Faulty Falcon

[Linuxbuilders]

Hi,
I am new here so just quickly I say hello to all.

I got Falcon, unit is faulty. I try to fix it, got all service documentation, etc. I have replaced real time clock and generally checked what I could. Unit does not power up. Now the clock mod is done in some funny way I haven't been able to confirm.
7408 is wired on "out" into tops of r216 and r221 and its "in" goes into r217. r216,221 bottom (from front side) are desoldered and left empty. When I fire up the unit I get 16MHz,32Mhz,25Mhz. This part seems to be ok, U34 got DE low (I think it is wrong) CPU hold signal is high, BERR is high too, Combo IC seems to be ok too, I get data and address toggling. I insert two pics of my screen. Any ideas what else I should check and what steps should be taken to fix it, if there is anyone with experience in it in MEL/AU I will be happy to meet too. Screen is rolling over up. Depending on reset it may be blank or as on pics. VGA is not plugged in, I use composite out to connect to TV. Floppy and hdd does not respond.

arati1.jpg

atari2.jpg

Thank you for help. The box comes from auction, sold as faulty. Apparently it used to boot up on occasions, sometimes with white screen.
Regards Mark

[Mug UK]

Welcome to the forum. I'm sure there's enough h/ware experts on here to help you with your problem.

You posted the same thing twice so I removed the one that didn't have the graphics showing up. Your next two posts will also be 'held in a queue' but once they're moderated, you're free to post as much as you want. We do this for all new visitors as it helps to hold back the flow of spammers that seem to descend on us from time to time.

[Linuxbuilders]

Thank you. Sorry for posting it twice, just realized that it is on hold back after I have pushed the button. If not a big problem could you move it to falcon/tt section please. Thank you.

[BoNuS]

This looks very very familiar to me See my topic, screen is almost the same...
http://www.atari-forum.com/viewtopic.php?f=27&t=16237

Still haven't my Falcon back

[Mug UK]

Thank you. Sorry for posting it twice, just realized that it is on hold back after I have pushed the button. If not a big problem could you move it to falcon/tt section please. Thank you.

Those sections are for when your Falcon is back up and running

[simbo]

confirm 68901 clk4 and output of DE to videl pin23
confirm reset pulse at powerup and halt signal toggles

i think your machine as the addressing is running where bonus sick falcon board has no stobes atall
would benifit from a diagnostic cart try a svga monitor if possible incase its a monitor mode issue

can you carefully using the right tool take out and clean the pins of the tos chip
if its socket is cracked this can be an issue also you can use a rubber band as a temp fix if this is the case
ive seen them with this added {lol:)}



you can look at the falcon repair guide its in the falcon hardware forum further down the forums main page

[Linuxbuilders]

thnx, will do. How hot CPU should be? It gets fairly hot and then falcon stops responding whatsoever. Did you get this other machine fixed? If yes so what was the problem? I do not have diag cartridge, I need to buy one. So far I have plugged PC monitor into falcon, (it works with my STE). I get out of freq range message. When CPU hot I get black screen, when cold I get as on pics. I have found that 7408 is just clock buffer on R217-> R216,221,222 - so I have buffered each trace separately from each other and as source I use R217. I will get 74244 and replace them just in case because I get VCC shorted on some legs (I am not sure it it is correct, as long as they in I have no way to test it anyway). Thnx

[Linuxbuilders]

confirm 68901 clk4 and output of DE to videl pin23

Clock4 got nice wave,

atari3.jpg

DE is odd, pic below...

atari6.jpg

atari5.jpg

confirm reset pulse

Not very clean wave, my oscilloscope is quite old so it may be a problem - pic below...

atari4.jpg

at powerup and halt signal toggles

This will check later on

i think your machine as the addressing is running where bonus sick falcon board has no stobes atall

Could u re fraze, I do not understand ...

can you carefully using the right tool take out and clean the pins of the tos chip

Do not think it is a problem here...

you can look at the falcon repair guide its in the falcon hardware forum further down the forums main page

I got original service materials but will look at it too, thnx

[simbo]

reset looks as if the circuit is jammed on reset
as its pulsing continualy and it should be the same as the halt

the reset button can have a kind of metal jam inside that comes from the switch sliding contacts fragmenting
but this is NOT the issue here
the circuit will just generate a pulse pulse pulse if the switch is shorted

although the Xhalt is aok
so no reason why Xreset is so strange a waveform
?? and Xhalt is ok as they are supplied on power up and reset
from the same timer chip so should be the same
via a buffer...!!!

