RPC with faulty keyboard interface

Arc/RPCs, peripherals, RISCOS operating system & ARM kit eg GP2x, BeagleBoard
Post Reply
drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

RPC with faulty keyboard interface

Post by drumm3r » Sun May 27, 2018 2:57 pm

Hi all

I recently bought a RISC PC 600 from eBay and am in the process of trying to get it going. I currently have it booting to a Supervisor prompt, but unfortunately the keyboard doesn't work at all (I've tried 2 with no luck).

I've checked continuity to the IO controller, and verified power, both of which are fine. I've also dug a bit deeper using a logic analyser and found that after a few slightly odd messages (registering several keypresses and releases, as well as the usual self-test result from the keyboard) the host is holding the PS/2 clock line low. As I understand it this stops the keyboard from communicating, but I'm not sure why it's happening. Also oddly when I boot the machine with no keyboard, I get an error stating this, so it's obviously been detected!

Below is what I see on the bus during boot:

Code: Select all

Keyboard: MAKE 'J' (0x3B)
Keyboard: MAKE 'K' (0x42)
Keyboard: BREAK 'H' (0xF0
Keyboard: MAKE Unrecognized Key (0x00)
Keyboard: MAKE Unrecognized Key (0xFF)
Keyboard: MAKE Unrecognized Key (0xF4)
Keyboard: BAT Power-On Self Test Successfull (0xAA)
Keyboard: MAKE [F3] (0x04)
Keyboard: MAKE [F3] (0x04)
Keyboard: MAKE [F3] (0x04)
Has anyone had issues like this? Not sure what to try next, as it looks like the IO controller is the culprit (holding the clock low) but I'm not sure why that might be!

Any help would be hugely appreciated.

Angus

Bat
Posts: 165
Joined: Tue Mar 13, 2018 1:17 pm
Contact:

Re: RPC with faulty keyboard interface

Post by Bat » Sun May 27, 2018 7:22 pm

Hi,
Have you checked the internal battery?
If it's good try leaving it disconnected for an hour then try again.
Apologies if you've already tried this.
Cheers :)

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Mon May 28, 2018 7:29 am

Hi

Thanks for the suggestion! Yes, I've replaced the battery and checked the RTC is running ok (there is I2C comms to the IO controller so I take that to be a good sign!). There doesn't seem to be any damage to the traces near the old battery either which is a relief

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Mon May 28, 2018 2:55 pm

Sorry - jumping in on my own thread again! Are RISC PCs particularly picky about the keyboard used? I've tested 2 - an old one I had lying around and a new (but cheap) one from eBay; although both show slightly different messages on the PS/2 bus while booting, neither work (same problem - the clock line is held low by the RPC). Is it just that I need a 'compatible' keyboard?

Thanks (again)

Ottly
Posts: 148
Joined: Tue Jun 10, 2014 10:34 am
Contact:

Re: RPC with faulty keyboard interface

Post by Ottly » Mon May 28, 2018 3:53 pm

Hi,

I have this problem with my RISC pc if the cmos isn’t working.

Only way the keyboard continues to work is if I hold down delete to reset cmos settings on power on. Red screen comes on for a couple of seconds then goes away and the machine boots. If I don’t hold down delete. The machine would boot to the supervisor prompt but the keyboard wouldn’t work.

So try powering on and holding delete until the machine starts to boot.

Scott.

User avatar
lcww1
Posts: 281
Joined: Wed Mar 15, 2017 11:16 pm
Location: Moon Base Alpha
Contact:

Re: RPC with faulty keyboard interface

Post by lcww1 » Mon May 28, 2018 5:42 pm

I second Ottly’s advice - and just to add that I’ve not found my RiscPC 600 to have any problems with the motley collection of ps/2 keyboards that I’ve plugged into it, so I doubt that the problem will be some subtle keyboard incompatibility - more likely to be the CMOS.....

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Tue May 29, 2018 3:16 pm

Thanks all for the advice. I'll try replacing the RAM/RTC part and see if that gets me anywhere.

Ottly
Posts: 148
Joined: Tue Jun 10, 2014 10:34 am
Contact:

Re: RPC with faulty keyboard interface

Post by Ottly » Tue May 29, 2018 3:44 pm

You might not need to replace it. I’ve just repaired 2 RISC pc motherboards that had suffered battery damage. They both would start with a red screen and then boot.

I looked at the circuit diagrams and found a number of tracks had been damaged. Repaired them all and now the cmos is working and no more red screen.

Does your RISC pc display a red screen prior to booting or not?

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Wed May 30, 2018 11:10 am

Hi Ottly

As far as I can tell it doesn't; I get a black screen until the machine gets to the Supervisor prompt (though of course I may just be missing it whilst my monitor syncs). Does that mean it isn't running any self-tests? I remember my A3010 going through the usual red/blue (I think) screens whilst booting, but the RPC doesn't look like it's doing the same.

