AtoMMC2 firmware 2.5

discussion of games, software, hardware & emulators relating to the Acorn Atom and Acorn System machines.
Post Reply
User avatar
sirmorris
Posts: 752
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

AtoMMC2 firmware 2.5

Post by sirmorris » Thu Sep 08, 2011 2:45 pm

Hi all,

V2.5 of the firmware here.

**EDIT** V2.6 is available further down.


This release finally brings subdirectory support.

There is a version of the ROM compiled to sit at E000 and if you have a ram/rom board this may be an option for you. You'll need a lightly-patched OS rom image too (supplied) which calls into the driver to auto-initialise it at startup.

If using the E000 version please, before switching, ensure that interrupts are disabled by entering:

*CFG A0


regards

Charlie
Attachments
release-2.5.zip
(52.42 KiB) Downloaded 168 times
Last edited by sirmorris on Tue Oct 04, 2011 1:44 pm, edited 1 time in total.

Lusciousthelock
Posts: 64
Joined: Tue Sep 06, 2011 9:12 am
Contact:

Re: AtoMMC2 firmware 2.5

Post by Lusciousthelock » Fri Sep 09, 2011 8:50 am

Hi sirmorris,

Downloaded this file this morning and checked my MMC's firmware which is 1.4.1 so guessing I've missed some updates. My manual says about copying 3 firmware files over, but seeing as their is quite a few files here... Do I just copy everything onto the card regarding the A000 version, enable bootloader and same as usual or do I need to do a bootloader update first etc?

Looking forward to subdirectory's :)

User avatar
sirmorris
Posts: 752
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

Re: AtoMMC2 firmware 2.5

Post by sirmorris » Fri Sep 09, 2011 9:14 am

The atom-firmware folder contains files to be blown into EPROM or ram/rom board's EEPROM.

if you're using the EOOO based firmware you need to also use the modded firmware, else you'll have to manually initialise the driver every time you press break.

Ignore the bootloader/bootupdater folder, this is always built but hasn't changed since 1.0.

The picfirmware folder contains files that the bootloader will pick up from the root folder on a freshly formatted sd card. There are a couple of different chips in the wild and each chip requires a slightly different firmware. The bootloader knows which to use so you don't have to. It's almost certain most people will need the '20' file. There used to be 3 files here, but from now there will only be 2.

The PIC firmware in this release should only be used with the supplied Atom firmware - the addition of subdirectories has introduced a dependency so please don't go using one without the other :)

Hope this clears things up. If not, don't hesitate to ask!

C

User avatar
BeebMaster
Posts: 2632
Joined: Sun Aug 02, 2009 4:59 pm
Location: Lost in the BeebVault!
Contact:

Re: AtoMMC2 firmware 2.5

Post by BeebMaster » Fri Sep 09, 2011 12:40 pm

Is this a new firmware upgrade for the MMC thing as installed in my Atom at Bletchley? Or summat else?

There was a flash card thing going on at the Lass the other week, but I think that was a different system.
Image

PhilYoung
Posts: 195
Joined: Sun Jun 12, 2011 4:55 pm
Contact:

Re: AtoMMC2 firmware 2.5

Post by PhilYoung » Tue Sep 13, 2011 5:21 pm

I've just upgraded to this release, it seems to have gone okay but a couple of questions:

<REPT> <BREAK> confirms that the Atom Firmware (A000 version) is v 2.5 A

However *HELP reports that the interface firmware version (and the bootloader) are version 2.2. Is this right, I would have though they'd be v2.5 as well ?

I think the firmware went in okay, I blew the EPROM first (with atommc2-2.5-a000.rom) then put the two FW files (atommc20.bin and atommc25.bin) onto a freshly formatted MMC card and powered up. Red 'error' led was on solid, green 'busy' led flickering until the leds went out, re-power and that was it.

Secondly, what sub-directory support has been added. The old version (which was 2.2) has sub-directory support to some extent with *CWD, but I can't see anything along the lines of *MKDIR or whatever looking through the Atom firmware.

Any thoughts ?

(edit) also *CFG stayed at it's default 'FF' if that make any difference with the A000 version

Cheers,

Phil Young

User avatar
sirmorris
Posts: 752
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

Re: AtoMMC2 firmware 2.5

Post by sirmorris » Fri Sep 16, 2011 8:57 am

Hiya!

