Computer Power Supply Converted to 13.8V for Radio

Do you enjoy making your own equipment or antennas? Discuss construction and design in here.
Post Reply
User avatar
Beakster
Super Member
Super Member
Posts: 261
Joined: 12 Jul 2007, 21:39
Location: Toronto, Canada

Computer Power Supply Converted to 13.8V for Radio

Post by Beakster »

Hey,

Picked up a new 400Watt PC ATX Power Supply and converted it to 13.8V using the instructions at: http://www.radiorampage.com/techpages/comppsu.php It was pretty easy, but I think it still needs a little more work.

Here's the PSU I got. $15 (Canadian) + Tax:

Image

Image



Case Removed:

Image

The capacitors in here can hold a deadly charge even after turned off. I always left it for a minute after turning off before opening. I also connected a load to the output, but kept switch on and AC input disconnected to try and use up some of the held charge. When handling I was careful not to touch anything on the AC side (top or bottom).

Here's the underside of the board:

Image


As per the tutorial I located the track that ran from the 5V wires down to the voltage control circuits. It was pretty easy and there was a bit of wire on the top of the board connecting it so I just removed that to prevent the 5V output controlling the total voltage.

You can see in the red circled area the underside of the jumpers connections:

Image



Red arrow here show's where I removed the jumper:

Image



I got a 1k trim pot and bent the pin 1 and pin 3 legs out, leaving the wiper pin straight. I put the wiper pin in one of the holes on the PCB left by removing the jumper. I was sure to use the voltage control side, not the side leading up to the 5V output wires.

Image


I connected a wire from the 12V section to pin 3 and from 0V to pin 1. I put some heatshrink on them too to prevent any short circuits. After that it was time to test it to see if it worked. I set the trim pot to the middle, bridged the green wire and a black wire on the ATX plug (to simulate the computer power button being on), connected a volt meter across the 12V (yellow) and 0V (black) pins of one of the molex connectors and powered up.

The supply immediately came to life and I was pleasantly surprised at how quiet the fan was for a cheap PSU. I slowly turning the trim pot until I got 13.8V on the volt meter.

Image


At one point I went a bit too high and the supply shut down. After this happened it took me a while to get it going again. You need to move the trim pot with the supply turned off, and leave it off with the power disconnected for at least a minute before it will turn back on. If the trim pot is still not within the require range it won't turn on and you need to adjust and wait another minute to try again. I almost thought I'd killed the supply but I found the sweet pot on the trim pot eventually and she powered back up.

Then I did some tidy up. I chopped off all connectors, desoldered all unnecessary wires and twisted the 6 yellow together and tinned the end, then did the same with 6 blacks. These go into a connector block which I use to connect to the radio's power cable. I also wired the green wire directly into the 0V section of the board.

Image


I then connected it all back up, and connected it to the radio. I spoke to some local contacts and compared the performance with the PSU and with a car battery. I was told I sounded clearer and louder with the PSU. I checked my power meter and it showed my radio putting out bang on 40Watts on SSB, rather than the 35/36 I was getting with the battery. There was no more noticeable interference using the PSU than with the battery and I specifically asked people I was talking to if there was any buzzing or anything on my transmit audio and was told there was not.

If I did experience interference/noise I was going to bridge some capacitors over the output like the tutorial suggested. The plan was to put them inside the PSU directly on the PCB by soldering them into some of the unused 12V and 0V holes, of which there are quite a few. I presume this PCB is also used for higher end supplies with more power and more connectors.

I was happy it was performing better than the battery I had been using and cost me less than $20 total and an hour or so of work. I stared doing some more testing though and measuring the voltage across the output when I was transmitting at different frequencies. I used FM at all times which shows as a sold 40W on my plug in power meter.

Sometimes I was getting a drop from 13.8 to 13.5 while transmitting. Occasionally it would go down to 12.5V just as I started. I tried this on a few different frequencies and noticed some times it dropped lower, much lower. There were times when the volt meter dropped all the way to 4V, but the radio still seemed to be performing fine. I put a 12V bulb from my van across the output too, and when the volt meter was dropping the bulb brightness did not change so I'm a little confused by this.

