I had started to  suspect U5E, and once Dave confirmed that he too was thinking it was the culprit, out it cam and I dropped in a socket and a new 74LS163...the video board fired right up.

I then plugged in the RAM boards (after fixing a couple of broken wires for the bank selects), and the CPU board that I fixed previously and I was able to load and run Galaxia over a serial connection to my PC. I then loaded my memory test program and that brings me to the next piece of good news...all RAM checks out ok. Surprising!

Now for the bad news. Up until now I was fixing the boards in my known good challenger 2P case. I then plugged all the boards back into the original machine, fixed a bunch more frayed wires, and fired it up again. This time it still showed the random characters...good! The bad news is that of the two power units that are installed in the machine ,only the rear one which produces the +5V signal on the backplane seemed to be working. The one mounted closest to the front of the machine, and which produces  the +9V line seemed very dead. I tested the voltages coming out of the actual transformer and it seem to be at about 10.2VAC. I tested the voltage from either side of the large capacitors and it reads 13.3VDC. By comparison the working 5V power unit reads 9.4VAC and 10VDC respectively. The output from the bad unit is 113millivolts. The voltages on the CPU board are also in the millivolt range. Unfortunately power supplies are something I know absolutely nothing about. If I can't stick a logic probe on it I'm a little lost.

After some probing around I did eventually track the problem down to the CA723CE regulator. I got a new one and replaced the dead one. Unconnected the leads before powering up, I tested the voltages, and now they were all within the correct range. I reconnected the leads, and powered her up. All seemed to be working fine, except the keyboard was completely dead.

Looking at the keyboard, I first tried replacing all the IC's with new ones as they were socketed and doing this was a no brainer. Unfortunately that did not solve the problem. I broke out the multimeter and tested all the diodes and they all seems to check out ok, next I moved on to the resistors and I eventually found a pull-up resistor that was rated at 4.7kohms but was measuring in the milli-ohms range (all other pull up resistors measured within spec) so I thought I was onto something. I had previously suspected something wrong with the 'shift lock' key. I traced this suspect pull-up resistor back to C0, which is one of the lines connected to 'shift lock', I was almost certain the resistor was the issue. Out came my trusty soldering iron and I removed the offending part. I didn't have a 4.7k resistor, but I did have a 2k which I though would probably work as pull-up resistors don’t usually have to be exactly the right spec. I was right. The keyboard worked!!!!

I will cover the repair of the disk unit in another repair log as the problems were numerous and as now the drive is still not operational.

Conclusion

Lots of problems with this computer. Although the machine was in pretty good physical condition, electrically it was in poor shape. Surprisingly the RAM cards were both ok. The complete list of fixes are:

•  505b CPU board - reseat all socketed IC's brought it to life.

• 540 Video board - Replace 12mhz Xtal and replaced a dead 74LS163 @ U5E.

• Power supply - Replaced a dead CA723CE regulator with an equivalent part.

• Keyboard - replaced a dead 4.7kohm pull-up resistor on the C0 line.

• Various - repaired many frayed and broken wires.

Having another working Challenger computer made the fix much easier than it would otherwise have been. The disk drive has been a serious head ache for me, and I am still struggling with it. Hoping I can get it repaired sometime in the near future, although the prognosis does not look good. However I really enjoyed fixing the main unit, and it ranks up there with one of the most satisfying I have done.

Much thanks to Dave at osiweb.org and Builldrom for their his help.

UPDATE: More problems. Look here for part 2.

I recently received another Ohio Scientific machine. This time a C4P. After my success with the C4-2P I'm confident I can get this working. Firstly I tackled the CPU board. As I had a complete second set of working boards from my precious Challenger 2P fix. This made the process of fixing the 4P much easier as I could just swapped the new board into the working system to see what worked and what doesn’t.

The 505b CPU board wasn't working, but I could see that it was trying to reset. Cleaning the legs on the socketed ic's seemed to fix that problem. Good start.

The 540 video board was totally dead. It differers slightly from my other working 540 bored in that it has the color circuity installed. The 540 board was not pulsing at all. I checked it with my logic probe and everything is either high, or low, with no pulse seen. I tracked down the problem to the main 11.79648 Xtal. I didn’t have a 11.79648 Xtal, but I did have a 12.000. Gently placing the legs of the 12.000 Xtal agains the legs of the dead Xtal  brought the screen to life, although it was just a thin strip of compressed random characters on the left side of the screen, Maybe 10 columns in total. In the Sams manual the 11.79648Mhz Xtal in the actually specced as 12.000Mhz. The OSI 540 Manual specs it at 12.08Mhz for color boards, and 11.79Mhz for B/W boards. Not sure what all this means exactly, but it does make me thing that a 12mhz Xtal would be fine as a replacement.

The next morning I went to replace the Xtal and realized that I had left the machine on for the past 10+ hours. Unfortunately as the screen is completely dead there was nothing to indicate wether the computer was on or off, except of course for the position of the switch. Anyway, the video bored seemed dead, and and touching the legs of the 12Mhz Xtal to the dead Xtal didnt produce any changes as it had done the night before. Probing around again, there now seemed to be areas where there was no pulsing again. My bad. Dave from osiweb.org had a very optimistic outlook on my mistake, and said it was better it happened now rather than later...I agreed.

I replaced the dead Xtal with the new one. Now to track down the new problem.

I tried to follow the the flow of non pulsing signals, but I just ended up going around in circles for about 2 days. I was getting very frustrated. I could go into detail here, but I don’t think it is important to the final fix. I did however discover in the process that by grounding pin 7 of the 74157 at U3C a screen full of random characters appeared although it seemed doubled giving me a column count of 74 characters! in that the left side is repeating on the right, basically doubling the column count. Thanks to Dave again I found out that by grounding pin 7 this keeps the COL_CLR line high. Normally, that line would reset the column counter and inhibit counting during the guard bands, as well as blank the video signal in the guard bands. By tying it low, I unblank the signal, and also no longer inhibit the counters after the end of the line. The *Q pin of the 7474 SR flipflop still will pulse high for 500 ns every 256 us, when the PRESET line (pin 10) pulses high, which is why the count is resetting. 

The narrow characters imply a 12 MHz column clock. This implies that pin 1 of U3C is high selecting the 64 column mode. The 32-char repetition could be due to resetting the count after 32 chars, which would imply a bad 74158 or due to pin 1 on U5E being stuck high or low. However, you say it's pulsing. All the pins

Unfortunately this was mostly gobbledygook to me as my skill does not reach so deep :-(

I did however try grounding pin 1 of U3C (and pin 7) I was able to get a screen of random characters with a column count of 37 characters!

Grounding pin 7 of U3C produces a a screen sowing a column count of around 74 characters.

Grounding pin 7 and pin 1 of U3C produces a screen showing a column count of 37 characters.

neoncluster.com