Apart from holding down delete during boot, are there any jumpers that reset the CMOS? What's your usual 'checklist' of things to run through whilst repairing motherboards?

Thanks (yet again!)

Ottly
Posts: 148
Joined: Tue Jun 10, 2014 10:34 am
Contact:

Re: RPC with faulty keyboard interface

Post by Ottly » Wed May 30, 2018 4:30 pm

I’m not sure re: running self tests on boot up. I’m guessing it must as it would be apart of the ROMs boot up code. The red screen remains on for quite some time, enough time for the monitor to sync, so I'm guessing your machine isn't displaying it.

I'm not aware of any jumpers to reset cmos. My machines do briefly display both the red and blue screens. I can get the blue screen to display if I remove all ram. So I'm guessing a blue screen indicates a RAM problem.

I'm by no means an expert. I've only recently fixed my 2 risc PC motherboards. Both had battery damage and the CMOS wasn't working due to damage to the tracks.

It’s a bit weird that your machine displays no red screen but only boots to a supervisor prompt. That would indicate that your machines hardware including CMOS memory is working, but its booting to supervisor prompt as the settings are not set correctly. Power on holding delete would normally rectify this, however your having keyboard issues. So I guess the problem could either be the CMOS is not working correctly, not keeping the settings or there is an issue with the keyboard interface. I'm suspecting you may have a keyboard interface issue as even with faulty non working cmos memory the machine would still normally boot if holding down delete. But your machine doesn't seem to respond to this.

But you never know, maybe the CMOS IC is faulty and requires replacing. Its worth a shot to replace it.

Do the keyboard lights at least flash on when the machine is powered on?

I've attached the schematic for the CMOS/Realtime clock, bottom left corner. Check the continuity of CMOS circuit with a multi-meter. You might want to also check the continuity of the keyboard interface also.
Acorn_RPCTRM_MainPCBCircuitDiagram1of7.pdf
(765.02 KiB) Downloaded 11 times

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Thu May 31, 2018 7:06 am

Thanks so much for your very detailed reply, Ottly! Hugely appreciated. I've got a new RTC/CMOS device arriving today and I'll continue to buzz things out (though haven't found any broken traces yet). The keyboard lights flash once immediately after applying power (but I think it's the keyboard that controls that), but not a second time which I think suggests the host isn't initializing it.

Thanks for the schematic - I've actually been using a copy I found elsewhere to do the debugging so far.


I'll report back if I'm successful!

User avatar
lcww1
Posts: 281
Joined: Wed Mar 15, 2017 11:16 pm
Location: Moon Base Alpha
Contact:

Re: RPC with faulty keyboard interface

Post by lcww1 » Thu May 31, 2018 10:22 am

From your descriptions, I think your RiscPC is getting through the POST sequence without error.

Ending up at the supervisor prompt with an incomplete set of ROM modules loaded can certainly be a symptom of a failing CMOS chip. See this old thread for a discussion: viewtopic.php?t=3376.

I had similar symptoms with an A5000 recently (booting to supervisor with a few modules loaded, unable to get to desktop, but keyboard working), and replacing the CMOS chip fixed the problem - merely booting whilst holding delete was not sufficient.

Presumably, on your RiscPC, the RISC OS keyboard module isn’t being loaded - the keyboard driver was moved out of the RO kernel and into a separate keyboard module from RO 3.5: http://www.riscos.com/support/developer ... mouse.html - this would explain why the keyboard might not work on a RiscPC with this type of CMOS problem whereas the keyboard remains operational on an A5000 under similar circumstances.

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Thu May 31, 2018 1:11 pm

Thanks for the information - very interesting. I received the new CMOS/RTC part from Farnell this morning, so I'll test it out when I get home. If this works I owe you all a beer...

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Thu May 31, 2018 4:41 pm

SUCCESS!!! (Well, almost...)

Massive thanks again to all those who've helped - I've just replaced the CMOS/RTC device and the keyboard now works, and I can start *Desktop from the Supervisor!

BUT... it seems I can't get the desktop to start automatically; even after holding down DEL or R during boot I end up at the Supervisor each time. Also once I'm in the desktop I can't open the ROM Apps (I receive the message 'Error when reading Resources:$.Apps - Directory 'Resources:$.Apps' not found).

I've soldered an AA battery holder in place of the old one, and am using a 1.2V NiMH battery in its place (I've also checked this voltage reaches the RTC; it does).

I'm going to fix this thing if it kills me...


Angus

User avatar
danielj
Posts: 6646
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: RPC with faulty keyboard interface

Post by danielj » Thu May 31, 2018 5:22 pm

If it's not remembering and you're sure the rtc ic is being powered, double and triple check all of the circuit connectivity/components around it, I've had the resistors/caps in that circuit crumble before when they looked fine on initial inspection.

d.

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Thu May 31, 2018 6:15 pm

Thanks! Having delved a bit further it looks like it actually is saving configuration data (video modes are saved between boots for example), though changing 'NO BOOT' to 'BOOT' I still get to the Supervisor prompt every time. Is there another option I should be setting?

