I'll give you a detailed step-by-step of what I did so you guys can see where I went wrong. :c

I pulled the cluster out and took it apart. Then I removed the speedo needle from the gauge. I removed the face plate and removed the numbers for the main odometer because the 10k mile number had turned halfway and it was really bugging my OCD. I reassembled the odometer and reinstalled the face plate. Here was my first mistake: I put the needle on a little bit, turned it up to the 10mph area to preload the spring a little, and then fully seated it. Then I put a little bit of grease onto the nose of the speedo cable protrusion that the speedo cable screws onto. When I reinstalled the cluster, my speedo indicated 25mph when turning 2000 rpm in 4th gear (should be about 40mph). My trip and main odometer are turning fine though, so I know it isn't a speedo cable issue, it's a gauge issue.

So I took it all back out again today. I removed the speedo needle and allowed the pin to turn all the way to rest and release the spring tension. Then I pushed the needle back on without adding any tension by turning it. I reinstalled everything and went for a test drive.

NOW IT READS 5 MPH FAST! I am comparing with the app Speed View on Android. When my speedo reads 30, I'm doing 25. When it reads 50, I'm doing 46 or 47.

Previous to this, my speedo was reading 10mph slow at 65. I am running 8 hole rims from a 96 C class. The speedo error with a correct functioning system should only be 2 mph at 40 and it should be 2 mph slow, not fast.

What did I do? I'm upset now and I have a bad case of W201 cluster arm. Is there a trick when putting the needle back on? Did the grease make it more slippery?

__________________
'84 190D 2.2 5MT (Red/Palomino) Current car. Love it!
'85 190D 2.2 Auto *Cali* (Blue/Blue) *sold*
http://badges.fuelly.com/images/sig-us/302601.png
http://i959.photobucket.com/albums/a...0/sideview.png

Anecdote: 14 yes ago I had a speedo shop replace my broken speedo cable.
They ended up making it a bit too long and it reads about 4 MPH faster than what digital street speed alerts claim.
I never bothered with it.
That's on my 300D.
Maybe yours will right itself with time ?

__________________
1979 300D 220 K miles
1995 C280 109 K miles
1992 Cadillac Eldorado Touring Coupe 57K miles SOLD
********************
1979 240D 140Kmiles (bought for parents) *SOLD.
SAN FRANCISCO/(*San Diego)
1989 300SE 148 K miles *SOLD

I wasn't aware the needle was even designed or intended to be removable...maybe you damaged the gauge.

__________________
-diesel is not just a fuel, its a way of life-
'15 GLK250 Bluetec 118k - mine - (OC-123,800)
'17 Metris(VITO!) - 37k - wifes (OC-41k)
'09 Sprinter 3500 Winnebago View - 62k (OC - 67k)
'13 ML350 Bluetec - 95k - dad's (OC-98k)
'01 SL500 - 103k(km) - dad's (OC-110,000km)
'16 E400 4matic Sedan - 148k - Brothers (OC-155k)

The length of the cable does not affect the speed displayed. The cable is rigid. Only the ratios of the drive gear and inside the gauge will affect the displayed speed.

__________________
-diesel is not just a fuel, its a way of life-
'15 GLK250 Bluetec 118k - mine - (OC-123,800)
'17 Metris(VITO!) - 37k - wifes (OC-41k)
'09 Sprinter 3500 Winnebago View - 62k (OC - 67k)
'13 ML350 Bluetec - 95k - dad's (OC-98k)
'01 SL500 - 103k(km) - dad's (OC-110,000km)
'16 E400 4matic Sedan - 148k - Brothers (OC-155k)

I think the needles are definitely removable. There is no way to get the faceplate off without pulling the needle. I won't dismiss the possibility that I damaged the gauge, but I was thinking it has to be something about that spring preload or lack of preload. Maybe I'm getting hung up on it.

__________________
'84 190D 2.2 5MT (Red/Palomino) Current car. Love it!
'85 190D 2.2 Auto *Cali* (Blue/Blue) *sold*
http://badges.fuelly.com/images/sig-us/302601.png
http://i959.photobucket.com/albums/a...0/sideview.png

If this model you drive is like the W123 Spedo, there is a white mark to the right of the resting pin for the needle. It is on the edge of the face plate between the pin and 6 O'clock position. Put the needle back on. lined up with this white mark, then lift the needle over the resting pin. This will tighten up the spring so it will read correctly.....we hope.

Charlie

__________________
there were three HP ratings on the OM616...

1) Not much power
2) Even less power
3) Not nearly enough power!! 240D w/auto

Anyone that thinks a 240D is slow drives too fast.

80 240D Naturally Exasperated, 4-Spd 388k DD 150mph spedo 3:58 Diff

We are advised to NOT judge ALL Muslims by the actions of a few lunatics, but we are encouraged to judge ALL gun owners by the actions of a few lunatics. Funny how that works

I have had to do similar things before. Sometimes you have to install the needle behind the rest and jump it over the rest to get it to read correctly.

__________________
1991 F250 super-cab 7.3 IDI. (rebuilt by me) Banks Sidewinder turbo, hydroboost brakes, new IP and injectors.
2003 S430 - 107K
1983 300SD - Tanoshii - mostly restored ~400K+.
1983 300SD - Good interior. Engine finally tamed ~250K.
Monark Nozzle Install Video - http://tinyurl.com/ptd2tge

Speedometers tend to lose calibration over long periods of time, because the magnets gradually lose strength. Your reading was low to start and tires and rims are critical to accuracy, so even the calibration mark probably won't get you the right reading.

It's a bit late, but I'll tell you how I do it. I have a small length of squared speedometer cable that I chuck into a drill. Before I disassemble, I spin up the speedometer and see where the needle reads. When I reassemble, I spin it up again without the needle, and while it's spinning, I drop the needle down at the correct reading and press it home.

You can still use this technique to save some time. Spin up the speedometer and note the reading. Since you know it's five mph slow, remove the needle and reposition it to advance the reading.

I should add that adjusting the pretension on the needle will get you calibrated at one particular speed. But if the tires have changed the total drive ratio, then the calibration will be off at any speed other than the speed you've used to calibrate the instrument. I usually calibrate at 60 mph, so any error will be either well above or well below highway speed.