Current Meter Recommendations

[am1]

Hi all,

I have been working on a project that requires me to measure the current across a 90 VDC motor of a medical device (similar to an elliptical) to distinguish between machine and user work. Currently I have been doing this with an AC current transducer (Automation Direct ACT050-42L-F) and an I/O link analog display for my PLC to read (IFM DP2200). I have been having a very difficult time with data collection/predictive modeling with the given system I have as there is a lot of noise in the data. For example if I start my device I will see a 6.40mA on my analog display. If I leave it running for 10 minutes and check again I will see 5.90 mA. Or if I stop the machine and start it a few hours later without changing any parameters/settings (speed, step length, etc.) then I will see a different reading, hence I have no reliable "baseline" to reference.

I spoke with Automation Direct on recommendations to add to the system design or circuitry to improve this/reduce noise; they believe the current transducer being used may be contributing to the noise due to the sensitivity of the inductance (with temperature, cable jacket, mounting, etc.) as these work like fluke meters. I was recommended to use a current meter or something similar for more sensitive/stable readings.

This brings me to my question, has anyone dealt with a similar issue with their current transducers? Almost all transducers have this style, so does anyone have any recommendations on alternatives? I found this AE001Current sensor with IO-Link (https://autosen.com/en/Condition-Monitoring/AC-current-sensors/Current-sensor-with-IO-Link-AE001) that is similar but seems to clamp down on the wire and had IO-link so an additional sensor wouldn't be necessary, however I am not sure how much of an improvement this would make. Any advise is greatly appreciated.

[Jeroen3]

How is this motor controlled?

[moffy]

"I have been working on a project that requires me to measure the current across a 90 VDC motor..."
If it is a DC motor how are you measuring the supposedly DC current through the motor, I assume that is what 'across' means, with an AC current sensor?

[am1]

I have adopted the project from another group, so have been using their selected components/circuitry without questioning it up until this point, which was a mistake on my part. The motor is on a DC drive (attached), and the transducer is on one of the motor cable wires (A1 terminal) which is for the motor's armature voltage (which is 90 VDC according to the motor datasheet), so this is likely one of my issues since its a AC transducer...

I spoke with the original designer for the system and they said that the transducer should be used to measure the input (AC) side and selected it because they have never used DC transducers for speed control motors. I am new to working with transducers, but would switching the transducer to the L1 wire change anything with the output I see currently from the sensor?

Or is the best measuring tool to achieve a more sensitive, stable current reading going to be a DC current transducer on the A1 terminal? (like https://cdn.automationdirect.com/static/specs/acuampdcs100.pdf)

[moffy]

Using an AC current transducer on L1 is better than on A1, but you will also be measuring the power consumption of the controller and depending on efficiency that could introduce a significant error. You would get better results using a DC current transducer on A1.

[am1]

Using an AC current transducer on L1 is better than on A1, but you will also be measuring the power consumption of the controller and depending on efficiency that could introduce a significant error. You would get better results using a DC current transducer on A1.

I made the switch of using the AC CT on L1 instead of A1, I am seeing slightly better/stable measurements from my transducer. I am getting a DC transducer to test out as well, and hopefully will see better results.

[moffy]

Hope it all works out well, you should at least get more reasonable readings, and thanks for letting us know your progress.

[am1]

I got a DC CT from automation direct (https://www.automationdirect.com/adc/shopping/catalog/sensors_-z-_encoders/current_-a-_voltage_sensors_(ac_-a-_dc)/dc_current_transducers/dct100-42-24-f) to try out, but the lowest sensing range is 0-50A which is a bit high, but I will see how it does.

I want to test it out but likely have a simple circuitry issue. I wired it as shown in the attached image and my sensor display was unresponsive. My wiring for the AC CT was correct; my AC CT only had a +/- output terminal, and the DC has +/- Power and +/- Output terminals. The wiring from analog sensor to PLC had remained the same since it worked well with my AC CT, so I would guess my wiring from the analog sensor to the DC CT has an issue. What have I mixed up? Thanks in advance!

For reference I am using an IFM DP2200 analog display and and X20DS4387 IO-link module with my DCT100-42-24-F and ACT050-42L-F. 1081 is a +24V supply from PSU, and 1090 is GND/-V from PSU.

