Hi guys,

I've been having some tacho issues where having my tacho connected to the -ive side of the leading FC coilpack, it goes erratic after 4500rpm. If I connect it to the -ive side of one of the trailing coils it only seems to read at half speed, and if I use the diagnostic output on the leading coil pack it works much better up until around 6-7000rpm where it starts to go a little funny and bounce around a little.

I was thinking of using the tach output function of the megasquirt, as the rpms show fine in tunerstudio and megatune. However I'm a little worried there might be something wrong with the ignition side of the megasquirt causing erratic spark at high revs.

Is this kind of thing normal to experience when connecting an aftermarket tacho? (SAAS 95mm tacho)

Should I just connect it straight to the computer and forget I ever had this issue? I'm running ms2 V3 with 2.0.1 firmware.

Any thoughts or suggestions will be much appreciated.

Regards,

Ben.

 

You should probably upgrade your firmware to start. There were some bugs in rotary trailing support in 2.0.1. That said those bugs wouldn't cause what you are seeing.

Stock FC coils are very noisy. One thing people do to get rid of the behavior you are seeing is to put a 50k resistor inline between the coil and tach input.

Running the tach using MS will also work if your tach is capable of triggering off a 5v logic-level low current output.

Ken

 

Hi Ben,

I pick up my tacho signal from the trailing coil pack out. Not sure what wire colour it is. But its certainly NOT half speed. I'm not picking it off the coil -ve, there is a wire for the tach signal out I believe.

Tacho bounce at high rpm will be dwell errors. I have read on another forum that the input dwell does not always give you what you expect.

I suggest using the same very old 2.10d firmware from ken. As you know my 12A is still holding together with good compression after 4 years. 2 years using this firmware now.

I need to make changes to support my current power level and more power and I just know something that 'should' work is not going to.

Set dwell to 3.3mS and 0.1mS minimum discharge. Disable all noise filters and route your loom like an OCD effected person

 

The only reason dwell would change is if the voltage from the alternator changes.

On MS2/extra, the dwell is normally VERY stable.

Ken

 

Thank you very much for your help guys, I will investigate these suggestions and post my results. I have very limited internet access at the moment - only at work really. I may in fact be running 2.1.0 rather than 2.0.1 (2.1.0RC3 or something perhaps?)

I will also have a double take of my trailing coilpacks and see if I can locate the tach out wire and I may also try Ken's suggestions of putting in a 50k resistor - in fact I might try this first.

I also understand I can build a circuit to give me a suitable input signal for the tach if it doesn't support the 5v low voltage option from the MS. This option however I cbf following

oh and Josh, I'm far too lazy to do anything in an OCD fashion haha.

Once again, thanks for your help.

 

That is good to hear. So for my own personal 'errata' records, which versions do have well understood discrepancies?

 

The dwell and spark in ms1/extra can have considerable jitter.

The dwell b/c it runs from the 0.1 ms section above around 2500 rpm, and the spark because there's 1 timer to do it and it's being done in software (so it has to fight with all the other interrupts to run) on a slow processor.

I would not expect a large amount of jitter on ms2/extra even though the spark pins are toggled in software there as well since we have 4 timers (between leading and trailing dwell and spark). There will be some jitter there though.

MS3 should show less than 1 usec of jitter if any at all since all the outputs are toggled on the XGATE coprocessor while all the major work happens on the main processor.

Ken

 

well the first thing I tried last night was putting a resistor in-line with the tacho signal. I had a 47k resistor floating around so I installed it and bingo! No more tacho problems.

Thanks a bunch Ken, although I don't quite understand why this value did the trick... it's beyond my comprehension. Ah well, it's working now and I'm happy.