After generally keeping an eye out for a 1989 or 1990 JDM KR speedometer for three or four years (and having no success) I got hold of a 1989 ZXR250 speedo instead. This has the same three wires (Brown, Black-Yellow and Pink) as the corresponding KR speedo for a speed limiter switch inside the speedo casing. I reasoned that there may be a chance that the speed limiter switch in the ZXR speedo would work in the same way as that in the KR version. For your general info, the following part numbers apply to OEM JDM speedo units:
1989-90 KR250C1, D1, C2, D2 - 25005-1418
1989-90 ZXR250A1, B1, A2, B2 - 25005-1427
Later ZXR250C (and D) models have a different part number again (25005-1480), but I haven't seen a wiring diagram for any of those models (so I can't confirm that any speed limiter switch wiring is the same as the KR and early ZXR models).
The ZXR speedo I obtained is obviously secondhand (about 33,500km on the odometer), but is in very good condition and seems untampered with.
The resistor which Spyros mentioned jogged something in my memory and I returned to the April 1989 KR-1S/KR-1R New Model Guide. After some in-depth translation, I discovered that this booklet seems to contain almost all of the information required to bypass the restriction. Here is a (translated) block diagram of the CDI unit from the booklet:
Sub-assembly No.10 is the speed limiter circuit inside the CDI unit. As you can see, this connects to the speedometer (via the pink wire). The speedo is depicted as a resistor in series with the secondary portion of an opto-coupler, the bottom end of which is connected to earth/frame/negative. Thus, it appears under 'normal' conditions that the pink wire is connected to earth via a resistor. From this, I inferred that when the speedo reaches 180km/h, the sector on the speedo shaft causes the opto-coupler to switch off, thus creating an open circuit between the pink wire and earth. The same effect should be created by unplugging the pink wire, which is what Virgie's experience has suggested.
Reading further in the New Model Guide, I discovered more interesting info (paraphrased from translation): When the voltage at the connection point with the speedo is outside the set voltage range, the ignition timing is fixed at 3 degrees BTDC from 10,100rpm onwards.
The ignition timing at 10,100rpm is normally about 13 degrees BTDC, so retarding it a further 10 degrees is bound to produce a noticeable difference in performance...as Virgie found.
With all the above in mind, I proceeded with testing the ZXR speedo, connecting the wires as follows:
Brown - to positive terminal of 12V DC power supply
Black-Yellow - to negative terminal of 12V DC power supply
Pink - to positive terminal of ohm-meter (i.e. a multimeter on the resistance range)
The negative terminal of the ohm-meter was connected to the negative terminal of the DC power supply (common with the Black-Yellow wire).
I have a laboratory-type 12V AC/DC power supply which I use for checking electrical components and wiring systems. The open-circuit output voltage was 14.5V DC when I checked it before the test.
The first problem I encountered was that I needed to spin the speedo drive shaft anti-clockwise (the right-angled 'gearbox' on the speedo unit does not reverse the direction of rotation as I had first thought). None of my rotary tools equipped with reverse functions would spin fast enough to indicate speeds over about 80km/h, so I resorted to a Dremel Multitool fitted with a nylon wire brush, driving the green plastic gear on the speedo shaft after the right-angled gearbox had been removed. By selecting a medium speed on the Dremel and varying the contact pressure, I was able to gradually increase the speedo indication from zero to about 230km/h (or at least where the 230km/h mark should be).
With the circuit connected, but the power supply switched off, the reading on the ohm-meter was infinity (open circuit). When the power supply was switched on (but the speedo shaft not yet rotating), the meter indicated about 18,000 ohms. This suggests that the opto-coupler switched on when the voltage was applied.
When the speedo shaft started rotating, the reading on the ohm-meter remained fairly steady all the way up to about 185-190km/h, at which point the reading changed to infinity. The infinity reading was maintained with further increase in speed (up to possibly about 230km/h) and then returned to about 18,000 ohms once the speed had dropped below the 185-190km/h point.
The result of the test seems fairly conclusive, suggesting that connecting the pink wire to earth via a resistor of about 18,000 ohms should bypass the restriction.
HOWEVER, remember that I have tested a ZXR speedo, not a KR speedo. The ZXR speedo is obviously different to the KR speedo in some (as yet unknown) way, but this may have nothing to do with the speed limiter switch. It would surely reduce the cost of the speedo for the maker (Nippon Seiki) and the customer (Kawasaki) if the same speed limiter switch was used on as many speedo units as possible.
Notwithstanding the chance to test a JDM KR speedo (or somebody pointing out a flaw in my testing procedure), I would suggest that connecting the '1275' CDI's pink wire to earth via a resistor of about 18,000 ohms should bypass the restriction and make the ignition characteristic the same as the export model '1287' CDI unit.
Despite the unknown factor of any difference between KR and ZXR speedos, it is unlikely that any damage could be done using a reasonably high resistance like that measured.
If Virgie or anyone else is willing to make the connection with a resistor as suggested and let us know what happens, we may be further along the way to closing the book on the matter of derestriction of JDM CDI units.