Anatomy of Ardic

Greetings from the non-insulated garage of my parents! I don't have any warm facilities to fix my car, so this will have to do. It was a nice warm day of -5 celsius when I decided that it is now or never (actually now or next summer) when I'm going to tear down the Ardic and see what's the problem.

Here's a nice post about removing and servicing Ardic. (Sorry, instructions only in Finnish).

Here's an explosion diagram of Ardic. Year is 2003, but seems to be identical to mine. Taken from Finnish Volvo forum, topic designated to Ardic problems..

Here's the big guy itself after removing the front bumper. Inferring from the rusted screws I would think this thing hasn't been serviced in years, maybe never during the 10 years this thing existed. There has been quite many previous owners and according to the owner history database, each of them had this car only little bit over 2 years.

After one hour of careful separation process in -6C, we can continue disassembly inside in room temperature! 

Glow plug, water pump and outer shell removed.

Testing the glow plug. Intact, as I suspected.

Inner shell shell removed, revealing chunks of soot attached to the walls 

Here's the combustion fan motor, working normally. Little bit to the left under the cap resides the flame sensor that is directed toward the combustion chamber. I managed not to take any pictures from right angle, but there was little bit of soot there as well, blocking the view of the flame and causing the main problem here. After some cleaning, the flame sensor reads ~8 Mohm when in dark and around 430 Kohm when teased with direct light from a flashlight.

Quite a bit of soot also under (actually over, since the unit seen here is held upside-down normally) the cup and the turbulator. There's a shadow under it too, but most of it is soot actually.

Here's the CPM. It's funny feeling seeing it now here, like meeting somebody in real life after you have spent weeks chatting online :)

We know each other so well already, so there was no need to be embarrassed. Let's take the cover off and look if there are any unhappy burned parts. None found.

Only after putting the whole thing together again and looking this image more carefully when uploading this picture, I noticed this burned looking solder joint little bit from the center to the direction of upper left corner. Weird, since the heater seems to be working now.  Maybe it was just the angle, light doing its tricks.

So, I cleaned the heater from all the soot and put the thing back together. Fingers crossed, I started the heater and behold, it works now! It seems that soot builds up as a result of imperfect combustion and eventually blocks the light/flame sensor. When this happens CPM thinks there's a problem with fuel delivery or some other functionality, and then stops the heater. Ardic does need service at least every two years, but some people service it annually, especially when there's a lot of short distance driving and the heater runs cold proportionally greater periods.

 I didn't touch the water pump since it seems to be working and I don't have any spare rubber parts should the pump need any of them changed after opening the thing to prevent leaks. Anyways, I'm going to buy a new pump next summer when I'm servicing this thing again, just in case. It's interesting to see how much soot buildup will occur during the winter months with my personal driving style and preferences. From there it will be possible to estimate how long a relatively safe service period would be. It would be possible perhaps to estimate this based on the voltage reading of flame sensor! Of course, this would require reaching a steady state, maybe after running the heater for one hour until all the temperatures reach equilibrium and then check the sensor voltage. The nearer it is to the 2.5 volt threshold (explained in previous post) when the heater is on, the more the there could be soot covering the eye of the flame sensor. I will have to check the reading soon when it is still clean.

UPDATE: The flame sensor voltage will fluctuate between 0.6 and 1.0 volts when the furnace has been cleaned. Boys, when it starts to climb over 2.0 volts near the 2.5V threshold, it's time to grab your wrenches and mops and start cleaning!

FFffuuuuuuu......

Succession of failures. What an oxymoron.

A week ago I continued testing the ignition of parking heater with diagnostic commands. Last time I checked, it worked without the key in ignition using VIDA, so I took another shot with ELM327. No go. Ok, started injecting the periodic keep-alive message to CAN bus and tried again, but without any success. Then I put the key into position I, and what do you know, the heater started. It DOES need the key in the ignition after all, so there goes my plan to implement the remote control using the diagnostic commands. <insert swear words and cursing here/> My previous success with VIDA must have been because of some kind of inherent timeout in the CEM. Diagnostic codes and commands work for some period of time after the key is removed from the ignition, but after a while (maybe because of security reasons) the CEM stops responding to them.

