Author Topic: My Variable DC Power Supply (30V/1A)  (Read 54986 times)

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1017
Re: My Variable DC Power Supply (30V/1A)
« Reply #45 on: August 25, 2013, 12:22:46 AM »
You get fans with a built in thermistor for temperature based speed control, or simply grab an old PC power supply with the temperature sensing fan and grab the tiny board with the thermistor and a transistor out of it and use that.

Nice cooling block there, even without a pump it will do well, just use radiator antifreeze in it and it will do fine, though the transfer rate will be much reduced by not having a pump to move it, it will be a lot lower in heat transfer capability with only a thermosiphon driving it.

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #46 on: August 25, 2013, 02:48:56 PM »
Back again with results. I mounted two old, 80 mm, 12 V, DC fans on the radiator and they work like a charm (third photo). I took the bridge rectifier too on the radiator and now it stays cool even without thermal paste. Basically everything that's on the radiator (three 2N3055's, one 2N3054 and the bridge rectifier) reaches a maximum temperature of 32 degrees Celsius. On the first photo you can see 28.3 degrees Celsius, the temp of one power transistor. Things change a bit on the second photo, because I lowered the voltage to 10 V and the temp of the same power transistor rose to 36.3 degrees Celsius; I say it's still reasonable.

OK, now what's up with the rest of the components? The values above were the result of a quite long (~1 h) stress test: I used a 24 V/100 W Philips incandescent light bulb as load. The maximum power drawn was around 100 W (23 V/ 4.3 A). The transformer got quite warm (reached 50 degrees Celsius at a given moment). The 20 W resistor went up to 51-52 degrees Celsius. The hottest of all was the 5 W resistor on the LCD: 70 degrees Celsius.

Plus here's an interesting situation: as you see on the second photo, I brought down the voltage to 10 V, current went down to around 2.64 A and a few moments later I heard a high pitched noise. It was coming from the small 100 uF/ 63 V capacitor. I touched it and it burned my finger. I don't have an explanation to this behavior which is not  present at higher output voltages.

It seems to me that all major components get hot in certain situations. I might need to drill holes on both sides of the case and mount one more fan on the top cover to blows cool air inside.

Finally, I would like to point out one more thing! This variable power supply has a big disadvantage: it can't stabilize the voltage. :( As I reduce the current with the help of the current pot, the voltage decreases too. Quite sad that I just noticed that. I don't know if there's a reasonable explanation to that, but in my head the voltage should stay still as I lower the current. Or I'm missing something?

As always, your thoughts are welcome! Cheers!
"Dubito ergo cogito, cogito ergo sum." Descartes

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1017
Re: My Variable DC Power Supply (30V/1A)
« Reply #47 on: August 25, 2013, 03:09:43 PM »
As you limit the current the voltage across the load will naturally decrease. The load is a relatively constant resistance, so if the voltage is say 12V at 1A it will only drop 6V with 0.5A though it.

The capacitor getting hot means it was either defective, had too high a ripple current through it ( which means the power supply is oscillating, check with a scope or by using a DVM set to AC volts and fed via a 220n isolating capacitor from the output), was a pirate part or was put in backwards.

Good to see the fans do help, though when mounting them it will be better to have a larger gap between the fan and the heatsink, a top and bottom spacer sheet glued to the heatsink and the fans spacing them about the thickness of the fans will increase airflow across the heatsink and lower temperature a lot more, probably by about 10C at the highest power dissipation. You might want to have a few holes in top and bottom of the case for airflow inside for cooling as well, or add a further fan to blow air through the case as well.

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #48 on: August 25, 2013, 04:08:19 PM »
Hi Sean! Thanks for the short explanation about voltage drop while controlling the current, though I'm gonna have to read more about that to fully understand it.

The capacitor was definitely not put in backwards; the polarity is correct. It could be faulty. To be honest, I snapped it off my previous J-31 PCB. With the current kit/PCB I received a 100 uF / 35 V cap and I thought what the hell, the higher the voltage the safer to use, so I desoldered it from the previous PCB and soldered it on the current one. It's either this or the power supply is oscillating. Since I don't have an oscilloscope, I will take it to a friend and check it there.

Regarding fans, I actually left space between them and the radiator. Please see the attached side shot; there is a 19 mm space. I will place a 120 mm or even bigger fan with low rpm on the top to blow fresh air inside the case. Poor power supply will look like the beast of Dr. Frankenstein when I finish it. :o
"Dubito ergo cogito, cogito ergo sum." Descartes

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1017
Re: My Variable DC Power Supply (30V/1A)
« Reply #49 on: August 25, 2013, 04:47:18 PM »
Not as bad as some I built, but looks like it will be with you for a long time.

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #50 on: August 26, 2013, 05:00:07 AM »
The first is always special. ;D

What's your opinion about the resistors below?
- OHMITE - 15FR100E - RESISTOR, 0R1, 1%, 5W
- BOURNS - PWR220T-35-R100F - RESISTOR, POWER, 0.1R, 0.01, 35W

At a first glance they seem more suitable than those huge cemented resistors that I bought, because the 5 W Ohmite is much smaller and and 35 W Bourns can be mounted on the radiator.
« Last Edit: August 26, 2013, 05:07:57 AM by birrbert »
"Dubito ergo cogito, cogito ergo sum." Descartes

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1017
Re: My Variable DC Power Supply (30V/1A)
« Reply #51 on: August 26, 2013, 02:47:04 PM »
The TO220 units definitely as they can use the heatsink you have and the existing fans. As they are close to the transistors anyway you will have shorter wiring.

