Hi folks,

I have an interesting problem I need to solve as quickly as possible.

I have a 0.3kW (ish) motor operating a spear valve on a Hydro site I am at right now.
Its on an Auma actuator but is old as the hills and Auma, although helpful, can’t provide any tech data, beyond what I a about to tell you.

The unit is rated at 48VDC, nominal with a start current or around 80A but it runs at circa 15A
Obviously this varies with load but it all I can give you based on my cheap DC clamp.

The unit is shunt wound with the field, circa 4A running, and armature, circa 11A running, separately switched with a pair of huge contactors.

The armature is always connected the same way around whilst the field is reversed depending on the direction we need to go.

I currently have a freewheel diode and a 10uF cap, with a current limit resistor across the armature and a 6uF cap with much smaller current limit across the field. Both also have a 0.1uf ceramic for good measure. (values base on parts bin contents rather than any calculation or logic.

Without any suppression the EMI/RFI is sufficient to stop a PLC a metre away. Thr PLC has a common ground but a separate 24V DC supply which includes a voltage stabiliser with at least 5s ride through.
The PLC reports ‘power fail’ as the stop reason, which clearly isn’t the case.

I know what I have now is not good or even approaching right…

What I need to know is what to fit, values, and where? But I have no idea how to go about calculating that.

I have attached a drawing showing the options I think I have and would like to know which, if any, are the best option configuration wise and what values I need to build them with?

Up against it here, will have to buy a box of assorted parts and spend time fiddling if I don’t have a better idea about what to do within a day or two. Any help would be very much apriciated.

Thanks,
Al

 

Makes one wonder if PLC inputs are high Z = EMI pickup issues.

Tackling issue from that end maybe do scope captures on PLC inputs,
like reset, power rails, etc..By doping this component values for fix
should be less cost, small size......

Regards, Dana.

 

Thanks Dana, but there are other odd issues with equipment, coms problems and unexplained reboots of metering kit...
I relay need to deal with the noise at source. Besides there are circa 24 inputs, all of which, other than a frequency counter, are filtered at 3mS on the physical input. The ground rail is common but the PLC is supplied from a separate 24V DC source and also has a voltage stabiliser.

Inputs are circa 10K but contain hardware filters and are also further filtered at a firmware level, set in the input config. If that were the only issue I could go to as much as 10mS pickup delay. PLC reports last stop as power fail, which it wasn't so it has to be noise internally I guess.

Not worried RE cost, it is all relative anyway, but I doubt 24 small filters would cost less than 3 snubbers or even 3 snubbers + 3 filters.

I do have a small handheld scope here but what I relay need is a storage scope with an external trigger... Of course I would also need to know what to look for!!

It is working right now, all be it with the odd coms blip/delay, what I need to do is make it better.... There are obviously still spikes floating about which means something will die sooner or later if I dint sort it out.

Regards,
Al

 

Could be lack of earth ground issue on the P.S.'s Taking all the DC power commons to Earth GND could be the solution.
If possible take all to a central Star point.
Max.

 

Thanks Max, but pretty much done that...
Gnd / 24V-0 / 48V-0 all common in huge distribution terminal.
Cable to motor is SWA, armour grounded...

It needs contact suppression anyway, just to give some mechanical life. Seems logical to eliminate, OK vastly reduce, noise that way as well

 

A couple of PDF's.
Max.

 

Thanks Max, looks like my evening reading is sorted... Great help
The second, Snubbers.pdf, looks to be the most relevant, from an application perspective but I will try and get my head around both.

What are your thoughts RE Freewheel diode, for the armature, and either MOV or TVS with current limit for the Field.
Obviously I cant use a single diode for the Field, given the the polarity changes and I am assuming that a TVS is similar to a couple of zener's back to back. I also read somewhere that using a MOV or TVS with a current limit will dissipate energy better than a TVS alone, which I think makes sense, either way eliminating over-current as a made of failure for the TVS seems reasonable. I guess the same may be true of MOV's but I don't know how low the lowZ phase is... More reading.

I appreciate that too much capacitance would likely hinder rather than help but I cant help thinking that a belt and braces approach may be a good plan.
I am assuming I can add Diode/MOV/TVS , whichever across the load whilst still having an RC around the switch? I assume this would only help but to be honest I don't know.

What I defiantly don't know is if an RC round the switch and also an RC in parallel with the load is going to help or simply oscillate and do more harm than good, although I suspect the latter.

Layout is on my original attachment... along with what I thought were all the options, just in case you missed it.

If I were to use MOV's/TVS's how much voltage headroom, no pun intended, would I need? System is nominally 48V but I have to consider charging voltage for the batteries, float only, so I wouldn't be happy with a threshold much below 58V.

Do MOV's and/OR TVS's stay on once the threshold is reached or turn off the same level as turn on?

Sorry ... None of this is stuff I have dealt with before. Good opportunity to learn though.

Cheers,
Al

 

Does the fault occur randomly or when an event occurs, e.g. Armature/Field Contactor(s) switching etc?
I would may be do some process of elimination and try to find out where the issue originates from,
Including substituting DC with battery power temporarily etc inc PLC for e.g.
If possible I prefer this method first to covering the problem with rafts of suppression etc. although it can cure to some extent I prefer to eliminate the source if possible or at least identify the culprit.
Max..
.

 

BTW did you apply any snubber or BEMF diode across the contactor coils, I have found these to be a problem in the past.
Max.

 

Defiantly not random...
Two dirty big contactors, as per drawing, 24V couls with diodes and switched buffer relays.
Each contactor switche both field and armature, the former reversed, and they are electrically and mechanically interlocked.

Interference only occurs when switching the load... Given that the spear valve is way too fast the on time can be ridiculously short, 350 mS minimum, yes I know, but the max is only 700 mS.

I have dangly bits on it right now, values defined by parts box contents rather than any calculation, what I need is a proper plan, even if that pan involve iterative testing.


Al

PS. essentially knocked off today and just opened a very nice single malt... Less sense as the evening progresses I suspect.

 

BEMF... Back EMF?

 

Back Electro-Motive Force.
It sounds as though that is a Bang-Bang control, seems to cry out for true servo mechanism type Spear valve.
Max.

 

I wish... 1/4 the speed 1/10 the headache.
PWM in some commercial package or even a small inverter and a conventional induction motor would have been a better plan but for now it is what it is.

If I cant stabilise it, as is, I may look at doing a mechanical mod to replace the motor but that is fought with potential mechanical issues.
In theory the gearbox, torque and positional switches shouldn't care what motive force is applied but given the lack of info its still a gambol.

I did wonder about SSR's rather than relays but 80A... who knows what wouldn't pop within a week.

My PLC will do PWM, directly, but an SSR at 4khz is still going to need some serious snubbing not to fry on day one.

 

It seems Max is resident legend. Thanks Max this helped me greatly.