Image Map Image Map
Page 4 of 14 FirstFirst 12345678 ... LastLast
Results 31 to 40 of 133

Thread: Frustrating Problem

  1. #31
    Join Date
    Jan 2010
    Location
    Central VA
    Posts
    4,750

    Default

    Agree on the "nothing but machine pin sockets, if possible!"

    Quote Originally Posted by Chuck(G) View Post
    Used to be Carbon tet or Freon TF for cleaning. Now long gone.
    Interestingly, there's still a lot of NOS small bottles of cleaning solutions that are 80/20 isopropyl alcohol and Freon TF. All of the ones I've got are just labeled as "head cleaner" or something else innocuous. I figured they were just alcohol since that's definitely what it smells like, but a lot of them list Freon TF!

  2. #32
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    33,060
    Blog Entries
    18

    Default

    I've got some of that, but now have only an empty aerosol can of TF. It was good stuff for cleaning--used by the tank-car lot in the old days in semiconductor manufacturing. It may still be legal in military/aerospace applications.

  3. #33
    Join Date
    Dec 2013
    Location
    Near Milwaukee Wisconsin
    Posts
    1,225

    Default

    Well.... I could not repeat the good system by heating the DMA area of my board. I could get any results using the heat. So the running system, although it lasted for an hour, must have been a coincidence. So back to inspecting the remaining IC's/sockets.

    I found that NAPA in deed sells Stabilant 22, but they want $110. I am investigating some other internet sources. So far the best price is around $80. Mike

  4. #34

    Default

    Quote Originally Posted by Mike_Z View Post
    I found that NAPA in deed sells Stabilant 22, but they want $110. I am investigating some other internet sources. So far the best price is around $80. Mike
    OUCH! I was just at one of my local electronics suppliers and they had the 15ML kit of Stabilant22A for $40Cdn or about $30US. However, their website clearly states "Product sold to Canadian Customers only".

  5. #35
    Join Date
    Oct 2016
    Location
    Dutchess County, New York, USA
    Posts
    64

    Default

    Besides using heat, have you tried cold? It used to be easy to find FreezeMist or the Radio Shack stuff, but as with most things not GREEN, it is almost impossible to find locally these days. But turn a can of compressed air upside down and spray onto a chip in question and it gets frozen with just a quick spritz. Cold solder joints and other bad connections often show up when frozen. I solved more than one problem with canned air.
    Crazy old guy with a basement full of Pentium 1 laptops and parts

  6. #36
    Join Date
    Dec 2013
    Location
    Near Milwaukee Wisconsin
    Posts
    1,225

    Default

    I'll see if I can find some cold spray, although maybe if I just left it out on the driveway it would get cold only 6 this morning.

    I cleaned a few more sockets and IC's, still no soap and I didn't spot anything that was tell tail. I thought that I'd take a break from cleaning and try some testing. Since the process seems to stop between the command and results phases of the FDC, I thought I'd look at how my 8259 PIC is working. I only use interrupts for the FDC, so that is another good reason, since the remainder of the machine seems to work just fine.

    I wrote a small loop and placed a HALT at the interrupt address. My 2K ROM monitor programs the PIC on boot.

    Code:
    000 010  HLT   /Interrupt address
    
    010 000  EI     /Enable interrupts
    010 001  JMP  /Loop wait for Int
    010 002  000
    010 003  010
    I removed the FDC and using a small home made pulser, I applied an interrupt. Seemed to work just fine. Then I thought the FDC would be sending multi interrupts at a fast rate. So I changed the program to loop through the interrupts.

    Code:
    000 010  JMP   /ISR
    000 011  100
    000 012  101
    
    000 100  MVI  /EOI
    000 101  040
    000 102  OUT
    000 103  100
    000 104  JMP
    000 105  000
    000 106  010
    
    010 000  EI     /Enable interrupts
    010 001  JMP  /Loop wait for Int
    010 002  000
    010 003  010
    Later I added a beep in the ISR routine. This worked, but occasionally the interrupt would jump to a wrong address. This seemed odd. I ran this test many times. Most times it worked fine, but once in a while it would jump to 000 020 or 000 034. Then occasionally it would jump to a higher address. As if the PIC call address was wrong or being interfered with. Maybe this is where my problem is? I tried to think of how my setup could do this and the PIC be OK, but could not. If this is what is happening with CPM, and incorrect CALL during an interrupt, that would definitely sour the milk. The next test I made was to try and see the PIC signals. I connected up the logic analyzer to look at the INT from the PIC to the CPU, the INTA from the CPU to the PIC and the 8 data lines. Triggering on the INT sig and using the program above

