Magnetic Field Sensors - Premature Failure

[StarCat]

Any Electronics Gurus, I can use some help on this one.
I have a machine that is eating up Mag Field sensors in one to two positions.
They are FESTO SMT-8M-A-NS-24V-E-0,3-M8D.
This is a 3 wire 24VDC device.
Brown lead is PSU+
Blue Lead is PSU-
Black Lead is switched leg-
They mostly serve as PLC inputs.

The last thing I did is roll through the control panel and tighten up some loose terminals. There was not enough time to take note of the loose ones being on the 24 VDC system, but this will be checked later. It was on a Saturday after being called in 2x while on 4 days so called vacation time.
This MFG [Festo] is known to have serious quality issues and another brand sensor will be tried in the short term.
As far as the history of this machine goes, we have had occaisonal failures of those sensors, but it has not been that frequent.
As a side note chaging away from that Brand of Sensor on another machine on the line stopped all failures cold, in another case when we were having failures that were very frequent.
Any assistance is helpful.

 

[gar]

190225-2223 EST

StarCat:

I suspect bad product, but it might be a misapplication.

I don't think you have a magnetic field sensor, but rather a proximity sensor.

See https://www.festo.com/net/en-gb_gb/SupportPortal/default.aspx?tab=2&q=574334 for the datasheet.

Make sure that the device is operated adequately within its specifications. You do not want to drive an unclamped inductive load. Look at your wiring for shorts, stay well within the current capability, and make sure leakage current is not causing a false output signal.

In to what state does the prox switch fault?

.

 

[__dan]

The prox switch is an extremely reliable device. The transistorized switching circuits used within their ratings can last forever.

I would lean to physical mechanical damage like impact damage or shock impact vibration from the pneumatic cylinder throwing. The three wires are 24 ga. If there's no visible impact damage, I would suspect maybe shock movement of the cylinder when it throws is eventually breaking the field wiring at the sensor. It may be possible to suppress mechanical shock vibration of the cylinder and wiring, or changing to a different brand prox.

The transistors are fragile if you can get a high enough voltage transient across them, but that's generally not what I suspect for the cause when they fail. I suspect they may be built cheaply to be a throw away device.

 

[StarCat]

Thanks

Thanks

The prox switch is an extremely reliable device. The transistorized switching circuits used within their ratings can last forever.

I would lean to physical mechanical damage like impact damage or shock impact vibration from the pneumatic cylinder throwing. The three wires are 24 ga. If there's no visible impact damage, I would suspect maybe shock movement of the cylinder when it throws is eventually breaking the field wiring at the sensor. It may be possible to suppress mechanical shock vibration of the cylinder and wiring, or changing to a different brand prox.

The transistors are fragile if you can get a high enough voltage transient across them, but that's generally not what I suspect for the cause when they fail. I suspect they may be built cheaply to be a throw away device.

Thanks for the replies guys. They are definitely MAG FIELD sensors and they are OEM on that piece of equipment, so they are the same part number device that came on the machine from the factory. They switch Negative 24VDC to various PLC inputs.
Service notes as follows:
A rash of fast failures starting around Dec.2018
Most failures will show output light ON, but will not switch power to the output lead.
One failure with both lights on, which is " Out of Position " indication even though "not " out of position.
Speaking to the OEM he is telling me that they can fracture if overtightened in the slot they are locked into.
Looking at the package, its a composite structure that is apparently glued or fused together.
I proposed painting them with clear nail polish yesterday on the brewmasters comments.
This brand showed a high failure rate on another machine and that problem was solved by changing to the " Numatics " brand in that case.
There are 2 obvious design changes that have changed the device from what they were a few years back.
They are in wet location subject to being wetted with both Beer, Water and CIP chemcals to a degree, but this has always been the case.
I have quality and reliability issues with several other products made by the same outfit, and I find the quality to be low and the price high.
I have been and will be moving away from using products made by that supplier whenever possible due to continual seeming degradation of the product and unreliability.
There is another brand sensor which costs about $90.00 USD each. This is roughly 40 dollars more pricey than the original equipment, but worth trying in this case.

