Has anyone managed to hack or replace the instrument panel?

[rcx348]

My car is fitted with an early dashboard, the smaller one with a shelf in front of the passenger seat, but my instrument panel is from a later model so it doesn't fit. I can't swap the instrument panel because the early ones have different plugs.

I'm thinking of fitting a universal digital instrument panel. I've done this before on several motorbikes, but on a car there's a lot more warning lights to deal with. Has anyone managed to connect universal warning lights to the existing wiring? Is it simple, eg one wire per warning light? Any help appreciated.

[fixkick]

ok custom car, 95 G16B 16v with 97 engine and matching 97 ECU. MPI engine, with unknown distributor. (non usa are unique)

it's very simple , unlike new cars with VFD/LCD or computer displays, etc.
most lamps (not blinkers. or high beam, or background) use a 12vdc ignition power feed.

you need to have a working speedo , with working VSS or the ECU will never work right.
just about any light will work, but for the ECU lamp CEL do not exceed the current stated. for the stock GE#18 bulb (transistor driven it is and tiny)
current max is. (ok ill paste my hard to buy parts page quote, here)
Instrument panel illumination lamps #158 (2) not Idiot lamps. This are wired to the Parking Dimmer circuit. ( these are T-3-3/4 wedge base)
Insrument panel warning (Idiot) lamps #74 T-1/3/4 wedge base lamp. 14v. 0.10 amps , many in the I/P cluster.
The CEL lamp is special GE # 18 is far weaker and must be so. T-1-3/4 . Do not use larger amps lamp as this overloads the ECU !
#18 Miniature Bulb Glass Wedge Base - 14.0 Volt 0.04Amp 0.56 Watt T-1-3/4 Miniature Glass Wedge Base C-2F Filament Design 0.13 MSCP. 5,000 Average Rated Hours
40mA lamp is very tiny, id not exceed 100 Ma or 0.10 amps.
not me.
now one more lamp?, the CHARGE lamp, if not the right lamp or missing or LED (bad idea) the alternator will fail to sell excite. (born dead, or 1/2 the time) #74 lamp is used stock. and needs hear 0.10amps (less may fail)
the other lamps just about any lamp works. (eg. oil, background lamps., blinkers highbeam, all are now 0.10 amp lamps. more ok up to .2amps )


Minimal lamps? and gauges?
Brakes. (any lamp)
Oil pressure loss (any lamp )
CEL (GE#18 only)
CHARGE (.1 amps or tad more)
VSS sender (100 % needed)
SPEEDOMETER
fuel, no need for fuel gauge
nor engine coolant temp?
just 6 wires to run off road, if oil pressure is not important? 5?

[rcx348]

Wow fixkick! Thanks.

I face two challenges: (1) VSS and (2) fuel level guage. The coolant temp, tacho, and speed sensor are all digitally built into most universal instrument panels. Fuel guage can be handled by a Koso Mini 3 universal fuel guage which can be adjusted to any sender, but it's ugly and expensive.

With the VSS, what would your approach be to get this signal to the ECU if using a generic instrument panel? I was going to try to extract the VSS from an original instrument panel. Or somehow generate the four pulses required for each turn of the speedo cable. But first I will see if I can find some generic/universal VSS.

---

BTW, most SJ owners I've spoken to, who have fitted a Vitara engine, seem oblivious to the need for a VSS!

---

Also anyone know how many turns of the speedo cable to turns of the wheel?

[fixkick]

the fuel guage is easy, you can buy generic gages. Summit racing.
universal instrument panels (it must be programmed to work) just like tuning a ECU.
and all that is expensive. why not just use the real Cluster. (its free, you have 2)
vss
well that is easy, the ECU wants a switched square wave. like all new cars have now. all .

again summit sells them.
you only need a , sensor on the drive shaft that picks up the turn.
the problem is that will click (pulse) way too fast the VSS is stepped down.
there are on 7ways to do that (the 5speed has 5.12:1 axle.
1: like suzuki did gears.
2: like new cars, it sends a pulse and the ECU programing matches that. (the ECU counts easy all computes counters are simple and easy)
3: use a mag HALL sensor (buy one) on the drive shaft. and then use counter off that line (or Pic or cheap processor to count it "Arduino" is just that)
4: take the mag senor and covert it to analog then scale it with analog amp, then covert that back to pulses 0v,5v,0v, repeat.
number 4 is crazy complex but cheap.
5: wild? Get at GPS box, that outputs speed, then using PIC processor or the like, scale it to speedo clicks.
6: cartoon like ? a wind anemometer, and scale it same as 5 (will lie ass off in windy days or 0 in a tail wind)
http://hackaday.com/2013/08/21/ultrasoni...r-station/
7: make a crude but cheap inertial navigation system, (using only speed) big waste but is on list (hack a day 2)?

