DPS5020 dual power supply build

[Dannyx]

Re: DPS5020 dual power supply build

Second side done

[stj]

Re: DPS5020 dual power supply build

actually rectifiers can short out - i'v seen it a number of times.

[Dannyx]

Re: DPS5020 dual power supply build

I was thinking of placing the secondary one before the bridge rectifier, so it's technically still AC. That is to protect the transformer in case the rectifier itself fails, which is unlikely, but NVM...

[stj]

Re: DPS5020 dual power supply build

just remember that unlike fuses, most circuit breakers only work on AC - or atleast they only trip where they say on AC.

[Dannyx]

Re: DPS5020 dual power supply build

I figured it would be wise to have some sort of short circuit protection on both the AC input to the two transformers and on their outputs. I was going to go with glass fuses, but then I thought THESE might be a good idea. They're adequate for the mains side no problem and I was planning to use them for the secondary as well, where I would of course up the current rating to around 20A, but then I noticed they can only hold 50v of DC , which is right on the edge of what the secondary outputs and I'm not sure it's ideal to push them to the limit like that, even though they'd probably never trip, unless the bridge rectifier becomes shorted somehow.

Then I though I'd go with THESE instead which can take an extra 10v of DC which would put them within the range of the transformer's output. Yeah, I know you're more likely to see those used on mains circuits, but they would be within spec for the low voltage side as well. They would be a PAIN to install though, since they would require cutting a BIG rectangle in the already thick and hard to machine metal, whereas with those pop-out thingies, I could just attack it with a 10mm drill bit, so that's another aspect to consider

[Dannyx]

Re: DPS5020 dual power supply build

@Dannyx
Did you forgot our previous discussion here...?
Plus stj post i've quoted gave you another hint (psu metal case)
and a relief concerning usb port to serial boards being optically isolated.

No, I didn't forget that. I went over the thread several times, since I was still a bit confused as to whether or not I should connect anything to the case. Running the two modules in series is not a priority to me ATM and I'm not even sure it could be done....can DC-DC converters like these even be wired in series ? However, keeping them separate to allow for a split-supply IS important, otherwise I would've just had a common GND at all times, including on the case.

WTH how did you manage to forgot the main power switch at the psu front panel...???

I didn't forget it, I just don't have it yet, so I don't know where it should go or what size hole it should have

[megaraider]

Re: DPS5020 dual power supply build

no, the usb to serial boards are optically isolated.
as long as they are not touching the metal case the pc wont be referenced to the psu's

@Dannyx
Did you forgot our previous discussion here...?
Plus stj post i've quoted gave you another hint (psu metal case)
and a relief concerning usb port to serial boards being optically isolated.

WTH how did you manage to forgot the main power switch at the psu front panel...???

[Your father, besides being good craftsman (my congratulations to him )
is a good guy to put up with his son laziness (shame on you )]

[Dannyx]

Re: DPS5020 dual power supply build

Ehh...no, it's actually the whole shaft that's wobbly. There's no actual bearing in there - the shaft just spins on the plastic shell of the drill, relying only on some grease to reduce wear, so after so much abuse, the plastic has started giving in, causing play in the shaft and chuck assembly. It's a cheap nasty thing anyway, no need to bother with it...

[stj]

Re: DPS5020 dual power supply build

strip the chuck and clean and grease it.
it's just 2 halves with the jaws seperated by springs in it.
maybe a spring cracked.

[Dannyx]

Re: DPS5020 dual power supply build

One side of the front panel is done. My old man, God bless him, did his best with the modest tools he had on hand, but it's not quite 100% despite his best efforts...some things just don't line up properly as you can see, but hey...beggars can't be choosers. He could really use a better drill press to unlock his full potential. The chuck on this one is a bit wobbly so the drill bit tends to slip off the intended target, hence why some of the holes are slightly offset there...

[stj]

Re: DPS5020 dual power supply build

the metal of the usb is frame-ground - it should be seperate from the 0v pin.
but that's not important, what matters is the side that plugs into the psu - that's opto-isolated from the usb side.

so you should be o.k.
check it - plug the adapter in and ohm between the usb socket and the ground on the psu module

[Dannyx]

Re: DPS5020 dual power supply build

But doesn't the outer shell of the USB connectors do just that ? Plus, isn't the black GND wire inside the USB lead common across all devices ?

[stj]

Re: DPS5020 dual power supply build

no, the usb to serial boards are optically isolated.
as long as they are not touching the metal case the pc wont be referenced to the psu's

[Dannyx]

Re: DPS5020 dual power supply build

Ok folks, I'm back on this project. I started working on the case today (or rather my old man did according to my instructions and tastes, as I'm a bit of a noob when it comes to machining metal stuff at this great precision ) and in the meantime I was gathering whatever parts might still be required. I've hit a bit of a snag however when it comes to those USB ports: remember the two modules can be connected via USB to a piece of software that controls them from a PC....pretty useless in the grand scheme of things, but if it has the option, I'd surely like it to work too. Trouble is, if I connect the two modules to USB, the grounds would become common through the PC or hub or whatever. It's not a problem if I'm powering something like a TV main board which requires two supplies (5 and 12v for instance), since the GND is common by default there, but if I want to go split-supply (+/- 12v for instance), that's not going to work, is it ? I found this isolator thing on Ali which according to the datasheet for the ADuM3160 chip it uses has separate GNDs and might do the job. Pretty pricey for my taste and I'd need two of them ! Am I at least looking at the right thing ?

[Dannyx]

Re: DPS5020 dual power supply build

Interesting. Transformers go a bit over my head as far as electronics theory goes - a LOT of math in there, but at least I got the basics One of the more complex electronic components, with a lot of stuff going on in that little heavy metal box - flux, current, capacitance, resistance, turn ratio....etc. it's got it all

[redwire]

Re: DPS5020 dual power supply build

It happens in all transformers, residual magnetism in the core. The only way to demagnetize is apply a lower voltage sine wave and then slowly dial voltage down. It's kind of like demagnetizing a screwdriver or cassette tape.

The formal name is "magnetizing inrush current" and it's why some transformers literally jump up in the air or blow primary fuses for no reason when they are energized, on the odd occasion. The core has a high (up to 90%) flux, just sitting there and then applying power the sine wave adds flux on top of that. I think some cores can saturate when that happens.

[Dannyx]

Re: DPS5020 dual power supply build

I see. Any way to "de"magnetize it ? Like shorting out the primary ?

[redwire]

Re: DPS5020 dual power supply build

The core is steel, it can stay magnetized for a very long time.

[Dannyx]

Re: DPS5020 dual power supply build

Ok, but how long can it remain magnetized like that ? It had been more than a couple of days when it happened

[redwire]

Re: DPS5020 dual power supply build

Big power transformers, they can be magnetized when you switch off power.
The flux alternates between N and S at 50/60Hz but say you unplug it when the core is at peak N. Then you plug it in for peak S, and you get a bigger than usual inrush current. The core has to flip magnetic polarity.
It just depends on the mains sine wave phase when you unplugged it and when you plugged it in again.