To Motorandy123.

Since the thread was deleted in the other forum, I`ll post my answer here.


If pin 6 haven`t been powered, you would have other problems to...dark BMU light, oil pressure, charge, low fuel warning (if low fuel), power to both instruments.

About counting the pins.....wiewed from the rear of the cluster...start top left and down, continue on top right...24 all together.

Inge K.

Well I have verified that I have connection from the GPI to pin 6 for power and to 14, 17, 24
for ground. I have power going to pin 6 with the key on. So I guess my GPI is dead. I'll take it
out and power it on the bench today. There's only a couple of CMOS chips on there not much to
fix it. I guess I disturbed things a bit when I opened it up. I might get some deoxit and go over
all the connections inside the dash. I had somehow gotten a couple of drawings that showed
the power on pin 24 and I didn't have a schematic of inside the dash. You cannot trace the circuit
very easily on the flex board and I don't have a good one to go by. The clutch switch is dead and
the GPI so I need to get one working to start it. I ordered a clutch switch but while that is in
shipping I'll try to get the GPI working. Thanks for all your help!

Motorandy123 wrote:Well I have verified that I have connection from the GPI to pin 6 for power and to 14, 17, 24 for ground.

??......ground is at pin 13, 18 and 23.

14=ground to BMU bulb.
17=power right turn indicator.
24=power instrument illumination.

I`ll guess you have measured ground through the bulbs.



About GPI, uppermost connector is ground...power is the opposite connector of the six.
Is it totally dead, no segments showing?

About the dead clutch switch, cut the wire... splice the ends together........then you are able to start, untill you get parts.

Oops, I was counting the pins like an IC
1 24
. .
. .
. .
. .
12 13

I know I can short the clutch to get it to go. I've ordered a new switch so will wait for that
while I figure out the other problems. I took the GPI out today and applied power to it and
nothing! I metered the transistors and they read OK. The LCD itself had some whitish stuff
around it so I removed it and cleaned everything with alcohol. Still no luck! The harness
shorted at the headset so could have spiked the power and fried the CMOS chips on the GPI?
I may have to buy a new one or make a relay replacement to get the neutral signal...

hmm interesting reading

currently playing with the gpi board and gear indicator display (thanks paul )

got a circuit and board worked out for a led display as direct replacement for the lcd indicator

just have to figure out the right amount of led current for sufficient brightness durring the day but not too bright for night work

"got a circuit and board worked out for a led display as direct replacement for the lcd indicator"


I like the sound of this

I used to work nights and found most LED's too bright at night. You need them dim at
night and bright during the day. I got rid of my Datel LED meter and got a LCD voltmeter
because it blinded me at night.

I like the idea of a gear indicator but am leaning towards "faking" a neutral
signal. If you look at this diagram;

bit: 2 1 0 display
------L L L 0 (N)
------L L H 1
------L H L 2
------L H H 3
------H L L 4
------H L H 5

If I use bit 0 and send +12V to the switch I can get a high in neutral and low in first (of course it
would be high in second). With that and the clutch switch I think it would be OK....

About the dead GPI...we did also have this same issue a ~couple of months ago.
I think that the lad who started that thread did find them cheap at Palo Alto Instruments (who also repair the GPI circuits).

In the other forum you also mentioned that you got a problem with instrument light and headlight.
I would guess that this is related to RHS switch assy, measure at the connector under the tank to confirm.

Dismantling RHS switch assy

Motorandy123 wrote:I used to work nights and found most LED's too bright at night. You need them dim at
night and bright during the day. I got rid of my Datel LED meter and got a LCD voltmeter
because it blinded me at night.