    Attachment 57877

    Here you can see the JUMP loop on the right. The 303, 000, 010 and then the middle the INT goes high, which is the interrupt. The CPU responds with three INTA's. And you can see the the PIC CALL 315 and the 010 000 interrupt address. This worked everytime, but so did my previous program when I only did one interrupt at a time. Later I want to replace the FDC and try anc trap this same data while booting CPM. Maybe I can see if an incorrect address is being sent by the PIC. Thanks Mike

    By the way this web site is also a little frustrating. When I type a long message it will kick me out. Sometimes it will save what I typed and sometimes not. Most times I have to type this in Word 6 and then copy it over.

  7. #37

    Default

    I can offer no further suggestions to this other than I'm having similar problems myself with a homebrew CP/M2.2 system that I built 36 years ago and I pulled out of loft after being put in storage 24 years ago.

    Last night I spent 2 hours on a 'detour' chasing a fault that vanished and I have no idea why. It will be back. What I did find supports all the other comments about bad connections. Probing a line with the 'scope found no signal until I used the end of the probe to scrape off the oxide. Plus, chips have been walking out of their sockets and needed pushing home again.

    Its all good fun and part of the reason we are here..

    Good luck

    Pete

  8. #38
    Join Date
    Dec 2013
    Location
    Near Milwaukee Wisconsin
    Posts
    1,225

    Default

    Well..... while in hot pursuit of a pretty good odor of the problem, my Televideo 920 monitor lost it's horizontal sweep. Aaahhh, the pleasures of owning and operating vintage equipment. Suppose I'll have a slight detour for a while. Mike

  9. #39

    Default

    As I mentioned in post #5, to deal with the sockets requires that an IC pin is soldered to a wire to make a handle so the pin (never anything else of any other geometry) is inserted into each socket hole (it pays to use contact cleaner at the same time, CO contact cleaner is fine) to feel the socket claw's tension and this also cleans the claw surface with two or three insertions and withdrawals. Craig chemicals pro gold works as a lubricant, even if it's not gold in most cases.

    As you have discovered there are essentially three types of socket claw designs, machine pin, and two types of dual wipe, the TI brand sockets make their connection side to side on the IC pin surface, most other dual wipe types are across the flat (thin) part of the IC pin.

    You might be surprised with the TI types how some socket claws have one side bent away, and are only contacting with any pressure on one side of an IC pin making them much less reliable.Or have lost Spring tension.Or were damaged when the IC's were inserted originally because the IC's pins were not initially bent toward each other to align with the socket's entry holes and entered the socket with too much side to side tension and at an angle. Some IC pins for this reason may have been forced down beside the claw , between it and the plastic socket housing instead.

    Of course with the TI socket type, the IC pin must be cleaned on the side edges. Each pin must be cleaned on both sides, if there is any trace of a grey or dark line on the IC pin from previous contact with the socket pin it must be removed (see post #5 again).

    As I noted it is a very long process to correct this issue of oxidation of the contacts between dissimilar metals of the IC pins & socket pins.

    One one of my disk controller cards went intermittent, even after I had cleaned all the IC pins and tested and cleaned the socket pins. So I removed the IC's again. Checked them all under high magnification, all ok. Re-checked all the socket claws for tension (TI types), wouldn't you know it I had missed just one, the claw had no tension (or maybe I damaged it inserting the IC), and that was all it took to disable the card.

  10. #40
    Join Date
    Dec 2013
    Location
    Near Milwaukee Wisconsin
    Posts
    1,225

    Default

    Pete, I have to agree, it is good fun. I've always enjoyed a good puzzle. This vintage stuff gives an old man something to think about and toy with. I have a handful of old machines and for the last 10 years or so, I've been helping a friend who owns a machine shop. He has a number of old machines that are run by 1970's computer equipment. He can not find anyone that will even attempt to repair them. All they will do is retro fit them with brand new control hardware at a monster price. Over the years I have been able to repair and bring back many of his machines.

    Hugo, what size wire do you use to test the tension of the socket holes? Looking at an IC, these pins are not exactly square. I would think you would want a piece of wire that is similar to the IC pin.

    I have noticed that some of the grabbers in the sockets are damaged. Those sockets have been replaced. And as Pete mentioned, connecting the test equipment to corroded pins can generate poor or no signals. I think that I will attempt you procedure for my FDC board, once I get my Televideo monitor repaired. Seems there is always something to look at.

    Thanks Mike

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •