Faulty Falcon

Hardware, coding, music, graphic and various applications

Moderators: Mug UK, moondog/.tSCc., [ProToS], lp, Moderator Team

simbo

Re: Faulty Falcon

Postby simbo » Sun Jun 06, 2010 4:21 pm

can you check pin6 of the 030 mpu for cpuclk signal
this is the clock you checked that doesnt change ???
the combel changes and makes this signal
and i see your saying the halt is low this means the 030 mpu is halted
bad reset circuit , double buss error , bad cpu
as i said wait for a diag cart FIRST !!
before more tests its a very precise tool for sure!!

you need to make a cable to an st/f/fm/e if you have one
and follow steps on page 67 do this anyway cart or not
maybe if you dont have an st machine you can borrow one
its handy...
you can use a pc a similar way.. if it has a comport

im sure its still something to do with the reset circuit ???

i dont think 50deg is a good temp for the mpu usualy in halt fc 1,1,1 at around 20-30 at most
however ive measured a few at around 50deg that are ok
and i think maybe some double buss crunch maybe going on

the added gals are fsm's finite state machines
and are use to interplerate or interproximate the buss with respect to the mpu porocesses
a bit like a node array controller setup ...
the nodes are the peripherals like dsp and ccs 68901 and memory and expansion buss
so ok you replaced the reset and halt inverter and buffer ic
so now check each chip is getting !reset and !halt
as the mpu is halted in its process
by the halt pin it outputs there is only three answers to this
combel videl or dma {no fc control as bonus machine has}

a falcon has a very intergrated chipset signals system that makes things very hard
without more input data to the equation supplied by the diag cart

we should have a diag cart to hire i think ... maybe there is a usefull tool here

we can all use without making one

above all
the tracking and thru holes between layers on a falcon pcb
are known to be week
senario can be as simple as a shard of solder hits !reset and + 5v somewhere along the line to all ic's

and a hole thru goes open circuit from the +5v current overflow to the joint
the chips reset isnt affected becouse it tolerates this easy
however the reset TTL chip dies and so does the track somewhere.. along the Xreset path to ALL ic's

a fixer realises the TTL chip is u/s and replaces it
the machine is still not working???

becouse the tracking is open circuit maybe the shard is also still there
and out with the mag glass and a fine scalple
literaly +5v psu for 50miliseconds
will open a thru ... to the reset chip alone
as it tires to sink the current it fails
but so does the tracks
make a map on paper of all signals at each stage

you should get on power on
an initial low high low on halt and reset

and inverted on halt and reset on all ic's
as Xreset Xhalt

then 150ms later a Xreset on Xreset line again !!
{but not a Xhalt} or you get a BERR further on in the gal
first!!! check all ic's supplied get this ...!!!


... :?:

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Mon Jun 07, 2010 10:11 am

Small development in case, I have checked all reset functions and it looks good, I get 5V and there is no easily detectable rubbish on it, I have manually tried to simulate it and seems to be ok. When I get coprocessor out then whole unit gets alive, my cpu clock gets switched, address bus starts strobing, and screen gets lines going up/down like in old tv when signal is lost. When I put 68882 back in the whole thing is frozen like ice. Is there any way to inform atari that coprocessor it out? Like jumpers or something like jumpers? Or is it detecting it on its own? I wait for my cartridge, is there anything I could do in meantime? My EBERR gets cleared too when 68882 is out. I suspect that some idiot did put coprocessor wrong way in. I could see marks on the socket. Is there any way to test it?

U31 de is 0V, reset nice 5V, on reset I get pulse, tiny hard to detect pulse from 0V to 5V. I think it may reset the circuit and that is why I get my screen funny...The pulse must be coming from one of ics, when I manually ground inverter it is still there. The pulse reflects the screen too. Each time I see it on scope the screen jumps a bit too, internal speaker ticks like metronom in like 2Hz freq.

When I take ram out it stops, all gets quiet, no activity on bus, clock is changing from 8 to 16.

Check this out, it may give a clue:

http://www.youtube.com/user/Linuxbuilders

My testing cartridge should be here this week so when I get it I will post my findings. thnx.

Cpu 40degs


thnx
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Mon Jun 07, 2010 11:53 am

can you check mpu for !halt and !reset at power on and reset button press
pins 129 and pin 64

if someone plugs a coprocessor in its auto detected and used no jumpers need changed
if it was plugged in the wrong way round
there is a good chance from what you say the mpu/combel is fried however dont write it off yet
they connect to the fpu chip

i take it the needed jumper is on the expansion buss???

simbo

Re: Faulty Falcon

Postby simbo » Mon Jun 07, 2010 12:15 pm

i had a good look at the videos
and it looks like the videl is now being addressed however through the tos loading process a reset occurs
and this is the tick tick

something isnt replying somewhere to the periferal query

maybe a gal/plc has an issue ..

you need to do a signals tick list again,, at pulse and at power on

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Mon Jun 07, 2010 12:59 pm

should it work without coprocessor? I have got one ebay, will be here by end of the week but not plugging until beast works again.
all halts, resets and berrs, etc are good, by checking all signals all looks normal, is it possible that ram is cooked and does it at some address? it would be possible that ram starts loading until some address and then kicks on and resets whole box. Any way to check this? I will get new ram soon (with whole testing set) but I am a bit scared to plug it in...
In hell we will meet again...

HitPoint
Atari freak
Atari freak
Posts: 63
Joined: Mon Jun 05, 2006 10:41 pm
Location: St.Petersburg, Russia

Re: Faulty Falcon

Postby HitPoint » Mon Jun 07, 2010 2:51 pm

Hey, i have same things on my falcon. Check all big chips legs. on my falcon COMBEL legs just slipped out on one side. surely thats the problem...

i attach image what i have.
i solder legs with microscope and check it. after all my falcon working again! ;)
You do not have the required permissions to view the files attached to this post.

simbo

Re: Faulty Falcon

Postby simbo » Tue Jun 08, 2010 1:00 am

Linuxbuilders wrote:should it work without coprocessor? I have got one ebay, will be here by end of the week but not plugging until beast works again.
all halts, resets and berrs, etc are good, by checking all signals all looks normal, is it possible that ram is cooked and does it at some address? it would be possible that ram starts loading until some address and then kicks on and resets whole box. Any way to check this? I will get new ram soon (with whole testing set) but I am a bit scared to plug it in...



a stock falcon is supplied without fpu
you just plug one it 'pnp'

i think as you say maybe its been wrongly inserted

wait for the cart it will show you the exact fault

simbo

Re: Faulty Falcon

Postby simbo » Tue Jun 08, 2010 1:01 am

HitPoint wrote:Hey, i have same things on my falcon. Check all big chips legs. on my falcon COMBEL legs just slipped out on one side. surely thats the problem...

i attach image what i have.
i solder legs with microscope and check it. after all my falcon working again! ;)



if for one am glad its fixed !!! :P

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Tue Jun 08, 2010 8:07 am

http://au.farnell.com/jsp/search/browse ... %2B1002963

Do you think I could use any of those to replace MPU if faulty?


mine is SC414230FE16, if not then where should I look for its replacement, I am trying to get old mac with the some cpu but not sure if I win it, next question if it is 30MHz MPU model can it work on slower speed too?
thnx
In hell we will meet again...

