CT60 memory test strangeness

Discuss CT60/CT63, CTPCI, SuperVidel and EtherNAT hardware here.

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

Post Reply
sety
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 112
Joined: Sat Nov 04, 2017 12:42 am

CT60 memory test strangeness

Post by sety »

Hi.

I've had this problem for as long as I've had the CT60 - If I let it complete the ST RAM check/start the TT RAM check it won't boot. It usually crashes with an access fault. I've lived with it simply by using a zero second boot delay, thus avoiding the RAM check all together. But I've had some free time over the break and I thought I would revisit it.

Anyway, I got it to boot into GEM after completing a RAM check and found that while it's in this state, it thinks it's an ST!

When I got it to boot it behaved exactly like an ST. It came up in ST Medium, I could select ST LOW but could not select ST HIGH (presumably because there was no SM124 connected). No falcon modes worked.

I've flashed on the later ABE_V7F and SDF_V7D firmwares but it didn't make much difference.
I'm using the boot firmware pm 1.03c

Has anyone experienced anything like this before? Screenshots below. Thanks!

This from sysinfo:
IMG_1513 2.JPG
This from the setup screen:
IMG_1518 3.JPG
Even when I try to boot MiNT it goes looking for the 68K kernel (which I don't have)
IMG_1520 2.JPG
You do not have the required permissions to view the files attached to this post.
ATARI FALCON/CT60 @99/25Mhz 512MB - ATARI FALCON/CT60e @77/16Mhz 512MB - 520STF 1MB TF68020 - 520ST unmodified (my very first computer) - A couple of useless XE's - and a really cool Atari desktop calculator :)
User avatar
wongck
Ultimate Atarian
Ultimate Atarian
Posts: 12908
Joined: Sat May 03, 2008 2:09 pm
Location: Far East
Contact:

Re: CT60 memory test strangeness

Post by wongck »

wow... you got a 3-in-1 ST, Falcon & CT60.

I have no idea why it is behaving like that but remove the CT60 and see if it still thinks it is a ST.
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
sety
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 112
Joined: Sat Nov 04, 2017 12:42 am

Re: CT60 memory test strangeness

Post by sety »

wongck wrote:wow... you got a 3-in-1 ST, Falcon & CT60.

I have no idea why it is behaving like that but remove the CT60 and see if it still thinks it is a ST.
:lol: I did try to load an ST game but it didn't work.

It only thinks its an ST if I let it complete the memory check in 060 mode. If I skip the memory check, everything functions normally.
ATARI FALCON/CT60 @99/25Mhz 512MB - ATARI FALCON/CT60e @77/16Mhz 512MB - 520STF 1MB TF68020 - 520ST unmodified (my very first computer) - A couple of useless XE's - and a really cool Atari desktop calculator :)
mikro
Hardware Guru
Hardware Guru
Posts: 2200
Joined: Sat Sep 10, 2005 11:11 am
Location: Kosice, Slovakia
Contact:

Re: CT60 memory test strangeness

Post by mikro »

This might be a deeper hardware problem - CPU type is usually recognised by catching an exception, for instance FreeMiNT uses "move.w ccr,dx" to distinguish between 68000 and 68010 and accessing the CACR (cache) register to distinguish between 68010 and 68020+.
misha76
Atarian
Atarian
Posts: 1
Joined: Sun Jun 17, 2018 11:44 pm

Re: CT60 memory test strangeness

Post by misha76 »

sety wrote: Has anyone experienced anything like this before? Screenshots below. Thanks!
My CT60e behaves the same: If I give him to test all the memory it crashes with an access fault.
GoldenGrahams
Atariator
Atariator
Posts: 27
Joined: Wed Dec 16, 2015 4:15 pm

Re: CT60 memory test strangeness

Post by GoldenGrahams »

I have this exact problem too with my CT60e. It boots fine if I abort the memory test before it enters the TT RAM test, however if I let it enter the TT RAM test then it will always boot with 'issues'. To begin with that 'issue' was an access fault, i.e. the machine 'crashed' and didn't let me do anything. I then re-flashed TOS on the CT60e which stopped the 'access fault' problem but instead I'm now only able to set 'ST compatible video modes'. If I try to select any other kind of video mode it just defaults back to the most recent ST compatible video mode that was set. Note that I have no issues whatsoever if I abort the Memory test before it enters the 'TT Ram test' phase. Has anybody else experienced this problem and have any ideas on how to resolve? It's definitely not a problem with my the memory stick for my CT60e as I have tried several 256mb and 512 ones. Thanks.
joska
Hardware Guru
Hardware Guru
Posts: 4694
Joined: Tue Oct 30, 2007 2:55 pm
Location: Florø, Norway
Contact:

Re: CT60 memory test strangeness

Post by joska »

I have a very similar problem with my CT60e. If I let it boot uninterrupted (i.e. letting it run all the memory tests and then boot) it will boot into some sort of ST-mode. ST-High resolution (red on white), if I attempt to start some Falcon-only software it complains that this is not a Falcon. After a reboot everything behaves normally.

