Strange problem..

md95 · Post by **md95** » Wed Jan 09, 2008 5:48 pm

You just have to be careful with going too large a capacitor,as it can filter TOO much of the signal. Jerry says the .1uf was fine, so i will choose to use the 470 and the .1uf cap. I should have those laying around..was going to do this tonight, but i cant find my multimeter..i need it to find/double check the proper resistor/pin/and ground on the MS2 card..guess it will have to wait until tomorrow so i can get my other meter from work(ehh its a much better one anyway

AbeFM · Post by **AbeFM** » Wed Jan 09, 2008 6:00 pm

I put my caps all on the proto area, not directly on the MS-II card or anything like that.

If you're using the "standard" MS-II pins, it's JS-10 for cam (1 pin before the 12V supply), and 3rd pin after that for the crank, but just use TSEL instead.

Still, I did everything to the wires that FEED those points. I don't think you want to put anything on the card! You want your caps one series resistor after the raw signal coming in. 0.1uf is correct (I think you'd said 0.01 before which is why I was confused).

I bet it would turn out. Then again, if you are up soldering EVERY night all your friends will forget what you look like.

md95 · Post by **md95** » Wed Jan 09, 2008 6:41 pm

AbeFM wrote:I put my caps all on the proto area, not directly on the MS-II card or anything like that.

If you're using the "standard" MS-II pins, it's JS-10 for cam (1 pin before the 12V supply), and 3rd pin after that for the crank, but just use TSEL instead.

Well i'm using a v2.2 board, so no proto area. Also, remember the JS10 input ON the ms2 card has that 1k resistor. Its on the card itself(looking at the MS2 card schematics, it's R9). JS10(v3 board), or JP1 pin 5(v2.2 board) goes to processor socket pin 17 of the 40 pin dip. From there, the signal goes through the on-card 1k resistor, then off to the MS2 CPU pin.
So the only way to get the cap between the 1k resistor and the CPU pin, is to do it right on the MS2 card. Jerry was saying when the cap was BEFORE that on-card 1k resistor, it didnt work.

AbeFM wrote: I bet it would turn out. Then again, if you are up soldering EVERY night all your friends will forget what you look like.

Heh, you should see my kitchen table..covered with MS parts, soldering stuff, wiring tools, you name it lol.. Working full time 6 days a week, keeping up a home, and a girlfriend to boot, i take any chance i get for my play time

I was too impatient to wait, i ran to autozone and picked up a cheap meter..only need it for continuity tests anyway

Time to get to work.

md95 · Post by **md95** » Wed Jan 09, 2008 8:44 pm

Ok, I replaced the 2nd trigger pullup to a 470 ohm. I then put the .1uf cap on the MS2 Card. R9 on the card is the correct resistor that goes from input pin 17 of the 40 pin dip(JS10 in, or pin 5 of JP1 header on 2.2 baord), and there was a tiny pcb hole just to the left of the resistor. I verified that this hole was after the resistor, and before the actual CPU input pin. Perfect. Since the hole was sooo tiny, the cap's lead would not fit. So i took a wire strand from some 16 gauge i had laying around, and it fit nicely in there, and soldered it to the card(of which was extremely delicate i may add), and then soldered the cap's lead to that thin wire. Not the strongest connection, but it should do ok. Then for the ground connection, i opted to use Pin 2 of JP1 header on the card. This pin is ground, so i thought that was the easiest to grab ground right there. Pin 2 on the 40 pin dip is also ground, so i could have soldered it to that small nub on that pin, but i thought this was easier. Now for the fun part.

I took the car out and drove a *good* 40 minutes on the highway. I started the datalog, and if a sync loss happened, i was ready to hit the space bar to mark it. Well, i must say, i never used the space bar.

I'm not going to claim 100% victory here, but before i re-modded the circuit, I would say the issue would have happened about 85-90% of the time, within 30 minutes of highway driving. I dont have any long trips planned for a while yet, so, until then i wont have 2+ hours of test time, but either way, it was a 100% successful run.

A big thanks to Jerry, Bruce, Matt, James, Ken, and whoever else helped solve this mysterious issue. I've been dealing with this for the past 10 months since this 4/2 CAS code's creation.

Moral of the story: Dont underestimate the power of a tiny capacitor!

I attached a pic of the MS2 card cap mod, its the best pic i could get with my camera. Dont worry, i'll insulate that cap with some heat shrink later..i was too impatient to do it lol..

Oh, i forgot to mention- R9 on the main site's schematics, shows its a 1k ohm resistor. Well, its not. Its a 331 ohm resistor(verified with the meter across the resistor). Its even stamped on the tiny thing. This is a newer blue card, so maybe they updated this resistor value?