I bumped the Atom firmware revision simply because I felt 2.2 was jinxed. Well, something like that ;)

The interface firmware didn't need to change and has remained at 2.2.

Subdirectory support is limited to using them for storage. All directory manipulation needs to be done on a desktop computer. There simply isn't enough program space in the interface uC to implement the create/rename directory functions :cry:

The CFG byte's value remains at 255. It's onlu really necessary to tweak it if you use the E000 firmware. In that case the interface's generation interrupts should be disabled.

I may be repeating myself here but I really need to know about any issues that people find with the firmware. Every Atom is unique and presents its own challenges ;¬) My pair are certainly challenging and until I get my Issue 5 built then I have to assume the random lockups, erratic starts, and other infuriatingly sporadic failures are part of this. Everything checks out in MESS, but I've learned never tot rust an emulator too much ;¬)

C

PhilYoung
Posts: 195
Joined: Sun Jun 12, 2011 4:55 pm
Contact:

Re: AtoMMC2 firmware 2.5

Post by PhilYoung » Sat Sep 17, 2011 9:27 am

sirmorris wrote:Hiya!

I bumped the Atom firmware revision simply because I felt 2.2 was jinxed. Well, something like that ;)

The interface firmware didn't need to change and has remained at 2.2.
Thanks for clearing that up.

Subdirectory support is limited to using them for storage. All directory manipulation needs to be done on a desktop computer. There simply isn't enough program space in the interface uC to implement the create/rename directory functions :cry:
Understood, that seems to work okay as well, with *PWD to move between previously created directories



The CFG byte's value remains at 255. It's onlu really necessary to tweak it if you use the E000 firmware. In that case the interface's generation interrupts should be disabled.
I've not got the means to try the E000 version, unfortunately, but thanks for providing it anyway. I suppose that this is the more logical place for a filing system replacing the DOS.


I may be repeating myself here but I really need to know about any issues that people find with the firmware. Every Atom is unique and presents its own challenges ;¬) My pair are certainly challenging and until I get my Issue 5 built then I have to assume the random lockups, erratic starts, and other infuriatingly sporadic failures are part of this. Everything checks out in MESS, but I've learned never tot rust an emulator too much ;¬)

C
As I mentioned by email, the only problem is that on 'BREAK' it only starts the MMC about 50% of the time. I've no idea why, it seems to come in short runs of working correctly and then not working. I don't think there are any variables to experiment with, I thought at one point it might have been down to either giving the BREAK key a quick stab or a longer press, but I can't get it to consistently work or fail. Lately I've suspected that it's not properly detecting the MMC card maybe.

It always works correctly on power up though, so it's probably down to Atom flakiness as you say.

By the way, I managed to fit the PIC board under the Atom board (so on the same side as PL8), it just fits if you take the top clamp part of the ribbon cable connectors off.

Cheers,

Phil Young

User avatar
sirmorris
Posts: 752
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

Re: AtoMMC2 firmware 2.5

Post by sirmorris » Tue Oct 04, 2011 1:43 pm

Dut to the excercising of a previously dark corner of my code by Kees' excellent JetSetMiner I have found and fixed a bug related to OSLOAD and OSSAVE.

If you call these functions they will likely fail unless the filename is stored at #140. This is normally the case when issuing commands from the keyboard or program, but if you call the vector from assembler then the wrong filename will probaly be used.
Attachments
atommc2-2.6.zip
(21.46 KiB) Downloaded 153 times

User avatar
oss003
Posts: 2797
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: AtoMMC2 firmware 2.5

Post by oss003 » Fri Nov 11, 2011 10:18 am

While refurbishing an Atom, I encountered a nasty problem causing the AtoMMC interface not to work. :(

The Atom worked fine but the AtoMMC interface could sometimes not read the SD-card and was not very stable.

I found out that it's very important for the voltage level not to exceed the 5,3V.
After adjusting the level to 5V it worked ok. :wink:

Greetings
Kees

User avatar
sirmorris
Posts: 752
Joined: Wed Feb 11, 2009 12:18 pm
Location: oxfordshire uk
Contact:

Re: AtoMMC2 firmware 2.5

Post by sirmorris » Fri Nov 11, 2011 10:25 am

This will be because of my use of a trivial voltage divider.

