Old Installation new laptop

[croc]

This car has been running fine since its one and only dyno session in 2004. It's possible that some minor changes were made in 2006, but I don't think a PC has been connected to the K-Pro since then. My trusty XP laptop was stolen last year, and the insurance company thoughtfully provided me with a Windows 7 one to replace it. After a year I still hate Windows 7 more and more with each passing week....

Last weekend there were some issues during a rally. In several stages, after 20 km, the engine died, but resumed running again immediately the key was turned off and then on. Later in the rally, it very definitely had lost power although it was running to full revs. After the event, all seems well again. But we cannot communicate with the ECU. Andrew Short's XP laptop will not even recognise the USB connection.

ECU is 37820-PRC-013, in a genuine JDM DC5 Type-R

This is the start of the calibration file I downloaded and saved in 2006:

;
; Hondata K-Manager Calibration
;
; Don't edit this file directly
;

[Application]
Version=1.2.0.3
SaveDate=17/04/2006 15:01:05

[ECU]
Hondata=0
Description=
ECUType=3
SerialNo=295
HardwareVersion=256
FirmwareVersion=256
SoftwareVersion=256
CopyProtected=0
Checksum=-1433250106
NumTables=6

I had an issue in the past where an upgraded S-Manager version managed to corrupt the whole calibration, and I don't want to repeat that. It became a very long day on the dyno. We want to datalog and dyno this car, but first we want to download and save what we now have in case we need to repair or replace the ECU. Would 1.2.0.3, even if we could get it, work in Windows 7? (Quite frankly I have no confidence at all with Windows 7, and I am sick and tired of upgrades, to every device I now seem to own, most of which cause major problems which were never there in the original versions...)

We have a rally series to try to compete in. What is our next step?

[Spunkster]

What shwos up in the Windows device manager when you plug in the K-PRo to the USB on the laptop?

[croc]

While you guys have been enjoying your summer holidays, we've been busy in the cold winter rain!

I took it to Torque Performance, and their XP laptop couldn't find the K-Pro either. I didn't see exactly what the XP computers came up with, but although they detected it, it didn't identify itself as a Hondata product. My Windows 7 one showed a yellow exclamation mark with a description something like like "USB-serial device", even after I uninstalled it and plugged it in again. Nevertheless at one point I seem to have been able to download some of the calibration. And then I couldn't again. Of course, while I think I downloaded and installed the drivers correctly, I'm not totally sure...

Andrew has another ECU to send you, so that one and our one will be on their way to you tomorrow.

Thinking ahead in case the ECU/USB issue is not why the engine was playing up, we are considering swapping the expensive Honda wideband sensor for the cheaper Bosch-type. Do you think this conversion is a good idea? The exhaust takes a battering from rocks and needs rebuilding every year, so the sensors will have had a few good shocks.

One more thing: to make the car mobile we fitted an EP3 Type-R ECU, and I have been looking at what that shows on a scan tool. Most of it looks OK, but I am puzzled that when stationary, the VSS fluctuates between a reading of 0 and random values of 7 or 8 and up to 12 km/h. While moving, the readings look normal and steady. And the reverse lock-out seems normal. When stationary but showing 12 km/h, you can still select reverse...

[Spunkster]

Send an email to support <at> hondata.com with the following and I will issue an RMA and give you further
instructions (include a link to this thread):

Name:
Address:
Phone:
Email:
Serial#:
Description of problem:

Please do not send anything without first obtaining an RMA#.

[croc]

Andrew from RallyAs is organising the repair and shipping. You will have heard from him. I'm not going to add to the confusion by getting two authorisations for the same ECU.
I really wish we could have datalogged it first.
Is the random VSS reading from the EP3 ECU while stationary a matter for concern?

[Spunkster]

It may be an issue if is doing this on a stock ECU.

[croc]

Ok I'll check all the grounds and if that doesn't fix it I'll see if I can get a good scope and analyse the signals.

[croc]

With the new, updated Kpro board in the ECU we can now communicate with the various laptops. But on engine start, the datalog went offline. The car has been running the same Champion C61 plugs for 8 years or so with no problems at all, including the dyno session, and they are still in perfect condition.

After a lot of checking grounds and shielding various components I remembered that the initial installation instructions contained a caution about using non-resistor plugs. Despite us not having any issues up until this year, I found four second-hand plugs and fitted them. Three old Splitfires and one even older Champion RC7YCC. Instant fix. I don't trust the Splitfies, which apparently cross-reference to the standard 7 heat range.

It took a surprisingly long time to find someone who stocked colder than standard plugs with a 5/8 hex, but eventually I was able to locate and order a box of NGK PFR8Bs.

On fitting them: - again the laptop disconnects from the Kpro as soon as the engine starts! Putting a shield over the top of the coils helps a little; grounding the ECU case helps a lot. Neither of these two had made any difference with the C61s, and the USB can't be plugged in with the ECU bolted in place.

Your link to the NGK website claims a 5k ohm resistance in the plugs but these measure between 3.6k and 4.1. (The Splitfires are 4.9 to 5.1k and the Champion measures open-circuit.)

Is this the problem? If so, how is it fixed?

[Spunkster]

The ECU case should not be grounded as nothing inside the ECU should be grounded to it. Have you tried this ECU in a known working car? Does this vehicle have a wire tuck?

[croc]

I can assure you that grounding the case almost eliminates the datalog drop-out. I assume that this is shielding the ECU internal components. The car is a JDM DC5, one of a limited run originally produced by Honda for the Japanese Production series, chassis number DC5-1011790. They had no internal trim, and although the wiring harness looks normal at first sight, they were made with only the wires needed for racing - we have had to add a heater/demister, horn and all wiring for these. All the engine, ELD and instrument wiring is there, but it doesn't have any reversing light, door switch or interior light wiring.

We have not moved or removed any wiring except that last year we relocated the battery to the rear floor. With the large diameter welding cable we used for the battery and engine ground cables, it actually cranks faster than it used to using the thin original cables. But the main cable does come through the firewall and runs close to the ECU.

I am about to try to track down something like the Radioshack "Snap-Together Toroid Choke Core RF Choke Catalog #: 273-104".
If I can find some toroids to go over the ignition coil connectors and the battery cable I will try them.

[croc]

When Andrew Short received the ECU back from America last week he put it in a standard DC5 and found no problems.

[Spunkster]

This would indicate it has to do with the wiring in this vehicle...perhaps the wiring is not properly shielded.

[croc]

Can't find any BKR8EIX, but have been told IRIWAY8 is equivalent. However at nearly NZ$200 a set it's an expensive experiment...

[croc]

What shielding would be needed?

[Spunkster]

I have never gone through a harness to find every wire that is shielded, but many of the wires and groups of wires are shielded. This helps prevent interference. Especially those sending signal back to the ECUs from sensors.