Hey, I'm far from qualified.
It is surprising that the problem is worse with a rise in humidity; ESD problems often get worse with dry conditions.

Kent on Aeroelectric had some good thoughts (cold solder joints); that general path might be expanded by looking at dust/dirt on the pc board(s) attracting moisture & forming conductive paths when humid. Ken on Aeroelectric also made a good point about not ignoring grounding issues, and using a (low) heat gun on/around the EM2 circuit board(s) could dry  out selective areas to see if that alters symptoms.

Finn makes some good points, too. I don't own an EM2, but I agree that it's hard to see noise on the injector pulse width/count lead scrambling anything but fuel data. It looks like *all* the displayed data is wonky. Where's the injector sense lead attached; near the EC2, or out at the injector? The injector impedance should be quite low (related to RF sensitivity), which should hold its control line at low impedance, as well. But the sense terminal in the EM2 is likely at high impedance, meaning its wire is a better 'antenna'.

That business of only grounding one end of a shield should not be taken as a universal truth. In this application, the EC2 and EM2 should be at the same ground potential, and you certainly aren't going to damage anything if you ground that shield to both the EC2 and EM2, to see if that changes symptoms.

Beyond that, we'd just be 'spitballing' about causes/solutions. Old PVC-insulated wire can absorb moisture in the insulation, but I wouldn't expect it to quickly track humidity changes once the insulation had absorbed moisture. If you implemented the injector disable switches, try disabling the primaries (leaving your sense line connected) to see if symptoms change. On the outside chance that the EC2's injector driver is somehow 'noisy', you might try moving the EM2 sense line to rotor 2 primary (pin 36?) to see if symptoms change. (I don't know if this would affect the data collection in other ways, but might be worth a shot.)

In the end, I suspect you need for Tracy to re-surface. :-)

Charlie

On 11/5/2019 9:01 AM, Finn Lassen finn.lassen@verizon.net wrote:

I'm sure that Charlie is much more qualified to answer this.
That injector lead could act as an antenna picking up all kinds of noise from the engine compartment.
Still I don't see how that would scramble the EM2 display.
I see two possibilities:
1) Much higher frequency (than 100 Hz) noise on that input line confusing the software. A noise filter should handle that. I don't know how it works, but try a ferrite bead over that line near the EM2 connector. And/or an RC low pass filter. Maybe 10 Kohm and 0.47uF or 1 Kohm and 4.7uF?? (1,000 HZ cut-off).


2) The noise from that line bleeds over to the data and control lines to the display. Try shielding all wires going to the display.

Finn

On 11/5/2019 4:12 AM, Todd Bartrim bartrim@gmail.com wrote:

Below is an email that I just sent to the AeroElectric list, but as many of you on this list also use the EM2 but may not be on that list, I thought I would send it along here too.

Charlie, I know you are on both lists, and are a frequent contributor on the AE list, I figure your insight to the EM2 could be invaluable??along with anything Electric Bob can add.

But I welcome any advice from anyone who may have also encountered this issue. I have sent an email to Tracy, but as he seems to be offline up in Colorado, I'm hoping to be able to solve it without bothering him too much.

Thanks, Todd

I have an issue with my engine monitor display where the display will scramble leaving me with a meaningless display as can be seen in the attached picture

The engine Monitor is from Real World Solutions and is the very early first edition EM2 (serial #003). I also have an early edition EC2 (engine ignition and injection controller).

RWS was Tracy Crooks company before he retired, and he still provides some support when he can, however he also is enjoying his retirement??offline so is often hard to get a hold of,??

so I thought I would pose the question here while awaiting a response so maybe I won't even have to bother him.

?? So here are the details of the problem. The engine monitor has 36? different parameters that it displays on 8 screens. There are 24 different sensors (pressure, temperature, level) with the remaining being provided through a serial data link from the EC2 along with an EFI sense which is connected to the #1 primary injector output from the EC2. I've determined that it's this last wire that is causing my grief.??

This EFI sense is measured for it's pulse width in order to calculate a surprisingly accurate fuel flow, RPM, HP, fuel remaining, MPG (for efficiency). But the noise on this line seems to be the cause of the data scrambling on the display. Simply pressing the cancel button (there are 4 user buttons to control display) will rest and clear the issue.??

?? ??But here's the weird part. It seems to be related to relative humidity. During hot dry, summer weather this display will rarely scramble, sometimes going up to an hour without an issue. But on cooler humid days I would have to reset it every few minutes which gets a little annoying. However right now as we are not enjoying the cold November rains before winter will arrive with cold dry weather, I'm finding that it is scrambling almost immediately after a reset. This makes no sense to me, but I've observed this pattern enough to be sort of certain that it is not a coincidence... but I could be wrong.

In any event, it matters not whether it is or isn't triggered more by the humidity, but what it did allow me to do is determine where the issue is coming from as before it was too random to allow easy troubleshooting but now as it is more consistent it was quite quick to narrow it down.

?? I've cut this line (which is shielded with only one end of the shield grounded) and installed a connector. With it disconnected, the EM2 display is rock steady (up to a 10 minute ground run) but with no RPM or any fuel, power, or efficiency data. Connect the line and it immediately will scramble. I estimate that the frequency of the injector pulse on this line would only be at max about 100Hz at 6000 RPM. I've considered that a simple low pass RF filter to ground may work as the frequency that I'm trying to measure is well below what the filter would remove, but would it still have an effect on the pulse width? which would be undesirable as that would effect the signal driving the #1 primary injector.??

?? I'm hoping that someone can suggest an easy fix so that I can implement it when I get back up to the airport this weekend, other wise if more data is needed to help find a solution I will take a scope up there and try to get some more detailed data.

??Right now our days are getting pretty short up north here so it's dark by the time I get off work, so I'm restricted to weekends only. Thanks for any help that can be provided.

Todd Bartrim

RV9 13Bturbo

Virus-free. www.avast.com

jlepgepifgicdmnj.png
EM2 display_scrambled.jpg