6040solder · Post by **6040solder** » Thu Jan 10, 2008 3:43 am

Fantastic that it worked, but :

Pleeeaaaase put some heat shrink over that soon. Thats scary.

AbeFM · Post by **AbeFM** » Thu Jan 10, 2008 10:15 am

I spent like two weeks at work playing with staking, potting, and conformal coating, head buried in NASA and MIL specs the whole time. Now that I've got it down, and I've got all the materials and support equipment, I keep getting tempted to do parts of my board - but I'm sure I'll want to change it so I'm resisting that temptation as well as I can.

FoundSoul · Post by **FoundSoul** » Thu Jan 10, 2008 1:30 pm

Excellent-- glad to hear this is working for you man, and great job adapting the mod for MS-2 users making it easy.

So far all is still going well here on this car also--

md95 · Post by **md95** » Thu Jan 10, 2008 2:16 pm

FoundSoul wrote:Excellent-- glad to hear this is working for you man, and great job adapting the mod for MS-2 users making it easy.

So far all is still going well here on this car also--

Sounds promising! How exactly did you do the cap mod on yours? Similar to mine?

FoundSoul · Post by **FoundSoul** » Thu Jan 10, 2008 2:41 pm

Mine was an MS1E unit on a v3.57 board. Cap goes from JS8 to GND.

md95 · Post by **md95** » Thu Jan 10, 2008 6:11 pm

FoundSoul wrote:Mine was an MS1E unit on a v3.57 board. Cap goes from JS8 to GND.

Ah, ok. I was under the impression it was a MS2. Oddly enough, i've done at least 3 MS1 Extra installs on 4g63's, and i've never heard the owners complain about this issue..I've ridden and tuned their cars for several longer drives, and never noticed it myself. Weird. 2 of the installs were with the 4/2 CAS on 1g's, and one was on a 2g 4g63 using the separate crank and cam sensors..didnt need the CAS at all, since those separate sensors provided the same signals as the CAS. I have not yet tried the 2g sensors on a MS2 yet though..would really like to see if it works.

bcrx7 · Post by **bcrx7** » Thu Jan 10, 2008 6:33 pm

Glad to see this. I reported this problem with the MSE1 HighRes couple of years back and never really got any good answers and just lived with it. Now I can fix it after the engine gets rebuilt (melted piston from a faulty BRAND NEW fuel pump).

Thx guys.

FoundSoul · Post by **FoundSoul** » Thu Jan 10, 2008 6:46 pm

This was the same wheel pattern as the 4G63 uses, but in a Miata-- could be the Opto sensor or wiring in the Miata is more susceptible to noise...

md95 · Post by **md95** » Sat Jan 19, 2008 7:18 pm

Just an update: I have not had one single issue yet since the cap mod on the MS2 card. I drive it to and from work every day, and had a few short highway trips as well. So far so good!

FoundSoul · Post by **FoundSoul** » Sat Jan 19, 2008 7:54 pm

That's great man! I've made the change on 4 cars now, all with perfect results so far. I held off release of a new MSPNP product for a while because I wanted to track this down, and I'm completely relieved that it's nailed and I can move on.... Release will be early Feb for the 96-97 Miata MSPNP MM9697.

Post by **jsmcortina** » Sun Jan 20, 2008 1:01 pm

One reason for a possible difference between MS1 and MS2.

MS2/Extra (2.0 code) is using an input-capture interrupt driven second trigger. Any noise is likely to trigger the input.

MS1/Extra is polling the pin every 0.1ms. With a short burst of noise you might get lucky and the code will miss it.

James

md95 · Post by **md95** » Wed Apr 09, 2008 2:13 pm

Just wanted to give a final update to this issue. It's 100% fixed. Had a 1.5 hour drive last weekend, each way, not one single issue. Ran beautifully. In fact, since my last posting on this topic, it's never missed a beat as well.. Case closed!

Keithg · Post by **Keithg** » Wed Apr 09, 2008 4:21 pm

md95,

You made that cap mod on the MS2 card. You also could have made it on the V3/V2.2 board as well and merely shorted that 1k resistor. I think I would have been more likely to have done that and put my low pass filter on the board instead of the MS2 card.

Heat flow the resistor off the MS2 card and short it with a small bit of resistor lead. That way your input circuit can be totally off the MS2 card. I did not catch this fact of the input resistor on the board. My 0.1uf cap is across the output of the VR input op amp as that is my 2nd trigger conditioner on this new board. Previously, I was using an input resistor, pullup resistor and a 0.1 cap on both an MS1 chip and an MS2 card. Both with similar results.

YMMV,

KeithG