Suggestions are presented for modification and use of "scrap" PC/compatible power supplies for use in providing power to low-power or medium-power mobile transceivers. These are ONLY suggestions, and are not presented as being fool-proof formulas for success.
Also, YOU must assume full responsibility if you decide to modify and use a PC/compatible power supply for anything other than its intended purpose (in a PC/compatible machine). Please, BE SAFE -- BE CAREFUL. These things can bite you!
After recently purchasing a 10-meter mobile transceiver to take advantage of the propagation opportunities caused by the current high level of solar activity, I found myself in need of a power supply to enable the use of that 10-meter transceiver, as well as the use of my 2-meter transceiver while stationary at my residence.
Being a person who enjoys the challenge of achieving maximum capability with the expenditure of a minimum of financial resource (O.K., I'm cheap!), I looked into my excess-equipment inventory (A.K.A. "junk pile"). I found that I had several excess 200-Watt (or greater) power supplies which were removed from old, scrapped PC/compatible computer chassis.
I have found used PC/compatible power supplies appropriate to this discussion in surplus-computer shops for as cheap as $5 per unit. Even if you were to purchase a brand new 300-Watt PC/compatible power supply, you should be able to find one for under $50 (sometimes quite a bit less). Compared to the price of traditional power supplies with similar power output specifications, this is a very cheap source of power supplies, even if purchased new.
Reasoning that my 10-meter and 2-meter mobile transceivers are capable of 25-Watt output, and the fact that they are solid-state designs, I concluded that the maximum power dissipation while being used in full-output, 100% duty cycle transmit mode (e.g., FM mode on both transceivers), would be approximately 30 Watts peak. This is nearly one order of magnitude less than the design output power of the PC power supplies. Also, these power supplies are rated as capable of delivering in excess of 8 Amps of current to the 12-Volt output, thus yielding a potential maximum 12-Volt line output power of 12 V X 8 A = 96 Watts. Some PC power supplies are rated for even more output current (10 or 12 Amps) on the 12-Volt output.
It is clear that such a power supply delivering a mere 30 Watts of output power (even continuously) would be extremely lightly loaded.
These PC/compatible supplies are of switch-mode design. As is detailed in the succeeding two pages referenced at the bottom of this page, you can see the features and benefits, as well as some disadvantages, of the switch-mode design. Key issues in the PC/compatible implementation are that the switch-mode design is of low power dissipation, is light-weight, and can be made relatively small in physical size.
In working with maintenance of PC/compatible computers, I have learned that most power supplies will not turn on, or generate output voltages (either on the 5-Volt or on the 12-Volt outputs) unless they are presented with a load on the 5-Volt output. There are a few exceptions to this, notably some older IBM brand supplies. Most PC supplies are designed to shut themselves off under a 5-Volt no-load condition as a safety feature (for the protection of the PC mother board and peripherals).
To "fool" the PC switching supply into turning on when power is applied to the 110-Volt AC input, it is necessary with most designs to place a dummy load resistor across the 5-Volt output. I have found that a load resistance sufficient to draw a current in the range of 200 mA to 500 mA will turn on most supplies. This would necessitate the use of a 25-Ohm resistance, 1-Watt dissipation minimum (drawing 200 mA); or a 10-Ohm resistance, 2.5-Watt dissipation minimum (drawing 500 mA). In some implementations, I have installed a small light bulb as the resistance. This also serves as a pilot light, and easily dissipates the power.
I have also found that for my use (25-Watt mobile transceivers), these supplies dissipate so little heat internally that the internal cooling fans are not needed. I therefore usually disconnect the cooling fan to cut down on audible noise.
In my implementations, I cut off all of the wires which exit the power-supply case except for one pair of wires which are the +12-Volt and Ground connections. The rest of the wires can be clipped off at the surface of the circuit board, or left as pigtails internally, being careful to not allow any exposed wires at the clipped-off ends to make contact with each other or the metal case.
Also, I drill a hole (or use an existing hole in the fan port after removing the cooling fan) to mount an OFF/ON power switch on the supply case. Be careful to trace the 110-Volt power input wires and switch the proper ones to be safe. Of course, also be careful to maintain properly-grounded wiring for safety, as well as for electrical noise-reduction reasons.
You might find that there are "birdies" or hash on some frequencies due to the swiching noise of the switch-mode power supply design. If so, try to separate the supply some distance from the radio and antenna leads. Also, you might consider using a balun in the power-supply output and 110-Volt power cord wires to reduce noise. Place said baluns as close to the supply as possible. A separate ground wire to the supply from your station ground might also help. I have not experienced any noise problems from the supplies I have tried with my radios, but must recognize that this is a potential problem.
The next pages represent a brief overview of the distinctive differences and relative merits of traditional pass-mode and switch-mode power supply designs. These pages are in Adobe Acrobat format, so you must have the free, downloadable Acrobat Reader to view them. Go to http://www.adobe.com to download the latest version of the Acrobat Reader.
Now, view the design distinctive pages...
If you have any questions, comments, or corrections to these pages, please feel free to e-mail me at KB0FPD@arrl.net.
Thanks to the Rolla Regional Amateur Radio Society club members, especially Peter (N0XZN, president, Rolla Regional Amateur Society) and Bill (W0OME) for encouraging me to give this presentation to the monthly meeting of the Rolla Regional Amateur Radio Society (7 August, 2000). Also, thanks to everyone on the 6 August 2000 weekly Microcomputer Net (14.325 MHz, 1600 UTC on Sundays), especially Sam (KJ4ZM, temporary net control for the 6 August 2000 session), Bruce (KG8YT), Bill (W9GUI), and Roy (KC0HAW) for encouraging me to make this material generally available. If I forgot anyone, please forgive me!