Start in neutral

[The Mighty Gryphon]

I'm going to guess that there is an output transistor in the cluster that has been toasted.  I sort of figured that the Bendix would draw more current.  The solenoid has to overcome the spring that disengages the starter gear when the starter is turned off in order to close the power contacts for the motor.  That is going to need a bit more current than 200ma.  Possibly on the order of 2-3 amps.

The clutch switch may have sent enough current to get the solenoid to sort of make contact, but not with enough force to get a good solid low resistance connection for the motor current. 

I think the load shed problem is related to the higher current flowing through the new starter and where the activating voltage drop for the load shed is coming from.  Are you taking it from the solenoid terminal or the terminal on the motor itself.  I would think that the voltage at those two points will be different.

[Tcapling]

Taking load shed from the motor side of the solenoid  same as the oem set up

[Tcapling]

I'm going to guess that there is an output transistor in the cluster that has been toasted.  I sort of figured that the Bendix would draw more current.  The solenoid has to overcome the spring that disengages the starter gear when the starter is turned off in order to close the power contacts for the motor.  That is going to need a bit more current than 200ma.  Possibly on the order of 2-3 amps.

THe clutch switch may have sent enough current to get the solenoid to sort of make contact, but not with enough force to get a good solid low resistance connection for the motor current. 

I think the load shed problem is related to the higher current flowing through the new starter and where the activating voltage drop for the load shed is coming from.  Are taking it from the solenoid terminal or the terminal on the motor itself.  I would thing that the voltage at those two points will be different.

After reading all the reply I believe you are 100% correct

[caveman]

Tcapling,

That is indeed an interesting project you have going on there!

I am curious if with that starter circuit you are getting the proper ground through the starter for the load shed relay?

Hope you keep us up to speed with this project.

[Laitch]

There are a few diesel creations out there; I only knew about the military KLR650. Good luck with this. Maybe a Smart Bike would help some riders from the get-go. 

[The Mighty Gryphon]

My bikes are already smarter than I am.

[rbm]

Seem to be a bit late to the party.

Make sure that the relay you have installed ahead of the solenoid has a snubber circuit across the coil primary.  This is simply a general purpose diode installed in reverse across the coil.  This snubber prevents electrical spikes travelling down the ground wire and taking out the transistors in the ICU.  If those transistors get destroyed, you'll have to replace the entire ICU.

As for the GPI board, the neutral / start enable driver transistor is probably bad and needs replacement.  This could happen during your testing where the attached solenoid tried to draw too much current.  Refer to the schematic for the GPI board here:  REFERENCE: Schematics for OEM Circuits

Look for the GPI schematic and click on large schematic link.  Find the Start Enable circuit.  V8 and V9 form a Darlington driver which turns on or off V10 when a neutral condition is detected.  In your case V10 is probably toasted and needs replacing.  It's a BD370 which are NLA meaning they are very hard to find but can be found on eBay.

[Laitch]

Factory battery turns it over quite well surprisingly. But  only so many times.

Sadly for me, i'm stuck back at the battery cranking amp/diesel engine interaction square. What does "only so many times" mean—2, 4, 16, steady for 25 seconds? Does the engine have a fuel/block heater?

[Tcapling]

Sadly for me, i'm stuck back at the battery cranking amp/diesel engine interaction square. What does "only so many times" mean—2, 4, 16, steady for 25 seconds? Does the engine have a fuel/block heater?

The engine has glow plugs, which are triggered by the ecu only when the coolant temperature is below a certain point. If I hook them up their use certainly reduces the number of times the factory battery will crank it at speed.
 While testing the various circuits I have cranked it with out starting (no fuel tank on the bike) an easy 20-25 times while still maintaining the higher crank speed at the end when tested direct. A 25 second crank is supported no problem.

For comparison sake, I have had a high CCA battery from my diesel truck hooked to it and behaves the exact same way.
BecUse if this I don't believe that the battery is the source of my current issue.

[Tcapling]

There are a few diesel creations out there; I only knew about the military KLR650. Good luck with this. Maybe a Smart Bike would help some riders from the get-go. 

There are definitely a few out there. The hdt klr was my first choice. But they are slow. No part of that would be enjoyable on anything faster that country roads.

The forum section of that site is where I took my inspiration for this project from. I am not the first to make a smart powered k100. There are at least 4 others running around in the u.k. and Europe.
There was even a Dutch (I think( company called evatrack that made a bespoke smart powered adv bike though it used a cvt instead of a transmission.

Mine will just be a diesel first at this particular rally.

[Tcapling]

Seem to be a bit late to the party.

Make sure that the relay you have installed ahead of the solenoid has a snubber circuit across the coil primary.  This is simply a general purpose diode installed in reverse across the coil.  This snubber prevents electrical spikes travelling down the ground wire and taking out the transistors in the ICU.  If those transistors get destroyed, you'll have to replace the entire ICU.

As for the GPI board, the neutral / start enable driver transistor is probably bad and needs replacement.  This could happen during your testing where the attached solenoid tried to draw too much current.  Refer to the schematic for the GPI board here:  REFERENCE: Schematics for OEM Circuits

Look for the GPI schematic and click on large schematic link.  Find the Start Enable circuit.  V8 and V9 form a Darlington driver which turns on or off V10 when a neutral condition is detected.  In your case V10 is probably toasted and needs replacing.  It's a BD370 which are NLA meaning they are very hard to find but can be found on eBay.

Tha KS for the very well put together info on the Internal circuitry of this dash. It made it extremely easy to zero in on the v10 transistor, which as you and others have suspected, has crapped the bed. Now to source a replacement and reassemble.

Thanks again

[The Mighty Gryphon]

That transistor is probably just a switching device that turns on when the transmission is in Neutral.

There are dozens of transistors that would work in it's place.  Maybe Rob could recommend one that would work.  I would guess that the operating voltage and current capacity would be the key specifications to find a replacement. 

You could also do a search for "BD370 substitute".

[The Mighty Gryphon]

Here, I saved you 2 minutes:

https://alltransistors.com/transistor.php?transistor=25149