Tyre Guage DRO – capacitance issues?

[Peter Cook 6Participant @petercook6]

I use a cheap tyre depth guage as a simple quill DRO on my SX1LP. Apparently it is a capacitance based measuring device.

It works very well – as long as I keep my fingers away from it. If I switch it on or zero it using my fingers I get 0.03-0.05mm "jitter" on the display. Zero it with a non conducting object ( plastic T slot cleaner usually) and all is well.

Take the device off the mill and it behaves perfectly as does another similar unit I use for measuring the depth of holes.

I suspect the issue is a capacitance effect between me and the ground plane provided by the big lump of iron to which it is screwed.

Can anyone explain in a bit more detail how these things work, and make any suggestions as to how I might screen the problem. A bit of brass or plastic between the device and the mill head makes no difference.

[Peter Cook 6Participant @petercook6]

[Michael GilliganParticipant @michaelgilligan61133]

I can do no better than direct you to the post by Hans U. Meyer. on this thread: **LINK**

http://www.pcbheaven.com/exppages/Digital_Caliper_Protocol/

If you need the technical details … his patents are available on espacenet

… but his closing paragraph is pretty blunt.

[ the inventor does not consider capacitative scales suitable for use on machine tools ]

MichaelG.

[Ian PParticipant @ianp]

There are millions of devices in use that use the same capacitive encoders for linear measurement. I have never heard of one being sensitive to any external (capacitive) influence.

Very noisy electrical fields might disturb them but none of the ones I have are even bothered by my high power demagnetiser, they just display the same readings whether its on or off.

Does you depth gauge work OK if its off the machine?

Ian P

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Peter Cook 6 on 02/04/2021 19:28:30:

[…]

Take the device off the mill and it behaves perfectly as does another similar unit I use for measuring the depth of holes.

[…]

.

Ian, for info. ^^^

MichaelG.

[old martParticipant @oldmart]

There are plenty of variations available, we have two at the museum, and have never had any problems, they are based on digital caliper technology.

Edited By old mart on 02/04/2021 20:53:22

[Jeff DaymanParticipant @jeffdayman43397]

You might try simply adding some plastic or aluminum spacers between the large metal mass of the machine and he scale, and use longer mounting screws. Distance to metal masses can affect readings in capacitive sensors.

Nylon or acetal spacers would probably affect readings less than PVC spacers, just FYI, PVC has affected capacitance in several linear measurement instruments I have worked on in the day job. Its' presence changed the dielectric constant in the sensor and caused the readings to be larger than they actually were. Nylon or acetal has less effect.

Edited By Jeff Dayman on 02/04/2021 21:19:49

[Robert Atkinson 2Participant @robertatkinson2]

Does it work OK if the mill is turned off at the socket? Fully unplugged at the mains?
I think the problem is interference with the mill bodywork being noisy. Your body is providing a return path.
You could also try firmly grasping the pillar with oe hand while pressing the buttons.

Robert G8RPI.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Robert Atkinson 2 on 02/04/2021 21:26:20:

…
I think the problem is interference with the mill bodywork being noisy.

…

Me too. Interference is highly likely.

For it's intended purpose a tyre depth gauge has no need of electrical shielding, so it probably doesn't have any. Expecting one to work whilst attached to an electrically noisy machine tool may be a step too far.

Try opening it up and check for missing components. The Lidl digital caliper I dismantled for experimental purposes had no by-pass capacitors fitted, even though PCB has pads ready for them. OK running on Its internal battery but enough hum was picked on the leads from an external power supply to confuse it, and I had to fit a bypass capacitor across the input traces to fix it.

Although the electrical supply safety earth will help drain noise away, from an RF perspective they are usually rubbish. Long cable runs connecting to an earth spike may be a low resistance sink at 50Hz, but are likely to be high resistance to the kilohertz noise capable of disrupting an unshielded DRO. The usual compromise is to run the sensitive electronics inside a metal box, and ground the box. Whilst the outer skin of the box is still hot with noise, with care noise doesn't penetrate inside. As Robert might explain further, the design of EMC proof enclosures and electronics is difficult!

I'm all for cost-effective home-made solutions, and this one is well-worth trying, but bear in mind re-purposing stuff not made for the job is always a bit risky.

Dave

[Michael GilliganParticipant @michaelgilligan61133]

Simple fact:

In 2012, the inventor of the capacitative scale concluded his post with …

“I wouldn't want them on a machine tool.”

.

My inference from that is that the successful implementations are fortuitous, and we should not be surprised by the failures.

MichaelG.

[Anonymous]

Some statements by SoD need clarification.

