KZ440 Speedometer Bounce

More
25 Nov 2020 09:15 #838916 by Grcko
KZ440 Speedometer Bounce was created by Grcko
An out of curiosity question. The speedometer on my 440 was accurate and steady up to about 60 mph and did a 5 mph (plus to minus) bounce over 60. Not a big deal It is what it is with the old speedometers. However, now that I changed the motor that has changed. It now has about a 1 mph (plus to minus) bounce throughout the whole range. Why would that change? I didn't do anything with the cable when I swapped the motors. And the cable only has about 2000 miles on it.

As Always,

1983 KZ440

Please Log in or Create an account to join the conversation.

More
25 Nov 2020 09:27 #838918 by martin_csr
Replied by martin_csr on topic KZ440 Speedometer Bounce
I would probably try lubricating the speedometer.

Please Log in or Create an account to join the conversation.

More
25 Nov 2020 09:50 #838920 by Nessism
Replied by Nessism on topic KZ440 Speedometer Bounce

martin_csr wrote: I would probably try lubricating the speedometer.


Can you please explain how? I've had the speedo and tach on my 750 open and can't figure out how other than for the odometer.

Please Log in or Create an account to join the conversation.

More
25 Nov 2020 10:55 #838925 by Warren3200gt
Replied by Warren3200gt on topic KZ440 Speedometer Bounce
The only place you can lube a speedo is light machine oil on the needle shaft bushes. This wont help re bouncing needle. The only reason the needle will bounce is if the cable is badly routed with tight bends or the inner cable is snagging through either lack of lubrication or frayed. To make the needle bounce the cable effectively flicks past a tight point in the outer caused by either of the above. As the inner cable comes to the tight point it slows and twists, as it goes past it it speeds up as i t eases and untwists . Only to repeat on the next revolution.
There is nothing in khi instruments that has a direct drive from cable to needle. The needle is driven by the magnetic field generated in the housing driven by the cable. This field is dampened by the oil filled inertia cylinder. These are the cylinders that leak out when a speedo is not stored upright. Once empty the speedo will read fast. They can be refilled but getting the correct weight of oil back into it is critical to accuracy.


Z1000J2 somewhat modified!

Please Log in or Create an account to join the conversation.

More
25 Nov 2020 11:18 #838929 by Grcko
Replied by Grcko on topic KZ440 Speedometer Bounce
I guess my question was why would it change? Seems odd to me.

As Always,

1983 KZ440

Please Log in or Create an account to join the conversation.

More
25 Nov 2020 11:34 - 25 Nov 2020 23:11 #838931 by Warren3200gt
Replied by Warren3200gt on topic KZ440 Speedometer Bounce

Grcko wrote: I guess my question was why would it change? Seems odd to me.

You probably dislodged some crap when disconnecting from the head and or possibly inadvertently gave it a less severe route when retightening back to the head.
All assuming there was no stickyness in the old drive shaft/pinion.


Z1000J2 somewhat modified!

Last edit: 25 Nov 2020 23:11 by Warren3200gt.

Please Log in or Create an account to join the conversation.

More
26 Nov 2020 13:39 #838999 by Nessism
Replied by Nessism on topic KZ440 Speedometer Bounce

Warren3200gt wrote: The only reason the needle will bounce is if the cable is badly routed with tight bends or the inner cable is snagging through either lack of lubrication or frayed.


This is not true. The speedo mechanism has an internal damping feature of some sort. If you open the speedo and physically rotate the needle it will return to 0 slowly because of the feature. I say this as someone that opened their speedo and cleaned it best I knew and tested it's operation, yet the needle wobbles through a 5 mph range even though the cable is new, well lubed, and not binding anywhere.

Please Log in or Create an account to join the conversation.

More
27 Nov 2020 01:58 - 27 Nov 2020 04:49 #839027 by Warren3200gt
Replied by Warren3200gt on topic KZ440 Speedometer Bounce
It is true. The cable rotates a magnet. The magnet sits in a inverted cup but doesn't actually touch the cup. On the bottom end of the needle shaft is a piece of steel. As the magnet rotates the magnetic field rotates the piece of steel. This piece of steel also doesn't contact the magnet but is driven by the rotational magnetic field. From the top of the piece of steel up through the cup is the shaft that is connected to the needle. Between the cup and the needle is a hair spring, as the needle shaft rotates it tightens the hair spring, when the cable rotation decreases the hair spring unwinds and returns the needle to zero. Mounted above the hair spring on the shaft is a capped cylinder with oil in it. This is the oil that leaks out if the gauge is not stored upright. Having the correct amount of oil is critical and is the "gauge adjustment" . A reduction in oil reduces the total combined weight of the piece of steel, shaft, cylinder and needle and which reduces the amount of total resistance to the magnetic field and the gauge will read fast. Over filling the cylinder will increase the total combined weight and increase the resistance to the magnetic field and make the gauge read slow.

