Atari pcb 74LS244 input outputs not matching

danpiraino1971

Active member

Donor 2014
Joined
Oct 27, 2010
Messages
1,403
Reaction score
17
Location
Bridgeport, New York
I have a Crystal Castles board where the 74LS244 at 2F for the program memory, has floating input pins (no high or low) and the output pins are stuck low. From what I read, inputs should match outputs. Both pins OE1 and OE2 are stuck high(according to truth table active is low). The output pins from the EPROMS to the inputs on 2F are also floating. Also 2J (74LS244) in the Bit-Mode read/write section also has input and out put pins that dont match. Inputs are stuck low but outputs are flipping. That one confuses me. Checked the /reset on cpu and its low which should mean its in reset. Watchdog is high, so that isnt on. R/W on cpu shows high, so if I'm correct its in read mode since the W has a line over it in the schematics and the R doesn't. (Line over something means active when not high). There are two 74LS244s at 1B and 1C. Their inputs come from the CPU and their outputs go through the address decoder section and on to the inputs of the EPROMs. 1B takes one half of the address lines and 1C takes the other half. 1Bs input and outputs match, their stuck low. 1Cs inputs and outputs also match, their stuck high. Both the OE1 and OE2 pins on these chips are flipping. Any feedback would be great. I think either 2F or 2J are causing an issue. I'm wondering since its in read mode if its having trouble reading the EPROMS.
 
Last edited:
This is where a logic pulsar would be useful. You could change the logic input and look at the output. This would tell you if the chip is working or not. You could change the inputs manually but unless you know what you are doing, I don't advise it.
 
This is where a logic pulsar would be useful. You could change the logic input and look at the output. This would tell you if the chip is working or not. You could change the inputs manually but unless you know what you are doing, I don't advise it.

Never used one but have thinking of getting one. Do you use a logic pulsar with power on to the board or off?
 
... inputs should match outputs. ...

This is only true if those outputs are enabled! Since both /OE lines are high, something else is driving the signals that you are reading on the outputs of those 244s and that is normal behaviour.

... Checked the /reset on cpu and its low which should mean its in reset.

If it is held low, there is no point in measuring any other signals until that is resolved or temporarily disabled. Sometimes one can disable the reset by pulling a pin out of a socket or clipping a pin on a chip in a way that is easy to resolder.

I had a Mappy stuck in reset. Pulled the chip that drove the reset signal, bent that pin out, reinserted the chip, and Mappy played just fine. Replaced the transistor in the reset generation circuit to fix the problem.
 
This is only true if those outputs are enabled! Since both /OE lines are high, something else is driving the signals that you are reading on the outputs of those 244s and that is normal behaviour.

If it is held low, there is no point in measuring any other signals until that is resolved or temporarily disabled. Sometimes one can disable the reset by pulling a pin out of a socket or clipping a pin on a chip in a way that is easy to resolder.

+1 to what this post said.

Sort out the reset line first. Nothing else matters much if the CPU is being held in reset.
 
+1...

Every line on the CPU is in an "unknown state" when stuck in reset.

The LS244/245 chips are VERY common on arcade boards and older computers... and VERY easy to misread.

The Enable/Direction lines on the LS244 (bus buffer on many address busses) and LS245 (bi-directional latch on many data busses) chips determine what the chip does. If they are in the wrong state then the outputs of the chip will be disconnected from the bus (hi-Z)...

Doesn't matter what those chips have for inputs or outputs if the CPU isn't running. First order of business: Fix the reset circuit.

RJ
 
So can I disconnect the pin on the cpu that goes to the reset circuit to see if it boots up without hurting anything?

Nope. This would leave you with a reset circuit that still isn't working and a CPU with a floating reset input pin.
You need to find out why the reset is being held low.
 
Different CPUs do different things when the reset circuit doesn't reset it upon power up...

For instance, a Z80 will act like it is dead.

If you want to try it, use a simple RC circuit in place of the reset circuitry.

Code:
 +5v
  |
  \
  /  1k
  \
  /
  |--------- reset pin
  |+
  = 10uf
  |
  |
 gnd
 
If it is held low said:
Ok I didnt find what was causing the reset. I just put the watchdog to ground to disable the reset temporarily. I probed the pins on the cpu and all the address lines are flipping but two(pin 18 and pin 25). One is stuck high and the other low. The high pin goes to 1B (74LS244), the input(pin3) and output(pin 17) on 1B are also high. The rest of the inputs and outputs are flipping. Then the low pin from the cpu goes to 1C (74LS244)input (pin 9) and output (pin 11) stuck low. The rest are flipping. Those address lines then go to the EPROMs.

Also theres a 74LS245 at 2E. The inputs(pins 11-18), that are coming from the outputs of the cpu, are all flipping. The outputs (pins 2 - 9) are all dead. OE (pin 19) is flipping DIR (pin 1) is held high. Then theres a 74LS257 in 2H, its in the bit mode read/write circuit. All the pins on it are flipping but the BITRD line which is dead.
I need to get a better understanding of what going on here. Do the stuck pins on the cpu mean the cpu is bad or can bad 74LS244s cause that to happen?
 
Stuck signals can be cause by any chip sharing that connection.
 
Back
Top Bottom