so given that U11 is used to invert reset to Xreset halt to Xhalt
and it's also used for the midi port {some dafto musitian has been in charge i think!!!}

its a fair chance just U11 is fried and some gates on it are damaged
its a 74LS06
and if its reset gate etc was damaged then the Xreset rail would maybe produce the waveform your getting
{floating damaged output gate rubbish and spurious clock pulses mixture as it looks above and your scope it aok
{nice one a guy with a scope!!! and a brain...for once...!! most users dont use scopes for some reason even when you can get a cheep 20mhz one for about 20 euro's etc or even use a logic probe with hold}}

the midi buffer U11 is a common problem with atari X and i would go ahead and replace it
you can pickup the reset pulse on pin 2 out pin 1 in
halt on pin 4 out pin 3 in
the in's are common

youll find its circuit on page 2 top middle

if the mpu gets a halt then it will still run as normal just not at the right time framing
so youll still get address strobes and other functions just no reset and halt
hence no boot
just garbage on the tv and no initial white screen or DE signal as this depends on a reset of the 68901 to state




what i was saying about mr bonus falcon thats here sick
is that the address strobe doesnt run and the machine never makes it out of reset
ive been at it now about 20hrs of time in total over about a year
changed a few ic's {as i have smd tools} but now stuck at the dma chip
not providing FC output correctly ive already found a faulty gal chip {replaced}
so im back at it on satarday lets see ..??

[simbo]

il will move this falcon topic over to the main 030 level hardware section tomorrow

so youll find it there

[Linuxbuilders]

Will do some more tests and let u know, thank you. I have no clue what this ROBOT does because I have never fixed one, so your help is very appreciated. Problem is that I have no idea what I should get on those pins, will go through ur guide tomorrow. 7406 looks like 7404, I got some 74HC04 in smd so I replace it and see. I do PICs so I am close to ATARI but like 20 years ahead and it is hard to step back . I hope we get this one working and create here very good topic for all people willing to fix their beasts. thnx

[simbo]

the picture you posted atari5.jpg is the XHALT pulse ??
if so the XRESET looks exactly the same


as for a robot im at a loss to know what you mean???
if i have made a typo dont panic
im dyslexic and dont often see errors ...

[Linuxbuilders]

Xhalt is nice 5V, Xreset goes bad, sometimes goes .84V and then we get black screen at all, sometimes goes 2.47V and then we get funny screen, I will replace the chip but first check what goes out when I get Xreset out of the chip, it is possible that some other chip is shorting XReset and it may not be U11 which is faulty, will post my findings, as well as this I will check timer 555 there but it does not look like faulty because XHalt and XReset are sourced from the some leg of it so I hardly believe in problem there. By the way just jb-d my IPad, it is rocking now I hope my Falcon will too soon , waiting for CT63 and CTPCI so it is kinda important to get falcon going before I get upgrades . I call my Falcon ROBOT...its purpose after fixing will be to drive my home automation system which I am building...

[simbo]

nice one ...

you may find as ive found before its simply a small shard of solder somewhere on the pcb
use a big mag glass and have a good look around each ic as you said before someone has fitted a buffer mod
maybe droped a blob etc..to the combel or videl
you can hunt for hours to find it and you find the tiny hair between two pins
sometimes its a bad decoupling cap {rare}

i would just carefully lift up pin 2 on u11 connect a 10K resistor to pin2 and ground {pin 7 or 8 on TTL chip}
and check it independant of the Xreset line and under load with the resistor
if its a clean waveform with normal voltage transition the chip is ok
and its further on

your right to be causious

if i need to find a short on the Xreset line
with pin2 lifted
you can use a small audio signal with the machine cold no power
set it to 1khz and 100mv p-p
then trace with the scope at each ic
where the audio is lowest at like 1mv p-p is the short
bit like how the telephone guy does it with cables
i simply use the pc audio card and a tonegen software
its no longer supported and is freeware
if you would like a copy pm me and ill email it or something
it allows for exact level from the pc
you can confirm with the scope its level

i use this

pcsoundcard >> cable rightchannel >>old pen >>glued in pin with audio
..................................ground connection crock clip...

and make a probe

[Linuxbuilders]

I have replaced U11, it was faulty, now I get nice 5V on pin 2 and 4, I get black perfect screen, DE is strange on U31, it is 2.01V and pulses to 2.02V in regular clock like 1Hz, scope shows what you see on atari5.jpg. Should I get there normal 5V or should it be some king of modulated strobe signal? I think it should be high 5V so any idea what can modulate this strange square wave on it? I have swapped U25 just in case to, How do I find if U31 is faulty? Any way to check it?