 

[gar]

190226-1003 EST

StarCat:

Is this device sensing an external magnetic? If so, then I would agree with calling it a magnetic sensor. If it uses an internal magnet, then a prox sensor is a better description.

I really don't know what device you have. It is too hard to track their part number to a circuit, but it appears to be a PNP output switching the + side. Thus, when ON the output lead is connected the the + supply, and when OFF the output lead is roughly an open circuit.

A good device should read little conductivity (high resistance) between + and output when no power is connected to + and -. The same will be true of a defective device that still has a good PNP output transistor. A defective device with a bad output transistor (shorted collector to emitter) would imply an externally electrically caused problem.

If the output transistor is not shorted, then most likely an internal circuit failure. This could come from overvoltage to + and -, or any of many other factors.

.

 

[milldrone]

Speaking to the OEM he is telling me that they can fracture if overtightened in the slot they are locked into.
.

The spec sheet indicates .6 newton meters 1 newton meter is .78 foot pounds , less than what you can muster with a small handled screwdriver

 

[__dan]

I have quality and reliability issues with several other products made by the same outfit, and I find the quality to be low and the price high.

Festo has a pneumatic cylinder set up like a linear slide. The seal travels side for the length instead of a rod through the end.

I wanted to know how they did that and there was a fully stocked shelf of the failed take outs, so I took one apart. It must have been inspired by their marketing dept. Then they get spec'd into factory engineered automation packages and locked in for replacement parts.

 

[StarCat]

Festo - An " Interesting company no doubt

Festo - An " Interesting company no doubt

Festo has a pneumatic cylinder set up like a linear slide. The seal travels side for the length instead of a rod through the end.

I wanted to know how they did that and there was a fully stocked shelf of the failed take outs, so I took one apart. It must have been inspired by their marketing dept. Then they get spec'd into factory engineered automation packages and locked in for replacement parts.

Gar, there are internal permanent magnets in the air cylinders that trigger the sensor at either end of the full sweep. They are locked into a slot and adjusted to center of ideal sweet spot. The cylinders are pricey too. These lines have been evolving since they came out. The newer ones have some improvements, gleaned by experience with what works less that ideal. Keeping the bearings on the Seaming mechanism for example is an exercise in adaptation after taking optical readings of the finished can seam. Its all quite abstract.

 

[gar]

180227-1407 EST

StarCat:

Your original description of this being a magnetic field detector is correct in contrast to what Festo classifies it as.

What positioning accuracy is required in your application? Is cost really the most important decision factor or is reliability of operation possibly more important?

.

 

[__dan]

If it's a cylinder mounted sensor for the magnet on the piston, those I would expect are (Festo) reed switches. I did not check closely the data sheet for your device. The Festo reed switch (cylinder mounted sensor) I've seen some early repeat failure. I've changed those that would break often enough. Not physically impact damaged but failed from mechanical damage of some kind.

You know it never happens, but that entire automated plant, maybe 4 - 5 years old now, is so poorly built, anti-engineered is the word I used, the owner has come to the same conclusion and is going to repurpose that space. They're building new from the ground up another automated warehouse to replace it, about four miles away and four miles closer to me, going up new this year.

 

[__dan]

Data sheet says prox switch but it is still possible it's a reed switch. Why I would ignore the data sheet and make my own guess, I have no clue.

From the picture, I think I have personally changed that same device repeatedly. Buy a bag of 10 -20 at least and expect to keep changing them.

I think we may have met before, ChemE on the old CS board.

 

[gar]

190227-2142 EST

__dan:

SMT devices are listed as "Magneto-resistive", and thus include electronics, and a solid-state output device (like a PNP transistor).

.

 

[__dan]

Yes, but is the switching action solid state or a mechanical reed switch device. The data sheet gives no clue I can see. This type I have seen described as a reed switch elsewhere, it's what I would expect or may be common.

Magnetic prox is a different device and switches on the presence of iron. All the data sheet says is prox.

Output of the sensor to the PLC can be transistorized. Usually they have a led for output state. I don't know if they run the reed switch bare to the PLC, just that they're described that way.

 

[__dan]

I see both types listed, magnetoresistive and reed. It's possible changing from one switching action to another could improve reliability. I did not look closely but it may be the same product family, for the same applications, the magnet on the piston.

https://www.festo.com/cat/en-gb_gb/data/doc_engb/PDF/EN/SMX8_EN.PDF

If you have the SMT, the reed switch device is an SME.