Options going forward are to use a converter chip (which I've tried before but shy from because of the cost/implementational fuss) or something I just saw on a new design ... zener diode clamps. Evidence is that they appear to work - could anyone with the hardware chops point out any flaws in this idea?

Ta

C

Prime
Posts: 2597
Joined: Sun May 31, 2009 11:52 pm
Contact:

Re: AtoMMC2 firmware 2.5

Post by Prime » Fri Nov 11, 2011 2:43 pm

sirmorris wrote:This will be because of my use of a trivial voltage divider.

Options going forward are to use a converter chip (which I've tried before but shy from because of the cost/implementational fuss) or something I just saw on a new design ... zener diode clamps. Evidence is that they appear to work - could anyone with the hardware chops point out any flaws in this idea?
Do you have a link to that design ?

I see no reason why Zenner clams would not work if an apropreate voltage Zenner where chosen, cirtainly at the very least it would help protect the card from voltages outside the maximum ratings, which I'm told can make the magic smoke escape..... Hopefully this didn't happen to Kees' card :)

An other approach might be to make sure that the whole of the AtoMMC only has 5V as an input, as I'm sure the PIC won't like it if the voltage goes too high, so maybe a single 5V Zenner accros the 5V rail which should clamp the output voltage of the PIC to 5V. Would prolly need a higher rated Zenner tho, still should be possible to estimate the total draw of the interface, I would doubt it's anywhere over 100ma.....

Cheers.

Phill.

User avatar
oss003
Posts: 2797
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: AtoMMC2 firmware 2.5

Post by oss003 » Fri Nov 11, 2011 3:22 pm

Prime wrote:Hopefully this didn't happen to Kees' card :)
:lol: No, the only thing what's smoking is my soldering iron .. :lol:

Before adjusting my power supply, I measured 5,6V on the AtoMMC interface.
The interface is working ok now without damage.

Greetings
Kees

User avatar
oss003
Posts: 2797
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: AtoMMC2 firmware 2.5

Post by oss003 » Sun Nov 20, 2011 3:35 pm

sirmorris wrote:If using the E000 version please, before switching, ensure that interrupts are disabled by entering:

*CFG A0
Charlie,

For CFG:
bit 7 is boot-loader enable
bit 6 is shift-boot sense (1 = shift must be held to boot, 0 = auto boot unless shift is held)
bit 5 is irq enable

To disable the interrupt, bit 5 has to be low so you have to type *CFG C0, right??

Kees

User avatar
1024MAK
Posts: 8095
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: AtoMMC2 firmware 2.5

Post by 1024MAK » Sun Nov 27, 2011 8:47 pm

sirmorris wrote:This will be because of my use of a trivial voltage divider.

Options going forward are to use a converter chip (which I've tried before but shy from because of the cost/implementational fuss) or something I just saw on a new design ... zener diode clamps. Evidence is that they appear to work - could anyone with the hardware chops point out any flaws in this idea?
I don't have an Atom so don't know much about the AtoMMC2.
Are you talking about using zener diodes to protect signal lines or to stabilise power supply lines?
Mark

User avatar
1024MAK
Posts: 8095
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: AtoMMC2 firmware 2.5

Post by 1024MAK » Sun Nov 27, 2011 8:57 pm

Prime wrote:An other approach might be to make sure that the whole of the AtoMMC only has 5V as an input, as I'm sure the PIC won't like it if the voltage goes too high, so maybe a single 5V Zenner accros the 5V rail which should clamp the output voltage of the PIC to 5V. Would prolly need a higher rated Zenner tho, still should be possible to estimate the total draw of the interface, I would doubt it's anywhere over 100ma.....
A current limiting resistor in series ahead of the zener diode would be needed (to save the zener from going short circuit and producing magic smoke).
Does anyone know the value of the current drawn by this the AtoMMC2 interface?
Zener_diode_5.1V_regulator.gif
Zener_diode_5.1V_regulator.gif (2.21 KiB) Viewed 1323 times
Mark

Edited to err, make the question in the last sentence clearer :oops:
Last edited by 1024MAK on Sun Nov 27, 2011 9:27 pm, edited 2 times in total.

User avatar
retroclinic
Posts: 3022
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: AtoMMC2 firmware 2.5

Post by retroclinic » Sun Nov 27, 2011 9:10 pm

With what load? With no load the current drawn would be virtually zero once the caps charged, as the zener would be under it's conduction voltage.

