Monitoring voltage & current on Motherboard?

[Ttaskmaster]

Afternoon,

Bit of an odd question, but are there any points on a motherboard (booted and fully running) where I can safely and permanently attach a voltmeter-ammeter unit and not short/blow up anything?
I know some MSI boards have dedicated "V-Check" points, but mine is an Asus X99 board - I have various empty headers, so I was just wondering....

The V-A meter unit I have in mind does 0-33v and 0-999.9mA-3.000A, and is independently powered by a 12v supply (ie Molex).

[matts-uk]

For what purpose? What are you trying / expecting to measure?

You can connect the meter negative to any ground and probe away at any exposed solder, pin. via, pad or track and get a voltage reading. Whether it means anything depends what you are looking for. Current readings (Amps) are taken by placing the meter in series, which is a PITA break in a made circuit, unless there is a test point. Easier to find a resistor in series, measure the voltage drop over it and convert with Ohms law.

[Ttaskmaster]

Looking to read voltage and current. Mainly for aesthetics, really.

I had a case from an earlier build that featured a bank of three such combo meters. Both values updated every 200ms.
These sorts of things: https://www.amazon.co.uk/gp/product/...?ie=UTF8&psc=1

However, the PC components I had back then were MSI ones with their trademark aforementioned 'V-Check' dedicated test points, so I couldn't really go wrong. They even came with a set of 2-pin connectors with female-ended wires specifically for sticking multimeter probes into. I believe Asus used to have something called Hotwire on some of their ROG boards that did something vaguely similar...

That's why I was looking for a mobo header or something to which I could permanently fix/plug some probe wires and get results displayed.

[matts-uk]

If it is purely aesthetic the easiest way is to break into a PSU connector. The mobo connector is probably the most interesting but sata power or a fan is easier to source an inline lead to cut into. The meter wiring is straight forward. I'll see if I can't come up with a schematic tomorrow. Gin O'Clock after a silly busy day around here at the moment. Seems most of our customer had no clue it is a bank holiday tomorrow.

[Ttaskmaster]

Well, I can certainly squeeze some probes into the 8-pin or 24-pin ATX power connectors easily enough. Same for SATA and Molex. I could even tap (bridge?) into a fan extension lead and save spoiling the existing cables.
Which ATX pins would give the most interesting results, do you think? Will they give fluctuating readings, or will they just show static values?

The meters themselves are still wired up from their previous installation. Thick black and thin red go to the incoming 12v power (Molex), thick red and thin yellow serve as the 'probe' plugs. Thin black has been removed - As per this: https://images-na.ssl-images-amazon....AC_SL1200_.jpg

There's a tiny red push switch on the back, which is marked 'key' on the PCB. I assume that's some kind of push-to-break? It didn't seem to do anything, that I could tell...

[DanceswithUnix]

Even at 12V, I don't think 3A is very useful range. That's 36W.

A current clamp on the 4pin/8pin EPS 12V connector would be interesting, but the voltage there should be a pretty constant 12V

The voltage out of the VRM circuitry would be interesting, but might be hard to find to tap into.

Honestly the best thing I can think of is a chassis fan. Simple to wire, low risk, and as the CPU load goes up the fan should be fed more power making the values increase.

[matts-uk]

In the sober light of morning. A fan header is probably your best bet. A 3 pin variable speed header would be ideal. The circuit below should work with a PWM but the voltage won't move around so much.

Looking at the PCB on the Amazon link, I can't make out the display driver chip number but the track running to the 'Key' button and then to ground looks like a reset line.

The Amazon description says the meter's current range is auto-ranging, 0 to 1000mA or 1A to 3A. A PC fan typically draws around 100mA to 300ma from what I recall.

http://emscom.net/images/meter.png

Ahh. No embedded images then?

[Ttaskmaster]

Thank-you guys!


I did get a bit of a result. The meters showed some values, at least, though very little movement once the initial readings had settled.