If NVDI is enabled during such a boot it will crash. If not, I get to the red on white ST-high desktop.

My Falcon is otherwise very stable in both 030 and 060 (95MHz) modes.
Jo Even

VanillaMiNT - Firebee - Falcon060 - Milan060 - Falcon040 - MIST - Mega ST - STM - STE - Amiga 600 - Sharp MZ700 - MSX - Amstrad CPC - C64
User avatar
tksm
Retro freak
Retro freak
Posts: 12
Joined: Fri Dec 20, 2019 3:35 am

Re: CT60 memory test strangeness

Post by tksm »

Hi.
First of all please check all settings in control panel CT60 CONF if that not help remove CT60 and boot Falcon without it then make all tests.
True problem is CLOCKPATCH if you wuld like use ct60 then please remove clock patch! and make sdma mod.
Check if your's modul (RTC) has battery ok.
Sometimes memory on ct60 causes issues (must be checked on pc and please clean pin's with isopropanol)
For me the best is Kingston 512 mb sdram.
Sometimes cleaning goldpin's on motherboard helps.
When I switch in 030 mode Falcon make memory test 14 mb then sysinfo shows ATARI FALCON 030 16 MHZ in 060 mode make 14 MB and 512 MB test without any problem's then everything work's ok.
emcclariion
Captain Atari
Captain Atari
Posts: 218
Joined: Wed Nov 28, 2012 6:37 pm

Re: CT60 memory test strangeness

Post by emcclariion »

MIne is the same, dont have clock patch, but my RTC battery is dead, but done use it
Atari TT Nova card Mach64, Lightning VME, 64MB of RAM, Mega STE 4MB, Nova mach 32 and Atari STE 4MB TOS 2.06 Netusb Ultrasatan 2, Atari Falcon CT60 512MB Netusb
GoldenGrahams
Atariator
Atariator
Posts: 27
Joined: Wed Dec 16, 2015 4:15 pm

Re: CT60 memory test strangeness

Post by GoldenGrahams »

Hi @tksm ,

- I've tried playing around with CT60CONF to no avail. Which settings in CT60CONF would you suggest changing exactly?
- My Falcon works perfectly with CT60e removed
- My Falcon does not have the clock patch.
- I've never heard of the 'SDMA mod'. What is it and why do I need to do it? Willy (seller of CT60e) advertised the CT60e as 'plug and play' with no such mods necessary.
- My Falcon's RTC has been tested recently, it keeps the correct time and date and works fine.
- I have tried several different sticks of SDRAM. Some 512mb, some 256mb, some CL2, some CL3. To no avail.
- Cleaning gold pins on Falcon motherboard hasn't helped.
- I get an 'address fault' crash when I try to run sysinfo on my Falcon in 060 mode, however it runs fine in 030 mode.
- I've also tried my CT60e board in my spare Falcon but I experience exactly the same issue.

Thoughts...?

I'm getting very close to returning my CT60e to Willy. However I'm concerned that on this forum lots of people seem to have a very similar issue. Also very strange that Willy says that no CT60e's have ever been returned to him!?!?
joska
Hardware Guru
Hardware Guru
Posts: 4694
Joined: Tue Oct 30, 2007 2:55 pm
Location: Florø, Norway
Contact:

Re: CT60 memory test strangeness

Post by joska »

Does it work correctly in 060-mode if you skip the memory test when booting? If it does, then it looks like this is "just" one of the CT60e's quirks.
tksm wrote: Wed Jul 08, 2020 12:35 am True problem is CLOCKPATCH
I doubt that. This behaviour is 100% reproducible on my CT60e, and it's definitely triggered by the TOS TT-RAM-test. I suspect it does something that f**ks up some internal CT-TOS settings/variables in RAM.
Jo Even

VanillaMiNT - Firebee - Falcon060 - Milan060 - Falcon040 - MIST - Mega ST - STM - STE - Amiga 600 - Sharp MZ700 - MSX - Amstrad CPC - C64
Rustynutt
Atari Super Hero
Atari Super Hero
Posts: 794
Joined: Wed Mar 21, 2012 7:38 am
Location: Oregon

Re: CT60 memory test strangeness

Post by Rustynutt »

Have yet to test an e extensively other than confirming they (3) work. Remember having issues with HD Driver kicking back an error when accessing NVRAM with a dead battery, but that's after initialization.

