Falcon Clock Patch is mandatory - Read why -

[Dlfrsilver]

Hello, to all the Falcon users who have not yet done the clock patch :

It appears that it has never been reported yet that before the clock patch appliance (exxos one or any other), you get data corruption when copying archives ZIP for example.

I remember pulling my hair off on trying to install Racer 2, and always get archive decrunch error, and also on other softwares.

Now, with the clock patch, i can process big files with no problem on my hard drive (Compact Flash) !

It looks like the Falcon has the data corruption like the first series of STE had, but here triggered by the "clock bug".

So, make the fix, it will ease your life !

[frank.lukas]

You can't generalize like that. My falcon, for example, ran wonderfully without any problems without a clock patch. Only when i used the DFB1 Booster i needed the clock patch.

[Dlfrsilver]

You can't generalize like that. My falcon, for example, ran wonderfully without any problems without a clock patch. Only when i used the DFB1 Booster i needed the clock patch.

Any demo using the DSP was crashing on mine, and i had repeated and serial corruption each time i copied, moved, transfered archive ZIP.

With the clock patch, the stability is now there, and the corruption problems are gone, too happy

[exxos]

My original falcon didn't have a clock patch. But when I was using the CT60 it was never stable. I only tried DSP stuff with STOS extensions, so any crashes or issues were likely blamed on software anyway. I sent my falcon to czuba and he did the clock patch etc and my machine came back more disfunctional and didn't boot up anymore. In the end I started looking into the clock patch issues myself. Also note my clock patch isn't the same as all the other generic clock patches. While the original versions can solve a lot of problems, mine is the only new generation of patch in existence.

The falcon clock patch has nothing to do with DMA issues on older machines like the ST/E. That was unstable bus issues rather than clock issues.

[Dlfrsilver]

My original falcon didn't have a clock patch. But when I was using the CT60 it was never stable. I only tried DSP stuff with STOS extensions, so any crashes or issues were likely blamed on software anyway. I sent my falcon to czuba and he did the clock patch etc and my machine came back more disfunctional and didn't boot up anymore. In the end I started looking into the clock patch issues myself. Also note my clock patch isn't the same as all the other generic clock patches. While the original versions can solve a lot of problems, mine is the only new generation of patch in existence.

The falcon clock patch has nothing to do with DMA issues on older machines like the ST/E. That was unstable bus issues rather than clock issues.

Ok, but then how to explain that before adding it, i had always problems of corruption on files moving, writing and decrunching, and once applied, everything is solved, no more file corruption ?

[exxos]

Ok, but then how to explain that before adding it, i had always problems of corruption on files moving, writing and decrunching, and once applied, everything is solved, no more file corruption ?

I'm afraid you lost me. The patch is likely needed on all Falcons, and indeed will solve a lot of such problems for people. That is why the patch exists.

I was just staying my original falcon didn't have a clock patch and it ran, but I cannot say it ran perfectly or correctly. Likely it didn't as I was unaware of hardware issues in late 1990s. Frank said his falcon ran without a patch, but as I haven't seen that machine I can only take Frank's word it ran correctly. I don't know what tests he has done or what software he has tried etc. Some machines might run without a patch. But I think it would be rare if a falcon ran correctly without a patch as so many falcons had a clock patch added and a lot of them still had issues.

As falcons are rare, it is impossible to say exactly what problems in general falcons have or don't have. It's not like on the ST where I can try 100 machines to get a "feel" of what general problems are and how to fix them. Some ST's can run fine, but do they really ? This is why some people are against my fixes because they either don't have issues, or simply not aware of them. It's just dumb luck, or ignorance. With millions of ST's out there and with faults being random, anything is possible.

Every machine is different and has its own quirks. This is why it's difficult to do fixes which work on every machine. My clock patch will solve all known issues I am aware of. IMHO even if machines does seem to work correctly, the fixes should still be done. It may work on a cold day, but not on a warm day for example. The list of problems / faults is endless.

[Dlfrsilver]

Ok, but then how to explain that before adding it, i had always problems of corruption on files moving, writing and decrunching, and once applied, everything is solved, no more file corruption ?