A single wire has inductance, so as the frequency of a signal increases the resistance of the wire stays constant, but the impedance increases. One might argue that the effective resistance increases due to skin effect, just to pre-empt the experts.

The subject of EMC covers a wide range of areas. The main ones are radiated emissions, immunity to radiation, conducted emissions and ESD. If a device is mains powered then there are also mains borne nasties to be checked.

Radiated emissions are fairly simple to control, with the exception of cables. I hate cables inside, or external to, a product. Immunity is also fairly simple, especially for some applications, such as consumer, where the unit doesn't need to continue working, but just has to recover without damage. In other applications, automotive for instance, the unit needs to work throughout immunity tests. It's no good if the ECU hiccups every time a spark plug fires. Conducted emissions are fairly simple to control with appropriate filters and/or shielding.

A big cause of EMC test failure is ESD – it has a habit of getting into places it shouldn't, through pathways that one hadn't thought about.

Andrew

[Les Jones 1Participant @lesjones1]

Looking at your picture it looks like the casing is all plastic. I think connecting one side of the battery connections to the mill metal work may solve the problem. Just in case there is already some connection via metalwork that I can't see I suggest making the connection via a capacitor. I suggest about 100 nF but the value is not critical. Also it should not matter which battery connection you connect to. I think there is a removable cover to connections for a remote readout to the right of the battery cover. (As positioned in your picture.) There will be 4 connections. two connect to the battery. (You can check which these are with a DMM on the resistance range.) The other two connections will be data and clock signals. The battery connections under this cover are probably the easiest to make a connection to.

Les.

[Peter Cook 6Participant @petercook6]

Thanks for all the thoughts, I did some experiments based on the ideas. Robert G8RPI's idea worked. If I touch the head of the mill with one hand while pushing the buttons all works fine.

The gauge works perfectly as a simple DRO when the machine is running or stopped. The readings correspond to the markings on the fine feed knob (without the backlash) over the full 25mm range of the quill (and the tyre gauge) as long as I keep fingers away from it. There is no sign of flicker in the readings when the mill is on or off, running or stopped.

However – the problem IS electrical interference, but not from the mill. By accident I experimented while the workshop lights were off. All worked perfectly. Switch on the overhead florescent tubes – problem reappears. If I hold the gauge very close to the florescent fittings the readings go haywire.

I can only conclude that if I don't touch the mill head while pressing the buttons I am acting as an aerial for the interference from the florescent tubes. So either a better guage to use as a DRO (one with a metal body?), some new light fittings for the workshop or simply touch the mill before zeroing.

Thanks again.

 

PS thanks Les I will try that next. A capacitor will be cheaper than new light fittings.

Edited By Peter Cook 6 on 03/04/2021 13:38:53

[John HaineParticipant @johnhaine32865]

Check your mill is earthed properly!

[SteviegtrParticipant @steviegtr]

Are you sure your lighting is earthed.

Steve.

[John HaineParticipant @johnhaine32865]

It may be light sensitive…

[John BaronParticipant @johnbaron31275]

Posted by Peter Cook 6 on 02/04/2021 19:28:30:

I use a cheap tyre depth guage as a simple quill DRO on my SX1LP. Apparently it is a capacitance based measuring device.

It works very well – as long as I keep my fingers away from it. If I switch it on or zero it using my fingers I get 0.03-0.05mm "jitter" on the display. Zero it with a non conducting object ( plastic T slot cleaner usually) and all is well.

Take the device off the mill and it behaves perfectly as does another similar unit I use for measuring the depth of holes.

I suspect the issue is a capacitance effect between me and the ground plane provided by the big lump of iron to which it is screwed.

Can anyone explain in a bit more detail how these things work, and make any suggestions as to how I might screen the problem. A bit of brass or plastic between the device and the mill head makes no difference.

Check and make sure that the machine/mill frame is truly earthed ! Any capacitive effect is likely due to any residual voltage on the machine frame.

FWITW, my mill didn't even have the earth wire connected to the machine in the control box ! The earth wire was just hung in mid air. Luckily it didn't touch anything that could have caused damage.

[MacolmParticipant @macolm]

I doubt if earthing the lighting will make much difference. This sort of susceptibility is common with high impedance logic circuits. If the capacitive scale is like most others, the positive of its battery is already connected to the metalwork of the scale. What is happening here is that the “operator” is an aerial receiving radiation from the lighting. The capacitance between his finger and a susceptible point in the scale circuitry is sufficient to induce a big enough "signal" to cause malfunction. If he “earths” himself to the mill metalwork, the unwanted signal is bypassed and its magnitude much reduced.

 

I doubt if it would matter whether the mill is earthed or not. There is plenty of signal path at the relevant frequencies to the mains wiring via stray capacitances.

 

Prevention would need an earthed screen, probably impractical in this case.

 

Edited By Macolm on 03/04/2021 15:20:22

Edited By Macolm on 03/04/2021 15:23:47

[Ian PParticipant @ianp]

Posted by Michael Gilligan on 03/04/2021 11:21:18:

Simple fact:

In 2012, the inventor of the capacitative scale concluded his post with …

“I wouldn't want them on a machine tool.”

.

My inference from that is that the successful implementations are fortuitous, and we should not be surprised by the failures.

MichaelG.

That a little cruel

Successful products are more likely the result of hard work by clever engineers

The inventor's comments were specifically relating to early examples of the genre, also apart from the ESD aspect (should not affect well designed products) (he mentions ground loops and I took it that was referring to DIY modifications involving external wiring and ground loops.

Ian P

[Ian PParticipant @ianp]

I'm sure Malcolm is right regarding why this particular unit is failing.

Digital capacitive based calipers and DRO scale invariably have 'Hardened' Stainless bodies and beams. The important capacitive action takes place on the rear of the display PCB and the PCB pattern on the moving beam (which is backed by the metal of the beam).

The important bits then are only a few thou apart and quite well screened in the assembled sandwich and usually impervious to any external influence.

In the case of the tyre tread gauge, the beam is usually made from plastic so the moving PCB track and the rear of the main PCB have no connected screening.

Not in current use now but I have butchered about six digital calipers to use as simple DRO's. non have ever so much as blinked. I do have a (plastic) tyre tread gauge but its travel is too short to be usefull.

Ian p

[Michael GilliganParticipant @michaelgilligan61133]

Ian wrote: The inventor's comments were specifically relating to early examples of the genre

.

Sorry, Ian … I disagree

He made it reasonably clear that the mass-market [capacitative] Chinese scales were based upon his design for Sylvac:

• Chinese calipers use very, very similar chips (and that's an understatement).

and that

• Calipers from Sylvac and other manufacturers (including Chinese ones) are now inductive …

.

I remain convinced that it's the capacitative principle that he was dismissing, not external wiring.

… but it doesn't matter; the job's done, and presumably Peter is a happy man.

MichaelG.

.

Edit: __ added reference, for clarity

Edited By Michael Gilligan on 03/04/2021 17:18:41

[Peter Cook 6Participant @petercook6]

Thanks everyone, Knowing the problem is 90% of the way to the solution. I will try some of the suggestions, and have ordered a "silver" version of the gauge to see if the plating helps. However now I know what to avoid it does the job.

Mill & lighting is all earthed.

The incident reminded me of one of my favourite quotations from Isaac Asimov

The most exciting phrase to hear in science, the one that heralds new discoveries, is not “Eureka!” (I found it!) but “That’s funny …”

 

Edited By Peter Cook 6 on 03/04/2021 20:43:19

[Robert Atkinson 2Participant @robertatkinson2]

Hi Peter,

Your earlier comment sounds correct. The lights are generating interference an you are an antenna bringin it close to the DRO. The mill body is acting as a "ground" return capacitively coupled to the DRO. By touching the mill with one hand you and the mill are at the same voltage and the interference has no effect.

I thought it was the other way around with th milll body being the intererence source and you the ground, bu the end effect is the same. The "metallic" cse version my well sort the problem.

Robert G8RPI.

[Peter Cook 6Participant @petercook6]

Just an update for anyone interested. I obtained a new "silver" tyre gauge and swapped the front part for the one I was using as the mill DRO.

 

The problem of the flickering display went away. Presumably the thin coating is sufficient screening to fix the problem. There is a side benefit. Although the devices look identical apart from colour, the new one does not loose its setting when it switches itself off, and comes back on when the quill moves. The old one reset to zero when it switched off, and had to be manually switched on again – so they are different implementations.

In swapping over I had to dismantle the original device and came across another interesting curiosity. The back of the LCD does not appear to have any electrical connection with the PCB.

The display has a strip of what feels like rubber along one edge. I cannot – even with a loupe – see any contacts on the rubber.

While the PCB has a corresponding set of pads onto which the rubber strip presses

I can only presume that the strip of rubber conducts from top to bottom, but not sideways so that matching pads on the LCD connect to the pads on the board. If so that is high tech (and potentially quite useful) material in something that cost me a fiver delivered – so what did it cost to make!!

Edited By Peter Cook 6 on 07/04/2021 17:21:58

[Michael GilliganParticipant @michaelgilligan61133]

Thanks for the interesting and useful update, Peter

… Regarding the rubber : I think you might search for ‘Zebra Strip’

MichaelG.

[Author]

Posts