Building a 13.8V 26A Power Supply
I had a 13.2VDC switched mode power supply (SMPS) go up in smoke on a recent camping trip. It was powering a thermoelectric (peltier device) cooler in our cabin. Was only running perhaps a 7A load and it is rated 15A continuous / 30A peak.
I use a similar power supply in my shack. It's very well-built and otherwise rugged. I also like it was made in the USA. That said, it's close to 20 years old and so I may be looking at bad capacitors. Probably time to look at a new supply.
Noticed the nicer 30A switching power supplies are around $100 and linear (big transformer) supplies are at least $150. Also noticed a lot of hams are using cheap switching supplies aimed at the LED display market. These are very similar to the PC power supplies most of us are familiar with, except they have a single ~12VDC output (adjustable). The old PC supplies had multiple outputs: 3.3V, 5V, 12V, sometimes -5V, etc.
First off, if you'd like to understand how a switch mode power supply works, this video is superb. His English is great and Czech accent is entertaining. This is a very similar 10A unit.
Thought I'd try one. Found a 30A one on eBay for $26 shipped.
Next I'd need a case. I had this 1991 external CD-ROM drive from a previous employer who was throwing them out. I long ago removed the drive, but kept the case, as it had a nice power supply and room in it. Made in Japan.
I was curious how much it cost new in 1991. A whopping $1098 list. That's $2035 today. For a CD-ROM. Now you know why this case was built so well.
In place of the CD-ROM face plate, I put an old PC tower faceplate. It's nicer looking than the case.
Still haven't figured out if I'm going to mount it this way or with the terminals at the front. Need to figure out where I'm going to put the DC connections - Anderson Power Pole connectors. Probably on the back, in order to plug up the holes where the IO port used to be.
Of course it has no "RF offset" knob like some of the radio-oriented SMPS units have. I assume that is a variable resistor to slightly change the switching frequency and thus produce a different set of RF birdies. Maybe I'll add one if I find the RF interference intolerable. The SMPS I currently use does produce birdies particularly in the 20m band.
Says it's rated 30A at 12V and adjustable 15%. So if I run it at 13.8V, it will be a 26A supply.
The rear of all these units is the same. I'm going to run that LED from the board to the front of the case. There's a green LED already there that's meant to be driven from the DC of the old unit.
I was on Amazon the other day and noticed a unit claiming to be 33A and 400W for $15 with Prime shipping. I couldn't pass it up. I figured a backup or 2nd unit would be great.
The $15 unit is on the left and the $25 original unit is on the right.
But upon examining it, it appears to have been made either by a copycat or at the same factory with slightly inferior or at least different materials. Most of it is exactly the same, but there are some important differences. Everything looks generic on the left while branded on the right. On the right, the big inductor has 2 toroid cores, 1 on the left. The main switching transistors are a larger package on the right. The only thing that looks "worse" on the right is the output capacitors - they appear a better brand, but are 16V and opposed to the 25V ones on the left. 16V is really close to the 13.8V I'll be running.
I use a similar power supply in my shack. It's very well-built and otherwise rugged. I also like it was made in the USA. That said, it's close to 20 years old and so I may be looking at bad capacitors. Probably time to look at a new supply.
Noticed the nicer 30A switching power supplies are around $100 and linear (big transformer) supplies are at least $150. Also noticed a lot of hams are using cheap switching supplies aimed at the LED display market. These are very similar to the PC power supplies most of us are familiar with, except they have a single ~12VDC output (adjustable). The old PC supplies had multiple outputs: 3.3V, 5V, 12V, sometimes -5V, etc.
First off, if you'd like to understand how a switch mode power supply works, this video is superb. His English is great and Czech accent is entertaining. This is a very similar 10A unit.
Thought I'd try one. Found a 30A one on eBay for $26 shipped.
Next I'd need a case. I had this 1991 external CD-ROM drive from a previous employer who was throwing them out. I long ago removed the drive, but kept the case, as it had a nice power supply and room in it. Made in Japan.
I was curious how much it cost new in 1991. A whopping $1098 list. That's $2035 today. For a CD-ROM. Now you know why this case was built so well.
In place of the CD-ROM face plate, I put an old PC tower faceplate. It's nicer looking than the case.
The 30A supply is sitting atop the case.
Has a nice standard power cord input.
This thing had a 50(?) pin Centronics SCSI interface.
The power cord input has a really nice noise filter built into it. Probably only see these on high-end A/V devices now.
I removed the transformer. This (linear) power supply board has some great parts on it, creating 5 and 12 VDC filtered for the CD-ROM drive. But I don't need it. So those devices will stay off or I'll desolder them and put them in the parts bin. What I do need is the front section. The switch is actuated by the front panel and there's a line filter / surge suppressor device. I'll use that to switch the AC.
Still haven't figured out if I'm going to mount it this way or with the terminals at the front. Need to figure out where I'm going to put the DC connections - Anderson Power Pole connectors. Probably on the back, in order to plug up the holes where the IO port used to be.
Of course it has no "RF offset" knob like some of the radio-oriented SMPS units have. I assume that is a variable resistor to slightly change the switching frequency and thus produce a different set of RF birdies. Maybe I'll add one if I find the RF interference intolerable. The SMPS I currently use does produce birdies particularly in the 20m band.
Says it's rated 30A at 12V and adjustable 15%. So if I run it at 13.8V, it will be a 26A supply.
The rear of all these units is the same. I'm going to run that LED from the board to the front of the case. There's a green LED already there that's meant to be driven from the DC of the old unit.
I was on Amazon the other day and noticed a unit claiming to be 33A and 400W for $15 with Prime shipping. I couldn't pass it up. I figured a backup or 2nd unit would be great.
The $15 unit is on the left and the $25 original unit is on the right.
But upon examining it, it appears to have been made either by a copycat or at the same factory with slightly inferior or at least different materials. Most of it is exactly the same, but there are some important differences. Everything looks generic on the left while branded on the right. On the right, the big inductor has 2 toroid cores, 1 on the left. The main switching transistors are a larger package on the right. The only thing that looks "worse" on the right is the output capacitors - they appear a better brand, but are 16V and opposed to the 25V ones on the left. 16V is really close to the 13.8V I'll be running.
Generally in electronics, physically larger means better heat tolerance. These switching transistors are almost twice as large!
This toroid transformer has 2 cores. Larger again generally means better heat dissipation.
The board of the $25 unit looks a little nicer than the $15 as well, better printed, more readable. And interestingly, the 360W $25 unit board has what appears the model # DC-400, presumable 400W.
I bought the $15 unit Prime, so I can return it, which I may. Or I may keep it as a spare or parts.
Now to assembly ... to be continued.
Comments
Post a Comment