On a stock Falcon, living with a dead NVRAM isn't an issue, did notice the e hits it for some reason.

The CT reads TOS from ROM before applying it's vodo, right? As well as peaking at NVRAM for screen, KB, and language.

Copywriter or not, a piss poor way of booting, if so. Should of just stuck a ROM and NVRAM socket onboard and been done with it.

Surely not a clock patch issue.
User avatar
tksm
Retro freak
Retro freak
Posts: 12
Joined: Fri Dec 20, 2019 3:35 am

Re: CT60 memory test strangeness

Post by tksm »

Sometimes power supply is the problem (pico psu) especially if you have cheap one.
I can only suggest that use PC ATX 24 pin power supply for test.
mikro
Hardware Guru
Hardware Guru
Posts: 2200
Joined: Sat Sep 10, 2005 11:11 am
Location: Kosice, Slovakia
Contact:

Re: CT60 memory test strangeness

Post by mikro »

GoldenGrahams wrote: Wed Jul 08, 2020 12:00 pmI'm getting very close to returning my CT60e to Willy.
Because you can't finish a boot time memory test which everybody skips anyway, wow.
However I'm concerned that on this forum lots of people seem to have a very similar issue. Also very strange that Willy says that no CT60e's have ever been returned to him!?!?
Maybe because people enjoy actually running software on it and not watching memory test. ;-)

Most likely it is a CT60TOS issue which hadn't been spotted before (it always fascinates me how all the newbies shout "it's a clockpatch issue!!!" by the first sing of any trouble ;)). I'll try to test it on my CT60/63 just to be sure but it's very unlikely it will behave differently.
User avatar
mpattonm
Hardware Guru
Hardware Guru
Posts: 505
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: CT60 memory test strangeness

Post by mpattonm »

mikro wrote: Thu Jul 09, 2020 6:44 am I'll try to test it on my CT60/63 just to be sure but it's very unlikely it will behave differently.
Indeed, I am sometimes seeing similar issue on original CT60 too. I doubt it is CT60e related.
Now reading this thread I am beginning to think the moments I have seen the issue were those I have left the Falc unattended and it went through uninterrupted boot process.
User avatar
mpattonm
Hardware Guru
Hardware Guru
Posts: 505
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: CT60 memory test strangeness

Post by mpattonm »

tksm wrote: Wed Jul 08, 2020 12:35 am True problem is CLOCKPATCH if you wuld like use ct60 then please remove clock patch! and make sdma mod.
IMO this is a pure nonsense.
tjam
Atari User
Atari User
Posts: 31
Joined: Sun Dec 09, 2018 10:54 pm
Location: Finland

Re: CT60 memory test strangeness

Post by tjam »

mpattonm wrote: Thu Jul 09, 2020 10:46 am
tksm wrote: Wed Jul 08, 2020 12:35 am True problem is CLOCKPATCH if you wuld like use ct60 then please remove clock patch! and make sdma mod.
IMO this is a pure nonsense.
As far as I know SDMA modification is just one of various clock patches (http://powerphenix.com/ct60/english/fitting63.htm). I have SDMA clock patch on both of my ct60 Falcons. CT63 made Falcon really unstable (crashed before booted to desktop), but after SDMA clock patch it has been rock solid.

At least one of my ct60 Falcon had similar problems with memory test, but I just disabled it from NVRAM... it's not really issue and probably not related to clock patch.
sety
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 112
Joined: Sat Nov 04, 2017 12:42 am

Re: CT60 memory test strangeness

Post by sety »

GoldenGrahams wrote: Wed Jul 08, 2020 12:00 pm I'm getting very close to returning my CT60e to Willy.
I doubt it's something that can be fixed in the lab, otherwise he would have just fixed the issue when he designed the ct60e. It's more likely that the problem is buried in abe/sdr or tos4.04 and I'm not sure if anyone has the sources for either, so it's unlikely it'll ever be fixed. Just set your boot delay to zero and forget about it. :)
ATARI FALCON/CT60 @99/25Mhz 512MB - ATARI FALCON/CT60e @77/16Mhz 512MB - 520STF 1MB TF68020 - 520ST unmodified (my very first computer) - A couple of useless XE's - and a really cool Atari desktop calculator :)
Rustynutt
Atari Super Hero
Atari Super Hero
Posts: 794
Joined: Wed Mar 21, 2012 7:38 am
Location: Oregon

Re: CT60 memory test strangeness

Post by Rustynutt »

mpattonm wrote: Thu Jul 09, 2020 10:46 am
tksm wrote: Wed Jul 08, 2020 12:35 am True problem is CLOCKPATCH if you wuld like use ct60 then please remove clock patch! and make sdma mod.
IMO this is a pure nonsense.
There is some merit in this, terminating the clock at the SDMA CLKEN or adding a capacitor tied to ground improves performance with the Mighty Sonic and Afterburner with Nemesis. Improves SCSI performance for sure at 20mhz bus.
Never tried running without an inverter on the main clock though. The FPU is successfully over clocked with it's own oscillator. The Mighty Sonic, Afterburner, and other devices seem unaffected using a direct main clock. It's the SDMA that ties things up in knots.

