made to add the heat 
fins. This happens all the time in engineering as to keep from having to 
completely re-engineer the part.

>And especially its first turn, probably more heated when tuned for
>the higher freq range.
>  
>
// Exactly, look at the low freq Ten-Tec coil we were talking about the 
other day. It had turned blue at the top of the winding at it's higher 
frequency end.

>I took few pictures of it, if anyone interrested.
>  
>
// Yea, I'd like to see these for ideas.

>73,
>Traian
>
>
>Will Matney wrote:
>
>  
>
>>Another idea I thought of for tubular coils, to keep them super cool,
>>would be to terminate one end through a hollow insulated standoff into a
>>pressurized chassis. This would force air through the tubing itself and
>>out the other end the same as the blower forcing air through the fins on
>>a tube. Just a thought =)
>>
>>Will Matney
>>_______________________________________________
>>Amps mailing list
>>Amps at contesting.com
>>http://lists.contesting.com/mailman/listinfo/amps
>>    
>>
>
>
>__________ NOD32 1.817 (20040719) Information __________
>
>This message was checked by NOD32 antivirus system.
>http://www.nod32.com
>
>
>
>  
>