There's kind of a standard for mechanical speedometer drives of 1000 RPM at 60 MPH, or 1000 revolutions per mile. I suspect that all recent CJ speedometers would be calibrated to that standard, but can't say for sure.
The drive gears in the transfer case need to be matched to the axel ratio. I suspect that the gears in the D20 case are matched to 4.11 or 4.56 axels, and you've got it in a Jeep with 3.08 or 3.26 gears.
Your reading is about 65% of actual distance (1700/2600=.654). 3.08/4.56=0.675.
I don't know if the speedo gears are interchangeable between D300s and D20s. If so, swap them. If not you'll probably have to go to a speedometer shop or Jeep dealer to get the right gear.
If I'm all wet on your axel ratios, there is a serious problem with the speedo, but since both speed and distance readings are off by a similar factor, I think the problem is in the drive mechanism.
As a test you could disconnect the speedo cable from the instrument and wrap a flag of tape around the end. Then slowly drive a measured quarter mile while a passenger counts the turns of the cable. It should be close to 250. If it's more like 160, you've found the problem.
The drive gears in the transfer case need to be matched to the axel ratio. I suspect that the gears in the D20 case are matched to 4.11 or 4.56 axels, and you've got it in a Jeep with 3.08 or 3.26 gears.
Your reading is about 65% of actual distance (1700/2600=.654). 3.08/4.56=0.675.
I don't know if the speedo gears are interchangeable between D300s and D20s. If so, swap them. If not you'll probably have to go to a speedometer shop or Jeep dealer to get the right gear.
If I'm all wet on your axel ratios, there is a serious problem with the speedo, but since both speed and distance readings are off by a similar factor, I think the problem is in the drive mechanism.
As a test you could disconnect the speedo cable from the instrument and wrap a flag of tape around the end. Then slowly drive a measured quarter mile while a passenger counts the turns of the cable. It should be close to 250. If it's more like 160, you've found the problem.
