• Please review our updated Terms and Rules here

IBM 5153 CGA Monitor with a loss of Horizontal Sync

simmiv

Member
Joined
Nov 2, 2017
Messages
42
Location
Australia, NSW
I've had this IBM 5153 CGA monitor for a long time. I think I checked it last about 10 years ago. It had a sticker on it saying "Works, OK". So, when I finally pulled out one of my IBM PC XT machines and got it working, I decided to hook up this 5153 Monitor to the non-IBM CGA card in the XT machine. It powered on fine and after some fiddling with the vertical and horizontal adjustment pots on the back I got it to be stable. That, of course didn't last very long. It began to floating and eventually I lost the picture more or less. I replaced the non-CGA card in the PC with an IBM CGA card I knew worked fine and had the same outcome. I then tried another CGA monitor I have, and old NEC JC1203 which works fine. It was rock stable and no floating occurring. Thus I knew the problem was with the 5153!

This is what it was doing:

At least when it was playing the nicest. Normally it would be skewed and floating all over the place.. Initially I suspected the sync chip, which is a Hitachi HA11235 IC. After a bit of analysis, I realized the problem was a loss of Horizontal Sync from the CGA card. So I hooked up my Oscilloscope and put channel one (Yellow) on pin 8 (Horizontal Sync) where the cable from the PC attaches to the main PCB and began tracing the sync pulse from the HA11235 IC with channel two (blue trace). I ended up on test point 21 with the following:

20201104_144145DVR.jpg

The blue trace on TP21 doesn't show the synch pulse as it should. The Sync pulse from the CGA card on the PC comes from pin 8 on the plug that attaches to the main PC board. It then travels via R222 (100 Ohm) resistor and then into Q204 transistor, where it is amplified. TP 21 is on the other side of that R222 resistor. So the Oscilloscope was showing the pulse strong and healthy on one side of the resistor, but not the other! See part of the schematic:

schematic.jpg

I began looking closely at that area of the PCB and this is what I saw..

20201104_150413DVR.jpg
The area looks pretty dusty and there is some old glue spread across it. The Orange wire is the Horizontal Sync pulse that is strong and good on the Oscilloscope, but the TP 21 on the other side of the resistor is pretty much not there! You can barely see R222 under the crust of the glue. One leg is visible on the lower side of the green cap C210. So, it was time to do a bit of cleaning and replace some of these things!
I proceeded to replace R222, some of the caps and additional resistors in the area. They had to come out anyway so I could get this glue off the PCB. After this operation, it looked a lot better.

20201105_120321DVR.jpg

All of the components tested fine out of circuit, but after replacing them and hooking the pcb back to the CRT it worked fine! The Horizontal sync was rock solid and on the Oscilloscope it looked like this..

20201104_145507DVR.jpg

So I can only assume the glue somehow, in combination with dust became conductive and was shorting out the Horizontal Sync pulse..

During this repair I also noticed that the case had some broken bits that had come off the top left and right corners.. i.e.

20201106_114538DVR.jpg

You can see on the top right and the bottom left where there are missing pieces of plastic. It's not surprising, because after 36 years the plastic housing has become a bit brittle. I managed to locate a couple of broken pieces, but two were missing. I glued the two I had, one on each corner and then decided to try Bicarb Soda with Superglue to fill the gap for the missing pieces.. Something I'd never done before!

20201106_162845DVR.jpg

I then proceeded to file it down and get it flush with the rest of the housing..

20201106_175010DVR.jpg

So far so good.. Now I only have to find matching paint so I can cover this up!!

So if you do have some sync problems on this monitor, it may not be the HA11235, simply some dust and crud!
 
Good job finding the fault.


With the type of PLL circuit for the H lock, the locking range is fairly broad. Once in lock of course the H hold control appears simply to affect the H picture phase, causing ambiguity about where to set the H Hold and H phase control because a number of relative settings appear to give the same result.

One way to do it is to disable the H sync (I did it on mine by disconnecting one leg of C213 with solder wick or you could just disconnect pin J201), then set the H hold so there is minimal picture drift left or right, though it might weakly lock on a stray signal, that gets the center frequency of the H scan osc ideal. Then re-connect the sync and then set the H phase for the best H picture position.
 
Thanks,
Yes I did to the Horizontal adjustment.. I used a .47uF capacitor to short TP22 to ground, which is on one side of C213.. This shorts the sync pulse to ground and the Horizontal Oscillator goes free running.. Then you adjust the Horizontal hold (R409) to get as close to correct sync as possible. It has been rock solid since.. I gather you came up with the change to enable this monitor to support EGA.. I've glanced through your paper and it is impressive work.. I don't think I have an EGA monitor and as I go through my collection of IBM PC's (Pc's, XT's and AT's) I may find one with an EGA adapter, so if that happens I would be interested in implementing this change into the 5153. Has the one you modified worked well with the change?
Simmi
 
Back
Top