Vida CEM swapping

ricked · Post by **ricked** » 28 Jun 2021, 02:05

RickHaleParker wrote: 27 Jun 2021, 14:25
ricked wrote: 27 Jun 2021, 05:43 " so im not sure what you dont understand about that... even incude the link to it
My bad your understanding of the English language is so perfect.

not every one does have "english" as first language, show some respect instead of bashing someone because you need it te be perfect what you read, jezus...

i did point out that the Chinese IC is junk, but it this it the attitude on this forum, i will pass en will not post anymore, if you want ban me go a head as i know enough now.

Post by **vtl** » 28 Jun 2021, 07:23

ricked wrote: 28 Jun 2021, 02:05 not every one does have "english" as first language, show some respect instead of bashing someone because you need it te be perfect what you read, jezus...

i did point out that the Chinese IC is junk, but it this it the attitude on this forum, i will pass en will not post anymore, if you want ban me go a head as i know enough now.

Gentlemen, calm down. Specifically in this topic of the forum the most part of participants are not native speakers, me including. It's not even a second language. I don't think the quality of language is that important on international forums as long as participants can get the idea.

ricked, you are welcome here and thanks for your contribution to the discussion. More than often I don't know what to suggest besides: "Hmm, it works for me", and it's great that someone else can recognize a problem in the circuit, like fake chips, which is apparently a big problem and was a primarily reason ащк the project migrating away from MCP2515, where fake chip is almost granted.

ricked · Post by **ricked** » 28 Jun 2021, 07:31

vtl wrote: 28 Jun 2021, 07:23
ricked wrote: 28 Jun 2021, 02:05 not every one does have "english" as first language, show some respect instead of bashing someone because you need it te be perfect what you read, jezus...

i did point out that the Chinese IC is junk, but it this it the attitude on this forum, i will pass en will not post anymore, if you want ban me go a head as i know enough now.
Gentlemen, calm down. Specifically in this topic of the forum the most part of participants are not native speakers, me including. It's not even a second language. I don't think the quality of language is that important on international forums as long as participants can get the idea.

ricked, you are welcome here and thanks for your contribution to the discussion. More than often I don't know what to suggest besides: "Hmm, it works for me", and it's great that someone else can recognize a problem in the circuit, like fake chips, which is apparently a big problem and was a primarily reason of the project migrating away from MCP2515, where fake chip is almost granted.

exactly... spot on

Post by **RickHaleParker** » 28 Jun 2021, 11:12

ricked wrote: 28 Jun 2021, 02:05 not every one does have "english" as first language, show some respect instead of bashing someone because you need it te be perfect what you read, jezus...

I did show some respect. There is nothing offensive about working out clarification from substandard English. Nothing I said degraded you in any way. It was all about the words not you.

However your response does imply that I have a mental defect. I took the high road and did not retaliate for it.

Post by **RickHaleParker** » 01 Jul 2021, 17:11

5ft24 wrote: 22 Jun 2021, 18:42 My P2 will not crack without them.

I think I got this figured out. The 120Ω in conjunction with the impedance of the twisted pair act as a Bandpass filter. Filters out reflection and noise. If you have long leads between the CAN transceivers and the CEM it is probably picking up noise and/or reflecting the signal.

I suspect if the hardware design places the electronics right at the OBD2 connector a lot of problems will go away.
A shielded OBD2 cable might make a difference.

Post by **vtl** » 03 Jul 2021, 09:50

Shielded cable is not really necessary, as it is a differential signalling. CEM to OBD-II port is not shielded, if I remember correctly.

What may help is to twist the pair from the OBD to CAN transceivers, so the noise would be picked up by both wires equally and be ruled out by the transceiver. CEM to OBD is twisted.

Post by **RickHaleParker** » 03 Jul 2021, 15:48

vtl wrote: 03 Jul 2021, 09:50 Shielded cable is not really necessary, as it is a differential signalling. CEM to OBD-II port is not shielded, if I remember correctly.

It is un-sheilded twisted pair.

What may help is to twist the pair from the OBD to CAN transceivers, so the noise would be picked up by both wires equally and be ruled out by the transceiver. CEM to OBD is twisted.

Twisted pair is needed to get the correct impedance. Signal cable construction is a precision science. That how a XΩ coaxial cable is XΩ of impedance regardless of length. Twisted pair is the same, impedance is the same regardless of length. However the longer the Twisted pair the higher the resistance and opportunity for interference and signal degradation.

Steal some CAN twisted pair out of a junk car or get a pre-wired OBD2 cable. You can find the pre-wired OBD2 cables for ~$6.00.

5ft24 · Post by **5ft24** » 03 Jul 2021, 15:52

i used cat6 cable was running 20 feet of it before I built the PCboard with the connector on the end

Post by **RickHaleParker** » 03 Jul 2021, 18:14

5ft24 wrote: 03 Jul 2021, 15:52 i used cat6 cable was running 20 feet of it before I built the PCboard with the connector on the end

CAT6 characteristic impedance 100Ω. CAN Bus characteristic impedance 120Ω.
Close enough that you could get away with it in most cases but why 20 feet? Two feet I could get but 20 ..

5ft24 · Post by **5ft24** » 03 Jul 2021, 20:40

It was the size of the chunk I had laying next to me with ends stripped and tinned. Everything was breadboard, so slapped it on and boom. Worked