I'm going to do some more experimenting tomorrow with it. There are a couple of videos on youtube of another guy who did something similar but used a different technique:

http://www.youtube.com/watch?v=S0JDuiTcMxs&NR=1

He changes the 12V regulator so he can get 30Amps out of it. My supply is rated for 15A at 12V so, will be good for 13A at 13.8V without modification I guess. The manual for my Magnum Delta Force says it uses 6 Amps max so that's plenty. He also removes the 3.3V and 5V regulators and rewinds the toroid with 18 gauge enamelled copper. I got some of this special wire and may do the same. I am not quite sure though if I need to put on the same number of turns, or if I should put on an many turns as I can. First I need to do some more investigations to see if there is a precise pattern as to when the voltage drop during transmit occurs and if it has any negative effects. So far I'm happy though as I can put the battery back in my van and the radio is working better than it did with the battery. :D
9-TM-333
26-TM-333
pioneer

Re: Computer Power Supply Converted to 13.8V for Radio

Post by pioneer »

the differenc in the volt meter readings and the actual voltage can be down to 'RF' upsetting the readings on the voltmeter, they dont like strong RF fields, this is why the lamp didnt dim as the voltage didnt actually drop to 4 volts, if it did your radio would have stopped working a long time before that.

A couple of de-coupling capacitors and a few ferrite rings should help stop this from happening.


The only reason more people dont use converted PSUs is because they can be VERY noisy, Interferance wise, especially on the lower HAM bands, to make use of them down there they do need to be well filtered.


Good luck with the supply . Nice1

Mark
User avatar
mattltm
Radio Addict
Radio Addict
Posts: 553
Joined: 03 Mar 2010, 20:42

Re: Computer Power Supply Converted to 13.8V for Radio

Post by mattltm »

Great write up. Good job :D

I done this last year but built it into a case and included a volt meter and pot to adjust the voltage on the front and create a bench PSU.
"Spooky action at a distance."
User avatar
forager
Super Member
Super Member
Posts: 210
Joined: 22 Jan 2009, 14:50
Location: Barnsley, South Yorkshire.

Re: Computer Power Supply Converted to 13.8V for Radio

Post by forager »

:D Nice read thanks for sharing. .
Matt
26-TM-150
User avatar
Spike
Regular
Regular
Posts: 88
Joined: 20 Feb 2011, 21:21
Location: North Bedfordshire

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Spike »

I've seen quite a few of the cheaper PSU's kill motherboards when regulation goes to pot, I would put a crowbar in there for peace of mind.
User avatar
Beakster
Super Member
Super Member
Posts: 261
Joined: 12 Jul 2007, 21:39
Location: Toronto, Canada

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Beakster »

Spike wrote:I've seen quite a few of the cheaper PSU's kill motherboards when regulation goes to pot, I would put a crowbar in there for peace of mind.
Sounds like a very sensible idea. I'll implement that at the weekend.
pioneer wrote: The only reason more people dont use converted PSUs is because they can be VERY noisy, Interferance wise, especially on the lower HAM bands, to make use of them down there they do need to be well filtered.
I am considering getting licensed and getting a HAM rig so I would like to use one of these cheap power supplies on the lower frequencies. Can you explain a bit more about what is needed to make it well filtered? Is it again a case of adding some capacitors across the outputs and some ferrite rings/maybe an inductor in series, or is there more to it?

Also how can I measure my noise level so I know if it needs reducing, and I can see it has reduced after any filtering changes are made? I assume an oscilloscope (which I don't have) is the only way?

Thanks
9-TM-333
26-TM-333
User avatar
crusty
Veteran
Veteran
Posts: 3229
Joined: 02 May 2008, 18:28
Location: East Lancs. UK

Re: Computer Power Supply Converted to 13.8V for Radio

Post by crusty »

Just for info. The better quality supplies such as this old 450w Thermaltake give a a more stable output if you can find them s/h or scrap. Weight is a good psu quality indicator. ;) Good luck with your project.

Image

(Partly butchered for scrap bin)
User avatar
mikecharlie
Top Poster
Top Poster
Posts: 1466
Joined: 21 Mar 2005, 12:48
Location: Scotland

Re: Computer Power Supply Converted to 13.8V for Radio

Post by mikecharlie »

Had a clearout in a cupboard, found an old Mitac PC, the old ones where the monitor sits on top of the thing.

I butchered up the old PSU in there for the ferrite ring and the plentiful supply of ennameled copper wire........! :lol:

I have salvaged some bits from the CRT monitor, and might have a fart about with the flyback transformer. :twisted: :twisted: :twisted:
Ive got a couple line deflection high -speed switching transistors, for the flyback driver.


I cannot throw anything like this out with ripping the guts out for usefull bits and pieces first. :lol:
Radio, wall of sound, coming up from my tower...
Radio, wall of sound, 24 hours of power...........!!!
User avatar
Beakster
Super Member
Super Member
Posts: 261
Joined: 12 Jul 2007, 21:39
Location: Toronto, Canada

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Beakster »

I have some old microwave transformers I used to make a welder. I'm thinking wrapping some power cable around one of them (with the windings removed) might make a nice filter.
9-TM-333
26-TM-333
User avatar
Beakster
Super Member
Super Member
Posts: 261
Joined: 12 Jul 2007, 21:39
Location: Toronto, Canada

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Beakster »

Today I changed the 12V (13.8V) Rectifier in the supply from a F12C20CT to a MBR3045CT. As far as I understand it then should make it safe for me to pull 30Amps from the 12V section of the PSU.