Trust me, I have rebuilt and or calibrated dozens of gauges. There is nothing in the gauge that can cause the needle to fluctuate unless the magnetic field rotation is fluctuating. The only thing that can make the magnetic field to fluctuate is the magnet rotation fluctuating which is directly driven by , you guessed it, the cable.

Edited to add, the shaft cylinder has a clear rubber cap on it and with a very fine needle syringe you can add or remove oil in the cylinder through the rubber cap to calibrate it.

This is not a Kawasaki unit but its the same principal.


Z1000J2 somewhat modified!

Attachments:
Last edit: 27 Nov 2020 04:49 by Warren3200gt.

Please Log in or Create an account to join the conversation.

More
27 Nov 2020 05:06 #839030 by Nessism
Replied by Nessism on topic KZ440 Speedometer Bounce
I bought brand new cables for my gauges, lubed them, and cleaned the gauges internally the best I knew how and they worked okay, but the needle wobbled through about a cm of travel so to speak. Bought two NOS gauges off ebay and now the needles are rock solid. Same cables.

It's clear that some sort of damping mechanism slows needle movement internally but I didn't know about the "oil cylinder" specifically. I suspect there is some sort of malfunction occurring in this device. What kind of oil and how to fill it, and to what level, would be useful information.

Please Log in or Create an account to join the conversation.

More
27 Nov 2020 05:53 - 27 Nov 2020 07:38 #839035 by Warren3200gt
Replied by Warren3200gt on topic KZ440 Speedometer Bounce
In your situation the only thing it could be is the transfer shaft from the cable to the magnet but I have never ever found that to be an issue. I would imagine a good clean and regrease of that shaft would have sorted it but nothing in the guts can cause needle wobble.
The oil cylinder is not a damper, it is an "adjustable weight" effectivey. The hair spring and magnetic field is the damper. If there was an issue with either of those the gauge would either not work or work but be consistently high or low but not wobble.
Regarding the oil cylinder, originally they had light machine oil in them, dont know if its called 3 in 1 in the US but thats whats its called in the UK.
To calibrate the tacho , strip the guts out the case, connect to the cable, start the engine then use a known to be accurate electronic tacho on the points. Start engine and add or remove oil in tiny amounts until both read the same.
Calibrating the speedo is not as easy. You need to be able to drive the cable at a known rpm, normally using the tacho drive cable into the speedo gauge. Then you need to do some maths depending on wheel diameter and speedo drive gear ratio's. Once you know those you can calculate how many engine rpm's equate to what the speedo should be reading then adjust the oil content as above for the tacho.
Thing is with both the tacho and the speedo they do not read directly linear. E.g. when the tacho reads 1k rpm the engine is doing 1k rpm after adjustment. However, as the rpms rise and the shaft bearing fiction is overcome the needle will read higher than actual. Not a lot , maybe 2 or 3 hundred rpm at redline but close enough for 70's production bikes. This is also the case for the speedo but they always came reading fast from the factory just to cover their ass's in case of lawsuits for getting nicked for speeding


Z1000J2 somewhat modified!

Last edit: 27 Nov 2020 07:38 by Warren3200gt.

Please Log in or Create an account to join the conversation.

More
27 Nov 2020 09:27 #839055 by Nessism
Replied by Nessism on topic KZ440 Speedometer Bounce
Springs are not dampers, but oil cylinders are.
Oil level doesn't strike me as a calibration element. It's a damping element. Once the rotational speed reaches a steady state the damping doesn't matter.

Please Log in or Create an account to join the conversation.

More
27 Nov 2020 10:42 #839062 by Warren3200gt
Replied by Warren3200gt on topic KZ440 Speedometer Bounce

Nessism wrote: Springs are not dampers, but oil cylinders are.
Oil level doesn't strike me as a calibration element. It's a damping element. Once the rotational speed reaches a steady state the damping doesn't matter.


Hair springs are dampeners of rotational forces and dampens the elasticity of the magnectic field. The total weight of the moving parts within the gauge govern how much or little drag the magnetic field has on the shaft foot. No two hair springs have exactly the same strength hence the need for a callibration element. The only part of the unit which can change the shaft total weight is the amount of oil in the cylinder. Even filling to the cylinder brim we are only talking about a few ml so wouldn't be enough to have any damping effect. It is more than enough to counteract any differences in hair spring strength though.
Every mechanical gauge ever built has some callibration measure built into it or no two gauges would ever read the same for the same input.


Z1000J2 somewhat modified!

Please Log in or Create an account to join the conversation.

Powered by Kunena Forum