[Techtoolslist] Atari Superman / Hercules Repair ~ Ram Read Fail

John Robertson jrr at flippers.com
Tue Jul 1 11:17:37 EDT 2014


On 07/01/2014 5:05 AM, William Stillwell wrote:
> (This is a cut and paste, image heavy, original post here:
> http://www.villagebbs.com/forum/index.php/topic,24850.msg205478.html#msg2054
> 78 )
>
>   
>
> ----------------------------------------------------------------------------
>
>   
>
> Symptoms were this:
>
> Didn't work , kept watchdogging... Hooked up Fluke, tried to do ram tests
> failed., then did usual, use the logic probe test all the pins, everything
> tested fine.
>
> So, I removed all the 2101/9101/5101 ram, and just put sockets in and put
> new ram in...
>
> OK, let us test again.. Fail.. ERRR..
>
> So, Manually write Address 0000, Read it, Pass, Try again, Pass, Tray again
> pass, do a ram test at 0000, fail..
>
> Scratch head some more.
>
> OK, lets test some other things, maybe something else will make since, Test
> ROM, Passes, flying colors, no issues...
>
> ok, let's make the LED(s) Light up ( D1/2/3/4 ) , write 0x1800 , 0x0f,
> Led(s) come on, Write 0x00, leds still on, hit repeat, lights go off.. Huh?
> Repeat rince.. Something Odd going on with /WRITE and /READ so poke around,
> all the logic chips test fine...
>
> OK, time to pull out the big gun, ( see pictures above )
>
> Signal I am looking for is in "RED" box.
>
> http://coinopflorida.com/tech/atari_superman/3.PASS_1.JPG
>
> Keep hitting "Repeat" ( Ie , Repeat Last operation on Fluke )
>
> Bam! the /AUDWR0 did not go low.
>
> http://coinopflorida.com/tech/atari_superman/4.PASS_2_F.JPG
>
> Hmm... So, its very random.. 1 out of 5 write attempts pretty much always
> fails.. But all the logic is correct.
>
> So, Going farther back in the circuit i was not suspecting something to do
> with Phase1 and Phase2 of the clock as you can see in the above picture it
> changes.
>
> Turns out on this board, Whenever "ROM" is being access, the CPU clock is at
> 1MHZ, everything else is 667Khz ( because the ram is slower and I/O is
> slower )
>
> So, i decided to check the Clock Signal in "High Speed" and "Low Speed"
>
> High Speed: Nice Clean 1Mhz Signal
>
> http://coinopflorida.com/tech/atari_superman/5.CLK_1MHZ.JPG
>
> Low Speed: Hmm.. That doesn't look right, its 667Khz but the wave form is
> all wacky.
>
> http://coinopflorida.com/tech/atari_superman/6.CLK_667KHZ.JPG
>
> So, the only part of the system that actually makes that clock signal is a
> 74ls160, so i remove it to test...
>
> http://coinopflorida.com/tech/atari_superman/7.%2074LS160.JPG
>
> Well, what do ya know, it fails...
>
> Pop in a replacement, and give it a shot.
>
> http://coinopflorida.com/tech/atari_superman/8.PASS_3_OK.JPG
>
> yahoo!!! nice good signal, every write attempt passes, also all the signal
> looks nice too, no shifting, exact same everytime.
>
> So, lets try the ram test again :)
>
> http://coinopflorida.com/tech/atari_superman/9.%20RAM_PASS.JPG
>
>
> I then proceeded to swap out all the defective dipswitches, and test as much
> of the hardware as i could without the machine, now off to the owner and
> pray it works
>
>
>
Nice description, I'm sure that may help others here.

Thanks!

John :-#)#
-- 
How to subscribe or unsubscribe from TTL 
http://seven.pairlist.net/mailman/listinfo/techtoolslist


More information about the Techtoolslist mailing list