[simbo]

hi i think the DE signal is normal
i would look now to why the reset went bad and what damage the faulty chip has caused to other bits

looking at U24 MC6850 now is a good idea as its connected to U11 directly it may also have suffered damage

if it has then it maybe isnt passing serialised data or initialising and jamming the bus or it's 'holding' irq low etc..
you can test this at pin9 u25 with the scope and see if ints are accuring}
{also check khz500 for a clock if i remember it comes from the combel to pins 3/4 U24}

this would cause a black screen if the chip wasnt shifting data
i know this will be a black screen becouse i use a nemesis that uses pin 3,4 on these controllers
if i leave 3/4 unconnected on either of them this is what i get

if you have one of these chips change it out
youll find one in any atari st also
however its an earlier revision but will still work ok

oh and well done so far

[Linuxbuilders]

got some 520STF so will re solder, let you know tomorrow, thnx

[simbo]

one thing i will add is you can chase your tail if your not carefully

i would replace both U24 and U11 again
if U24 has suffered it could already again have damaged U11
then you find the new U24 again gets damaged
so ... concider changing it again first before adding power

[Linuxbuilders]

chips replaced, xhalt, xreset ok, de is funny as before, but now I get xbeer low 0V so there is some progress, steps to find which one is doing xbeer low?

this is pic from the screen now, sometimes it is black sometimes as u see on the pic. I have sty520 motherboard for parts, very limited source however some chips I can reuse. I would rather identify wrong ones instead replacing all of it. Any advice how to narrow faulty parts? VGA says that is out of range.

atari7.jpg

[simbo]

what you need to do now is check the signal daisy chain BG or bus grant to see which ic has the bus

you also need to check BR BGK bus request and bus grant acknowlage on the ics on the bus
when one request BR you should get a BG from the 68030
it depends at what part of the tos load up to ram stage or init that its breaking at
as different ic's will bus master at each stage...

you can watch most signals from the expansion port or the gals
i would now check all data lines are high when there is no reset or halt

if there is a duff ic it may hold just one data line on the data bus
do this test with no power using a multimeter
just ground the black probe and test with diode range all data lines at the cart port or ram
they should all read about 1000 or more

do you have
F030 Hard Guide 2.1
Atari_Falcon030_Service_Guide.pdf

there is a flow chart in the service guide for tos loading process and bootup
chip by chip

a quick test is to check cpuclk it will be 8mhz at reset and tos sets it to 16
if you turn the time base on the scope right up youll see the wave at 8 or maybe 16mhz
youll see it change easily

and check DTACK signals are there atall??

a good idea is to use a tick list on paper for all signals


if you look at page 61 of the service guide youll see the boot flow
basicaly you need to check the reset line
when you reset you will get another reset if this echo isnt there the 68030 isnt making it out of reset


i am serious when i say that at this point i would use a diag cart from best in usa
you find them on ebay or via his eshop or email him
the cost is about 80$ or less + shipping and any tax
maybe 100-120$ however it will tell you exactly what's up with it

http://dev-docs.atariforge.org/files/fa ... 4OCT92.zip
here is a dev doc to build a simple one
or
http://www.best-electronics-ca.com/falcon.htm

[Linuxbuilders]

Thnx Mate, just on the way to buy one, will do all as follows over a weekend and will let you know, thanx for help. It is a different thing to design a thing but fix another thing it is a skill which you get with experience, which I do not have with this equipment.
What I have got is Service Guide Part Number c303062-001 October1, 1992. It contains all schematics and lots of tech data. Will compare with the one you are talking about. Could you PM me where I can get the one you are talking about? If you need a copy of mine ask, will do so and post to you.

Regards.

[simbo]

http://dev-docs.atariforge.org/

here in the falcon section same place as the code above

you can find a torrent online with lots of atari stuff

the torrent looks stable

just copy the name of the .pdf above into google and youll see the torrent link
1.048GB file

[simbo]

http://cgi.ebay.co.uk/Diagnostic-Test-F ... 1e5c361f3a

i found the international ebay link for the guy who sells it for $40 +post

youll still need the cart but its easy to make one

[Linuxbuilders]

That is where I buy all things for atari, this guy is rock solid and cheaper then best. My testing cartridge and all other things are already on the way. Thnx

[Linuxbuilders]

Checked all lines on exp port, all 700 around. When i fire it up i never get any change in clock from 8 to 16. Address lines 4,3,2 are always L.
U42 pin3 always H, pin 5 and 7 H and L on reset, pin 9 and 12 and 14 and 16 and 18 always L, U56: BERR- H XROM3,4 - H BR and BG0 and BG1 and BGACK - H BMODE - H, U36: BR - H BGI - H BGA - H BERR - L RESET - H, U55: BERR - L HALT - H RESET - H BGACK - H BG - H BR - H, when I put my metter on 20v range on r234 I get strobes on the screen, look like vertical lines strobing in different places, I think videl wakes up because of disturbed clock.

There is one thing I have found:
U38 reset is low, I think it needs to be low because it comes from U25

And in the end GAL U63 holds EBERR down, when I get it out EBERR goes high on other chips. What would be a reason and what this GAL does?

Replaced U42,43,26 just in case, no change

u55 getting very hot. (I would estimate around 50'C)

This is what I have got once during my testing:

atari8.jpg

Will add more if find anything.