At first I tried removing the other 2 rectifiers for the 3.3V and 5V, but the supply wouldn't power up without them. The must be connected to the voltage monitoring part of the supply and I think may even be used to power that part of the supply so I put them back in. I'm not taking any power from the 3.3v or 5v sections so I think its safe to take the full 450Watts of the PSU through the 12V section now. I need to get some thicker cables to carry the 13.8V once I am taking 30Amps too.

I've decided not to bother with the crowbar circuit. I've found that by turning the trimmer resistor I installed up until the voltage goes above 14.9V it causes the PSU to shut down, so I think the over voltage protection is already sufficient.

Before:
Image

After:
Image
9-TM-333
26-TM-333
User avatar
Beakster
Super Member
Super Member
Posts: 261
Joined: 12 Jul 2007, 21:39
Location: Toronto, Canada

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Beakster »

I've made a good breakthrough with this now :D

So what I had above was fine and it worked. I had changed the rectifier to one that would support 30A. The 5V and 3.3V outputs still worked but I wasn't using them. I tried removing the rectifiers for them but then the PSU wouldn't turn on so I had left them in place. The only problem that really existed was that when I put a load on the output the voltage dropped more than I was happy with.

Measuring the voltage when load was applied was confusing at first as my volt meter went all wonky and gave readings which were far too low to be true. My method for measuring the voltage was to test the voltage across the output to the PSU when my radio was transmitting (which should use 6 Amps). The wonky voltage readings were caused by RF feedback upsetting the voltmeter. I later found that so long as I was transmitting on a frequency where I got 1:1 SWR the voltmeter functioned correctly. As the SWR increased the accuracy of the Voltmeter decreased.

Anyway, I found that when transmitting my voltage dropped from 13.8V to around 12.2V. I decided that wasn't good as when I use more power (I want to use this PSU for 20A+) the drop in voltage would be unacceptable as it would probably go so low that the radio wouldn't work. I then endeavored to find a way to remove the 3.3V and 5V parts of the circuit as I believed removing them would stop the voltage being reduced on the 12V output (based on what I heard the guy saying in the youtube video: http://www.youtube.com/watch?v=rkApRg5VWnc).

The first attempt to remove the 3.3V and 5V rectifiers caused the PSU to no longer switch on. I examined the PSU board a bit more and decided to find out more about the IC on it. I googled SD6109 and easily found the datasheet for it. This datasheet was a massive help and enabled me to really understand what was going on. It even has example circuit diagrams.

It seems this IC designed specifically for these kind of PSUs. It is incharge of controlling the switcher to produce the correct output voltages as well as monitoring for under-voltage and over-voltage conditions and switching off the PSU when they occur.

It has a 3.3V, 5V and 12V input which it uses to monitor the respective voltages. After removing the rectifiers for the 3.3V and 5V these voltages no longer existed, hence the IC was seeing 0v on the 3.3 and 5v inputs, which is below its allowable minimum, so it turned the PSU off.

So far as I understand it, separate to the main transformer/regulation circuit that supplies our output voltages, there is also a smaller transformer/regulation circuit that supplies the standby voltage. This allows the PSU to appear to be switched off (standby/sleep mode), and yet your computer can activate it with a key press, or on receiving a LAN packet.

This 5V standby supply is what I decided to use to fool the IC into thinking there was a 3.3V and 5V supply (even though the rectifiers were removed so there wasn't):
The 5V standby supply goes to pin 20 of the IC.
The 5V voltage check is on pin 3
The 3.3V voltage check is on pin 2

First thing I did was to disconnect pin 3 and pin 2 from the rest of the circuit. I was going to have to supply voltage to them from a different source and I didn't want that voltage going back through the circuit. Disconnecting them was easy and they both connected directly to a resistor before going anywhere else. I desoldered these 2 resistors which disconnected them, and also left me with a perfect spot to solder wires into for the new voltages I had to provide.

Next I soldered a wire to pin 20 where I knew there was a good 5V supply (the standby supply used to power the IC). I ran this wire to a trimmer resistor, which also connected to ground. The wiper was then joined to pin 2 and another wire went from the 5V connection on the trimmer to pin 3. This means that pin 3 of the IC is essentially connected directly to pin 20 so has a perfect 5V supply. The trimmer resistor is working as a voltage divider, and can be configured to give pin 2 exactly 3.3V.

