RE: [CAD_CAM_EDM_DRO] Re: Major encoder trouble, need advice.

Hello Henrik,



Perhaps it's time to borrow an oscilloscope and check to see what all your
signals look like. A picture of what is happening would help a lot. Many
x-electronic techs keeps scopes around for such purposes.



Good luck,



Bill N



_____

From: CAD_CAM_EDM_DRO@yahoogroups.com
[mailto:CAD_CAM_EDM_DRO@yahoogroups.com] On Behalf Of Henrik Olsson
Sent: Thursday, September 25, 2008 10:57 AM
To: CAD_CAM_EDM_DRO@yahoogroups.com
Subject: Re: [CAD_CAM_EDM_DRO] Re: Major encoder trouble, need advice.



Hello again,

I've now tried powering the encoder with a separate 5VDC power supply, a
1000uF capacitor across the powersupply terminals and a 1uF MKT cap right on
the encoder. That didn't help....

I checked the encoder and there is NO termination across the signal-pairs on
the encoder, perhaps that's why it seems to work "better" (but still not
working) when I add "my" termination in // with the SN75115's internal.

Is it possible that the lack of termination at the "encoder end" of the
cable is causing this even if I add the resistor-cap termination 300mm away
from the encoder? If so, why the heck isn't it inside the encoder where it
should be...(?)

Moving on, I rearanged my wiring so that I now have a 'true' star-point.
I'll try with words to describe how it's setup:

* Heatskink of UHU connected to GROUND/EARTH/PE - one single screw serves as
my start point here.
* Motor supply (130VDC) negative connected to starpoint.
* Motor housing connected to starpoint.
* Motor cable shield connected to starpoint and *only there*.
* Logic supply for UHU (15VDC) -ve connected to starpoint.
* Encoder supply (5VDC) -ve connected to startpoint.
* UHU motor supply input -ve connected to starpoint
* UHU logic supply input -ve connected to starpoint
* Separate +5V feeds the encoder, it does no longer get its power from the
UHU.
* Encoder -ive connected to starpoint
* Encoder cable shield connected to starpoint.

At first I didn't have the encoder GND and shield connected to star-point
but instead to GND in the connector on the drive, it didn't work. So I moved
both of them to the star-point but that actually made it worse - many more
invalid transitions in a short time.

Now, as I see it the ONLY ground-loop that may exist is the onboard GND-link
between logic level GND and motor supply GND.

I've ordered a couple of 26C32's but they'll probably won't arrive untill
after the weekend.

Best regards,
/Henrik.





[Non-text portions of this message have been removed]