Unfortunately, I also managed to blow one of the meters because I had a dumb moment and forgot that, in this instance, the red wire goes to the black terminal... Why they don't just do colour matching - I'm sure there's a good electrickeral reason, but it's beyond my ken!

At least now I know this will work when wired correctly.

[matts-uk]

I had a dumb moment and forgot that, in this instance, the red wire goes to the black terminal...

Oops. Lucky the meter blew before the wires caught fire! Your meter was rated for ~3 Amps. A 400W PSU can put 10 times that on the 12V rail.

Why they don't just do colour matching - I'm sure there's a good electrickeral reason, but it's beyond my ken!

It would be confusing for people who have a clue

Electrical flows are similar to water, except the motive force is electromagnetic rather than gravity. If you want to measure Voltage (pressure differential) the meter is connected in parallel (+ to +, - to -). If you want to measure Amps (current) flowing through circuit the meter is connected in series ( + to -, - to +).

Similar to a water flow meter, the internal resistance of an ammeter needs to be low to prevent interference with the reading. When you connected the ammeter + to the 12V + you by-passed the resistance of the load (the fan) and created a dead short to ground through the meter. ATX PSUs contain short circuit protection but not fast enough to protect a sensitive meter from an order of magnitude difference. Boom!

At least now I know this will work when wired correctly.

You might protect the ammeter by inserting an in-line 500mA fast blow fuse into the circuit.

A much better circuit would involve separating the metering from the PSU rail with a transistor - and that opens up all sorts of Arduino driven possibilities.

Anyhow. Happy to hear you gave it a try.

[DanceswithUnix]

Perhaps next time you can rig a fan up to a weather sock and just do the metering purely mechanically

Edit: Actually, that could be pretty cool. I always fancied a case vent like the flappy exhaust cap on the top of American trucks. But with hopefully less smoke coming out.

[Xlucine]

Thank-you guys!


I did get a bit of a result. The meters showed some values, at least, though very little movement once the initial readings had settled.

Unfortunately, I also managed to blow one of the meters because I had a dumb moment and forgot that, in this instance, the red wire goes to the black terminal... Why they don't just do colour matching - I'm sure there's a good electrickeral reason, but it's beyond my ken!

At least now I know this will work when wired correctly.

No obvious blown fuse/fusable resistor on the PCB?

[Ttaskmaster]

Oops. Lucky the meter blew before the wires caught fire! Your meter was rated for ~3 Amps. A 400W PSU can put 10 times that on the 12V rail.

Yeah, was...!
Two of them do actually have reverse polarity protections. Murphy's Law decided I'd start by connecting the only one that didn't... and the blow was pretty much instantaneous. Touch-pop. I'm not wholly convinced the contact wasn't simply accidental, either, in such a tiny space with only a few mm of room for the probe ends.

Either way, I made a mistake. I feel stoopid, but I still own it.
Got to try and find a new meter with the same display layout...

If you want to measure Voltage (pressure differential) the meter is connected in parallel (+ to +, - to -).

And that's exactly why my head went down the basic "Red to Red" routine...

If you want to measure Amps (current) flowing through circuit the meter is connected in series ( + to -, - to +).

"Uh. Well now I know that... "
[/Jayne Cobb]

You might protect the ammeter by inserting an in-line 500mA fast blow fuse into the circuit.

Or just pay more attention when wiring them up next time!!

A much better circuit would involve separating the metering from the PSU rail with a transistor - and that opens up all sorts of Arduino driven possibilities.

I did start looking at Arduino stuff, but that's waaaaaaaaaaaaaay outside my area. I even got a Picadillo 35T, but couldn't get my head around even the basics of UECIDE.

Nah, most of what I want to do with this build already comes set up and built (by professionals ) on PCBs. I just have to make up the surrounds and wire it in, maybe swap some LEDs around.

[Ttaskmaster]

No obvious blown fuse/fusable resistor on the PCB?

Nothing obvious, no.
Everything is teeny-tiny surface-mounted components anyway and nothing I could manipulate with my limited skills!

I'd hoped that little red push button was a reset of some kind, but even when the meters are working it doesn't seem to do anything.