DMC60 calibration

[quin]

I am using a DMC60 controller with the PWM signal coming from a PLC.  I get a flashing red signal which I assume means invalid pulse width. The calibrations instructions call for using a joy stick. I tried simulating the joy stick by sending 3 successive signals: pw 1.0ms, 1.5ms and 2.0ms. The pulse height is 5volts, the period is 10ms.

Need suggestions please

[quin]

Rather guess at the exact characteristics of the transistor, here is how it operates:  When the PLC output is active, it conducts current from the 5v supply thru the load, thru the transistor to ground.  I have the option to cycle the ontime of each period to affect PWM.

The element by the red arrow is a diode.  It is there to protect the transistor when we turn off an inductive load such as a relay coil.

[quin]

Thank you Michael,  we are going to use the buffer.

Quin

[malexander]

Hi Quin,

The calculation you did to arrive at 151uA assumes infinite input impedance on the CAN transceiver inputs. The datasheet for the CAN transceiver says that differential input between the inputs can be anywhere from 10K to 100K and does not list a typical value. Additionally, it says that the common mode input for the inputs can be between 10K and 30K. For simplicity lets assume that the common mode input impedance is 10K, that the differential input impedance is also 10K, and that you've connected the black wire to the same ground potential as the CAN transceiver. The effective input impedance then becomes 5K. The minimum voltage to threshold is 0.9V. Do we exceed that voltage with 5V flowing through the 33K resistor? (5/(33+5))*5 = 0.66V. Nope we don't exceed 0.9V in this case and therefore do not meet the minimum threshold required for a logic '1'.

In your case I would strongly recommend using a buffer to create a low impedance output from your PLC. That should prevent any problems that were being caused by driver strength.

Thanks,
Michael

 

[quin]

Hi Michael,

Let me try to restate your post to make sure that I understand what is happening.

My PLC output with a PWM feature, produces a zero volt signal for 1 to 2 ms out of every 10ms period.  This 1 to 2 ms segment is the “on time”.

With 24v at node “A”:

When the PLC output, which is a sinking output, is in-active (which I call logic “0”), there is .434mA available to flow to the DMC.  When the output is active, there is zero current flowing to the DMC since node “A” is essentially at ground potential as is the black wire.

With only 5 v at node “A” and the PLC output in-active, there is only .151mA (5v/33k) available to the DMC.

I assume that .151mA difference from zero is not enough for the DMC to distinguish between the start of and the end of the off-time. Hence the “invalid input PW” error signal.

Question, since I already have observed some stuttering of the motor, would it not be wise to use the op amplifier as you suggested and thereby make sure that I always provide a more solid definition between on-time and off-time to the DMC?  I don’t want any field issues.

Thanks for your patience.

Quin

[malexander]

Hi Quin,

Based on the drawing and your oscilloscope capture approximately 424 micro amps of current ((24-10)/33000) flowing through the output when driving a logic '1'. The output impedance of the driver (~33K) is too high when driving a logic '1' and that's why it doesn't work when you don't have the 24V supply attached to node A. The output impedance when driving a logic '0' is much lower, very close to 0 ohms. In any case, driving it this way won't cause any damage to the DMC60 so you can continue to do that.

If you want to eliminate the the 24V connection to node A then you will need to place a high impedance non-inverting buffer between the output of the PLC and the white wire and power it from your 5V supply. I wouldn't bother to do that unless you observe erratic behavior while operating the DMC60 with the 24V supply connected to node A.

Thanks,
Michael

[quin]

Hi Michael,

Here is a scan of the input signal accross black and white wires.  The scan is of the neutral position, 1.5 ms. the amplitude is ~10volts.  Period is 10ms. The ontime moves smoothly to 1ms and to 2ms.  No spike visible.  Motor runs smoothly.  I did notice that sometimes when the probe is attached the motor stutters but no visible disturbance on the scan.  Will the DMC60 be bothered by noise in the field?  Do we possibly need some filtering?

I received additional information about the internal wiring of the PLC output circuit.  It is attached.  Note that there is a 33K pull up resistor in parallel with the diode.   This explains our 10 volt output.

Quin

[malexander]

Hi Quin,

We tested it all the way down to 1V with a signal generator (Agilent 3500B) and it worked fine so I don't think it has anything to do with the 5V voltage threshold. I'd love to see a scope capture of the input pins just to understand what's really going on here.

In our testing the ground of the signal generator was not tied to the ground of the DMC power supply so that shouldn't matter. I don't actually understand how the diode even conducts current in this scenario because the way you have it drawn the anode should always be at a lower voltage than the cathode, and therefore should not conduct. In any case, the CAN transceiver that the wires connect to is rated to operate in 24V systems and the inputs can tolerate that voltage so it should be fine.

