suddenly the axis can only move high speed

dear all,
yesterday I was wondering why on my ISEL Milling machine with a rack tool changer (8 tools in a row) the Z-axis moves down as soon as the M6 remap macro reaches the M6 command. With some more reading I figured out that the entry in the ini File caused it becaus it said
[EMCIO]
TOOL_CHANGE_QUILL_UP = 1.
Ok, I thought I sorted that, commented this line out and restarted the machine. When pressing the referencing button, the Z-Axist jumped to the roof, followed by the X-Axis and the Y-Axis each hitting the hard limits. I dind't know that such small DC servo motors can accelerate so much. I was completely puzzled becaue I was expecting normal speed movements as it had done it for the past few days. Now, beween yesterday and today I have not knowingly changed anything - so how could that happen?
I switched off the control, the machine, started it up again but each time I want to jog an axis it moves 10x as fast as supposed and also at least 10x as far. Most of the time, the servo drive then comes up with an error. I reloaded an init and hal from a couple of days agon just in case I did somethin really stupid last thing yesterday but no improvement.
When i look at the pulses for the movement on the oscilloscop they change their spacing according to the speed I select - double the speed  shows twice the pulse repetition rate. It happens on all 4 axis so I doubt that the servo drives are faulty but all channels on the Mesa 7i96s sounds unlikely to me too.
On thing I noticed was when I use the MDI command to move the Y-axis slowly with say G1 Y10 f50 , an error message pops up that this command would cause a movement of the X-Axis beyond negative limit?? When I do that with Z i get the same. Why would I get a complaint about a X-travel when I want to move Y or Z?
I honestly don't know where to start looking
Has anyone ever come acrosse such a weird behaviour?
Rainer

the scale is unchanged (400). I use leadshine DCS810 Servo drives for 60V Isel DC brushed motors. As I mentioned, all 4 show the same behaviour. Is there a way to calculate the expeced axis speed when I know the pulse rate? With steppers I could calculate 1 Step=1.8°, then with the scale and pitch of leadscrew arrive at a number. How do the servos do it? Somewhere they must convert the Step information to "so much rotation per step"?