G80 Control Card - Vector Memory access question / ongoing woes .....

xxyyzz

Member

Donor 2012, 2015
Joined
Jan 19, 2012
Messages
31
Reaction score
34
Location
Illinois
G80 Control Card - Vector Memory access question / ongoing woes .....

All,


My problem (and this is where I am looking for some G80 expertiese), is that I have an X/Y - Control Card,
which even after recently replacing the X/Y interconnect cable / sockets, continues fails with the following ' faults ' :

(1.) Upon power up - screen shows random lines (drawn / blanked / drawn all over the place).
(2.) Connecting Output(s) to an O-Scope yields same problem / random lines being drawn.
(3.) Depressing the ' Test ' switch on CPU - Card - indicates - repated U27 Ram - Fault
(4.) Replacing every RAM (2114) chip - U21 -> U34 - has no effect / no change,
(5.) Removing RAM (2114) chips - has no effect - U27 Fault continues - as if Ram is not present.
(6.) Problem is isolated to the Control Card itself, (verified after Control card swap).
(7.) Problem seems to point to ' inability ' to either write to or perhaps access Vector Ram by the CPU Card?


In an effort to try to resolve this issue, I have also gone in and replaced the following Parts(s)... (previosuly)....
(please note: These were replaced one at a time - Board Powered after each instance - with no change).
(I have also added some notes - based on my limited knowledge of this particular ' Board ' )

U2 - Color Word is Read & Applied to RGB / DA - Converters (U5 and U1 - T) and associated Diodes/ Resistors.
U5 - Draw - Blank On / Off -
U3 - Compensates for Electron Beam - Deflection -
U10, U11, U12 - 74LS161 <- Vector Address Counter - Circuit - Loaded with First Address of the Line Instructions -
U14 - CPU DataBus <- D0 - D7
U15, U16, U17 - 74LS191 <- Vector Line Length Counters
U18 - Signal End - 74LS10 <-
U24 - U31 <- 2114 (RAM) -
U33 - Advances - 74LS283 <- Program Counter - by 10 counts - (if instructed Not to draw) -
U48 and U49 - 74LS244 <- Multiplexers - Allow CPU to Address Video Memory - U24 - U31
U50 - 74LS20 - Decoder <- Selects 1 of 15 outputs - goes >Low< & stores Video Words - in memory one at a time.
U51 - Sequences U50 <- through its 15 - count - cycle. At count 14 - Pin 16 of U50 - goes >Low< - to - DRAW.
U33, U34 and U22 Allow Character Words to be taken from Memory as the Program Counter Addresses
them via U36, U49, U35 and U21 -

Looking any information on how ' RAM ' on the G80 - Control Card is initially accessed (upon power up), by the CPU.


At present time - U38-Pin9 - MemR - indicates as ' Good '? High -Low-High-Low, etc...
U42 - Pin 3 - MemW - indicates a constant ' High ' as if Memory Write is not possible?
U24 - U31 - (Memory Chips) - Pin 10 = continue to indicate a constant ' High ' .


Again, Appreciate any additional information anyone can offer.

Regards,
-xxyyzz
 
Finally got around to checking Data Sheet associated with 2114 (Vector) Rams.
(should have done this previously, but you live and you learn).

Pin 10 = WE = Write Enable - Chooses between Read and Write Modes.
Pin 8 = CS = Chip Select - Enables Read and Write Operations.
Writing occurs Only at the time that CS and WE are both ' Low '


Well at least that exaplains why I have no (proper) activity - Pin 10 = ' constant ' High.
However, I have yet to figure out what is causing this, in the first place.