[PCB.Wiz]

Hi all,

I have been working on a project that requires me to measure the current across a 90 VDC motor of a medical device (similar to an elliptical) to distinguish between machine and user work. Currently I have been doing this with an AC current transducer (Automation Direct ACT050-42L-F) and an I/O link analog display for my PLC to read (IFM DP2200). I have been having a very difficult time with data collection/predictive modeling with the given system I have as there is a lot of noise in the data. For example if I start my device I will see a 6.40mA on my analog display. If I leave it running for 10 minutes and check again I will see 5.90 mA. Or if I stop the machine and start it a few hours later without changing any parameters/settings (speed, step length, etc.) then I will see a different reading, hence I have no reliable "baseline" to reference.

Those are not large changes, are you sure the sensor is really the problem ? Have you verified with a better current reading method ?
A motor will start cold, then warm up, so the lubrication and winding resistance will change, even if the load does not.

[am1]

Those are not large changes, are you sure the sensor is really the problem ? Have you verified with a better current reading method ?
A motor will start cold, then warm up, so the lubrication and winding resistance will change, even if the load does not.

This was a general example, but yes I will will see a 0.5-1.5 mA difference as device runtime changes. I understand that there is a degree of natural variability, but I wonder how I can work around this and/or if there is a better metric I should be assessing for more repeatable/predictable outputs to make differentiating system vs user more clear? The system is a medical device that have footplates similar to an elliptical on a linkage system. The motor moves the footplates, so as you pass through a revolution, the current reading will change as it cycles. Currently, I am detecting patient effort/work by assessing the current drop (if work is done, the current will drop). Ex: when the device is running and a patient is passive, the current is in a range of 7-9 mA, when the patient is actively doing work, I would see a current range of 5.5-7.5 range as it passes through its cycle. But this value/range will shift with some tolerance based on how warmed up the motor is, so I would ideally like a more predictable metric to assess.

I am trying to switch from a AC CT to a DC CT, but am facing the circuitry issue I posted in my last post. Any input and advice is greatly appreciated!

[moffy]

You could try reversing the direction of the cable through the DC current sensor, they are direction sensitive.

[robert.rozee]

hi am1,
    i'm not 100% sure what you are trying to achieve, but you may like to have a look at isolated hall-effect current sensors made by Allegro Microsystems:
https://www.allegromicro.com/en/products/sense/current-sensor-ics/fifty-to-two-hundred-amp-integrated-conductor-sensor-ics/acs780

a couple of options that appear on ebay are:
https://www.ebay.com/itm/233508445684  (a module using the ACS780)
and
https://www.ebay.com/itm/295452714079  (a bare device, the cheaper ACS756).

these devices are pretty simple to use. they require a 5v supply (Vcc) and have an output pin that normally sits at Vcc/2 when no current is flowing through the sense pins. a positive current flow will then cause the output pin to linearly rise above Vcc/2, while a negative current flow will cause the output pin to linearly drop below Vcc/2.

i'd suggest you start out hooking up the output to an oscilloscope so that you can see the general shape of the waveform output your motor/generator is creating, before then creating some software to process the data into something meaningful.


cheers,
rob   :-)

[ejeffrey]

DC hall effect current sensors are just kind of intrinsically drifty and low accuracy at low current.  They are also sensitive to orientation and background field.  The allegro ones are a bit better than non contact sensors.  However if you are going to interrupt the signal wire, why not use a shunt resistor and instrumentation amplifier (or just a voltmeter)?  What is the maximum current you need to sense?

[am1]

I am using a MTPM-001-1L18 motor and MGC403-11 motor driver. The F.L Amps is 10.4A although we likely are not close to that (max rpm we run motor at is 40 rpm and not ran at full loads).

Thank you all for the recommendations. I have tried reversing the cable direction without luck. When I was working with the AC CT, I would see the screen display ~4.00-4.05 mA when the machine was turned on and motor was not running, then when I start the motor I would see it update with live current draw. After swapping to DC, the display remains off when powering the device and when running the motor. I am open to trying different sensing options, but I would like to get this sensor in to make some initial comparisons between the AC and DC CT.

Since the display-to-PLC wiring should be fine, should I be using a different wire lead that I am unaware of for display-to-transducer wiring? Thanks again everyone.

