Falcon IDE problems...

Hardware, coding, music, graphic and various applications

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

Dal
Administrator
Administrator
Posts: 4178
Joined: Tue Jan 18, 2011 12:31 am
Location: Cheltenham, UK
Contact:

Re: Falcon IDE problems...

Postby Dal » Wed Oct 20, 2010 10:08 pm

Right U25 changed - No improvement...
U24 changed - No Improvement...

There is another 6850 at U52 - should I change this also?

...I bought two MC6850's just in case :wink:
Mega"SST" 12, MegaSTE, STE: Desktopper case, IDE interface, UltraSatan (8GB + 512Mb) + HXC floppy emulator. Plus some STE's/STFM's

simbo

Re: Falcon IDE problems...

Postby simbo » Wed Oct 20, 2010 10:29 pm

the only thing that is left past the two buffers is the combel ic

as a custom ic and directly connected to ide
i would now look to it and check its signals using a scope

its 4 pins
believe me when i say leave the machine till this is done
and you need to fork out $45 for a new one + post
and an smd rework unit to replace it

sad news... :| and a hard job ... i have done a few times before

Dal
Administrator
Administrator
Posts: 4178
Joined: Tue Jan 18, 2011 12:31 am
Location: Cheltenham, UK
Contact:

Re: Falcon IDE problems...

Postby Dal » Thu Oct 21, 2010 10:32 am

Ok cheers Simbo, I have sent you a pm...
Mega"SST" 12, MegaSTE, STE: Desktopper case, IDE interface, UltraSatan (8GB + 512Mb) + HXC floppy emulator. Plus some STE's/STFM's

simbo

Re: Falcon IDE problems...

Postby simbo » Thu Oct 21, 2010 10:32 pm

if the Xdiskint signals isnt present from the dma chip
pin 141

youll get no ide interrupts

i see in a previous post that you have a signal here at u24 page 6 and page 2 of the schematics
pin 2 of u24

not sure that its the combel or the dma
needs a scope to see whats going on
with both ic's

i will hedge my bet's and say the dma is the most likely
as the combel shows no signs of other damage

the dma at pin 141 maybe is the issue

it will be high and go low on a dma disk transfer event
to marry with the ideirq that give an interrupt request to the 68901
irq stack then the 68030 will handle the transfer as its event will flag...

as it says error and not no reply then i can assume that the disk is there
but the machine cant marry with it
ie it knows something is there
maybe just no dma interrupts if no interrupts it will reply once
then lock

you may find it helpfull to look at the berr signal and load up the driver
see if there is a buss error generated {exceptions in dma wont cause bombs if the machine expect something
and gets something it wont berr}

if you get a berr generated its most likely the dma
if no berr mostly the combel

Dal
Administrator
Administrator
Posts: 4178
Joined: Tue Jan 18, 2011 12:31 am
Location: Cheltenham, UK
Contact:

Re: Falcon IDE problems...

Postby Dal » Mon Oct 25, 2010 3:09 pm

OK Diag cart arrived today so quickly went up to the poorly Falcon on it's sick bed and run it up. I performed the IDE diagnostic and got this result:
IDE Diag.jpg


Was looking for a document that would tell me what the address F00039 refers to (DMA??)
You do not have the required permissions to view the files attached to this post.
Mega"SST" 12, MegaSTE, STE: Desktopper case, IDE interface, UltraSatan (8GB + 512Mb) + HXC floppy emulator. Plus some STE's/STFM's

User avatar
wongck
Ultimate Atarian
Ultimate Atarian
Posts: 12780
Joined: Sat May 03, 2008 2:09 pm
Location: Far East
Contact:

Re: Falcon IDE problems...

Postby wongck » Tue Oct 26, 2010 12:04 pm

Dal_1978 wrote:OK Diag cart arrived today so quickly went up to the poorly Falcon on it's sick bed and run it up. I performed the IDE diagnostic and got this result:


From what I know. the CT63 firmware came with a diagnostic just like the cart.
You just press CONTROL + ALT + LSHIFT + RSHIFT before the boot sequence.
http://forum.atari-home.de/index.php?ac ... pic=6990.0
(page is in German... so use Chrome or Google)
This depends on the version of firmware you have.
I think 1.03c has it.
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

Dal
Administrator
Administrator
Posts: 4178
Joined: Tue Jan 18, 2011 12:31 am
Location: Cheltenham, UK
Contact:

Re: Falcon IDE problems...

Postby Dal » Tue Oct 26, 2010 3:59 pm

Oh yeah! Wish I realised that before I shelled out £20 on the diag cart from 16/32! Nevermind...

I get a slightly different result from the diags on the CT60:

EB Access Error. Access Address: FFF00039 Program Counter: 00ED2ECA

This message appears twice against a red screen.
Mega"SST" 12, MegaSTE, STE: Desktopper case, IDE interface, UltraSatan (8GB + 512Mb) + HXC floppy emulator. Plus some STE's/STFM's

simbo

Re: Falcon IDE problems...

Postby simbo » Tue Oct 26, 2010 11:23 pm

error EB is a bus error generated by the mpu or the combel
becouse a device fails to reply

you will need a scope and look at the dma pin 141
before you go any further
if you dont have one go to a TV or radio amature shop there are many in england and just a couple in scotland

they will have a scope and i am sure can test one pin is there or not

i would take along some cream buns as a bribe
the dma must generate and interrupt together with the ide if you dont get this signal at the mfp {68901 }
from u24 output pin then the device will appear not to be there
not sure this would stop the falcon seeing it

access error is usualy interrupt related but could equaly be the combel ic
more tests needed

take a laptop with a copy of the service guide on it
and your cart
leave the ct60 at home

simbo

Re: Falcon IDE problems...

Postby simbo » Wed Oct 27, 2010 12:28 am

http://cgi.ebay.co.uk/Pocket-Oscillosco ... 230a88b70a

here is a nice scope ready build but uncased

looks very well made ... for £30~

simbo

Re: Falcon IDE problems...

Postby simbo » Wed Oct 27, 2010 12:33 am

F00039 is in ram ,... its at a rom location address the instruction used to read and write is making the access error
live at so the address is shown where the routine went wrong lives..

this is a normal error to show the ide isnt being accessed 'half there'

i think myself i would replace the dma ic but this is a close guess

User avatar
wongck
Ultimate Atarian
Ultimate Atarian
Posts: 12780
Joined: Sat May 03, 2008 2:09 pm
Location: Far East
Contact:

Re: Falcon IDE problems...

Postby wongck » Wed Oct 27, 2010 12:22 pm

Dal_1978 wrote:Oh yeah! Wish I realised that before I shelled out £20 on the diag cart from 16/32! Nevermind...



Sorry I did not read this thread completely as it got a bit too hardwaredy and writing got a bit too long....
It was the big red picture that caught my eyes.... I saw that before sending my CT63 to the hospital.
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

Dal
Administrator
Administrator
Posts: 4178
Joined: Tue Jan 18, 2011 12:31 am
Location: Cheltenham, UK
Contact:

Re: Falcon IDE problems...

Postby Dal » Wed Oct 27, 2010 12:58 pm

Too many words hurting your eyes Wongck?? :lol: :lol: :lol:
Mega"SST" 12, MegaSTE, STE: Desktopper case, IDE interface, UltraSatan (8GB + 512Mb) + HXC floppy emulator. Plus some STE's/STFM's

User avatar
wongck
Ultimate Atarian
Ultimate Atarian
Posts: 12780
Joined: Sat May 03, 2008 2:09 pm
Location: Far East
Contact:

Re: Falcon IDE problems...

Postby wongck » Wed Oct 27, 2010 1:16 pm

Dal_1978 wrote:Too many words hurting your eyes Wongck?? :lol: :lol: :lol:

:mrgreen: long posts does.... just loose track of the topic.... so I give up reading it & skip :roll:
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

simbo

Re: Falcon IDE problems...

Postby simbo » Thu Oct 28, 2010 12:06 am

its a point to note that from the address above

you can note what part of the access went wrong

by looking at what ins opcode is run at that address

39 will be part way through the access attempt

so knowing how a op code routine is set out
will tell you what access ins went wrong

knowing ide ins range on op will help

this is easy

i have a 40 way pci digital i/o card and can see the atari bus via the cart port
on capture.... even generate my own interupts and just monitor data flow in RT

Dal
Administrator
Administrator
Posts: 4178
Joined: Tue Jan 18, 2011 12:31 am
Location: Cheltenham, UK
Contact:

Re: Falcon IDE problems...

Postby Dal » Thu Oct 28, 2010 6:53 am

Next time I'm up in Glasgow, you need to show me some of this stuff! Would love to see how this is done...
Mega"SST" 12, MegaSTE, STE: Desktopper case, IDE interface, UltraSatan (8GB + 512Mb) + HXC floppy emulator. Plus some STE's/STFM's

simbo

Re: Falcon IDE problems...

Postby simbo » Sat Nov 20, 2010 3:50 am

dont worry darren

your falcon is in good hands
and the combel has been fitted
i still have a few issues
so i am doing pin to pin and track pad to pad tests just now
208 pins .26mm pitch ,,, isnt much fun
it will be back with you soon all being well
have an issue just now with coms to the videl and mpu
but i will resolve this tomorrow morning .. expected bad pin... had a bit of work with TT;s and magnum cards
so end of next week its done and tested and i hope its back too you m8 a couple of weeks
after full test fixture tests... and a burn in of the new combel ic
and i can and will offer regrets for the forum staff for screwing up your member account {andy!}
this will be resolved soon
feel free to phone me any time
except when i am a bit pissed {friday feeling and two hard days and nights soldering before}
been working on frank lucas TT's also on its mpu section and ram L1 errors like my TT
got mine working atlast was the mcu another hard ic to hand eye solder...
and tested the magnum cards ...

so ill get back to your machine sunday ... planned...!!!
excuses excuses ... but i need a break from it take your time is the answer... here...!
its not the first time ive had a chip DOA from best..!
but there is no proving it
i will order another if needed...
:cheers:

simbo

Re: Falcon IDE problems...

Postby simbo » Fri Nov 26, 2010 6:24 am

ok i have decided that this new combel was dead on arrival
ive tripple checked all solders pad to pad, pin to pin, pin to pad

and there is no shorts no open
and it doesnt work

so its a gooser


ill order another from mr best...

simbo

Re: Falcon IDE problems...

Postby simbo » Tue Apr 26, 2011 12:25 pm

progress

ok the latest combel is in place
and i will continue to test pin to pin and pin to pads
at the other end
so tracks can be tested also
its a thankless chore to do this however needed it is { a yoda chore}
i am almost ready to power it up
ill double/tripple check each pin to pad and pad to pin
before power!!
so dull day now as the sun died
tomorrow is another day and can save you 45$ costs etc

so for the next few days ill be quiet and working
got a TT to repair also
possible a book to read off the rom from ?? :cheers:


got a 68060 mpu from dal this morning for his ct60 board
i also made up a ctcm board
as rodi doesnt supply them now

i also have two spare made up pcb can be easily populated as i have 2 spare ic's etc

so if someone needs a CTCM clock board
let me know
cost is £15 post... ea

EvilFranky
Atari Super Hero
Atari Super Hero
Posts: 870
Joined: Thu Sep 11, 2003 10:49 pm
Location: UK
Contact:

Re: Falcon IDE problems...

Postby EvilFranky » Fri May 20, 2011 10:14 am

I may be interested in one of those CTCM boards for my CT60...if I can find a reasonable priced rev.6 68060 :roll:

SofiST
Atari Super Hero
Atari Super Hero
Posts: 591
Joined: Tue Apr 28, 2009 12:25 pm

Re: Falcon IDE problems...

Postby SofiST » Fri May 20, 2011 12:37 pm

simbo wrote:F00039 is in ram ,... its at a rom location address the instruction used to read and write is making the access error
live at so the address is shown where the routine went wrong lives..

this is a normal error to show the ide isnt being accessed 'half there'

i think myself i would replace the dma ic but this is a close guess


All in area $F00000 - $F00039 is not RAM but IDE registers on attached disk (or CF card) . Read available DOCs, please . There is a plenty of them around.
And Falcon IDE (as others too) goes not via DMA. On Falcon IDE logic is integrated in Combel chip.

EvilFranky
Atari Super Hero
Atari Super Hero
Posts: 870
Joined: Thu Sep 11, 2003 10:49 pm
Location: UK
Contact:

Re: Falcon IDE problems...

Postby EvilFranky » Fri May 20, 2011 12:47 pm

Isn't there some way for the Falcon to get hard disk -> DMA via the Blitter?

I'm sure I read something on here about it :?:

simbo

Re: Falcon IDE problems...

Postby simbo » Sat May 21, 2011 7:15 am

or i can just finish off replacing the combel
as the older one had issue
and this i would feel would also effect other parts of the machine
as the ide control register on the combels die was fried
and should be a WARNING to others to connect the ide port right and not 4pins down
as this connects 5V to the combels ide control register...pins
:::::::::::::::::::::::::
..^.........................^
as this is what happend to dals i think before he got the machine

ive almost finished charting each pin to each place
after checking all pins for shorts and opens
it usualy takes about a week to chart 172 pins out
and test the bus for stuck pins etc with no power
using two needle probes and a strong light or better the sun {if i ever get to see it here in glasgow}


then its on to warm tests with the cart and serial ..

simbo

Re: Falcon IDE problems...

Postby simbo » Sat May 21, 2011 7:28 am

SofiST wrote:
simbo wrote:F00039 is in ram ,... its at a rom location address the instruction used to read and write is making the access error
live at so the address is shown where the routine went wrong lives..
this is a normal error to show the ide isnt being accessed 'half there'
i think myself i would replace the dma ic but this is a close guess


All in area $F00000 - $F00039 is not RAM but IDE registers on attached disk (or CF card) . Read available DOCs, please . There is a plenty of them around.
And Falcon IDE (as others too) goes not via DMA. On Falcon IDE logic is integrated in Combel chip.




the address i mentioned is also a ram location
even the bold ppera knows that ALL memory in an atari is an addressable ram location
therefore so are the registers!!!!

your splitting hairs and i dont see your point

i already know exactly what is wrong with this machine
you see ppera this is yet another example why if you dont have anything to add
DONT you just inflame people like me when you fail to read a topic and just blabber the last post
and 'your opinion'
i like the way you add that the DMA chip has nothing to do with ide
it just shows how disconnected you are within a falcon or any st
the dma is responsible with all actions of hdd access to memory even the ide ...
it will be granted the bus to make a shift then the shifter withinthe videl
will move the data to or from ram using the combel mad register
ide to scsi defo uses the dma
the dma is still always incontrol even if other ic's make the shift of the actual data
and is responsible for address strobe and mpu actions via the FC function control and the busgrant chain
you never know it will make a change if i dont have to clean up a topic ...

SofiST
Atari Super Hero
Atari Super Hero
Posts: 591
Joined: Tue Apr 28, 2009 12:25 pm

Re: Falcon IDE problems...

Postby SofiST » Sat May 21, 2011 8:33 am

EvilFranky wrote:Isn't there some way for the Falcon to get hard disk -> DMA via the Blitter?

I'm sure I read something on here about it :?:


Yeah. But it has no sense at all - not faster than via CPU. Only on ST(E) is faster.
If you thought that doing via blitter may fix the problem - hardly. Still, logic for selecting IDE port on Falcon works on same way (connecting at given address area), just transfer self goes via blitter, what is on driver SW to handle.
But I can imagine that adding extra IDE logic on Falcon will work - same as add-on adapters for ST(E). And it can work even better than internal, if uses more standard timings, according to newer ATA specs.
Last edited by SofiST on Sat May 21, 2011 8:49 am, edited 1 time in total.

simbo

Re: Falcon IDE problems...

Postby simbo » Sun May 22, 2011 10:21 pm

st softy
please keep your defluxial comments to your self !! next time


Social Media

     

Return to “Professionals”

Who is online

Users browsing this forum: No registered users and 1 guest