 

[gar]

190227-2239 EST

__dan:

See second page of https://www.festo.com/cat/en-gb_gb/data/doc_engb/PDF/EN/SMX8_EN.PDF . All SMTs are listed as "magento-resistive".

.

 

[StarCat]

Comments

Comments

190227-2239 EST

__dan:

See second page of https://www.festo.com/cat/en-gb_gb/data/doc_engb/PDF/EN/SMX8_EN.PDF . All SMTs are listed as "magento-resistive".

.

Gar and Dan you guys are the best.
Isn't it funny how its next to impossible to get at the truth these days?
I have for sure given up on attempting to read Siemens DOCs.
I also have zero use for metrics or the people that are attempting to replace HP with watts. Don't like, need or want that kind of instrusion or nonsensei.
The sensor position is critical and is set at the ends of the cylinder travel by how the lights come on. This allows the timing critical actions of the PLC to work as they should. Reliability far outweighs price. I ordered 2 units from Turck as of 2 days ago. I also found another equivalent yesterday:
https://www.balluff.com/local/us/products/sensors/magnetic-field-sensors/
These are classed as T-slot units.
Thanks for all input.
Dan I especially like your Anti-Engineering comment. I have been dealing with this for 30 years.
Its the fact that everything has been monetized that drives Engineering downward. At some point, nothing hardly will work in theory......

 

[RumRunner]

Gar and Dan you guys are the best.
Isn't it funny how its next to impossible to get at the truth these days?
I have for sure given up on attempting to read Siemens DOCs.
I also have zero use for metrics or the people that are attempting to replace HP with watts. Don't like, need or want that kind of instrusion or nonsensei.
The sensor position is critical and is set at the ends of the cylinder travel by how the lights come on. This allows the timing critical actions of the PLC to work as they should. Reliability far outweighs price. I ordered 2 units from Turck as of 2 days ago. I also found another equivalent yesterday:
https://www.balluff.com/local/us/products/sensors/magnetic-field-sensors/
These are classed as T-slot units.
Thanks for all input.
Dan I especially like your Anti-Engineering comment. I have been dealing with this for 30 years.
Its the fact that everything has been monetized that drives Engineering downward. At some point, nothing hardly will work in theory......

Notwithstanding some sort of manifesto[forgive pun] in above responses which are informational.

Those catalogs presented are not what you would expect in your application. Those are air cylinders but could also be hydraulic.

You alluded to beer brewing industry.

The cylinder being referred to is not the one shown on catalogs presented by other posters.
This cylinder is the heart of the seamer or closer that apply the lids to the cans. Some facilities call them seamers others call them closers.

The turret assembly that holds these cylinders require critical adjustments that only a few technicians are allowed to make adjustments. . . they are specially trained.

To OP: I don't know what brand of machine [close/seamer] you have in your facility.

The one my company usually services [PLC/sensors] is the Angelus brand -- which most technicians call the cadillac of seamers. They were the first closer/seamer that used PLC program on their controls.

Although other manufacturers like American Can and Continental Closers had followed suit in PLC programming later on.
The sensors we installed on Angelus Seamers are either Pepperl+Fuchs or Turck. . . they are both made
in Germany.

Their mounting are different from T-slot configuration. Can be purchased in either PNP or NPN transistor output.
These seamers/closers almost dictate the production of highest quality product.
A machine that spits out 1800 cans per minute. . . the production crew cannot afford to have leaking container/cans that can get into the supply chain.

At this speed-- a lot of beer will go into the sewers in a matter of a few minutes if these seamers/closers did not function properly and are left unchecked. I've heard production foremen getting fired or reprimanded for putting out a trailer-load of bad cans.

The last replacement we used was TURCK. Almost the same price as Pepperl+Fuchs.

As a leftover from some service call we did years ago...I installed one in my garage to indicate whether the garage door is open or closed.
It's a TURCK _yellow casing, cylindrical. It's been there almost twenty years.

Furthermore, if your application is not critical as I mentioned above, a generic reed switch will serve your purpose. (I think)