PDA

View Full Version : XT floppy problem, motherboard issue



framer
November 9th, 2013, 09:48 AM
I've had a motherboard issue with my XT. I can get only one floppy to be work. To make a long story short, both drives work and if I reverse SW1 7&8 or turn them both off I can use drive B but C becomes D.

I used the switch test program @ http://www.minuszerodegrees.net/5160/misc/5160_test_sw1_via_basic.htm and step 9 is suppose to be 82 is showing 18 and does not change if just switch 7 is turned on. So I know I got an issue to check out. I checked the switch and it seems to be fully functional. So somewhere something broken????

Now I'm lost in how to troubleshoot beyond this point.

I did have a spare board so the unit is working again.

I would like to repair it if possible. It is useable with a single drive or using the hard drive as D.

framer

Stone
November 9th, 2013, 10:13 AM
Yes, it's useable with a single drive and the HD as D:.

modem7
November 9th, 2013, 01:17 PM
Most of the SW1 reading circuitry is shown below. The circuitry reads either switches 1/2/3/4 or 5/6/7/8.
The green arrow shows the line that does the selection, low for 1/2/3/4, high for 5/6/7/8.
The four orange arrows show the outputs (representing either switches 1/2/3/4 or switches 5/6/7/8).

Potentially the cause is in the circuitry shown, or the chip that reads the outputs, U29 (the 8255A).

You were able to test switch 7, which suggests to me that you have a multimeter, which in turn suggests that you can measure DC voltage. You may even have a logic probe.

Try the following:
1. Close switch 7.
2. Get to BASIC.
3. Enter: out &h61,&h48 (selects switches 5/6/7/8 for reading).
4. Measure pin 14 of chip U41 (a 74LS244). You should read a low (about 0 volts).
5. Open switch 7.
6. Pin 14 of chip U41 should now read high (in this case, about 5 volts).

If the above test worked, then putting that together with the results of the test on my web site, suggests that chip U29 (the 8255A) is faulty.

If the above test failed, then we need to do some more diagnosis.


http://www.minuszerodegrees.net/5160/misc/5160_sw1_reading.jpg

framer
November 9th, 2013, 04:52 PM
WOW! I'll test it in the morning. Yes, I have a logic probe.

Thanks

framer

framer
November 10th, 2013, 07:49 AM
The test worked. U29 is the problem. Can I piggy back another chip on top or will it need to be removed?

framer

modem7
November 10th, 2013, 10:12 PM
The test worked. U29 is the problem.
Before going down the 8255A replacement path, I think it would be good for you to verify continuity between 14 of chip U41 (your measuring point) and pin 16 of the 8255A.


Can I piggy back another chip on top or will it need to be removed?
Removed.

Consider a socket for the replacement 8255A.

Chromedome45
November 11th, 2013, 06:38 AM
If you go the route of removing the 8255a consider using a pair of diagonal cutters and cut each pin of the 8255 individually as close to the body as possible. Then pull chip out of the way. Then using a soldering iron to heat each pin and remove them. Use a solder sucker or solder wick and clean out each hole individually. Afterwords insert a 40 pin socket and solder in place. Then install new 8255a. Much easier than trying to desolder the whole chip and eliminates the possibility of lifting a trace. I do it that way and have never had any problems.

framer
November 11th, 2013, 10:13 AM
If you go the route of removing the 8255a consider using a pair of diagonal cutters and cut each pin of the 8255 individually as close to the body as possible. Then pull chip out of the way. Then using a soldering iron to heat each pin and remove them. Use a solder sucker or solder wick and clean out each hole individually. Afterwords insert a 40 pin socket and solder in place. Then install new 8255a. Much easier than trying to desolder the whole chip and eliminates the possibility of lifting a trace. I do it that way and have never had any problems.

I figured to take my dremel to the pins. Would it be sick to solder the socket to the pins? OK, I'll remove them if I must.

framer

framer
November 11th, 2013, 04:56 PM
Before going down the 8255A replacement path, I think it would be good for you to verify continuity between 14 of chip U41 (your measuring point) and pin 16 of the 8255A.





verified continuity

framer

Chromedome45
November 11th, 2013, 05:41 PM
If you use a dremel be careful might go to deep and cut a trace! But hopefully it's not the 8255 that's the problem

modem7
November 12th, 2013, 12:57 AM
I do not know where you are getting a replacement 8255A from, but it too could be faulty (in a different way). That is one of the reasons why I suggested the use of a socket.


But hopefully it's not the 8255 that's the problem
I am sure that he is hoping that the 8255 is the problem cause, because if it is not, then he has more work to do.

If anyone disagrees with the diagnostic process I used, and conclusion, then please 'put your hand up'.

Chromedome45
November 12th, 2013, 02:43 AM
Hence my suggestion of removing it in the 1st place. Cutting out the old chip is the best way. Then put in a socket and get another 8255a. I might have one or he could get one through from Jmaeco or some other part distributor. Maybe get 2 if one is bad as well? up to the OP.

No Modem 7 I don't disagree with you at all. I think your right on target. Your diagnostics confirmed my XT board was good. Had a bad exapnsion card with shorted tant's on the 12v and 5 v lines.

framer
November 12th, 2013, 03:08 PM
I got a lot of 4 NOS ones on fleabay for $15.00 including shipping. I got a socket so I'm set there. I should have everything by the weekend.

framer

framer
November 14th, 2013, 04:46 PM
Well the new socket's in and I tested every pin for continuity with where in was suppose to go. My old eyes hurt from tracing the schematics. The 8255A should be in tomorrow. Can't wait...


framer

framer
November 16th, 2013, 11:10 AM
Success!!. Early checkout everything seems to be working. The switch test came out normal.

U29 was bad.

Thanks modem7 for your effort to get me on target.

framer