PSU ground connection

[jimb]

I bought a DC power supply on ebay from a seller who customizes them and asked them to use a 4-pin XLR at the cable end of the output, with pin1=GND and pin4=5v (that's what the input of the device I want to power says at the back). The PSU has the right specs for my device, but when I connected it and turned it on, my device didn't power up.

Looking at the back of the PSU, the cable is coming out of a coaxial socket with "+" in, "-" out, and I started suspecting that the "-" hadn't been wired to the chassis ground as I expected (perhaps mistakenly, but I don't know any better). I brought out the voltmeter and surely enough, both read 5v (pin1 and pin4). Also checked that the used XLR pins correspond to the "+" and "-" of the coaxial plug - all good there. I hope my device hasn't been damaged. Please correct my logic if I'm off the mark.

Opening up the PSU, the coaxial output's terminals are clearly visible, as is a chassis ground further down the PCB (I'm presuming that's a chassis ground because it's standing by itself on the PCB without any traces around it, and a screw attached underneath it that's fixed to the chassis). My instinct is to desolder the coaxial socket's "-" from the terminal on the PCB, and use a short insulated cable to connect the socket's "-" to the chassis ground (not the AC socket ground). How wrong am I?

Thanks guys!

 

[davenn]

hi welcome to EP

a bit hard to follow your description

Looking at the back of the PSU, the cable is coming out of a coaxial socket with "+" in, "-" out, and I started suspecting that the "-" hadn't been wired to the chassis ground as I expected (perhaps mistakenly, but I don't know any better). I brought out the voltmeter and surely enough, both read 5v (pin1 and pin4). Also checked that the used XLR pins correspond to the "+" and "-" of the coaxial plug - all good there. I hope my device hasn't been damaged. Please correct my logic if I'm off the mark.

show a photo or 2 and where you measured +5V
also show a photo of your device that is getting the power, clearly showing power connections and polarity indications

Opening up the PSU, the coaxial output's terminals are clearly visible, as is a chassis ground further down the PCB (I'm presuming that's a chassis ground because it's standing by itself on the PCB without any traces around it, and a screw attached underneath it that's fixed to the chassis). My instinct is to desolder the coaxial socket's "-" from the terminal on the PCB, and use a short insulated cable to connect the socket's "-" to the chassis ground (not the AC socket ground). How wrong am I?

another photo or 2 required to show us what is going on


Dave

 

[jimb]

Thank you, Dave. Hopefully it'll make more sense with the pics.
The device's power input that this PSU was bought for:



The PSU's output:



The cable between the PSU and device, checked for correct wiring to thepositive and negative of the coaxial socket. Measured 5v on each used XLR pin (pin1=5v, pin4=5v):



Inside the PSU, showing the output socket and what i think is the chassis ground post (the three legs in the middle of the image, leading up to the screw on the chassis below it:



A shot of the PSU with requested spec 230v 50Hz input, 15W 5VDC regulated:

 

[jimb]

I take it the only return path would be the AC ground, but getting 0v from the PSU instead would be a better option.
Just thinking out loud...

 

[Minder]

If I read your right, you are concerned that the -ve side is not connected to earth GND.?
If so it is customary to isolate the output of a P.S. and then the user has the option of whether to earth ground the common or not.
M.

 

[jimb]

Thanks Minder. I don't know if it is or how to test this. If it's not, would that lead to my machine not powering up? It worked with another 5vdc 1amp supply feeding it pin1=gnd and pin4=+5v.

 

[Minder]

Whether the -ve is grounded should be incidental, if you measure from +5v to chassis ground and it reads 5v then the common is grounded, otherwise not.
It should not be the cause on non-working.
M.