There is a way to spoof the other modules into thinking that the key is in the ignition by overwriting the continuous status updates made by CEM about the key position, but there's no workaround when it comes to CEM itself. It would be like breaking into you own car. So I folded in front of this another unforeseen obstacle, took the path of less resistance and ordered the AEM module. It's going to cost little bit over 200 euros (including software updates to CEM and AEM itself), sure, but I didn't see any way beyond the key problem and I'd spent enough time tackling this one.

BTW. Guys in King of Thrones are wrong. The winter isn't coming. It's freaking already here! Few days ago the temperature went down to -22 celsius ( -7 fahrenheit)!  And that resulted in the next system failure...

The parking heater stopped working. The day before this total failure it already gave a hint for the upcoming problem. For some reason it hadn't reacted to previously set timer, but did go off when starting it manually.  Next day, I had set the timer again, but when I approached the car, I could see that everything was still frozen. Attempted starting the heater manually, and it did run for a minute or two, but then stopped again. Tried it for the third time, but the same thing happened.

Now after three unsuccessful start attempts the CPM goes into lock-down that can only be reset with VIDA or other expensive diagnostic device. Poor man's choice, Torque + ELM327 won't do, since they won't be able to query CPMs fault codes and reset them. So, my efforts developing the Sardine CAN and playing around with VIDA weren't total waste of time. Lucky me, except that I was visiting my parents' house and I didn't have the device with me, so I had to resort to firing up the engine in -22C since there isn't any additional heater installed in the car.

Except now the car would start. Motor would cough for a second or two, but then the start motor just continued to spin. Now I was starting to get pissed off.  This sounded more like a battery problem or frozen fuel lines, since a problem in either one of those could affect both the motor and the heater. There was no low voltage light on the DIM lit,  but I didn't have my multimeter to check the battery voltage. The artic diesel variety that is offered here in Northern Finland is supposed to handle at least -29C temperatures without gelling/waxing and clogging the fuel filter and the fuel lines. Anyway, I didn't have any way to pinpoint the problem at this stage. Luckily I managed to start the engine after covering it with blankets and letting two powerful heat blowers warm it up from below a bit. Even that took over 2 hours.

So I got home, connected my dear Sardine CAN, fired up VIDA and started snooping around. First, the fault codes:

Now this is interesting. There's no way simple fuel clogging problem would result fault codes piling up as they do in this list. Especially CEM-6C48 is meaningful in this context. It would indicate there was a communication problem between CEM and the transponder implanted in the ignition key. It could be caused by many different reasons (two different keys next to each other in the keychain, electrical interference, wrong key etc), but end result would be the same: Fuel delivery is inhibited to fuel injectors and motor wouldn't start. However this wouldn't have any effect on the heater, since its fuel input is controlled by heaters own fuel pump which is independent from any security measures related to engine. Anyways, fault codes this many in so many different modules would be unlikely to occur for reasons indicated in each fault code. There must have been some kind of electrical connectivity problem because of the weather, most likely in the CEM or one of its harnesses. If this ever happens again, I will have to dig into this, but for now, I reset the fault codes and wait if any of them should pop up again.

Now that the fault code for heater (CEM-5F4F : Too many unsuccessful start attempts) was also been cleared, I was free to try starting up the heater, but with this time I had VIDA to guide me with troubleshooting.

Glow plug, combustion fan as well as water and fuel pumps were working normally, but the flame sensor read constant voltage. According to VIDA, voltages from 2.5 to 5.0 are interpreted as "no flame" and voltages below 2.5 volts mean there is a flame in the heater. The CPM waits for a while for the diesel to ignite, but after 1-2 mins it gives up, stops the heater and increments the error counter. So, either the flame sensor has broken down or is covered up, or the diesel fuel doesn't reach the heater. Anyway, the problem most likely isn't going to go away magically and repair shops usually ask between 400-600 euros + parts for the repair, so I think you're going guess where I'm going with this..

Next: Ardic tear-down!