4MB Upgrade & 16MHz Booster progress

[exxos]

Post delivered on a Sunday

Mark

, nope, just took me a couple days to remember to post here about it

[exxos]

booster-dev.jpg

Been really busy these past days, but finally got the dev-system built. Those who brought the dev-pcb, they have also been shipped.

This board has everything in sockets. So easy to swap out/change pull ups. Also every IO pin (other than address & databus form CPU) has a 100R resistor to reduce the nasty ringing on signals. They also in sockets (DIL ones) so they can be changed easily or swapped out for other values.

4 layer PCB, middle 2 layers are VCC and GND. It inherently has about 915pF of capacitance, so this will help with HF bypassing across the board. A couple of bulk capacitors and a lot of de-coupling caps all over the place. May be overkill, but I'm seriously not wanting to end up buggering about on this board.

4096 ROM like on V2, TOS104 & TOS206 on board. 68HC000 Freescale CPU & ATF1508 PLD (1504 will probably be enough though for now) Jtag programmable so I don't have to keep faffing about removing chips in and out the board every 5 seconds to reprogram them. The ATF has awesome amounts of IO capability and speed. So no more "it won't fit in the 22V10" type annoyances.

Across the front is access to every IO line on the board for easy connection to Logic Analyser. I got tired of spending time bodging sockets on sockets to have the GAL's removable and LA pins on top, So this is all as simple as it gets now. They are all labeled up as to what signal is on what pin. Another annoyance of keep having to refer to the schematic and working out what signals are on what pins. So all these improvements mean we mean business to making progress this time

The unused header at the top of the board is for the expansion bus, that will come later and will offer things like fast-ram, which will run at whatever speed the CPU is running at (my aim is 32mhz but hope to push higher eventually). So far I have been struggling to get past 16mhz, but 32mhz I really think is solvable in due time. At least on this board I don't have to worry about signal noise or regulation issues on the supply rails. Signal noise is mostly what I seem to be spending time on fighting on past boards rather than spending the time actually developing the code with Rodolphe.

Hopefully will have some more news in the not to distant future. Updates may be slow, but rest-assured these boosters are still being worked on

[Sheldon]

Hello,

Very nice new and job.

Regards

[DarkLord]

Looks great!

[troed]

Arne tried replacing that 2mhz line with 4mhz, and it caused the same problem as to when the DE mod was not done at all.

The DE mod should work. [snip]

I was searching for something completely different and stumbled upon these newsgroup posts from 2001. Posting here since they indicate another way of solving video whilst raising the speed of the whole system:

>> * A potentially most beneficial approach would be the one which involves
>> raising the clock frequency of the entire system to 10MHz or 12MHz.
>>[...]

>> noone can quote -exactly- the modification needed for video circuits.
>> I have a general
>> idea of how it might work, but a tested schematics would be so much
>> easier- so if you find one let me know
>>
>
> Hmmm... I think that such projects are very hard for realize, even if you
>find all ST magazines involved.
Not necessarily. I played a little with that project and for a start raised
the system clock to 10MHz (40MHz oscillator). By declaring external
syncs to TOS and by supplying the correct (or I thought so) sync pulses
obtained by dividing the 500kHz signal, I obtained a standard-resolution
picture on a SM-124. From that to an extended resolution it should not
be very difficult to go. (My attempt was not completely successful as the
picture trebmbled slightly (about 10 pixels) left-right, so I did not get
everything exactly right- or else at higher CPU speed the picture
started too early after a hsync pulse nd the monitor maybe couldn't
handle it). It was all done with several 74F and 74HCT chips.

I didn't have time to finish it, but I prepared some connections on
the motherboard to make everything more reliable, and will try again
with a modified circuit. My problem is finding reliable information
on pixel clock rates and line width- somehow the numbers I have do not
add-up correctly to 35.7Khz .

https://groups.google.com/d/msg/comp.sy ... teg3QybIwJ

[exxos]

Thanks, I had seen something about 8mhz to 10mhz, though the overclock was a double clock of the entire system, IE each speed in gembench was 200%, though you need faster DRAM of course. But the problems were all basically solved with the video. The main problem was it wasn't stable as the MMU generates the clocks, and 4mhz to 8mhz, the MMU couldn't deliver a clean enough clock. Somewhere I was talking about replacing all the clocks to solve it, but this "upgrade" I am not even going to start until all the other problems are fixed with the boosters. Once everything is running at 32mhz or whatever, then the ST ram mods will be done, and the whole system should double in speed again, as everyone knows, ST-RAM is the main bottleneck. Somewhere there are images I posted with 200% listed, I think Arne might have posted some also.