I'm afraid you lost me. The patch is likely needed on all Falcons, and indeed will solve a lot of such problems for people. That is why the patch exists.

I was just staying my original falcon didn't have a clock patch and it ran, but I cannot say it ran perfectly or correctly. Likely it didn't as I was unaware of hardware issues in late 1990s. Frank said his falcon ran without a patch, but as I haven't seen that machine I can only take Frank's word it ran correctly. I don't know what tests he has done or what software he has tried etc. Some machines might run without a patch. But I think it would be rare if a falcon ran correctly without a patch as so many falcons had a clock patch added and a lot of them still had issues.

As falcons are rare, it is impossible to say exactly what problems in general falcons have or don't have. It's not like on the ST where I can try 100 machines to get a "feel" of what general problems are and how to fix them. Some ST's can run fine, but do they really ? This is why some people are against my fixes because they either don't have issues, or simply not aware of them. It's just dumb luck, or ignorance. With millions of ST's out there and with faults being random, anything is possible.

Every machine is different and has its own quirks. This is why it's difficult to do fixes which work on every machine. My clock patch will solve all known issues I am aware of. IMHO even if machines does seem to work correctly, the fixes should still be done. It may work on a cold day, but not on a warm day for example. The list of problems / faults is endless.

In fact, the 13.000 Falcon 030 sold by Atari were prototypes machines, unfinished and rushed machines sold as it is in shops.

My Falcon ran before the patch, but limitedly. Daniel Hedberg demo's Joydemo was crashing each time i launched it. He also advised me to add the clock patch to stabilize the machine.

About the ST, i have cornered the problem about hard drive corruption. The STE released in 1989 and 1990 are beta machines with beta motherboards, and STE made in 1991 have the final version of the motherboard and have no problems with hard drives.

One of my friends (a musician) have one of those STE, and the ultrasatan works straight on his machine. Mine were made in 1989 and 1990, and they need the 68HC000 because they corrupt the hard drives systematically.

But all in all, i do agree that it's very complicated to make fixes that work in a uniform manner on a specific ST, due to the numerous MB revisions existing out there.

About using my Falcon 030 after applying your clock patch, it's fully stable by cold and hot weather. I have the luck these lack days to make it works with temperature as low as 15°C (day temp!) and around 32-35°C during the afternoon (i got very hot weather here lately as well).

I can confirm for my own case that after a full recap, adding a compact flash, 14mb of ram, adding your clock fix, my machine works exactly as intended with no spurious crashes or freeze.

[Dlfrsilver]

Interesting informations found here on https://mikrosk.github.io/clockpatch/ :

How do I know I need it?

There are two or three typical symptoms (as mentioned, typically surfacing in higher resolutions):

- SCSI read/write errors, corrupted data
- crackles during sound playback (corrupted data between SDMA and DSP and/or DAC)
- Floppy disk read/write errors (especially in Falcons with accelerated bus)

As you can see, the worst possible scenario is recording audio, using DMA run it trough the DSP and write output to a SCSI disk in 720x512x16bit resolution.

Atari used this tool to verify fixed Falcons – if you see green bars from left to right, you’re good. Red bars mean trouble.

[Dlfrsilver]

And the last interesting part :

"How do I know it worked?

Well, the first good sign is that the Falcon still boots. Then you can do various tests:

- Copy a big zip/lzh/arj archive to your floppy/SCSI hard disk and run the unarchiver’s “verify” option on it
- Copy a lot of small files into a directory on your floppy/IDE/SCSI disk and then compare the directory trees with TREE_CHK from Kobold

set a very demanding resolution, like 864x640/256 and try direct-to-disk recording with a huge sample to SCSI hard disk – you should hear it on playback without any crackles"

[Rustynutt]

You can't generalize like that. My falcon, for example, ran wonderfully without any problems without a clock patch. Only when i used the DFB1 Booster i needed the clock patch.

But was that mostly using the Falcon with NOVA ATI?
There the bus is relieved somewhat. No VIDEL processing to slow everything down.