User avatar
wongck
Ultimate Atarian
Ultimate Atarian
Posts: 12789
Joined: Sat May 03, 2008 2:09 pm
Location: Far East
Contact:

Re: Faulty Falcon

Postby wongck » Tue Jun 08, 2010 11:20 am

simbo wrote:
HitPoint wrote:Hey, i have same things on my falcon. Check all big chips legs. on my falcon COMBEL legs just slipped out on one side. surely thats the problem...

i attach image what i have.
i solder legs with microscope and check it. after all my falcon working again! ;)


if for one am glad its fixed !!! :P


legs are so small 8O .... man you guys are good [smilie=greencolorz4_pdt_01.gif]
My Stuff: FB/Falcon CT63 CTPCI ATI RTL8139 USB 512MB 30GB HDD CF HxC_SD/ TT030 68882 4+32MB 520MB Nova/ 520STFM 4MB Tos206 SCSI
Shared SCSI Bus:ScsiLink ethernet, 9GB HDD,SD-reader @ http://phsw.atari.org
My Atari stuff for sale - click here for list

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Tue Jun 08, 2010 11:59 am

This is actually still large size, I can do it with use of special tools.
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Tue Jun 08, 2010 12:06 pm

if your mpu turns out to be gubbed
i can send you one i know is aok

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Tue Jun 08, 2010 1:14 pm

thnx, will wait then for cartridge and will see, I will return favor or pay. Will see. I have found on ebay this,http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=220530250903&ssPageName=STRK:MEWAX:IT
which would be okey but postage price is just too much, other thing is that I do not know if I can clock them down.
In hell we will meet again...

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Thu Jun 10, 2010 10:10 am

Got cartridge, it is atari falcon diagnostic v1.276

It does not work with my ste at all, when i put it into falcon i get rubbish, need step by step instructions, i did crosover rs cable to terminal, set it for 9600,8,1,n it fires up e0 and then dies. When i have cartridge in it i get some movment and rubbish on the creen and then it stops. Please let me know step by step what to do and how and then i put video on utube what i get. Thnx
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Thu Jun 10, 2010 10:28 am

the complete guide to how to use the cart is in the service manual
you downloaded
page 67>

follow the steps there first {dead machine uut}

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Thu Jun 10, 2010 10:58 am

Did it, no go. Should it work with st? I wonder it the cartrigde is working. It does not click on st or ste...

When cartridge in the clock does not change to 16mhz, when out then it does.
On xrom 3,4 i get wave on H. Not sure why.

So there is one conclusion, if unit is dead then even holy mother will not help in body of any cartridge.


Disconnected reset from cpu, no more ticking, when connect it back all goes crazy again. Reset is clean when cpu out, when cpu is back then it makes reset low like in 2hz, i think it is cpu cooked.
It ticks when clock goes from 8 to 16mhz, clocks goes up when reset is put back to mpu, so it looks like u56 is driving the clock up down and all over again, but it must be very fast because I can't see it on scope. So it can be u56 cooked or cpu responds to it, it detects cpu/mpu and then gets it lost and resets clock again. Hard to guess which on is bad.
But when reset taken out the clock stays 16mhz. So it is possible that mpu is just crazy and when 16mhz applied then it gets unstable. Does it make sense?
Last edited by Linuxbuilders on Thu Jun 10, 2010 9:17 pm, edited 1 time in total.
In hell we will meet again...

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Thu Jun 10, 2010 12:39 pm

When I put mpu reset on 10k to pos and look at scope i see the some strobe to L, so it is definitely mpu restarting itself, now is a question: how do I check if it is mpu crooked or some chip is doing rubbish on data/address bus and mpu just restarts itself because of it? It does not look like ram bad, i have swapped it and i just get different pattern on the screen. Thnx


I have soldered leds into expansion port to see where it gets stack:

http://www.youtube.com/user/Linuxbuilders

and this is what my terminal shows from serial port:
IMG_1386.JPG


over a weekend will do proper board with leds to detect address and data and see where it is freezing or restarting.
Any suggestions are welcomed.

thnx
You do not have the required permissions to view the files attached to this post.
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Thu Jun 10, 2010 9:19 pm

Linuxbuilders wrote:Did it, no go. Should it work with st? I wonder it the cartrigde is working. It does not click on st or ste...

When cartridge in the clock does not change to 16mhz, when out then it does.
On xrom 3,4 i get wave on H. Not sure why.

So there is one conclusion, if unit is dead then even holy mother will not help in body of any cartridge.


Disconnected reset from cpu, no more ticking, when connect it back all goes crazy again. Reset is clean when cpu out, when cpu is back then it makes reset low like in 2hz, i think it is cpu cooked.
It ticks when clock goes from 8 to 16mhz, clocks goes up when reset is put back to mpu, so it looks like u56 is driving the clock up down and all over again, but it must be very fast because I can't see it on scope. So it can be u56 cooked or cpu responds to it, it detects cpu/mpu and then gets it lost and resets clock again. Hard to guess which on is bad.
But when reset taken out the clock stays 16mhz. So it is possible that mpu is just crazy and when 16mhz applied then it gets unstable. Does it make sense?


it should i agree work with the st level X so there is a prob with it
ask in email to the best see what he says...

however with the falcon if the cart will just not work and as you maybe found an issue with the fpu
the mpu 030 is gubbed...

and as its connected mostly to the mpu maybe equal to how the fpu is also connected to the combel
maybe one or both are gubbed

confirm the cart in stX and then see...

the falcon is far more intergrated in terms of embeded units and also complex control signals

inspect and clean especialy the cart port

youll be AMAZED how many stX and one or two falcons in the last 20 or more years ive found a shard of wire in
accross the bus

its lucky the cart port sometimes !!! has current limits

most people dont think too look

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Thu Jun 10, 2010 9:41 pm

Now we have long weekend so I will give it proper cleaning however I doubt that it is the problem. Can you let me know what would be conditions to get mpu from you and where should I look for other chips? The man has confirmed that cartridge will not work with STx and only falcon is able to read it and it was tested before shipping. When I get proper address and data indicator I will post results. Thnx
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Thu Jun 10, 2010 11:59 pm

Linuxbuilders wrote:Now we have long weekend so I will give it proper cleaning however I doubt that it is the problem. Can you let me know what would be conditions to get mpu from you and where should I look for other chips? The man has confirmed that cartridge will not work with STx and only falcon is able to read it and it was tested before shipping. When I get proper address and data indicator I will post results. Thnx



ill be back on after the weekend
im off for a few days holidays

cya soon

dont worry just pm me your address ill send you the mpu
i know for sure its ok
youll prob need to tidy pins and resolder carefully

50 deg is too hot and as the address strobe it running you should get fc0 1 2 control lines
these are both related to mpu combel and dma and ticking like this is a sure sign is either combel or mpu
as rom 1 ,2 3,4 dont seem to be made maybe its the combel
however lack of pre processes point to the mpu
even the service manual says this is whats wrong page 67 first

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Fri Jun 11, 2010 7:06 am

Ok I think I have isolated problem, when mpu is connected to reset I get LOW strobe over it, it resets whole system, 2Hz is fast enough to stop whole booting process, that is why I get a tick because snd chip restarts itself. I have checked clocks and when 16Mhz is on then I get strobe over it of 8Mhz which indicated that combel restarts and sets clock back to 8 and then 16. This strobe is present on DOTCK and CPUCLOCK, both are not related to each other and strobe time is slightly longer on DOTCK. So what I did I have buffered reset for CPU so it is isolated from rest of the system. This gave me clean clocking and no more ticking. When I check reset from mpu on my buffer I get LOW strobe there coming out of mpu reset pin. So I think my mpu is faulty and probably rest of the system is ok (I hope). When cartridge is in it locks itself on A4F0 address when 14MB ram is in, when 1MB ram is in then I get different address. So I think it starts executing program and locks up when mpu hangs.
There is only one question, is it possible that MPU is executing reset vector and sets reset pin low for a purpose? Is reset pin working in 2 directions or should it be only an input?

