Measure and display approx 1,100 DC amps

I need to measure and display on a 2004 LCD the current on a spot welder up to almost 1,500 amps. The output voltage is only approximately 2.5VDC. I have not seen any kind of sensor that would even come close to working. I'm using a UNO R3, ATmega328P-PU to control the spot welder and would like to use it with a display for the current. I'm using #2AWG welding cable for my leads and thinking if I measure the voltage drop across about 18 inches of one lead I should be able to get about 200 millivolts. I will have to do some testing and measuring to get the exact length but it looks like 18 inches is going to be close. My plan is to have it at a point I get exactly 200mV, that way if all else fails, I've found a digital meter I can use and it uses 200mV for full scale reading. My thoughts are, if I filtered that 200mV with a LC filter I might be able to feed it directly to an Analog input. Since this is a spot welder and I will be reading directly off one of the leads, I'm sure there is going to be some major spikes during the weld pulse and will need to be filtered pretty good. One of my concerns is, since the Analog inputs are 0-5VDC, is 200mV enough signal or will I need to look for a way to amplify it. This will lead to another problem, I have pretty much been out of the electronics world for 25 years, since I retired (yea, I'm and old fart) and while I do still remember a lot of the basic stuff, I not the sharpest knife in the drawer on the newer stuff. After all. I was still working on a lot of vacuum tube equipment when I retired. The next major hurtle is I'm familiar with the Arduino UNO R3, I've been using it as a PID to control my coffee bean roaster for a couple of years, I'm not much good on programming.

So, Is there anyone able and willing to help, or is this one of those you wouldn't touch with a 10 foot pole?

Oops, a bit of a typo error and I don't see a way to edit a post. I don't think the voltage straight off the secondary of many transformers is going to be DC. I need to change those DC comments to AC and I know that means I'll have to take steps to make my 200mA AC a 200mA DC signal.

If it's AC then it becomes a little easier (maybe) Use a conventional CT to convert your 1500A via a 5000:1 transformer to 0.3A. Drive this current through a 10ohm burden resistor to generate 0-3V AC. From there use a precision rectifier to produce a 0-3volt DC voltage. After smoothing you'll get near enough a 0-5V DC signal for your Arduino.

Alternatively measuring the mains input current (circa 15A) might be a lot easier using readily available hall sensor devices.

Don't use an LC filter, that's a resonant circuit if not loaded, use an RC filter. So long as the R value is high enough you have protected your Arduino - but you will need to keep the sensitive electronics well away from the high current path (a metre or more if you possibly can). A hall-effect sensor might be a better way (electrically isolated).

Yea, it's looking like if I want to display my weld current, it's going to be that separate meter. Not sure why I had my head up my butt thinking I would have a DC signal. I guess because I was using my bench power supply playing with different methods of how I was going get and measure the signal voltage and that was all DC. The meter I was looking at used a DC input, so now I've got to my shunt voltage/current to a filtered 200mV DC. I guess I can rectify it just like any AC, using the end with zero potential as my negative and the end with my 200mV as the positive with some type of rec between them, even if they are still reading the same piece of wire. Never tried that so that's one I will have to play with.

Anyway, thanks for the suggestions. I really don't care about reading the input current with a Hall devise unless there is a formula I can use in the Arduino that would read the input and convert it to display it as my output current. My concern using a separate meter is, I'm only looking at a 100ms weld pulse and that's going to barely be a blink on a meter. Using the Arduino, I'm thinking I can keep it one the display longer and even have the option to press a button to manually reset the display.

At such large currents, a linear hall sensor close to the wire might be sufficient. Magnetic field at 1 cm from 1000A wire is around 200gauss

A linear 2.5mv / gauss hall sensor will output +/-500mV see page 3 "Measuring High Currents" examples here: Allegro PDF

If you know the turns ratio of the mains transformer you could measure the primary current and infer the secondary.. higher volts - so you'd need isolation - but much lower current.

regards

Allan

It would probably pretty easy to calculate the primary. The secondary is only three turns of 2AWG welding cable with approximately 12 inch leads. Total cable length with turns through core is approximately four feet. The transformer is not powered on until the secondary is shorted together through two 1/4" copper rods turned down to .060" at the tip onto a zinc or zinc plated metal tab so you are probably looking at maybe 0.5VAC but about 1100 amps, and probably 2.5VAC no load if it were to ever be powered without the tips shorted. I just got my parts from Mouser and building the control board now, and hopefully can get some exact data when I can actually make some test welds with it. Problem is, I recently had basil joint surgery on my left hand (and I’m left handed) and it’s in a hard cast with my wrist and most of my hand immobilized, and soldering all those small components is proving to be no fun at all. I couldn’t even work on it today because trying to do too much last night, has my hand swollen and hurting.

I’ve looked for CT’s but finding one is proving difficult. they are either for a lot less current, or cost more than I would plan to spend if I wanted to build 20 of these things. Even thought about getting a ferrite toroid core and making one but that many turns through a doughnut is not my kind of fun.

Even considering measuring the primary and trying to do some funky math that would give me something that looked like a secondary current.

I’m also giving serious thoughts to saying p**s on it because I don’t really need it, just thought the thing would look better with a meter in it.

Make your own CT....

Wind eg ten turns of insulated wire round the wire carrying the current - input or output - and measure it's volts.

Calibrate from there.

You could use a clamp type current meter as your reference

regards

Allan

Allan, thanks. I was going to try that when I get it working but since I'm going to have way less than 1 volt to work with, I was thinking it's probably going to a pretty healthy number of turns to get me enough voltage for the Arduino's analog input. Will even look at using a voltage doubler or trippler if it looks like I can get enough of a signal to work with.

Even thinking about ordering one of those off ebay for a couple bucks that will go around the cable and rated for a lot less current, then see if I can use some math or additional amplification and make one of those work. I think when I get the thing together and can start doing some actual testing, I can use of a little redneck engineering and get a signal I can use, just figuring out what to do with it in the Arduino will be my major challenge then. I'm not too swift on the code for those things.

allanhurst: Make your own CT....

Wind eg ten turns of insulated wire round the wire carrying the current - input or output - and measure it's volts.

That's not a transformer. The primary and secondary have to share a flux path.

For $15 I found a 1000 Amp hand held clamp on meter from China on ebay. I will order that and see if the internals will give me everything I need to make a meter for my spot welder. If not, I've pissed away way more than that on a lot of projects that didn't turn out. If it does, then I will get off dirt cheap.

Hi MarkT...

That's not a transformer. The primary and secondary have to share a flux path.

They do - the wire it's wrapped round... else clamp meters wouldn't work

You have a 1-turn primary -whatever path it takes (see Green's theorem)

regards Allan