I'm using a DMC-2183 with 1.0s firmware. All of the following is with stepper motors.

I'm trying to make a Linear Interpolation (or Vector) path using the E and F axes, and operate a microdropper device at regular intervals along this path.

I've connected the trigger for the microdropper to the "step" output of the H axis. I set axes E and F to Linear Interpolation mode in the S axis, and slaved the H axis to the S axis. The speeds in question are relatively low -- say 20 Hz or so.

In general, the setup works well. However, whenever I start the LI move, the H axis step pin spits out a bunch of spurious pulses before it starts sending out the expected ones. This does not happen when running the H axis alone, or when slaving the H axis off of a non-vector axis (for example CA).

The amount and timing of the spurious steps seems to vary with how long the path is. Gantry mode on or off has no effect. Here's my test code:

[code]
GAH=S; 'slave H to S
GRH=1; 'gear ratio is 1:1

LM EF
LI ,,,,0,15
LE
BGS
[/code]

I looked at the output of the "step" pin on my scope. What I expected to see was 15 pulses, since the LI moved 15 steps and the gear ratio was set to 1. What I get is a bunch of steps in very rapid succession, followed by the correct 15 steps. Strangely, the frequency of the steps starts high (2700Hz) and then goes down to about 500Hz -- sort of an inverse acceleration. Sometimes there is one burst, then a short pause, and then a second burst.

Any idea why this is happening?