Written by: David W. Ishmael - WA6VVL 2222 Sycamore Avenue Tustin CA 92780 (714) 573-0901 daveishmael@cox.net home email dave_ishmael@qscaudio.com work email
Latest Revision: 07 December 2005
The following article appeared in the Mar.'95 Issue#71 of Electric Radio and was written by David W. Ishmael – WA6VVL.
During a recent 80M SSB QSO with my KWM-2, I found myself constantly changing frequency, albeit in small steps, since I didn't have an RIT to compensate for the KWM-2's initial warm-up drift. My KWM-2's frequency stability met the "100Hz during any 1-hour period following a 20-minute warm-up" specification, but the initial warm-up drift during that 20-minute period was annoying. Using a digital thermocouple meter with a type T thermocouple, I measured the KWM-2's internal temperature rise. At 115 Vac line voltage (+288 Vdc from the 516F-2), the measured internal temperature rise was 35 - 40°C. At normal room ambients, internal temperatures of >60°C were common. After an hour or so of operation, the top cover was hot to the touch. I started to consider adding a fan to the KWM-2 to reduce its initial warm-up drift. In addition, Walt Hutchens, KJ4KV's letter in ER#64 had planted a "seed" in terms of reducing the heat generated inside the KWM-2. After reading Bill Kleronomos, KD0HG's great article "Electron Tube Survival Primer" in ER#66, I made a few measurements and ordered the fans.
My measurements indicated that there was sufficient clearance in front of the amplifier compartment to mount two "micro-boxer" fans side-by-side. The fan(s) I selected was an NMB model AC 2412PS-12W-B30, 115 Vac, with a specified unrestricted airflow of 10 CFM @ 3000 RPM. Each fan draws 4W (44 mA). I purchased the NMB fans from Newark Electronics @ $11.26/ea. plus shipping/handling. I chose the NMB AC fan primarily because I didn't want to build a separate DC supply to power the fans - it was just more convenient to "plug them in".
A few comments about fan selection:
The fan bracket was made from a scrap piece of 5-1/4" x 8-1/4" 0.062" aluminum. The holes for the two fans were punched with a 2-1/4" Greenlee chassis punch. The bracket was cut down into an L-shaped piece using an Adel nibbling tool to fit the existing 4-40 mounting holes on the top cover of the KWM-2's final amplifier compartment while clearing the cover's existing ventilation holes. Longer 4-40 x 9/32" screws were used to install the fan assembly to the top cover.
Wire-type fan guards are used on both fans and the guard on the right-hand fan just clears the support bracket for the top cover. Square plastic fan guards won't clear the support bracket with the existing bracket dimensions.
The power for the two fans is provided by a separate power cord to the 115 Vac. A slide-switch mounted on the fan assembly controls the fan's power. The connections to the slide-switch and AC cord were insulated with heat-shrink tubing to minimize accidental contact with the 115 Vac. I considered modifying the 516F-2 power supply to supply the switched 115 Vac to the fan assembly and eliminating the slide-switch. I finally decided to keep the fan assembly separately powered. Right now, the extra line cord and switch is OK. If a toggle-switch is used instead of the slide-switch, the maximum clearance above the bracket is about 9/16" before it hits the bottom of the top cover.
New fan assembly added to the KWM-2. This is a drop-in addition. No modifications are made to the KWM-2 and the fan assembly can be added without taking the KWM-2 out of its cabinet. The fan guards are optional.
Bottom view of the KWM-2 fan assembly.
I chose to mount the fans with the air flow down. I tried both directions and got similar results, so I left the air flow down. With the fans installed and running, there is still a small temperature gradient across the top of the KWM-2 after an hour or more of operation, but overall, the improvement is surprisingly good. The top cover above the fans and amplifier compartment is almost cool to the touch, warming as you reach the left-hand side of the rig. More importantly, frequency drift is negligible during long periods of receive. The initial 20-minute warm-up drift has been reduced to a tolerable level.
The NMB fans cannot be heard during receive at normal audio levels. During transmit or station mute, however, the fans can be heard, but I don't think that it is objectionable.
This was a worthwhile mod that took about four hours and $35 to install. This was an absolutely drop-in modification - no mods were made to the KWM-2 nor did it have to be removed from its cabinet to install the fan assembly. The KWM-2 runs about 20°C cooler with a corresponding improvement in warm-up drift and long-term frequency stability. Tube life and reliability should also improve.
My thanks to Warren Ruland who converted my pencil drawing(s) into a CAD drawing for this article.
Selected References:
Graph explanation and notes:
It's been over ten years since I wrote the original article for Electric Radio. The article has generated an immense amount of mail, both "snail mail" and email. In that time, I have built and tested a number of fan assemblies for my KWM-2. I do have a few final thoughts:
1) My solution to the "problem" of cooling my KWM-2 is somewhat arbitrary. This includes the fan selection including the fan voltage, bracket dimensions, the decision to include a slide-switch, etc. The number one consideration is that ANY FAN UNDER THE LID IS BETTER THAN NONE. All the design considerations start with that initial assumption. How to get there is up to you and depends on how much time and effort you want to invest. This is a win-win situation.
2) Voltage-doubling off of the 6.3 Vac provides an opportunity to use a "gazillion" 12 Vdc fans that are available both new and surplus. My latest trip to an amateur radio swapmeet found relatively hi-performance 60mm 12 Vdc Sunon fans for $2.50/ea. Lower performance 60mm fans were $2/ea.
Top View of KWM-2 bracket w/Sunon KDE1206PTS3 12 Vdc fans.
Bottom View w/6.3 Vac voltage-doubler PCB. The LM317T is below the PCB.
The purpose of this graph is to show the cooling effects of three different configurations using two Sunon KDE1206PTS3 15.9 CFM 12 Vdc fans.
3) Don't overlook the use of a single 80mm 12 Vdc or 24 Vdc fan. A single 80mm 24 Vdc fan assembly was built on single-sided 0.062" PCB material. The fan assembly is completely self-contained, weighs 5-1/2 oz. with the fan guard, and uses a Panasonic FBA08A24H1T 43.5 CFM 24 Vdc fan powered from a voltage-doubler using 1,000uF 35 Vdc caps powered from the KWM-2's 6.3 Vac @ J24 pins 8 and 9. Of all the fan assemblies that I have tested over the years in my KWM-2, this is the highest performing combination to date, and among the quietest.
Pencil drawing of a possible layout for the PCB/bracket for an 80mm fan.
In closing, thanks for all your comments (and thanks for purchasing my brackets). The original Electric Radio brackets are no longer available.
73, Dave – WA6VVL Tustin, California
The following table lists 230 60mm (2.36”) fans for use in the Electric Radio fan bracket from 27 different suppliers. 229 of these are 12 Vdc fans. This table is a bit “dated” since the last revision was 12Nov98, but I think that much of the data will still be useful, especially for surplus fans.
