Why it is important to mea­su­re on phy­si­cal lay­er in CAN bus ?

Our world is ana­lo­gue. Digi­tal is a human inven­ti­on to sim­pli­fy many things. Smo­ke signals bet­ween moun­tain peaks, signal lights bet­ween ships, Mor­se signals on air – all the­se alre­ady were digi­tal signals. As long as digi­tal works, it is invin­ci­ble – for ins­tan­ce in pre­ci­si­on of trans­mis­si­on. But even when you trans­mit digi­tal signals, you need an ana­lo­gue medi­um below it. Such a medi­um is fail­u­re-pro­ne. Smo­ke signals fail in darkness, signal lights are useless in fog­gy con­di­ti­ons, Mor­se signals are going to be inau­di­ble when inter­fe­ring signals are too loud.

On a fiel­dbus and in elec­tro­nics in gene­ral, vol­ta­ges are defi­ned to rep­re­sent digi­tal High and Low. Becau­se ever­y­bo­dy knows that vol­ta­ges are ana­lo­gue values, an ille­gal ran­ge is defi­ned whe­re no signal is sup­po­sed to be. To get a cer­tain amount of trans­mis­si­on relia­bi­li­ty, in nor­mal con­di­ti­ons the vol­ta­ges are defi­ned in a distan­ce to this ille­gal range.
To signal a vol­ta­ge you need a wire as medi­um. What works easy in a flash­light, gets dif­fi­cult on high fre­quen­ci­es and long trunk lines very quick. Then we talk about capa­ci­ties and fre­quen­ci­es, not only about resis­tors and swit­ches. You will need more than good know­ledge and a mul­ti­me­ter to get use­ful and repeata­ble statements.

Some­bo­dy could say that at pro­blems in a CAN bus com­mu­ni­ca­ti­on, error frames and repe­ti­ti­on of messa­ges occur. This is not qui­te cor­rect. We have to say that at a point whe­re error frames are on the bus, we can not speak from a hitch-free digi­tal trans­mis­si­on any­mo­re. The bus can still run becau­se of the work of the safe­ty mecha­nisms in CAN, repea­ting not cor­rect trans­mit­ted messa­ges. At this point only a litt­le wor­se­ning would lead to a com­ple­te break­down of com­mu­ni­ca­ti­on. That’s why we have to sta­te that per­ma­nent error frames (on a for­mer good run­ning bus) are an indi­ca­tor for a near­by total fail­u­re. Some­ti­mes this level of wor­se­ning never hap­pens and the bus just stops working.

Errorframe CAN

What hap­pens ? Machi­nes are aging, wires get­ting britt­le and break, con­nec­tors get­ting wet and cor­ro­de, com­pon­ents in sen­sors and actua­tors are sub­ject to aging effects. The distan­ce of vol­ta­ges to the ille­gal ran­ge is shrin­king, fre­quen­ci­es are shif­ting, capa­ci­ties degra­de signals. Besi­de of aging, the inter­fe­rence resis­tan­ce plays a roll on this. What works gre­at on a lab table, may not work besi­de other machi­nes or mobi­le on the road.

How can I gage, how good the digi­tal trans­mis­si­on works, when the only thing I know is that it works ?

We should mea­su­re, how the distan­ce to the ille­gal ran­ge of vol­ta­ges is – how much wear reser­ve is left – how much inter­fe­rence resis­tan­ce we have for the work in the field. We should mea­su­re, if the qua­li­ty of signals is worse than at the other day or if it is as good as new.

This works only on mea­su­ring, eva­lua­ting and com­pa­ring ana­lo­gue values, deep down on the CAN phy­si­cal lay­er. You can do this the clas­sic way using an oscil­lo­scope. You will need expert know­ledge and expe­ri­ence. A device that eva­lua­te the mea­su­red values ins­te­ad of just sho­wing a visua­li­sa­ti­on of a signal, is much more hel­pful. When you can get a direct com­pa­re with an older mea­su­rement or ano­t­her mea­su­rement point wit­hin the bus, even trou­ble shoo­ting will be much easier.

For this pur­po­se are the GEMAC fiel­dbus tools. A tech­ni­ci­an is able to say some­thing about signal qua­li­ty in short time. Even to go through an accep­tan­ce test can be done in minu­tes and gives clear state­ments, which can easi­ly be prin­ted. On trou­ble shoo­ting it is pos­si­ble to quick­ly iso­la­te whe­re a pro­blem is. The out­co­me of main­ten­an­ce and repairs can be eva­lu­t­ed in terms of numbers.

What do you know without mea­su­rement ? Only that it works or not. Without a mea­su­rement device you stay with the digi­tal “Zero or One”-statement. With a mea­su­rement device your eyes are ope­ned to a who­le new world, the ana­lo­gue world we are living in.

CAN signal fading

view-measurement-one-stationCANtouch Hand

How can I gage, how good the digi­tal trans­mis­si­on works, when the only thing I know is that it works?”

CAN-Bus-Tester-2---Montage-Koffer-1000x1000Ansicht

Print Friendly, PDF & Email