Years later (I'm so far behind), going to try using an MC88LV926DW to distribute clocks to various IC's. The idea is perhaps centrally locating the IC and provide essentially a zero delay clock source.
That said, it's also possible that isn't what the cause is, and it's a communication fault between COMBEL and SDMA, requiring a delay, not the trace itself. You have to admit, routing of the SDMA clock leaves a lot to be desired.
I'm sure you've ran across a lot of areas to improve with your new layout.

The CT60 has this ability, but I don't know the specs of how the external bus clock is output, Rudolph didn't utilize the MC88LV926DW AIRC.
Think that's way at first he supported removal of the clock patch in his installation guides.
User avatar
mpattonm
Hardware Guru
Hardware Guru
Posts: 505
Joined: Mon Oct 21, 2002 8:52 am
Location: Czech republic
Contact:

Re: CT60 memory test strangeness

Post by mpattonm »

Rustynutt wrote: Sat Jul 11, 2020 2:32 pm The CT60 has this ability, but I don't know the specs of how the external bus clock is output, Rudolph didn't utilize the MC88LV926DW AIRC.
Think that's way at first he supported removal of the clock patch in his installation guides.
Have a look at the CT60 schematic (page 3): http://powerphenix.com/ct60/CT63_schematics.pdf
Rather than simply patching system clock, he generates clock signals from an onboard 50MHz crystal and injects them onto Falcon mainboard.
Ergo yes, if you use/plan to use CT60 as Falc mainboard clock source (aka "fitting with solders"), any other existing clock patch becomes redundant and has to be removed.
But in the other case, removing properly working clock patch harms system stability and I call it a nonsense.
stormy
Atari Super Hero
Atari Super Hero
Posts: 757
Joined: Tue Jan 26, 2016 12:39 pm

Re: CT60 memory test strangeness

Post by stormy »

Here's an analogy to this situation:
Man buys a turbo charger for a ford Falcon.
Man hasn't serviced the car since 1990.
Car blows up.

You need to service old computers, recap their PSU, motherboard, replace nvram batteries etc etc. If your Falcon is in the same state it came from the factory 30 years ago, the last thing to blame would be the modern modification.

Also why are you using an old boot ROM? Here is v1.05 viewtopic.php?f=97&t=35594
Falcon CT60e 060 - 256mb ram - Phantom bus and DSP accelerated // Atari TT - Thunder and Storm IDE 64mb ram - Lightning VME - USB LAN - ATI Mach64 2mb
Rustynutt
Atari Super Hero
Atari Super Hero
Posts: 794
Joined: Wed Mar 21, 2012 7:38 am
Location: Oregon

Re: CT60 memory test strangeness

Post by Rustynutt »

mpattonm wrote: Sat Jul 11, 2020 4:15 pm
Rustynutt wrote: Sat Jul 11, 2020 2:32 pm The CT60 has this ability, but I don't know the specs of how the external bus clock is output, Rudolph didn't utilize the MC88LV926DW AIRC.
Think that's way at first he supported removal of the clock patch in his installation guides.
Have a look at the CT60 schematic (page 3): http://powerphenix.com/ct60/CT63_schematics.pdf
Rather than simply patching system clock, he generates clock signals from an onboard 50MHz crystal and injects them onto Falcon mainboard.
Ergo yes, if you use/plan to use CT60 as Falc mainboard clock source (aka "fitting with solders"), any other existing clock patch becomes redundant and has to be removed.
But in the other case, removing properly working clock patch harms system stability and I call it a nonsense.
I don't disagree with you. Was pointing out the cap/resistor mod, one not covered in some of the patch data sheets.

I have a CT60, CT63, and 3 CT60e, the latter not including circuitry to use it as a clock source. For my interest, the MC88LV926DW is an easier solution as the mod also works with my AB040.
Totally agree, if a Falcon is working fine, drop the CT60 in and run it. If issues arrive, it's highly unlikely it's due to a working clock patch.
Post Reply

Return to “CT60 / CT63 Area”