You'd need to measure the current drawn by any load placed on the output. You can then work out what the voltage drop across the resistor would be, and what power it would dissapate.

Ideally, you'd replace the 100R resistor with a fuse, that would then allow you to draw as much current as the fuse allowed, and if the voltage goes over 5.1v, then the fuse blows (or trips if using a resettable one).

EDIT: SirMorris, are you talking about the voltage divders for the 3.3v? You can use a zener clamp on the supply to those, or one of the 74LVT range of buffer chips that have dual voltage rails.

Mark.
Image

User avatar
1024MAK
Posts: 8095
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: AtoMMC2 firmware 2.5

Post by 1024MAK » Sun Nov 27, 2011 9:24 pm

retroclinic wrote:With what load? With no load the current drawn would be virtually zero once the caps charged, as the zener would be under it's conduction voltage.

You'd need to measure the current drawn by any load placed on the output. You can then work out what the voltage drop across the resistor would be, and what power it would dissapate.

Ideally, you'd replace the 100R resistor with a fuse, that would then allow you to draw as much current as the fuse allowed, and if the voltage goes over 5.1v, then the fuse blows (or trips if using a resettable one).

Mark.
Err, maybe I could should have worded my post a little better :oops:
The interface circuit that I meant is the AtoMMC2.
I am well aware that the circuit in the diagram I posted will not actually regulate if the input is at or below 5.1V. I posted it to show how a zener diode could protect a sensitive circuit. The value of R1 (at 100R) is just an example, as until the nominal load current is known (i.e. that of the AtoMMC2) it is not possible to work out a suitable value.
Also it is wise to understand that low (current) value fuses often are about 10 to 100 ohms when measured.
Thanks for keeping me on my toes Mark! =D>
Mark

User avatar
danielj
Posts: 6684
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: AtoMMC2 firmware 2.5

Post by danielj » Thu Oct 17, 2013 7:54 am

oss003 wrote:While refurbishing an Atom, I encountered a nasty problem causing the AtoMMC interface not to work. :(

The Atom worked fine but the AtoMMC interface could sometimes not read the SD-card and was not very stable.

I found out that it's very important for the voltage level not to exceed the 5,3V.
After adjusting the level to 5V it worked ok. :wink:

Greetings
Kees
Okeycokey, my AtoMMC is playing up (strangely not in SDDOS mode, just AtoMMC mode) - how did you tweak the voltage to 5V exactly? It does seem to be hovering up around 5.3V... I'm currently using the disk pack PSU... I can't really see an obvious way :?

Is there some way I could add a trimmer to the circuit without melting anything?


d.

Prime
Posts: 2597
Joined: Sun May 31, 2009 11:52 pm
Contact:

Re: AtoMMC2 firmware 2.5

Post by Prime » Thu Oct 17, 2013 10:55 am

danielj wrote: Okeycokey, my AtoMMC is playing up (strangely not in SDDOS mode, just AtoMMC mode) - how did you tweak the voltage to 5V exactly? It does seem to be hovering up around 5.3V... I'm currently using the disk pack PSU... I can't really see an obvious way :?

Is there some way I could add a trimmer to the circuit without melting anything?
Daniel, what are you using to measure the voltage with, I did find that my meter gave me more accurate readings than the voltage measurement using the Rigol scope (was about 0.2V higher on the Rigol), just mention this as I know you have the same scope.

Cheers.

Phill.

User avatar
danielj
Posts: 6684
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: AtoMMC2 firmware 2.5

Post by danielj » Thu Oct 17, 2013 11:32 am

I was using a meter, but it is a cheapy one. I was going to scope it tonight just to make doubly sure..

I'm getting the classic "SYN" and the directory list not getting right the way to the bottom with the AtomMMC firmware and flakey loading. That said SDDOS seems mostly to be fine :? go figure!

d.

User avatar
oss003
Posts: 2797
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: AtoMMC2 firmware 2.5

Post by oss003 » Thu Oct 17, 2013 12:06 pm

danielj wrote:I was using a meter, but it is a cheapy one. I was going to scope it tonight just to make doubly sure..