I like the idea of a gear indicator but am leaning towards "faking" a neutral
signal. If you look at this diagram;

bit: 2 1 0 display
------L L L 0 (N)
------L L H 1
------L H L 2
------L H H 3
------H L L 4
------H L H 5

If I use bit 0 and send +12V to the switch I can get a high in neutral and low in first (of course it
would be high in second). With that and the clutch switch I think it would be OK....

already on the gpi board ...but you could fake up a triple input gate and an output drive transistor, pretty easy ...with a diode to isolate the gpi output to the clutch lever switching

about the led display ...still thinking on ways to current limit night operation ...some ideas ..just working it out to do

On the headlight, the left hand switches were melted so the PO got a used one off EBAY, well that one was bad! It didn't do high or low beam as the center contact was sunken. I heated it up and pushed it in a bit.
Now I have contact on high only as the one I pushed in went too far and it holds the contactor up in that
position. So I'm going to add a relay for the headlight and trigger it off the high beam line.

I'm still looking for a elegant solution to the neutral signal. Do you think something like this would work?

basically no ...

the switching of the neutral lamp is to earth not b+ (corrected )

maybe the transistors could turn on a relay perhaps, if you gave them bias and picked up a current limmited feed out the common to the gpi switch, the relay needs to handle up to the lamp wattage ...but make sure with something a bit more like 1 amp +

dont forget it switches other stuff as well

I was looking at this diagram and it looks like we feed +12V to the neutral light and the starter relay so
I figured 200ma max. The original GPI only has a few tiny transistors...



Hmm this shows we need to ground the neutral light....

yes your right ,,,dont know what i was thinking

but one of the transistors used on the gpi is a bd370 ...good for 1.5 amps max

so its got to illuminate the globe and power the relay ...lets say with head room .5 to .75 amp max ? (globe = 3 -4 watt /12 volts somewhere around .3 amp ? ) not sure about the relay current but lets say under .1 amp maybe ? ))

dont know if the labeling on the circuit is correct ...might be just confusing as it is possibly overlaid on the circuit says charge rather than neutral indicator or just forgot to label the neutral globe

so the 2n2222 are good for maybe 200 to 350 ma (.2 to .35 amp ) max ?

I just looked up the specs on the BD370B (1.5 Amps!). I have reverse engineered the schematic for the
neutral light driver on the GPI. There's not much there but they are using some funky parts.
I don't think I have any similar ones in my junk parts bin but it looks like fixing it might be
my best bet. I'll have to meter around a bit more to see what's bad...

yeah the cmos dont cost much , possibly under 6 bucks out here for both ...retail

plus the trannis ...maybe 10 bucks all up ?

i would have a poke around the zenners though and the curent limit resisitor before it ?

Having trouble finding a transistor that I can buy 1 or 2 of. Doesn't cross to american parts. Do you have
any suggestions?

um look up bd139 good for about an amp ...but different package flat pac style (sorry they are common as anything out here ...) trying to think of something that radio shack sells ...my memory isnt that good from 20 years ago now

ill go have a look at their catalogue brb

nte 383 possibly ? npn with about 1 amp should work ?

I have a few tiny relays maybe swap it out for one of those?

About light problems: the reason to that I suspected RHS switch assy (or connector) is that you have lost both headlight and instrument light, which both (and park+tail)is feed from RHS sw.a.

A problem with the LHS sw.a. would only affect the headlight.

But of course, it could be more than one fault involved in this.

If your park+tail also don`t function, but do function with the ignition switch in the park position..........you also got a problem with the RHS sw.a. (or connector).

I've designed a circuit to simulate the starter enabler when in neutral as well as decode the TGPI and display the gear. I'm at the breadboarding stage and have the final PCB designed. It can do either function or both functions depending on how the board is populated with components. I'm seeing many posts recently both here and on other forums where people are needing such a separate TGPI enabler circuit. In any ase, when the breadboard is debugged, I'll make the plans and documentation available. In addition, I'll commission some PCBs to be made up where members can buy from me or order one themselves.

My circuit is below:


The 4025, BD140 and 1n4001 constitute the Neutral detection circuit that will enable the starter and light the neutral lamp. One could populate the PCB with just those components to get just that function. The 4011 and LED is the gear postion indication.

P.S. the starter relay draws 540mA and the neutral lamp about 160mA for a total of 700mA approximately.

Have you figured out how to dim the display? A light sensitive resistor to make it bright
during the day and dim at night would be nice!