Thanks,
Michael

[quin]

Michael,

I talked to the PLC company.  There is no option for a different driver for the output. They interface with your type of motor controller all the time with no problem.  I also asked the effect of having 24 volts on term "A".  They said that this in effect provides a stronger signal than having it at 5v. and that possibly your controller is marginal at 5v.

So the only way it will run is per my diagram.  I also tried connecting the 24v grd to the 5v ground and it runs ok.

My questions:  Do you see any long term problems wiring it per the diagram?  Do you prefer the grounds being common or separate?  Separate seems to me to provide some protection from power side problems affecting the PLC or the DMC.

[malexander]

Hi Quin,

What you are describing sounds like it would work well for driving a relay but I'm not sure that's going to work for driving the input of the CAN transceiver (MCP2562-E/SN) that sits between the wires and the MCU. It sounds to me like current only conducts during the on time. I don't know what the input lines are going to look like during the off time. This is definitely a time when I'd get out the oscilloscope and probe the inputs, but since you don't have access to one that's not an option. Does your PLC have the option of using a push-pull driver for the output?
 

Thanks,
Michael

[malexander]

We have MOV's as part of the H-bridge circuit so it shouldn't be necessary but it also won't hurt.

[quin]

Michael,

It is an isolated convertor.  Please mark up my drawing to show how you recommend it be wired.

Also what about the MOV?

[malexander]

Hi Quin,

Typically the black wire is connected to a ground reference and the white wire is driven with a PWM signal referenced to that ground. In this case I'd expect the ground reference to be the negative side of the 5 volt supply. From your drawing it looks like you may be using an isoalted convertor so there is no common reference. Is that correct? If so that may be why it's behaving in an odd manner.

Thanks,
Michael

[quin]

Hi Mike,

 

The attached diagram is how the DMC is wired to successfully operate the motor.

 

The PWM signal is supplied to the DMC via a sinking transistor output from the PLC.

 

Terminal “A” of the PLC was inadvertently connected to 24v to make the DMC operate.

For all conventional outputs, the PLC  terminal “A” would be connected to 5v.

 

We are very satisfied with how it operates the motor when not connected to a mechanical load and we expect it will function OK when we test at full load.

 

However we are hesitant to start using this in multiple commercial applications without an understanding of why it works (wired as it is).  Without this understanding we can not evaluate what external field conditions could cause a system failure.

 

Also is it necessary to have an MOV at the motor terminals or do you have adequate protection internal to your DMC?

 

 

 

 

[quin]

Hi Mike,

When I refered to the PLC program and a 12v motor:  This was entirely different project where the PWM signal vaaried the on time of the power transistor and thereby controlled the motor speed.  Your DMC60 was not involved.  I was just trying to illustrate that the PLC program functions correctly.  This program also has a ramp feature for start and stop.  When I run it with your DMC60, the motor is very erratic, trying to stop and start on the ramp up and ramp down.  Even at full speed it is very erratic and sometimes the DMC shows a over temp fault.  There is no load on the motor on any of the above tests.

I have another DMC60 and will try it tomorrow pm and let you know results.

[malexander]

Hi Quin,

It should work with a 24 volt motor/supply. When you say "This PLC program works fine with a 12v motor" is that with the DMC60 connected to a 12 volt supply and the motor connected directly to the M+ and M- leads of the DMC60? I'm a bit confused about your last sentence... the DMC60 has a full bridge inside of it and should pass all power provided to your motor. There shouldn't be any need for an external power transistor.

Thanks,
Michael

[quin]

Thanks Jon,

With a period of 15ms and a pw of 1.5ms, all 4 lights flutter rapidly with multiple colors.  I forgot to mention that I am running a 3/8 hp brushed DC motor with a 24 volt power supply.  Will the controller handle this.  Wiring is correct. Don't have access to a scope.  This PLC program works fine with a 12v motor.  The PWM signal is used to cycle a power tranistor in the ground leg of the motor.

[jpeyron]

Hi @quin,

I reached out to one of our design engineers about this thread and they responded:

"If they are going to drive it from a PLC where they have absolute control of the output duty cycles should not need to calibrate. Instead, they should try sending a continuous signal with a 10-20ms period, and a positive duty cycle of 1.5ms just to see if it changes the LED state. Ask them to make sure that they connected their signal source to the DMC60 correctly. It needs to have ground connected to the black wire and signal to the white wire. If that still doesn’t work then I suggest using an oscilloscope to capture the signal and make sure that It looks correct."

thank you,

Jon