AC ammeter for measuring VDF output current 4

[abbas2005]

Hi everybody;

There are several 400V, 3Ph, induction motors equipped with Variable Speed Drive (VFD) mounted in LV MCC room. The output frequency in normal operation of the motors will be about 20Hz~50Hz. I want to buy/finalize Ammeter to be mounted in Local Control Station (LCS) of the motors. As the ammeter will measure the output current of the VFD's, I need to have good accuracy in 20Hz~50Hz frequency range. Could you possibly explain how can i achieve this.

Thanks

 

[davidbeach]

A remote display from the VFD might provide a more useful indication; power might be more meaningful measure than amps.

 

[Skogsgurra]

If there are several motors on each VFD, you will need separate transducers - one for each motor.

If there is one VFD for each motor, the best way is to use an analog output on the VFD and use that signal (you can usually have 0-10 V or 4-20 mA) for the instruments on the local Control station. The built-in measurement works very well, has a high accuracy and works across all the frequency range, down to 0 Hz.

If you prefer to measure with separate transducers, you can use ordinary current transformers. They work quite well down to 20 Hz, but you should probably use CT:s that are designed for around twice the burden you would choose for a 50 Hz application. This is to avoid possible saturation at the lower frequency range.

You could also use a LEM (Hall Effect) or similar module. They work all the way down to DC.

Rogowski coils may be used but they are better at higher frequenies and you have to pay attention to the lower band-width limit.

A shunt resistor will not work so well, mostly because there is a lot of EMI that is difficult to handle even with the best isolation amplifiers.



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jraef]

I agree with Gunnar (Skogsgura), if you already have a VFD and one motor on it, the output current measurement it provides you will be more accurate than any external metering system you can buy (or at least afford). 99% of modern VFDs now on the market provide a feature called Sensorless Vector Control and although that feature lacks official definition and can be done in many not so honest ways, even the cheapest of them require very accurate current measurement inside of the VFD to make it happen, and it's a very simple internal process to place that information in reach to the user either through a prorrammable analog output, or via a communications port like Ethernet or some other Field Bus protocol. If you are using a VFD that does not provide any of that functionality, that means it was either purchased as cheaply as possible (meaning accuracy is irrelevant to someone in your organization), or too old and should probably be upgraded anyway.

But, as Gunnar also said, if you have multiple motors behind one VFD and you want accurate individual current values, that's a different scenario and his advice is sound (as per usual). If you are not adept at integrating something like an LEM or Rogowski Coil, you may want to buy them as "current transducers" rather than raw components. So that's an advantage if just using CTs since it's likely you are already familiar with them.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[ScottyUK]

Beware that some LEM transducers don't like being mounted on conductors which rapidly change potential relative to earth. I'd like to think LEM have got this sorted by now, but on some earlier (1990's / early 2000's vintage) types it was necessary to put an earthed electrostatic shield around the primary conductor otherwise dv/dt coupling superimposed a voltage-based signal into the transducer output signal. I spent a few puzzled hours trying to work out how a circuit with a very large inductance could have such a high di/dt as the transducer was displaying before realising that the apparent behaviour was impossible and later proving the transducer was at fault using one of the expensive Tek current probes as a reference.

 

[itsmoked]

Most of the drives I've taken apart actually have LEM transducers in them so I suspect they may have gotten them sorted Scotty.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[abbas2005]

Many thanks for all explanations.

Kindly note, there is one motor on each VFD.
As per project specifications, Local Control Station (LCS) will have Start/Stop push buttons, Local/Remote selector switch and an AC ammeter without any power supply. One multi-core cable from LCS to LV MCC has been considered.
I think for current indication from VFD output, a separate cable and also power supply is needed.
So it is preferred to achieve good accuracy by use of a Current Transformer and AC ammeter.

 

[itsmoked]

And, we're trying to tell you that, specifically, that is bad idea that will be not very accurate.

There are loop instruments that use 4 to 20mA as the signal and you can get LCD meters that use the 0~4mA to power themselves. So separate cables, use the VFD's analog output in 4-20mA mode and use a handful of electronic LCD meters.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[itsmoked]

Here's an example:

Ebay 2 Loop powered meters

Keith Cress
kcress -

http://www.flaminsystems.com

 

[GroovyGuy]

1) Does your ASD not have a 4-20mA analog output that can be configured for current?
2) Does your ASD not have an integral display that indicates current?


"I have not failed. I've just found 10,000 ways that won't work." Thomas Alva Edison (1847-1931)

 

[ScottyUK]

This is what happens when a contractor blindly follows an ill-conceived project specification. The client issues a poor specification, and the contractor misses the unreasonable requirement at the time of bidding. Then the unreasonable requirement becomes legally binding in the contract.

You can't get decent accuracy using a CT on a VFD output, and even if you could get a suitable CT the VFD output isn't a sinusoid so the moving coil meter will mis-read.
The control cable isn't really suited to a 4-20mA signal, although if you are using 24V DC control you might be OK. 110V AC or 230V AC control sharing a multicore with instrument signals is very bad practice.

Your choices seem to be either run in a suitable instrument or comms cable, or request a deviation to the spec to omit the field ammeter. Why do you want a field ammeter anyway? Motor current is a lousy indicator of what is happening in the process, and the Mk. I eyeball can determine whether or not the motor is running.

 

[LionelHutz]

If you just want to meet a spec, then put a 5A secondary CT on one of the phases and drive a 5A analog meter with it.

Otherwise, get an analog meter that uses a 4-20mA input with the scale you want and drive it with the analog output from the VFD. You'd just need to run a shielded 2-wire conductor to the local control panel for the 4-20mA.

 

[ScottyUK]

Ha-ha, that made me chuckle. It will be as much use as a chocolate frying pan, but it meets the spec.

 

[jraef]

Running the CT output wires themselves to a remote station runs the risk of what happens if a wire or terminal breaks. Open circuit CTs are VERY dangerous.
It's one of those "Just because you can doesn't mean you should" scenarios.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[Skogsgurra]

Scotty, you may have missed that motor current is as sinusoidal as you could wish. It is voltage that cannot be measured without a low pass filter.

A CT works, but will probably saturate at low frequencies. That's why I recommended a CT with about twice the design burden. The OP may not have a possibility, or budget, to put an extra cable from VFD to operator panel. That would otherwise be the simplest, cheaoest and most accurate solution.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ScottyUK]

Good catch Gunnar. Thanks.

 

[jraef]

I would use something like this. If the circuit is already installed, you can use the split core version that allows you to clamp it around wires without removing them.

http://www.nktechnologies.com/ac-current-transducers/atr-current-transducers/

Then take the 4-20mA output and run it to a loop powered meter like what Keith said, or even just get an analog meter with a 20mA input and have a custom scale made that says 0-100% current or the actual current range based on the motor. Digital meters tend to be more rugged however. I've used these little CUB4 meters from Red Lian, they are small, NEMA 4X (watertight) and very rugged.

http://www.redlion.net/product/cub4-loop-powered-process-meter-reflective-display

"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[ScottyUK]

For something a little more 'industrial' and ready-mounted -

http://www.beka.co.uk/loop_powered_indicators.html

 

[jraef]

Good point. At 400V 50Hz the OP is likely closer to your world than mine...


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[abbas2005]

Many thanks for all explanations,
There is conflict between contract obligations, budget limitation and technical requirements.
So i will request a deviation to the specification, to omit CT and the field ammeter or extra cost for high accurate local metering via VFD output.