[arf]

Possible, though I just assume its a RAM bottleneck slowing things down. I guess if GEM apps were run in Alt-ram and the CPU could access alt-ram at 16mhz, as well as TOS at 16mhz, then I would assume it would give a huge boost in GEM app speeds. I think AlanH was doing some Alt-ram boards a while back, but I will have to ask him if they are likely to be capable of running at 16mhz.

Any news here? Would be great to see a 16 MHz board plus a MonSTer board, where the MonSTer runs with 16 MHz - RAM access, but also IDE would greatly benefit, I guess.

[exxos]

Possible, though I just assume its a RAM bottleneck slowing things down. I guess if GEM apps were run in Alt-ram and the CPU could access alt-ram at 16mhz, as well as TOS at 16mhz, then I would assume it would give a huge boost in GEM app speeds. I think AlanH was doing some Alt-ram boards a while back, but I will have to ask him if they are likely to be capable of running at 16mhz.

Any news here? Would be great to see a 16 MHz board plus a MonSTer board, where the MonSTer runs with 16 MHz - RAM access, but also IDE would greatly benefit, I guess.

I already sell basic 16mhz mods in my store which support 16mhz ROM , I don't know about 16mhz support with other projects, your have to check with the people who make the stuff. I do plan on fast-ram boards, but that is a long way off as need to get the booster running at higher speeds like 32mhz first, then it would be 32mhz fast-ram.

http://www.atari-forum.com/viewtopic.php?f=15&t=28059

Fast-ram would be a board which would fit the expansion header on the current dev-system and future V3 boosters. Nobody seemed interesting in developing stuff for it anyway. Rodolphe designed a fast-ram board for the V3 a long time ago, but we are pushing for 32mhz speeds so anything else will just have to wait.

[exxos]

Not really a update as such. Though I hope to get working on this again in about a weeks time. Rodolphe has been working on the 020 and making good progress there, though we are both at the same point that we need to solve the stability issues at 32mhz. We hope to push the CPU to the limits and 50mhz is our target speed. ROM will also be accessed at the higher speeds.

As mentioned before, the V3 has a expansion bus so this paves the way to true fast-ram at whatever the CPU speed is (32mhz+). Rodolphe already design a ram card last year so its just the 32mhz CPU which is holding things up. Rodolphe has already drafted a lot of the code so we just need to start testing things out. We both have been super busy these past few months, so the booster work had to be put on hold until recently.

One of the main hold ups is noise on the motherboard signals. This drove me nuts for some months, but the dev-system as posted previously, has resistors on all main IO pins with the option to add/remove/change them. So we can start with a stable platform this time which caused 90% of headaches over the past year.

I still have a couple of the dev-system PCB's if anyone wants to look towards beta testing code or thinking towards developing addons for the V3 series.

As a side note, I plan to discontinue the V1 series booster in favour for the V1.5 series in the new year. The last batch of V1's will be made over the next couple of weeks.

[arf]

Thanks, I had seen something about 8mhz to 10mhz, though the overclock was a double clock of the entire system

Do you mean that: http://www.stcarchiv.de/stc1992/09_gehtdoch.php -- a speedup of the whole ST bus system, overclocking it from 8 to 10 or even 12 MHz.

I know a few users running such a system since years.

[exxos]

Thanks, I had seen something about 8mhz to 10mhz, though the overclock was a double clock of the entire system

Do you mean that: http://www.stcarchiv.de/stc1992/09_gehtdoch.php -- a speedup of the whole ST bus system, overclocking it from 8 to 10 or even 12 MHz.

I know a few users running such a system since years.

Nope, double overclock of the MMU and ST-ram. Arne pointed me towards the mod many posts back, but I have not done much with it yet as just not got the time. It works, but needs a re-do of the ST clock's as its not stable. http://atari-forum.com/viewtopic.php?f= ... 00#p265199

This gives 200% speed boost as the CPU can RAM can run at double speed. Now imagine that with a 32mhz overclock on the CPU and ROM and probably get over 300% speeds. ST-ram is always the bottle neck , so whatever speed we get out of the CPU (hopefully 32mhz-50mhz) then whatever those results are will likely double once ST-ram is running at double speed. This is why I got "distracted" into my 4MB "plug in" upgrade http://www.exxoshost.co.uk/atari/last/4MBMMU/index.htm as a faster ST-ram chip was needed. But this ST-Ram mode will not work on the STE sadly. Ultimately the STFM will be better for overclocking, its once reason I have not started work on any STE boosters yet. While the STE is a "better" machine, its worse for modding.

The V3 series has a huge amount of IO ports, so once the V3 is working how we want it, the next step is st-ram boost. So you can probably see the "big picture" of how this is all going to fit together now.....