On a different note if anyone finds themselves near Chippenham, Wiltshire there's a beer waiting as a thanks for all the help!


Angus

User avatar
DutchAcorn
Posts: 1899
Joined: Fri Mar 21, 2014 9:56 am
Location: Maarn, Netherlands
Contact:

Re: RPC with faulty keyboard interface

Post by DutchAcorn » Thu May 31, 2018 6:40 pm

The RTC chip does two things; it stores CMOS settings and functions as a real time clock. It now holds the configuration settings, but does it also remember time and date? I have seen an A5000 boot to Supervisor because of an RTC error so it could be helpful to check if the connections to pins 1 & 2 of the rtc chip are ok.
Paul

User avatar
danielj
Posts: 6646
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: RPC with faulty keyboard interface

Post by danielj » Thu May 31, 2018 6:44 pm

*CO. Desktop

?

But have you done a power-on holding down delete since you replaced the IC? That should clear up any muckiness. sorry, didn't read properly!

d.

User avatar
lcww1
Posts: 281
Joined: Wed Mar 15, 2017 11:16 pm
Location: Moon Base Alpha
Contact:

Re: RPC with faulty keyboard interface

Post by lcww1 » Thu May 31, 2018 10:24 pm

Great progress drumm3r!

Now, regarding your missing ROM apps - are you running RISC OS 3.5 by any chance?

Also, one pitfall to be aware of is that it’s possible to run into trouble with the HD after a CMOS reset on a RiscPC with RO 3.5/3.6 - make sure you do a *configure ADFSBuffers 0

Ottly
Posts: 148
Joined: Tue Jun 10, 2014 10:34 am
Contact:

Re: RPC with faulty keyboard interface

Post by Ottly » Fri Jun 01, 2018 5:47 am

Excellent news. Well done! Those pesky CMOS chips do cause a world of pain if there not working don’t they.

The reason why the apps aren’t working is with the later versions of RISC OS the apps were relocated from ROM to hard disk to be soft loaded I believe.

You might also want to check all the Roms are plugged in. *ROMMODULES to display them, can reintialise them with *RMREINIT <rom name>
And check the CMOS Settings. *STATUS.

If there not set right then it won’t automatically boot the desktop. If you want the *STATUS settings from a working machine let me know and I can post some pics of mine up.

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Fri Jun 01, 2018 6:43 am

Thanks all!

Ottly - good tip; I'll check what is unplugged this evening and try re-initializing. I didn't know about the move of the apps to hard disk (I'm still used to RISC OS on my old A3010, which seems a bit simpler!). I'd really appreciate seeing the settings from a working machine, so if you could post a list of the *STATUS settings that would be fantastic.

lcww1 - Yes, it's RO 3.5. I'll configure the ADFS buffers tonight to see if that improves things.

User avatar
lcww1
Posts: 281
Joined: Wed Mar 15, 2017 11:16 pm
Location: Moon Base Alpha
Contact:

Re: RPC with faulty keyboard interface

Post by lcww1 » Fri Jun 01, 2018 9:07 am

The ROM apps were removed from RO 3.5 but returned to their rightful place from RO 3.6 onwards. I’d suggest upgrading to RO 3.7 or 3.71 anyhow, in due course.

In terms of which ROM modules you’ve got loaded on RO 3.5, you can check the output from *ROMMODULES against the list given on this handy website: https://www.marutan.net/db/products.php?kProduct=10

Depending on what you find on your HDD, you may need/want to install a fresh !Boot system - there’s a helpful guide to doing that on a RiscPC here http://www.snowstone.org.uk/riscos/docu ... ewold.html

Lastly, there’s a long stardot thread here viewtopic.php?f=16&t=10172 covering some similar RiscPC CMOS issues, which may be of interest, though I didn’t see any obvious solutions to your current startup to supervisor issue there. I think the thing to check now is whether all the ROM modules are being loaded at startup, and proceed from there.

Ottly
Posts: 148
Joined: Tue Jun 10, 2014 10:34 am
Contact:

Re: RPC with faulty keyboard interface

Post by Ottly » Sat Jun 02, 2018 6:06 am

Attached is the STATUS settings from my RISC PC after powering on with DELETE held down.
status.txt
(1.22 KiB) Downloaded 22 times

drumm3r
Posts: 11
Joined: Sun May 27, 2018 12:46 pm
Contact:

Re: RPC with faulty keyboard interface

Post by drumm3r » Tue Jun 05, 2018 7:28 am

Apologies for the delayed response; busy family weekend!

Thanks again, Ottly. I went to check and update the configuration and to my dismay found the machine won't boot again! Different fault now - I get a very 'grating' beep from the speaker (as opposed to the nice soft beep during a standard boot) and no video - the monitor complains about an unsupported resolution.

Back to square 1! Looks like I have more work to do...

Post Reply