Now the IC is receiving the correct voltages on its 3.3V and 5V voltage check lines so it won't have an error condition even though the 3.3V and 5V rectifiers have been removed. After these changes I powered up the supply but it did not turn on. I turned it off and let it sit for 5 minutes to let the capacitors discharge. I set my original trimmer resistor down a bit (the one which I turn to get 13.8V) as I figured with the 3.3V and 5V rectifiers gone I might be getting a bit more voltage through the 12V output. I also made sure my new trimmer resistor used to get the 3.3V was just past the mid point.

Sure enough it worked and the supply powered up. I connected the voltmeter up and set the output to 13.8V using my first trimmer resistor. I then connected the volt meter to the second trimmer resistor and turned it until I got 3.3V. I then turned the first one up slowly monitoring the voltage output. It got as high as 14.9V before the supply shut down. This was good as it confirms that the IC is still doing its job with the 12V input and shutting down if a significant over voltage situation exists. Good news as if the PSU goes wrong the radio will be protected from dangerously high voltage.

I turned the trimmer resistor back down, left it for a minute and powered it back up and set it to 13.8V again. Then I connected up the radio and tried my load test again. This time when I was transmitting the voltage dropped from 13.8V to 13.6V! That is a much smaller drop than when the 3.3V and 5V rectifiers were in place (12.2V).

I am now pretty happy with the supply. Once I get a new radio which uses more power I will test to see how much of a voltage drop I get on transmit. If I find it is too much I will remove the 12V toroid and remove all the windings on it. It also has windings for the 5V line on it, and the 3.3V is on a separate toroid. The 3.3V toroid could be removed but its not connected due to its rectifier being removed so it does no harm there. The plan is to rewind the 12V toroid with a thicker cable, but the same number of turns. It looks like its already wound with #18 wire, so I think the thing to do is wind it with a double strand of #18 wire, or a single strand of wire twice as thick. So far as I understand it this should make the drop is voltage under load even less.

Anyway, its been a fun project so far. Most of the time I spent was figuring stuff out. If I had to do it again I think it would take me 45 minutes and cost $20 including the PSU, 30A rectifier and trimmer resistors. Not bad for a new 30A 13.8V PSU.

Here's a few pics of the latest modification:

You can see the new trimmer resistor I installed here. The black wire goes to ground. The red wires from the 5V standby on pin 20 to the trimmer and to pin 3, and the orange is the 3.3V wiper output to pin 2.

Image

Image


You can see here only the 12V output rectifier in place.

Image

Image
9-TM-333
26-TM-333
User avatar
northern35s
Veteran
Veteran
Posts: 3780
Joined: 02 Dec 2009, 16:09
Location: Blackpool

Re: Computer Power Supply Converted to 13.8V for Radio

Post by northern35s »

A good write up and a good read, this is good gen for anyone considering doing the same ;)
Mobile DXing from the car and on foot
jenny1

Re: Computer Power Supply Converted to 13.8V for Radio

Post by jenny1 »

Now I know that computer power supply it will be converted to 13.8v for Radio.
User avatar
Guzzy
Veteran
Veteran
Posts: 4969
Joined: 24 Feb 2005, 02:33
Call Sign: 26TM002
Location: Kent

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Guzzy »

Great write up. :)
User avatar
Beakster
Super Member
Super Member
Posts: 261
Joined: 12 Jul 2007, 21:39
Location: Toronto, Canada

Re: Computer Power Supply Converted to 13.8V for Radio

Post by Beakster »

Just a bit of an update. I've made a few changes to the PSU now. Firstly I replace the wires with some nice fat cable that can handle 30 Amps easily. Looks much better than the yellow wires all taped together. To do this I had to drill the PCB to make bigger holes for the wire.

I added some decoupling capacitors too. I'm not sure if this was necessary but it seemed like a good idea. The big fat electrolytic at the bottom in the pic is what I added. There is a little green one and ceramic one in there which i added too, but you can't really see in the picture.

To the output wires I added a couple of ferrite chokes and a pair of 25Amp fuses.

Most importantly though, I added an AC line filter which I got at my local surplus store for $3. I noticed I was getting quite a bit of noise on 20M and 40M, however I didn't realise at first. I have been comparing the PSU with a battery and when I did my comparison tests although I disconnected the radio from the PSU to connect it to the battery I did not switch off the PSU. This meant that even though the radio was powered by the battery it was still getting the noise from the PSU. This noise is emitted on the AC power cable and then uses that and the houses wiring like a big antenna to transmit the noise.

When powering the radio from the battery I was able to see the S meter go up by 4 or 5 S points when I turned the PSU on. I went down to my local surplus store and got this Corcom filter. It is a power socket with built in filter so it replaces the old AC power socket and fits in nicely. This has made a great difference to the noise level. The S meter on the radio shows the same noise when running from battery and running from PSU, and when turning the PSU on while running from battery there is only the faintest difference in static noise but it doesn't even register on the S meter.

Image

Image
9-TM-333
26-TM-333
Post Reply