[exxos]

DEV-SYSTEM STARTS HERE

http://www.exxoshost.co.uk/atari/last/1 ... ex.htm#dev

Myself and Rodolphe have made a start with the new dev-system board. I'm not going to post every little update on here as its getting time consuming. So most updates and progress will be posted over on my site. Of course this work parallels with the 020 work with Rodolphe as a lot of work is common to both projects. So its worth keeping a eye on on both pages for progress.

[Atari74user]

Exxos: As a side note, I plan to discontinue the V1 series booster in favour for the V1.5 series in the new year. The last batch of V1's will be made over the next couple of weeks

Just to check, the V1 is not switchable is that correct?

If that is correct, the V1.5 may be suitable to me, however again the Fast ROM is an option, and not mandatory?

The only reason I ask these questions is I actually don't want to tamper with my TOS 2.06 switcher already installed in my STFM, everything is stable and backwards compatible. However a 16MHz V1.5 switcher would be a welcomed addition, along with a Blitter. For my set up it would seem a pretty stable combination, as everything could be switched back to original TOS running at 8MHz without the Blitter on if needed.

[exxos]

Just to check, the V1 is not switchable is that correct?

If that is correct, the V1.5 may be suitable to me, however again the Fast ROM is an option, and not mandatory?

The only reason I ask these questions is I actually don't want to tamper with my TOS 2.06 switcher already installed in my STFM, everything is stable and backwards compatible. However a 16MHz V1.5 switcher would be a welcomed addition, along with a Blitter. For my set up it would seem a pretty stable combination, as everything could be switched back to original TOS running at 8MHz without the Blitter on if needed.

The V1 is a very simple booster, while it can be switched with some work, it was never really designed to be switched and work with any other mods, including the blitter. There are workarounds, but V1.5 is better in general now. It has a lot of fixes which V1 does not.

You don't have to use fast-roms no. You would just tie the booster CE line to 5V and thats it. Though if you have 100ns or faster ROM's then it would probably work in fast-rom mode anyway. Theres no problems in using or not using the blitter on V1.5 either.

[Atari74user]

That's cool, thank you very much for the response! I am sold on the idea.

[exxos]

FINALLY!! 32mhz CPU, 32mhz TOS (With waitstates ) and working with the blitter. See my site for some more info on the past couple days progress.

32mhz_blitter.png

http://www.exxoshost.co.uk/atari/last/1 ... m#DEC92015

[wongck]

Whoa!!!
Fantastic. This will be like a SST.

[exxos]

Whoa!!!
Fantastic. This will be like a SST.

Now its over to Rodolphe to finish the 020 side of things 020 is a faster CPU and has instruction cache so will be awesome to see what 32mhz 020 will be like

I think a little more can be pushed out of TOS speeds, but need to build a variable delay for that, but will see how things ago

[exxos]

Now I am pleased to say I have pushed the ROM's as fast as possible.. and here below are the results with 68ns wait state and 36ns wait state. If there was such things as 10ns ROM's then we could push the speed fully to 32mhz, but 45-ns is hard to find. 55ns seems is easier but still hard to find.

1WSvs0WS.jpg

[exxos]

I'm pleased to say I have smashed the ROM speed limit from before and now running 32MHz CPU and 32MHz ROM

32mhzROM.jpg

[DarkLord]

Those are great results Chris. Congratulations on your success(s) so far.

I'm curious to see when you get the 68020 working how close (or better,
considering the ROM speedup), it will compare to something like the Pak
68/3 boards.

Thanks for all the hard work!

[exxos]

Those are great results Chris. Congratulations on your success(s) so far.

I'm curious to see when you get the 68020 working how close (or better,
considering the ROM speedup), it will compare to something like the Pak
68/3 boards.

Thanks for all the hard work!

I can't remember if you have a pak3 ? If so, maybe you could test GB5 so have its benchmark..

[DarkLord]

Yes, I've got the Pak 68/3 board with a 40mhz 68030 and 40mhz 68882 in my STacy.

It's running TOS v3.06.

Umm, remind where I can download GB5 and I'll give it a shot.

[exxos]

Yes, I've got the Pak 68/3 board with a 40mhz 68030 and 40mhz 68882 in my STacy.

It's running TOS v3.06.

Umm, remind where I can download GB5 and I'll give it a shot.

here http://atari-forum.com/viewtopic.php?f= ... +5#p275140

[DarkLord]

Got it. Have to get some sleep now - just finished a close to 60 hour work week (12.5 hrs last night)

I'll try it later tonight though and see what kind of results we get. Should be interesting (fun!).