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #52 on: September 12, 2013, 04:50:43 AM »
I'm a bit confused about the value of the resistors. blankfield recommended 0.1 Ohm ones, but if I=V/R, then wouldn't this type of resistor increase the minimum current or mess up something else? ???
« Last Edit: September 12, 2013, 05:45:50 AM by birrbert »
"Dubito ergo cogito, cogito ergo sum." Descartes

dr_p

  • Jr. Member
  • **
  • Posts: 64
Re: My Variable DC Power Supply (30V/1A)
« Reply #53 on: September 12, 2013, 09:30:47 AM »
You are measuring voltage(for current limit) "before" those 0.1ohm emitter resistors, so they don't influence things. Well they do slightly - they cause an extra voltage drop, but the op-amp takes care of things, opens up the transistor some more, and in the end a 1A current will still drop 100mV on the current shunt.

But keep them low value, because you lose output voltage on those babies. If you were to use 3 pieces of 10 ohm resistors, you effectively add a series 3.33 ohm resistor to your load. That means that at 3A you drop 10V on just those resistors. So your 0-15V power supply is now only capable of 5V output at 3A, but if you draw less current, there's less voltage drop on the resistors and it's capable of more output voltage.  Also, If they are too big in value they limit the maximum current capability for the same reason - too much voltage drop on them, no more voltage left to drive the output at the desired current.

I hope I'm not getting you confused.
« Last Edit: September 12, 2013, 09:36:20 AM by dr_p »

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #54 on: September 18, 2013, 03:58:12 PM »
Not fully understood yet, but I'm trying. In the meantime the latest parts arrived and I soldered/wired everything together. Please check attached photos. So, it's not nice, I know, but I tried to do my best. I definitely didn't plan that last PCB with the 5 W resistors. I will be doing wire management later, but with these PVC insulated cables it's hard. Maybe I'll try silicone cables next time, but those are expensive...

It's still not 100% ready, because the main switch is not wired yet, there's no power for the fans, the back case cover has to be installed too and I put no fuses. Please advise in this regard. What kind of fuse and fuse holder do you recommend? Should I put on both places: the primary and secondary of the transformer? I'm a bit tight on space.

At least now I can work on the solar cell phone charger. Yay! ;D I already charged some Li-Ion batteries. :)
"Dubito ergo cogito, cogito ergo sum." Descartes

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #55 on: December 06, 2013, 03:31:51 PM »
Hi Ladies and Gents!

Almost three months since my last post, yikes! ???

A lot has been going on, hence the huge time-out... family stuff, renovations in the apartment and work all at the same time = no chance to tinker with the power supply. I believe Martin is one person on this forum who roughly knows what I'm talking about.

But, winter came this week with temperatures around -10 degrees Celsius, which hopefully brings along days of peace and quiet. I'd like days like that anyway, because for me Christmas is not Christmas when there's rush; instead it should be all about family, nice mood and force-meat rolls in cabbage leaves. :)

I'm about to order a few parts off eBay, small but important parts and assemble the PSU for a review on an oscilloscope. I think that once I hit that point I'll have a working unit (Thanks to all of you, again and again!), but in the same time I'll know what kind of changes need to be done to improve it.

Thank you for your patience and I promise to keep you posted more regularly from now on. Cheers!
"Dubito ergo cogito, cogito ergo sum." Descartes

SeanB

  • Administrator
  • Hero Member
  • *****
  • Posts: 1017
Re: My Variable DC Power Supply (30V/1A)
« Reply #56 on: December 07, 2013, 01:57:19 PM »
Those look very nice, I miss them.

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #57 on: January 05, 2014, 12:08:50 PM »
I installed and activated DesignSpark. After 5 hours of work I came up with this (please see attached files; exports of the work, but I am happy to share the .pcb file with anybody upon request).

The original design helped a lot, but I had to make quite a few changes to it to represent my personal build. This is just the main board. I have to make two smaller ones too (i.e. bridge rectifier with the filtering capacitors, plus the one with the four power resistors).

Now, the big question is, whether this is something Martin had in mind for the oscilloscope competition?
"Dubito ergo cogito, cogito ergo sum." Descartes

dr_p

  • Jr. Member
  • **
  • Posts: 64
Re: My Variable DC Power Supply (30V/1A)
« Reply #58 on: January 07, 2014, 09:28:13 AM »
as a very rough rule, I use 1mm (40mil) of track width per every 1 Amp. If I recall, your design goes up to 3-4A, so either go 3-4mm of track width on the final current path, or (if the PCB is done) add solder or even a 0.5mm copper wire to the track, to lower it's resistance and current caring capability.
I see you thickened the traces a bit, but: trace leading to trimpot PR1 doesn't need to be that thick, it's not caring current. However, the input capacitors traces are caring current, since they supply power when the 50Hz sine is near 0V.  Also, slightly larger traces for the output cap, even if it's only 50uF. 

birrbert

  • Full Member
  • ***
  • Posts: 149
Re: My Variable DC Power Supply (30V/1A)
« Reply #59 on: January 07, 2014, 12:27:34 PM »
Mersi dr_p! I will make changes accordingly. I'm still trying to figure out how to make things more precise in DesignSpark. I didn't like the 3D view; it would need more details like connecting pins and such.

As you may have noticed, I changed the default colors to colors that are more pleasant to/forgiving with the eyes. The default combination of black background and vivid colors was making me dizzy.
"Dubito ergo cogito, cogito ergo sum." Descartes