Regards,
-xxyyzz
 
OK here is the latest:

Checked Voltage at Pin 18 - 2114 - (All Ram Chips) = 4.86 VDC.
Checked CS - Pin 8 - 2114 - (All Ram Chips) = Constant ' High '
Checked WE- Pin 10 - 2114 - (All Ram Chips) = Constant ' High '

According to the ' 2114 ' Truth Table ' - possibilities include - '
-----------------------------------------------------
CS --- WE --- I/O ------ MODE
-----------------------------------------------------
H --- X ---- Hi-Z ------ Not Selected
L --- L ----- H -------- Write 1
L --- L ----- L -------- Write 0
L --- H ---- Dout ------- Read
------------------------------------------------------
(the dashes I used were for spacing purposes)


Checked Activity on A0 - A7 = Activity on All Address Lines - except A2 - Pin 7.
Checked Acttity on I/O 1 - I/4 = Activity on All I/0 - lines.

Current problem is that ' A2 ' Address line seems dead -
No activity (either high , low or pulsing).
Logic probe indicates a barely audible Click.
Will Trace ' A2 ' back to its source.

Was wondering if anyone else has ever run into this with ' Vector Ram '?


Thanks
-xxyyzz
 
Traced ' A2 ' - Pin 7 -
Connection goes back to several LS244's - U20 / U35 / U48
I.E. U48 Pin 14
I.E. U35 Pin 14
I.E. U20 Pin 7

Thinking that Vector RAM is not the issue (proven to myself previously - wanted to be sure).
I Removed the following RAM / Pressed Self Test (button on CPU):
I.E. Removed U27 - Self Test Fails - indicating U31 - ad infinatum.
I.E. Removed All RAM - Self Test Fails - indicating U27 - ad infinatum.
(in either of the above, ' A2 ' continues to remain in the same ' state ' even with no Ram in place).


Replaced LS244's - with no change.
Glanced at LS244 - Truth Table.

Going forward will look past LS244's, provide an update.

Wondering if anyone else has ever run into any of the ' Ax ' lines being in this same state?


Thanks,
-xxyyzz
 
Ok - here is the latest (in what I call my ' ongoing excercise in frustration....lol..).. seriously

** After Tracing ' A2 ' (pin7) out to from Each Memory Chip - (pick a chip - any chip)- I noted:

(1.) ' A2 ' branches Off in sevaral Directions (I will use ' U27 ' in this instance).
(2.) At Chip Location - ' U48 - LS244 ' Pin 14 - there is no issue / problem is not present.
(3.) At Chip location - ' U48 - LS244 ' Pin 06 - there is no issue / problems is not present.
(4.) At Chip location - ' U35 - LS244 ' Pin 14 - Problem continues - Logic Probe indicates no activity - just click - click.

(5.) At Chip location - ' U35 - LS244 ' Pin 6 - there is no issue / problem is not presemt.

(6.) At Chip location - ' U20 - LS244 ' Pin 7 - Problem continues - Logic Probe indiates no activity - just click - click.

(7.) At Chip location - ' U20 - LS244 ' Pin 13 - Problem continues - Logic Probe indiates no activity - just click - click.

(8.) At Chip location - ' U10 - LS161 ' Pin 13 - Problem continues - Logic Probe indiates no activity - just click - click.


I have not yet had a chance to Trace out past ' U10 - LS161 ' .

However, I have already Replaced ' U20 ' / ' U35 ' / and ' U10 ' and ' U12 ' with no change.

Also:

Looking at ' U10 - Pin 13 - it is the Output of ' QB ' and the Input of ' QB ' is Pin 4.
Looking at at ' U10 ' - Pin 4 - it connects to ' U12 - Pin 4 ' = ' QB ' .
Tracing ' U10 - Pin 4 ' - it connects to ' U21 - Pin 15 - LS244 ' which outputs to Pin 5.
Continuing on - ' U21 - Pin 5 ' connects to ' U26 - Pin 1 ' which has no issues(from prior testing of All Pins on All 2114 (Memory) chips).
etc....

I will keep tracing this out and provide an update.

I may be wrong, but I am starting to think that something may be holding Address line ' A2 ' in an ' ambiguous ' state - I.E. - somewhere between' High ' and ' Low ' rather than an actual Defective Address Line? Hence the ongoing failure of ' U27 ' .

I say the above, since after replacing the ' supposed ' defective parts? (ok not really defective) I find no change.