all have a cost. time, money. or buy it.
if i had to run CLUSTERLESS> id do 2 or 3. (id make a dirt cheap axle magnet pickup) and scale it to 1026 rev. per mile.

the stock car uses a reed switch,.
and has many gears stepping it down
remove the cluster, then clear lens. see the face plate , black, see on the bottom the clicks per mile , ? this is it.





where to start is my page.
http://www.fixkick.com/sensors/speedo/un...speed.html

see this text?

The VSS clicks 4 times per Speedo cable resolution , check this an DVM ohm-meter or continuity tester ,or simple test light.
The Speedo heads states the rev per mile, and is a hidden lable, inside:
5speed : sticker " 1026 Rev = 1 mile" Rev= revolutions (lens off front of cluster ,see markings)
working backwards from 1026,
the Rear prop shaft turns, 3.8 times more per mile. or 3898.8 rev/ mile , and the rear axle (going straight) turns 5.125 less. at 760 Rev/mile
see, my TIRE Rev per mile tables and math here.

---

summit sells a speedo end cable sensor. use that?

did you see my speedo gear table, it even shows the error. stock
1003 clicks against 1026 perfect. 5speed , i do not know the other transmission cluster dial markings. only the math per table.

1026 turns per mile and 4 click.
the ECU does time, so is of no importance. here.
just clicks per mile.

axle ratio. + speedo ratio (gears)
the speedo cable clicks once per turn as the photo shows. the switch is direct drive, but the speedo is unique for each trans. as is its scaling.
the ECU knows what car, and what head is run and knows how to scale the clicker,
at 3.8x the drive shaft speed and know the axle ration so knows distance. (tire sizes correct)

---

hacks
http://www.alliedelec.com/search/product...U=70119253

this is a dirt cheep hall.
epoxy a magnet to the drive shaft up front.
then make a sensor to pick it up.
or better for 6 bucks packaged
better is this .....
http://www.mouser.com/ProductDetail/Hone...%252bNo%3d

DIY comes at all levels, $1 buck do it all the hard way or????

---

the last one will trip at 350 to 275 Gauss. with magnet away its zero. so will work. its very with 75 Gauss diffs.
Honeywell, as seen on Aircraft

---

honeywell, 1 wire ground, 1 wire power 1 wire output, to 5v pull up resistor then to processor (5v) input.
now a simple program in the same, to count down, this line by about 4:1, 3.8: is perfect.
and then toggle the output of (your dirtcheap processor here) to putput pin and then ECU VSS pin. a 5vdc square wave it must be.
i do not have the ECU input schematic for this pin, but i think its got just a weak pull up to 5v.

i do things backwards, (savvy) i take a working ECU.
then connect my scan tool'
then i connect my pulse gen and make a square wave and connect that to ECU VSS pin
in then use my scope to be sure my guess at matching up this signal are correct. a 5vdc signal.
i then see the scan tool show say 50mph .
ok that works,
next i make my own VSS system with these facts.

as you know , Suzuki does not publish this data.
i must measure it, and find out on the bench.

read the books from. Greg Banish.
its tuning book covers this. he tells you his tech, needed to make the whole dash work. using just what i tell you above.
for many things. and he said was the most expensive of all. (so many )
some are illegal, fooling the odometer . case in point.

but this one is easy.
it's just a switch that must click right.
the ECU needs a specific input levels must be right, 0v and 5v and if your device needs a pullup resistor (the honeywell must) then that too.
once the levels are correct we only need the right pulse rate.

