HID Ballast and CAN BUS

[MH3.0R]

Ric,

Thanks for the insight and I guess we do need some perspective.

I take your point about the 1MHz to 200Mhz test frequency range. My concern is that high frequency noise generated by swiching transients (from stray inductance and capacitance) can induce any range of frequency or component or harmonic therof. A 16Khz arc welder has even created interference in a CAN bus system (see http://ieeexplore.ieee.org/Xplore/login ... ber=744676).

So, in my opinion, I am not convinced that there will be no interference of a ballast or two placed in the engine bay on a vehicle that has a CAN Bus electrical system, regardless of the frequency range of the device. I think the level of interference is based around the cleanliness of the switching voltage waveform. If high quality components are used in the filtering, then we may be less likely to suffer any observed interference (things that stop working).

Another of my main concerns is the interference that we lay people do not see, that is not evident in the day to day operation of our vehicles. This pertains to the safety systems that utilise CAN bus communications, for airbags, VDC, EBD, Door unlock, etc. These cannot be tested until they are needed and I would like to be comfortable in the knowledge that if I ever do need my safety systems to work, that I do not have an after-market component that is not Electro Magnetically Compatible with the OEM systems.

Still open for further discussions and all opinons are respected and are adding to our global body of knowledge.

Kind Regards,

Mark.

[MH3.0R]

Ric,

Just in from the USA Legacy GT BB,

http://legacygt.com/forums/showthread.p ... ost2305752

Interesting commentary from another engineer.

Also found the link to KEMA, the Dutch EMC E4 certification body for vehicle electrical systems at:

http://www.kema.com/products_and_proces ... ty/EMC_E4/

I have also requested some suppliers to provide me with a copy of their E4 certification so I can validate. If I can, then all my concerns will vanish around those OEMs that have genuine E4 EMC compliance.

I welcome any further discussion?

Kind Regards,

Mark.

[BlackFlatSix05]

Can you tell me where you got your info from regarding 04 onwards Liberty's being CAN BUS ?

I think you will find only 06/07/08 onwards Subaru's are CAN BUS.

Happy to be corrected though.

[tangcla]

So, in my opinion, I am not convinced that there will be no interference of a ballast or two placed in the engine bay on a vehicle that has a CAN Bus electrical system, regardless of the frequency range of the device.

Ferrite clamps on all wiring then?

[MH3.0R]

Can you tell me where you got your info from regarding 04 onwards Liberty's being CAN BUS ?

I think you will find only 07/08 onwards Subaru's are CAN BUS.

Happy to be corrected though.

The Legacy Service manual, in the areas pertaining to Electrical, section 13 I think. There is a section about CAN BUS communication. The service manual in question covered the Gen IV (MY04 onwards), but not sure if it is accurate for all later varients.

Another thread that I examined in the recent months described the experience of a budding racing car enthusiast who modified his own Nissan GT with HIDs (part of a sponsorship deal) and the result was all light bulbs blown, but with all fuses intact. Apparently the result of High Voltage and High Freq leakage. Have not heard of any other cases along these lines, though I have not really been searching either.

Kind Regards,

Mark.

[BlackFlatSix05]

I'm only new to Liberty's having come from WRX world and according to Romraider ecu definitions CAN Bus vehicles are the later 06 pignose and 08/09 Ugly Betty models.

I assume that would be the case for the Liberty 06 onwards.

[MH3.0R]

Your assumption is not well founded and the pages from the Liberty/Legacy GenIV Service Manual indicates otherwise. Section 13, part 15 clearly shows CANBUS communications between all the following sub-systems:

Combination Meter;
Body Integrated Unit;
Steering Angle Sensor;
ECM;
TCM;
VDC;and
ABS.

Unless you can find a better reference?

This is the prime reason for me not installing the HID units until now. The MOD monster in me wants to do it, but I do not want to have any interference (HV or HF) in any of these systems that could potentially degrade the function they perform. Pay a premium for the Liberty safety features, and if I need them one day, I want assurance they will be working fully.

Kind Regards,

Mark.

[tangcla]

The body module is definitely CANBUS, I can confirm this with my experimentation (and troubles!) in getting the rear wiper motor wired up in sync with the correct pulse.

[tangcla]

Pay a premium for the Liberty safety features, and if I need them one day, I want assurance they will be working fully.

Just plan your accident to happen when it's daytime and your headlights are off

[MH3.0R]

Pay a premium for the Liberty safety features, and if I need them one day, I want assurance they will be working fully.

Just plan your accident to happen when it's daytime and your headlights are off

Hahaha! Yeah, but it will probably be with you because you have the HID mods!!!

[tangcla]

Hahaha! Yeah, but it will probably be with you because you have the HID mods!!!

Yeah... my CAN-tampered car will want to collide with yours

[BlackFlatSix05]

My assumption might not be well founded but I'm not convinced my 04 Liberty is CAN Bus. I found a list of Subaru vehicles that are here :

http://www.subaruforester.org/vbulletin ... 07-a-6220/

[Ric]

I suspect they were using CAN bus inside the car before then, but only connected it to the OBD2 connector from those dates.

[parso_rex]

Hmm I think you might be taking the idea of interference a bit too seriously. There are many very good filtering systems already implemented in the factory ECU and additonal modules. On ECU you can see a bank of electolytics that do a very good job of filtering high voltage/frequency transients.
I can tell you if interference were a problem then every car with an Innovate LC1 installed would have dropped dead by now and there are plenty of later model Subaru's around that don't have any issues. AS mentioned above the communications method used has built in error correction so unless the whole thing was somehow affected for a long period of time its not going to be an issue. Another thing is the CAN bus when used in automotive apps is using a balanced two wire set with differential inputs and outputs which in itself gives it another level of immunity to interference


Here's something that was giving me grief a while back. This is the pulse that is fed back into the cars electrical system from an LC1.

This is using my CRO (pfft yeah I know who doesn't have one these days LOL )viewing the waveform as it appears on the battery terminals believe it or not. One would think that the capacitive properties of the battery would be sufficient to effectively filter the pulse this is not the case, you need a capacitor that is fast enough to catch the pulse.


Heres the same waveform viewed at the same place but with the addition of a nice cheap and fast 4700uF electrolytic placed across the LC1's supply rails

I've got some more info which I can add later but bascially what I'm trying to say is that the factory systems are actually very good so I'd hate to think what beasty has the ability to cause it grief

[Lukovitch]

^^ what he said. A properly designed system should be able to cop just about anything thrown at it.

I work with robotics, they have CPU/DSP/FPGA's on the same PCB as multi channel servo drivers and no issue - because they are well designed/isolated/filtered.