If anyone here is an expert on Address Lines (G80 - control card) or any other ideas, please feel free to chime in.

-xxyyzz
 
Have you measured resistance between the chips for that missing signal?

NOT just continuity, but actual resistance?

My R/S multimeter will beep for anything <300 ohms. Well, a bad trace can still show up with a "beep" on it which can lead to situations like this.

If you have good continuity but the signal degrades then either the chip driving it is weak or a chip receiving it has an internal issue and is dragging down the voltage on the line.

RJ
 
Thanks -

I'll be honest I never thought of measuring the actual resistance between the ' chips ' as you had indicated.... Great Idea ....

Yeah I am likewise using an ' Old ' (believe it or not still works) R/S - Meter from 1989 - '' Micronta '' something xxxx model# (I forget). (ok useless information) but I will add that mine beeps for anything 200 ohms or less (continuity).

Anyhow, I will check / report back with findings (results).

-xxyyzz
 
I may be wrong, but I am starting to think that something may be holding Address line ' A2 ' in an ' ambiguous ' state - I.E. - somewhere between' High ' and ' Low ' rather than an actual Defective Address Line? Hence the ongoing failure of ' U27 ' .
Not uncommon -- any failure where two chips are driving the bus, one driving high and the other driving low, can result in some indeterminate logic level. For example you try to drive +5v on a line that contains a chip that has failed by shorting-to-ground: you won't reach 5v, probably won't see 0v either, most likely something in-between.

Do what you've been doing; probe the address and data bus with your oscilloscope and make sure they "look good" meaning you see each line pulsing to a full logic-high and logic-low. Sometimes a single line sticks out as bad and you can work your way through the problem.

This process is a lot easier with a Fluke 9010 tool because you can replace the CPU and tell it to toggle a single address or data line. You can go one-by-one verifying that each one toggles properly.
 
Not uncommon -- any failure where two chips are driving the bus, one driving high and the other driving low, can result in some indeterminate logic level. For example you try to drive +5v on a line that contains a chip that has failed by shorting-to-ground: you won't reach 5v, probably won't see 0v either, most likely something in-between.

Do what you've been doing; probe the address and data bus with your oscilloscope and make sure they "look good" meaning you see each line pulsing to a full logic-high and logic-low. Sometimes a single line sticks out as bad and you can work your way through the problem.

This process is a lot easier with a Fluke 9010 tool because you can replace the CPU and tell it to toggle a single address or data line. You can go one-by-one verifying that each one toggles properly.

Check the power on the +5 voltage a low voltage sometimes will not let a high actually go high. check voltage at different points, sometimes you get farther from connector and voltage will drop.
 
First of all '' BIG - Thanks '' to everyone for their input / feedback - regarding this issue....

I was finally able to find / fix my problem with Address Line - ' A2 ' .

There is what appears to be a 'totally good ' trace between Pin 4 - LS244 (U4x) and the ' via ', which connects through from TopSide of the Board to the Bottom (and so on) until it makse its way to the 2114 Ram.

Problem was actually between Pin 4 - LS244 and the the ' via' itself.

Even though the Trace looked good, it would, occasionally loose connection which would cause Vectors to be drawn randomly along with a continuous - U27 - RAM - failure.

After soldering in a piece of wire problem solved.

On an unrealted sidenote:
This whole mess began weeks ago (and forced me to quickly learn how to troubleshoot this board).
Initial problem had to do with Missing the Color Green in the upper 2/3-rd's of the screen.

This too has since been Fixed ..

Unfortunately I wound up replacing multiple parts so do not necessarily know what resolved the missing ' color ' issue.


Regards,
-xxyyzz
 
Wow.... but thanks Appreciate it.

By the way,
I will eventually write something up (from my notes / personal exprience) - post it once done.
However it will be strictly for G80 Control Card, since this is all that I know of the G80 - so far.

Going forward I need to get a much better understanding of the Timing Card, etc...


Regards,
-xxyyzz
 
Back
Top Bottom