Shunt Resistor feels cheap

[pentium]

I want to install an ammeter onto a 240v 40A branch circuit so I bought this and it recommended for a shunt resistor I install this in series on one side of the phase (Remember, this is Canada so were split phase = +120/-120v) but it looks REALLY freaking cheap for something that needs to push a hell of a lot of juice through.
Mind you, I work with DC a lot so I'm used to heavy guage cable and connectors. AC wiring always throws me off because all the things you can get away with in code. IS this actually what I'm supposed to do, or is this just overrated crap from the land beyond the sea?

[eccerr0r]

Re: Shunt Resistor feels cheap

All a shunt resistor needs is to be low resistance and can dissipate the heat it generates. If it trulyis 75mV RMS and 50 amperes, it'd dissipate less than 4W. Now I can't really tell how big that thing is, but if they are #6 screws as a guidance, I can believe it can dissipate 4W fine - it's bigger than those ceramic resistors we've seen a lot.

However, I'd look for something like the clamp-on meters - using a current transformer is probably a better bet for AC. However if you have the sensitive electronics for using a hall effect sensor, it's even better...

[kaboom]

Re: Shunt Resistor feels cheap

Get one of these:
http://www.flex-core.com/split-core-...ansformers.htm

Don't forget to load it correctly (burden) for accurate readings.


If this is only a 240V load, you may use a single meter in either leg. However, if there are any 120V loads in the '240' device, you'll need two meters, one per leg.

[pentium]

Re: Shunt Resistor feels cheap

only[/b] a 240V load, you may use a single meter in either leg. However, if there are any 120V loads in the '240' device, you'll need two meters, one per leg.

Yeah, I realized that after I posted. :/



Oh well. I'll just add a second meter and draw in a line saying not to pass a certain point.

[redwire]

Re: Shunt Resistor feels cheap

Don't use a series-shunt ammeter with your electrical service! A few reasons:

The meter and shunt resistor are not properly isolated nor safety rated for the voltage, neither is CSA/UL approved. Even then, both must be housed in a proper metal cabinet. If you are trying to loop in and out of your electrical breaker panel, that would not meet Building Code on the 240V service side.

You would instead attach a current-transformer (split-core type) onto the wire of interest (right inside the panel) then run the CT leads out to your panel meter. The CT provides safety and insulation from the panel voltages. Then you need to calculate the "burden resistor"/shunt to match the CT to your panel meter.

Recommendation: don't connect anything from China to your electrical utility. There have been some house fires from the unapproved components. There's too much juice there- if anything cracks or fails it's a big boom.

Popular is to use decent (approved) CT's with a low-cost energy monitor. There still is a danger with the voltage-monitoring aspect, as line voltage goes to the monitor.

Note a panel meter will not read correctly with inductive loads (motors), as it does not compensate for power factor.

[kaboom]

Re: Shunt Resistor feels cheap

Several things.

Load diversity can mean that even if you have a 20 and 30A bkr on the A leg in the subpanel, and two 15s and two 20s on the B leg, you still aren't guaranteed to be over a certain amount. Or under, for that matter. You really need to do a load calc.

What loads are you feeding from the subpanel?

What wire size is the feeder from the service to the sub? I bet the bkr is sized for the now-too-small feeder, and there's too much in the sub.

You have to use a four-wire feeder for the sub. Run both legs thru the watthour meter, and tie the can to the grounding conductor. This is either a green wire if in pipe, or the concentric in your SER cable. The grounded conductor, neutral, runs straight thru the meter can and is landed on the neutral bus in the sub.

Why is there a kWh meter, anyway, in the sub's feeder?

Neutral and ground are only to be connected at the service.

Keep any mickey-mouse junk out of, and away from, your service, subpanel(s), feeders.

And again, you need a four wire feeder to that sub.

[pentium]

Re: Shunt Resistor feels cheap

The meter and shunt resistor are not properly isolated nor safety rated for the voltage, neither is CSA/UL approved. Even then, both must be housed in a proper metal cabinet. If you are trying to loop in and out of your electrical breaker panel, that would not meet Building Code on the 240V service side.

The idea was to enclose the shunt and ammeter circuitry in its own dedicated steel box with proper electrical conduit connecting everything together. My parents wouldn't allow it if it was all just screwed to the wall, especially my engineer father.

What loads are you feeding from the subpanel?

Mainly electronics ranging from a projection TV to a PDP-11/84.

What kind of current transformer do you recommend? ebay has these for as cheap as candy but have no clue if they would accurately work with the analog ammeter I have.

What wire size is the feeder from the service to the sub? I bet the bkr is sized for the now-too-small feeder, and there's too much in the sub.

No cable has yet been purchased. The plan was to buy that last.

Why is there a kWh meter, anyway, in the sub's feeder?

I was running Silicon Graphics systems in month-long spans last time I lived with my parents. They almost murdered me when they got a $1200 hydro bill and we were splitting the hydro 50/50 back then. I secured a meter and box and the plan is to use it to calculate how much power I'm using so they know how much of that bill is from ME.

[redwire]

Re: Shunt Resistor feels cheap

There's another bit of confusion to look at.

I realized your panel meter SCALE is 0-50A AC (linear) which means it is DC movement and designed to connect to a power-measuring unit, and act as a readout only. Probably 50mV or 1mA DC for full-scale.
Panel meters have permanent magnets and need DC to move the pointer. For AC applications, like VU meters and AC ammeters, you need a rectifier somewhere. This diode bridge voltage drop makes the scale non-linear on the low end, see pic. That's the give away if a meter has internal diodes.

Test the meter, put a resistor (a few kohm) in series with a DC power supply or 9V battery and see how much current makes it read full-scale. Flip polarity and see if the needle goes -ve. It shouldn't if it has internal diodes, which I doubt...

[eccerr0r]

Re: Shunt Resistor feels cheap

I was wonderring about this - there have been AC meters for a long time. It's possible to make a meter without permanent magnets - like they do for AC motors.

If someone wants to try the 9V, start with at least a 100Kohm to 1Mohm resistor and work your way down. Some meters are just that sensitive.

I was an idiot and used a 1.5V battery on a sensitive meter and fried it. Oops. Oh well, the meter had static problems I couldn't resolve, the static on the face would affect meter movement...

[ben7]

Re: Shunt Resistor feels cheap

I've seen one person who had a setup where a DC ammeter was the only option. He had used a bridge rectifier in series with one of the AC lines, to pass the AC current. With the DC + and - shorted out by the shunt, he could use a DC meter to measure an AC current. I have done this too before

[redwire]

Re: Shunt Resistor feels cheap

Oddly, I've never seen a bonafide AC panel meter- they must be expensive or huge if they exist. You have to generate torque for at least 90 degrees of movement, and panel meters are moving coil, so no brushes allowed to flip the field.

Using a bridge rectifier, you loose 600-1200mV which is huge with a 50mV shunt!

[eccerr0r]

Re: Shunt Resistor feels cheap

Looking at google they do exist but probably expensive only because of economies of scale. Basically a spring is used as the "magnet" and low hysteresis ferromagnetic metals are used (I'm not sure what would happen if you run DC through it, it may magnetize and "destroy" the meter until you run DC in the other direction for the exact same amount of time until it demagnetizes.)

Definitely cannot use diodes with currents through the shunt we're talking about here... Even with schottkys the losses through an appropriate shunt would be huge...

[tom66]

Re: Shunt Resistor feels cheap

If anyone wants some, I've got a massive surplus of Taehwatrans TS-10L current transformers. 85A, 3000:1. For £3 each plus postage. Much safer than using a shunt resistor and panel meter imo.

[pentium]

Re: Shunt Resistor feels cheap

BuMP.

Yeah, when the shunts come sabotaged from the factory I'll just use the other method.

[kaboom]

Re: Shunt Resistor feels cheap

BuMP.

Yeah, when the shunts come sabotaged from the factory I'll just use the other method.

If you're talking about the notch, that looks like 'resistor trimming' to me.

[pentium]

Re: Shunt Resistor feels cheap

I shouldn't be drinking this early in the afternoon.

[pentium]

Re: Shunt Resistor feels cheap

Yeah, ignore me on that last jab. The notches are part of the shunt resistor normally. You essentially get it in a spool, cut it into pieces and put terminals on either end.
Anyways, I feel slightly more safe using them after I repalced the plastic backing with bakelite. More resistant to heat and fire.


Anyways, Kaboom, here's what the whole thing looks like assembled: