Talk Morgan Forums Morgan Sports Cars Maintenance & Servicing GDI Gauges Intermittent Fault

Update

Had a couple of runs out with just the Speedometer connected to the SAT-BUS and it has not dropped out once, which makes sense if it doesn't use the CAN-BUS.

For the second run I connected my Torque Pro to the OBD port and logged the Revs and Water Temps, I wanted to see if they were dropping without the gauges/SAT-BUS involved. A review of the log shows no issues, no blank records and no delays between entries. This needs to be repeated before giving it any significance though.

Next runs set for Sun & Mon. and I'll connect the Rev Counter back up to see what happens.

I'm afraid I haven't got anything constructive to add Mark, but I'm loving the learning process that's going on here. I'm sure you'll track the fault down in the end with such a methodical approach.

Update

I re attached the Rev Counter and Water Temp Gauges.

From a cold start the Rev counter does not register any Revs for a short while (5-10 seconds) but then works fine for the rest of the trips. This is new, it has always been instant and still is if started soon after switching off. Torque Pro shows the Revs during this period.

The water temp gauge appears to work fine but the scale starts at 60 degrees so wouldn't register straight away. Again Torque Pro registers the temp from start up.

There has been no observed drop in any of the Gauges, I increased the frequency of the Torque Pro Log entries to half a second for the recent runs and there was 1x instance of Torque Pro reconnecting to the ECU, the drop was only 1 record so half a second, which doesn't feel significant but someone on here may put me right.
I'm not expecting m/any comments at this stage, it just helps to write down where I'm at and may help others in the future.

I don't plan to add the Fuel Gauge for a while, I want to be as certain as I can be that the problem has gone before adding it, and due to Grandparent duties during the 6 weeks holidays it may be a while before we get any significant runs out.

Regards

Hi Mark,
I was stuck in traffic on the A1 the other day and I had the exact same problem that you recorded at the start of this thread. For me the issue is purely temperature related. As soon as the car was moving and things began to cool down everthing came back to full operation. It happened three times as the temp increased or decreased.
My car has not been out in the rain this year.
Cheers
John.

Hi John,

Sorry to hear you have the same issue, it's quite annoying.

Temperature is a factor with mine, it's more frequent when the car gets hot but mine can also have the issue when travelling at a steady 50 on an A road. It will be interesting to see if/how your issue develops, mine started very infrequently but over time became almost inevitable. On an 18 day tour of Spain in June it happened all but 1 day.

Thought you may be interested in how I'm going on.

I have just reconnected my fuel guage and have a trip to Harrogate planned for Sunday back Monday, so this will determine if the fault is fixed or if it's something to do with the Fuel gauge/sender or something. It will take quite a few more trips before I'll be convinced it's fixed.

I did only 2 things that might have affected the issue, before running the car with selective gauges
1. Cleaned the Engine/Chassis harness connector.
2. When I found and tested the splices I didn't have any loom tape so I didn't bind them back to the loom, which may have relieved some stress on them.

Anyway it may be wishful thinking on my part but I'll find out fairly soon.

All the best.

There has been no re-occurrence of the issue!

Done quite a few day trips and 5 days touring Pembrokeshire, I recorded all the trips with Torque Pro and the logs show no drops that we may have missed visually, so as confident as I can be that it's resolved

I'm confident that the issue was a CAN-BUS problem within the Chassis Harness but as per my previous post the cause is either
1. a dodgy connection within the Chassis/Engine Harness connector OR
2. an issue with the CAN-BUS wiring Splices which sit just in front of the same connector.

I'll tidy the wiring back up at some point and report back here if the issue returns (or not).

Thanks again for all the help chaps.

Regards

H Mark

Just got round to reading your last post. Congratulations !! This result was in no short measure down to your logical and methodical approach and sheer dogged persistence. well done.

I have been meaning to comment on wrapping a copper screen around the CAN BUS cables. With great respect to other posters this is something you should definitely not do. At the risk boring every one I will try to explain why.

There are two types of signal transmission lines balanced and unbalanced, an example of an unbalanced line is as follows. The transmitting end has one cable connected to ground, the receiving end also has one cable connected to ground, so you only need one signal wire connected between them. A bit like a headlamp bulb only needs one wire, the return is through ground.

In an unbalanced type lets say the transmitter sends a signal by increasing the signal wires voltage to one volt, this travels down the wire and arrives at the receiver pretty much at one volt. All is well however, if the signal wire passes close to an electric motor or ignition cable a voltage could be induced within it lets say it does and its one volt. This arrives at the receiver, which thinks it's a signal and so we have interference. To stop this happening the signal wire is enclosed within a woven copper screen all along its length and the screen is connected to earth. The interfering signal does not get past the screen and goes harmlessly to earth.

The second type of transmission line is a Balanced Line as in the case of CAN BUS. With this type there are two wires neither of which are connected to ground and the receiver responds to the DIFFERENCE between these two wires. In a balanced line, with no signal, both wires sit at the same nominal voltage. Lets say it is 4 volts, to send a one volt signal the transmitter moves one line up to 4.5 volts and the other line down to 3.5 volts. The receiver responds to the DIFFERENCE i.e. 1 volt. These two wires are twisted together so if they run close to a motor or ignition cable both of them get the same interference signal. So with this type of line with a one volt interference signal they would BOTH rise by 1 volt i.e. to 5 volts. The receiver sees no difference and therefore does not respond to the interfering signal. Hence no screening needed.

Producing a transmission line by screening one conductor (unbalanced ) or twisting two conductors ( balanced ) produces something called a transmission lines characteristic impedance, which in the case of CAN BUS is 120 ohms. It is important that the BUS is terminated at both ends with 120 ohms to match the characteristic impedance. If this is not done then reflections along the line can occur causing data corruption.

So, if you are still with me, if you wrap a copper screen around a CAN BUS line you will change its characteristic impedance and it will no longer be 120 ohms and therefore not properly terminated by the 120 ohms resistors that are fitted either end of it and that can cause reflections so you could end up worse off

Sorry for the long post, both my adult daughters have said to me “Dad, when you give me one of your explanations would you mind if I record it on my mobile phone, so that if I am having a sleepless night I can play it, I will be gone in two minutes”.

Good Luck
Regards Tony

Very interesting Tony , I'm genuinely not bored by this interesting tech stuff 'cos you never know when it may come in handy

Twisted pair transmission has been around since the introduction of telephones in the late 19th century and we can thank Alexander Graham Bell for its discovery. The mutual cancellation effect produced was adopted for computer (LAN) communications at their outset and later CANBUS.