---

the ECU needs VSS to calculate load
load on the 16v is MAF flow, engine RPM and SPEED and vacuum based, the fueling maps go richer for more load. (conservation of energy laws)
The injector rates are by that. (at least)
There are fuel tables for this (VE) that are indexed by Load.
if the index is wrong, the engine may go lean and burn the exhaust valves or detonate (ping) this car has no knock senor so no , fail point protection there.
some ECU to go failsafe mode for loss of VSS , just for those reasons. (warranty)

[fixkick]

there are many clusters. but most are just extra lights
and are tuned for transmission type.(axle)
extra features.
cruise.
4wd
Power lamp (OD) 4speed at.
RWAL ABS failure.
MPH,or KMH
Cal/ usa fed. and zillions more by country..... (LAWS, some even limit mph top)
one some cars there are holes for missing lights, i shine a flash light in front lens and see them all. omg cruise lamp missing? sure its not opted.
the clusters are here.
huge lists
the MPH dial, spring inside, is tuned for your clicks per mile. but the VSS clicks 1 per turn, so that never changes. but the gears do.

http://www.fixkick.com/sensors/speedo/cl...frack.html

---

my ecu 1996 has just one pin 7, vss pin , YELLOW.
the other end goes to VSS reed, and its other end. goes to cluster ground
the ECU sources current to the switch. so is probably a 5 to 10k (10,000) ohm resistor in the ECU to +5vdc rails.
so, any open collector driver can drive this. input.
so could most 5vdc computers totum pole outputs drive this line pin 7 yellow wire.
http://cdn.instructables.com/FEN/AP8Z/GL....LARGE.jpg

see page 26 here, cluster page,
http://www.fixkick.com/ECU/1996/Full96%2...-power.pdf

[rcx348]

Thanks. Ok, so no easy solution for the VSS. You've just prompted an obvious solution which I have missed... to modify the 1995 instrument panel so that it can plug into the 1997 wiring harness.

So I need to calculate the ratios to check that the speedo cable RPM is correct.

All running gear (engine, gearbox, ECU ...) is from 1997 car, which had 4.8 ratio diffs. I have kept the 5.1 diffs from the 1995, so prop shaft is spinning 1.0625 times faster relative to wheels. But tyres have have changed from stock 27.3" to 28.9", decreasing speed of prop shaft by 1.0586. Net difference is that the prop shaft is spinning faster by 1.0037 compared to wheels (basically same speed). So we can ignore difference caused by using non-1997 diffs and non-stock tyres.

So only question now is whether using the 1995 instrument panel on a car that is effectively 1997 makes any difference. I guess it doesn't because the VSS always generates four pulses per turn of the speedo cable, and the speedo cable is turning at the same speed as the ECU expects right?

So I just need to make the 1995 instrument panel plug into the 1997 wiring loom which means re-configuring the wires and connectors....?

[fixkick]

the cluster will work only pins different,
for sure speedo. VSS and Oil, and charge, and CEL. all lamps are same by function,

all that matters is speedo cable spins at right rpm but if off , its not going to be noticed, Just the needle and your eyes)

thats it, 1995 instrument panel plug into the 1997 wiring loom.
nothing more.
your sound like your having fun with your right, Bravo that !

[rcx348]

Getting on with this comparing the two consoles to see how to arrange the wires to make the 1995 cluster work on the 1997 wiring. Any idea why there seems to be more than one earth pin in the consoles?

[fixkick]

sure , the have extra grounds, tie them together.

if you use an ohm meter , on each ground pin, they will all read zero ohms'
if not they must be tied together, and for sure wont hurt , only make it better.
as they say, you can't have too good of ground, on suzuki's 10x that. (bad grounds and the usage of body, ground that should never have been)
like the classic pump ground, omg
the suzuki lacks a star ground , copper topology. ( like newer or better cars)

[rcx348]

Ok. I've got much further. The diagram in post #2 matches exactly to my 1997 instrument panel. Except that pins 6 and 36 on my clocks are joint together.

On the 1997 clocks, as seen in the diagram, pin 12 is a diode from the brakes fluid level warning light. On the 1995 clocks there's also a pin with a diode, but it comes from the brakes fluid level warning light AND the parking brake light. Will I need to cut this extra link?

With the dials, coolant temp, tacho, they each seem to have three pins. An earth pin, a pin which joins all three dials plus the VSS (I think) (pin 6 in diagram), and a third pin which is unique to each dial. Same on both panels, so that's clear.

The turn signals seem to be configured differently on the two panels. Some more study needed here.