I'm getting the classic "SYN" and the directory list not getting right the way to the bottom with the AtomMMC firmware and flakey loading. That said SDDOS seems mostly to be fine :? go figure!

d.
Hi Daniel,

strange, they use the same routine for accessing the SD-card. Maybe it's an SD-card problem? Sandisk works fine for me. You can also try to format FAT or FAT32, just have to try which works the best.

In my power supply there's a pot-meter to adjust the voltage. I don't think there's one in the Disk pack.

Greetings
Kees

User avatar
danielj
Posts: 6684
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: AtoMMC2 firmware 2.5

Post by danielj » Thu Oct 17, 2013 2:17 pm

I'll try a couple of different cards and see if it changes anything, and have a proper poke with a scope later... Struggling to find something smaller than 4gb!

d.

User avatar
roland
Posts: 2943
Joined: Thu Aug 29, 2013 8:29 pm
Location: Born (NL)
Contact:

Re: AtoMMC2 firmware 2.5

Post by roland » Thu Oct 17, 2013 5:01 pm

oss003 wrote: In my power supply there's a pot-meter to adjust the voltage. I don't think there's one in the Disk pack.
That's correct, the disc pack has no easy way to regulate the output voltage. There isn't even a resistor that could be replaced. It's a simple design with two voltage regulators and some capacitors. The only way to regulate the voltage is to add or remove 2114's :D
256K + 6502 Inside
MAN WOMAN :shock:

User avatar
danielj
Posts: 6684
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: AtoMMC2 firmware 2.5

Post by danielj » Thu Oct 17, 2013 5:51 pm

roland wrote:
That's correct, the disc pack has no easy way to regulate the output voltage. There isn't even a resistor that could be replaced. It's a simple design with two voltage regulators and some capacitors. The only way to regulate the voltage is to add or remove 2114's :D
Haha :) I've just changed from generic card to Kingston card... All good now. Strangely it seemed to take a minute or so to "warm up" then seems absolutely fine, voltage measured at the card was 3.29V

Two things I've noticed generally which are probably useful to note: although you can disable atommmc with ?#bffe=4 (or 0, can't remember which off the top of my head) if diskromen is connected, dos can be selected but the drive won't work. You have to have diskromen disconnected and work that way round.

If you initialise atommmc with ?bffe (which really is no hardship), branquart seems to be locked on a single rom - no idea what, an unlock is necessary to get everything playing nicely again!

D.

User avatar
CMcDougall
Posts: 6346
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland
Contact:

Re: AtoMMC2 firmware 2.5

Post by CMcDougall » Sat Oct 19, 2013 3:19 pm

danielj wrote:my AtoMMC is playing up - how did you tweak the voltage to 5V
:-k :!: :?: *maybe*/hopefully [-o< that's whats wrong with mine :?
I've just taken 3days away from my Atom/MMC (as was going off my head! :x ), in a massive luxury static/lodge caravan thing 8)

Mine is all ran off a master compact PSU/Disc box/monitor stand which my cheapo (£3) multimeter says it spits out 5.36/7v. It seems to be quite accurate as battery's etc are at right reading too :)

cheers for the 'bump' Daniel, now don't have to solder on a new PL8 etc etc..... :D

What is the best 'spot on' 5v adaptor to use then?
I am thinking of using my old train set/ scalextric set adaptor next (plus less space!)
ImageImageImage

etaoin
Posts: 1
Joined: Sat Oct 06, 2018 12:34 pm
Location: Netherlands
Contact:

Re: AtoMMC2 firmware 2.5

Post by etaoin » Mon Oct 08, 2018 11:18 am

I know I'm a bit late to this game, but I recently dusted off my trusty Atom that I bought new somewhere mid 80s, and fixed some stuff to get it up and running again. Looking around for things to bring it into the 21st century a bit, I came across AtoMMC2 and this thread as the most likely place to obtain it.

Is AtoMMC2 still available anywhere?

User avatar
oss003
Posts: 2797
Joined: Tue Jul 14, 2009 11:57 am
Location: Netherlands
Contact:

Re: AtoMMC2 firmware 2.5

Post by oss003 » Mon Oct 08, 2018 8:38 pm

etaoin wrote:
Mon Oct 08, 2018 11:18 am
Is AtoMMC2 still available anywhere?
You'll have to ask Charlie (sirmorris).
You can find contact information about a lot of Atom expansion here: viewtopic.php?f=13&t=8168&p=86259&hilit ... ard#p86259

Greetings
Kees

Post Reply