I have found where to get chips too: http://www.best-electronics-ca.com/custom-i.htm

So will order some and swap it and see.

http://www.faqs.org/faqs/motorola/68k-chips-faq/

this is one interesting faq:
http://patpend.net/technical/68000/68000faq.txt
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Mon Jun 14, 2010 11:33 am

best is a quircky guy to deal with for sure {}good luck
his minimum order is 50$ parts {not post}
so his mpu combel etc are $45 so he is 'wide' here

i got two ic's from them and ended up paying my goverment £40 in tax

terrible
i would have ordered one at a time if i had known and avoided the tax
however his min order is the problem for sure here

ive added you to skype
give me a vid/audio call later...

im on most days this week

two heads are better than one
i feel an audio conversation is cool here
and the mpu is the cause i feel

as the combel is setting itself to 16mhz then the mpu calls a reset
this is normal
some reading is needed to see where the issue is FIRST
before replacing the mpu
not sure its the mpu myself
as i say the fpu connects both to the mpu and the combel ic
a missing signals from a burned gate can be enough
given the fpu could have been inserted wrong then something either combel or mpu is gubbed

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Mon Jun 14, 2010 12:03 pm

I have some doubt in FC2, FC1 and 0 is strobing like 3.8V down and up. FC2 is always up. I have disconnected all chips but gal and it looks like gal keeps it at this level, gal off and FC goes up to 4.9V. when reset pressed FC1 and 0 goes down and up and FC goes like 3V and then back to 3.7V. It looks odd too me...
when I take out FC2 of mpu all hangs, when I take it out of DMA, it starts and hangs, when I take it out of Combel then it goes as normal. (If I could call it this way), when gal is out all hangs too.

this is state of exp slot with cartridge in:
1 -l 2-l 3-h 4-h 5-h 6-h 7-l 8-h 9-3.8V 10-l 11-h 12-l 13-l 14-h 15-h 16-h 17-h 18-h19-h 20-h 21-h 22-l 23-working 24-h 25-h 26-l 27-l 28-l 29-l 30-l 31-l 32-l 33-l 34-l35-l 36-l37-l 38-h 39-h 40-l 41-h 42-h 43-l 44-h 45-l 46-l 47-h 48-l 49-h 50-l
when cartridge is out I get the difference in:
4-strobe 5-strobe 6-strobe 7-strobe 10-strobe 11-strobe 21-strobe all address lines-strobe

so it gets stuck when cartridge is in, the some like when fpu is in so I think my fpu may be ok and something what is shared/executed initially fails when those chips are in.
In hell we will meet again...

User avatar
Linuxbuilders
Atari freak
Atari freak
Posts: 54
Joined: Wed May 26, 2010 1:48 pm
Location: Melbourne AU
Contact:

Re: Faulty Falcon

Postby Linuxbuilders » Tue Jun 15, 2010 9:21 am

http://www.freescale.com/webapp/sps/sit ... %20Sheets&

Reset is bidirectional and those strobes over this pin may not necessarily mean that mpu is faulty. It may reset itself and rest of the circuit because of illegal value of data/address on the bus. I think that here we need a testing tool which will be plugged into address/data/control bus and stream over rs232 what happens on it. I need to think about it, it would be easy to do on pic, very simple firmware and circuit. I think without exact knowledge what is happening on those bus lines it will be very hard to pinpoint where is the problem.
In hell we will meet again...

simbo

Re: Faulty Falcon

Postby simbo » Tue Jun 15, 2010 9:37 am

as i said in the first few posts
you need to check the bus grant signals on each ic
to see which one is in grant when the reset pulse happens

the ic granted the bus isnt replying to the ic granting it

check out the pin functions list for the dma controller 2.1.6.1 {in service guide} and 2.1.1 mpu pin functions list


also use the cart with the fpu removed and see what the error code is
refer to the error code list on page 91 onwards


check mfp pin 8 for activity incase rs232 isnt right

looking at section 3 of the trouble shoot a dead unit
it says if the cpu is not halted it should be reading instrucions from the rom on the cart
and address and data lines should be toggling
if not replace the cpu


Social Media

     

Return to “Professionals”

Who is online

Users browsing this forum: No registered users and 5 guests