The main failure on these terminals were:
a) a fried power switch (too tiny contacts burning off after some years)
b) a self-unsoldering hi-voltage transformer. Gets a little hot for the solder tin they used in a flow-through solder bath ...
Now  - and severely burned screens.  Sometimes you can see the host entry mask already without having the thing powered on :-)

Western Digital Video (Now Paradise)
   OEM Driver Software Library
XGA-1
XGA-2
Actionmedia II
NEC Multisync Graphics Engine 
8514/A
Matrox- This is the latest OS2 and NT 3.5 driver for Impression, Impression Plus, Impression Lite, Ultima Plus, and Ultima models. I don't know if it supports the MCA cards.
IRISVision by SGI (Run by Roger Brown)

Collection of IBM Monitor Fixes (by Frank Reid) Thanks, Jim Shorney

ATI Graphics Ultra Pro (and some other ATI cards)
Artist XJS (5080)

Connect as follows:
Pin                Pin
 1 ----^v^v^v^v---- 6
 2 ----^v^v^v^v---- 7
 3 ----^v^v^v^v---- 8
10 ---------------- 11 (shorted, in words)

Note the original unit has shrink tubing on the resistor leads-Not a bad idea.
   An alternate technique would be to just plug in any old monitor on the base video VGA output. Of course that takes up a bit of physical desktop.

Will The he Real ATI GUP Terminator Stand Up?
From Carlyle Smith
   The resistors provide a simulated load to the RGB signal circuits, a condition apparently required by the M-Motion Adapter. Brad thinks that this terminator may also work to satisfy the ATI GUP logic requirements. This, in fact, may not be so.
   In stark contrast, I checked the terminators supplied  with the ATI Graphics Ultra Pro 16-bit ISA and MCA versions, and they are identical. They are wired like this:

Pin               Pin
  1 ---------------- 6
  2 ---------------- 7
  3 ---------------- 8
10 --------------- 11

In words, the system still thinks that an 8512/3 display is attached, but that the color signals are shorted to their respective grounds. I have no idea what the system or GUP logic decides about this.  Maybe the 75-ohm load terminator and the dead-shorted terminator may be interchanged. Maybe not. There is an easy way to find out. QED. 

Dead Short? Not My Pepperoni-Tuna Pizza!
Peter Fires Back
Hi !
   To add another bit of info: The ELSA XHR Gemini/2 858 card comes with the same 75-Ohms resistor terminator that Brad described. Luckily I have the Gemini manuals and this defines the use of the terminator as follows:
   Usually the card is designed to work as "2 cards / 2 screens" solution. Under e.g. AutoCAD the normal VGA displays the texts and help screens, while the Gemini only displays the hi-res graphics using a Fixed frequency monitor.
   Assumed you have a multisync monitor you could attach this to the Gemini and plug the terminator to the standard VGA. The Gemini/2 has VGA Loop Through capability without additional components, while the other (ISA/EISA) Geminis require a VGA feature connector cable or the Gemini VGA loop through Add-On board (Rem: the Gemini/2 uses the AVE connector on the MCA).
   With the GMISETUP the card must be configured as VGA loop through / single screen installation. That's all. 

   Personally I want to add that a *short circuit* (Null Ohm) between any of the RGB color signals and their according GND returns is a bad idea. I would not recommend to use a "plug" that shorts the video card outputs to GND with no resistor between. This could cause damage of the video output drivers ... and if that video card is a planar-Video you will probably need a new board
afterwards.

Peter-
The 9527 is a DDC2B type monitor and *may* have some problems with signalling the proper display-ID to the video-card. It uses pin 15 as DDC clock (usually monitor ID3) and uses ID0 (pin 11), ID1 (pin 12) and ID2 (pin 4) as well, with pin 5 as RxD/TxD, which is Digital GND / Self-Test on many "common" type monitors. 
   I think your problem is based on that. I also vaguely recall that there was a sort of "fix" or technical recommendation from IBM concerning this compatibility problem.

There are basically two types of "video support slots" available:
1. BVE = Base Video Extension
2. AVE = Auxiliary Video Extension

The Base Video must be present on any machine. Most machines have the base,video on the planar and therefore offer only slots with AVE - which enables *additional* cards to utilize the base video for synchronized output or low-resolution / text modes. (Famous notorious example: IBM 8514/A adapter)

Machines like the Mod. 95 which have no planar video offer both: BVE and AVE. A BVE-capable card like the XGA / XGA-2 *must always* be present to supply the base video functions initiated by the machine BIOS. A second card in an AVE-slot can be present.

On a Mod. 95 the Slot #5 is the BVE-slot to which a BVE-capable card must be present (short SVGA, XGA, XGA-2), Slot #7 is the AVE-slot, to which additional cards fit like IBM 8514/A, ATI, Spea or Matrox cards.