[robert.rozee]

DC hall effect current sensors are just kind of intrinsically drifty and low accuracy at low current.  They are also sensitive to orientation and background field.  The allegro ones are a bit better than non contact sensors.  However if you are going to interrupt the signal wire, why not use a shunt resistor and instrumentation amplifier (or just a voltmeter)?  What is the maximum current you need to sense?

the main reason for the Allegro sensors is galvanic isolation; i am not too sure that 'am1' has too much expertise in working with the voltages involved (115V or 230V AC and 90V DC). given this is a medical application there is wisdom in ensuring that measuring equipment is kept well and truly isolated from these voltages.

the next reason is that the waveforms being measured, at least at the DC motor end, are likely to be complicated. hence my suggestion of using an oscilloscope for display so that 'am1' can get some initial feel for what is being measured. an AC current transformer at the motor is likely to miss the low-frequency stuff (a person operating pedals attached to the motor through gears), while measuring the line current is 'miles away' from where the events of interest are happening.

as for the accuracy of the Allegro sensors, from the datasheet i see they are typically within +/-2%. remember that the current conductor and hall sensor are extremely close together within the device package, unlike a wire floating around within the core of a CT. i see accuracy being the least of the problems that 'am1' needs to get past.


cheers,
rob   :-)

[am1]

Using the AC CT to measure the line current has worked ~alright~, I am able to detect current drop when effort is applied in most scenarios, but having more sensitivity and clarity in my readings would be great, especially to pick up on more "lighter effort" levels of work being put in. Although I will still have low accuracy, I would like to get my DC sensor to work in my system to see how my output readings will vary. Does anyone have an idea on what I need to change in my wiring from my AC to DC wiring in the images from my previous posts?

I have done testing using an oscilloscope to see the system natural behavior when running empty and when no work or effort is applied by user to the system (roughly sinusoidal). And observe the dip in the current when effort is applied (clear drop in current or slightly lower current values with a smaller amplitude, based on amount of work done).

an AC current transformer at the motor is likely to miss the low-frequency stuff (a person operating pedals attached to the motor through gears), while measuring the line current is 'miles away' from where the events of interest are happening.

You described my issue perfectly. I have not used anything like the Allegro sensors before, so I am not sure how/where these would go in my system to get a more isolated reading. I want to isolate/ignore the "noise" (the systems natural behavior) and only look at what is being done to the system. I will look into these some more, in the meantime if anyone has input on the DC CT circuitry that would be a great help as that is the roadblock I am at currently.

[am1]

I just tried to wire my DC CT like my AC CT, and removed the +/- Power pins, and my analog display read "OL" open loop and started to blink "SC1". I wired my +24V and GND wires back to the PSU terminal blocks and get the blank screen again on my display. Should I be using my PLC module for the CT power (brown wire) and the PLC module ground for the CT ground (black wire)? Then what would be my output + and - wires? I know this is likely a simple mix up but help is appreciated.

[moffy]

I think you should power the DC current sensor from the IFM DP2200, L+ and L- supplies, the +ve output of the DC current sensor, according to the IFM manual, should go to +24V or L+ and the DC current sensor analogue output (-) should go to the IFM DP2200 analogue input.
P.S. There is a possibilty that the +ve output of the current sensor goes to the analogue input of the IFM DP2200 with -ve output not connected (this assumes power comes from the IFM DP2200) but the datasheet for the DCT100 is vague, you should contact the manufacturer for clarification.

[am1]

I think you should power the DC current sensor from the IFM DP2200, L+ and L- supplies, the +ve output of the DC current sensor, according to the IFM manual, should go to +24V or L+ and the DC current sensor analogue output (-) should go to the IFM DP2200 analogue input.
P.S. There is a possibility that the +ve output of the current sensor goes to the analogue input of the IFM DP2200 with -ve output not connected (this assumes power comes from the IFM DP2200) but the datasheet for the DCT100 is vague, you should contact the manufacturer for clarification.

I have still been unsuccessful with this transducer. I get a "nPrb" display on my display which is for no sensor recognized. I attached the wiring diagram that both the CT sensor manufacturer and the analog display manufacturer confirmed were good. I can confirm that the DCT is receiving +24V and I have tried reversing cable direction. I also tried breaking the circuit to test the output on the sensor. I have +Power, -Power, and -OUT connected to the DCT like shown in the diagram, and connected my multimeter in series between the white wire(4-20mA) and the +OUT terminal,  I get a reading of 1.6 mA DC. Have I received a faulty sensor if my outputs are not in the 4-20 range? Thanks in advance.

[moffy]

If as you have said you have confirmed the correct wiring with the sensor manufacturer it could well be faulty. I am sure the manufacturer can advise you how to test the current sensor.