help request with my RiscPC
Pages: 1 2
Pitteloud Stéphane (8213) 25 posts |
ok, no problem. I will try myself until tuesday. |
Rob Andrews (112) 164 posts |
You only need to boot the computer with the shift key held down to get over the error this will get you to the desktop then you need to investigate what is causing the error by opening up !Boot, to do this you hold the shift key then click on !boot, this will open a filer window with lots of folders. |
Pitteloud Stéphane (8213) 25 posts |
Hi Rob i first solved the problem with removing the LAN card, then I could boot a little bit further. But I think I’m doomed with this card :-) Now, the boot process seems to go well, but it take ages to come on the desktop. sometime, the desktop eventually show up, sometimes a nice error message popup to say that the disk is not present. The exact same hard disk boot fine and fast (well, as fast as the internal IDE can do) to the desktop. |
Rob Andrews (112) 164 posts |
hi which slot in the back plane are you pluging in the ide controller? is it a two slot back plane? |
Pitteloud Stéphane (8213) 25 posts |
yes, it’s a two slot backplane. I’m in the upper slot, but I tried the lower slot too, with same result. |
Rob Andrews (112) 164 posts |
switch off take out the back plane and plug the podule into the motherboard to rule out a fault with the back plane then restart |
Pitteloud Stéphane (8213) 25 posts |
This is physicaly not possible : the Podule slot is 32×3 female connector, on the motherboard the podule slot is a PCI like style slot. Not physicaly compatible. |
Rob Andrews (112) 164 posts |
sorry your right have you tried it in the bottom slot? |
Pitteloud Stéphane (8213) 25 posts |
yes, I have tried both slot, no luck. I’m under the impression that the podule board can’t talk correctly to the hard disk, The boot begin, take ages, eventually come to the desktop, or sometimes tell that the drive is unreachable, and I can’t go further. Same disk drive work flawelessly on the internal IDE. |
Rob Andrews (112) 164 posts |
there are some drives that will not work in the risc pc it been a long time since i had mine up and running so i cannot recall if it was seagate or connor drives that had problems what drive do you have? |
Pitteloud Stéphane (8213) 25 posts |
I’ts a seagate. I try one last think : make an image of the hard disk on a compact flash, try to boot the Compact Flash on the internal IDE to prove that it’s good (I know that it will avoid the LAN to work, but I just want to see if it’s boot) and then plug it on the Podule. If it doesn’t work, i will then assume that the podule has a problem. EDIT |
Pitteloud Stéphane (8213) 25 posts |
I tried with both discs attached, a copy of the hard disc on a CF card attached to the IDE podule, and the original hard disk on the internal IDE. When boths hard disk are connected, the System begin to boot on the compact flash that sits on the podule, and finish to load the desktop on the harddisk that sits on the internal IDE. When I remove the internal IDE, the boot begin on the CF that sits on the podule, wait a while, an go to the desktop and ask me "please insert disc “hard drive 4”. Now I’m asking me : do I need to change something in some boot file to tell the system where to look for the startup files needed for the desktop ? I have the feeling that at some point, the OS try to find somes files in the original location, meaning the internal IDE hard drive. If you look at the bar, I have 2 Disk 4. |
Andrew Rawnsley (492) 1445 posts |
The message is from ADFS (ie. internal IDE interface) not ZIDEFS, so there’s obviously a reference in your !Boot to files on ADFS::HardDisc4.$ If I were you, I’d look in !Boot.Choices.Boot.Desktop file (shift-double click your way down), and also the files in the Tasks sub-folder (same location). Replace any references to ADFS accordingly. Note that where possible, replace references to ADFS::HardDisc4.$. with Boot:^. |
Pitteloud Stéphane (8213) 25 posts |
ok, thanks for the tip. If I understand right, the filesystem itself, aka ADFS or ZIDEFS is a part of the path of the files? are there anothers files where a reference to the ADFS can be hidden? |
Andrew Rawnsley (492) 1445 posts |
I think you may need *configure IDEdiscs 0 to get rid of the :4 It’ll probably also speed things up if it isn’t looking for non-existent drives. |
Andrew Conroy (370) 740 posts |
Or if the drive is plugged into the ZIDEFS interface, you want to do a |
Pitteloud Stéphane (8213) 25 posts |
ok, thanks to all your infos, now its is working. I can reach the desktop and have the whole system loaded without error. But…. when the desktop appear, the desktop wallpaper need about 33 seconds to appear, this take only 10 seconds on the internal IDE controller. I have the feeling that the disc access are much slower. Isn’t the IDE podule supposed to be faster than the internal IDE ? |
David Feugey (2125) 2709 posts |
Not always. With SCSI yes, as SCSI to SCSI transfers are possible.
Here is the problem. This interface is made for much older and slower hardware. |
Rick Murray (539) 13851 posts |
It depends – doesn’t the internal IDE mostly run in PIO mode? Not a big deal for harddiscs of the A5000 era, but by the time the RiscPC came out…
Actually, looking at the hardware, I think the problem is that you’re basically using a lot of I/O, latches, data transfer, and processor time to make what is essentially a bit-bashed interface. |
Pitteloud Stéphane (8213) 25 posts |
Ok, thanks for the clarification. then this interface is pointless for me. I asked if this interface is faster than the internal IDE, I was answered that yes, it’s a bit faster, but not so much. I wanted this interface for 2 reasons: 1. being able to use a compact flash and the LAN interface together (I realised after many hours that my connection problem was there only when I use a compact flash on the Internal IDE, and that with the regular hard disk the lan card work well) and 2. having a faster interface than the slow 1.6 Mb/s internal IDE. |
Andrew Conroy (370) 740 posts |
In the tests which we performed at work on this interface we did see a slight but measured speed increase over the regular ADFS interface on the RiscPC, hence I said “it’s a bit faster, but not so much”. It’s a Bank Holiday and I’m at home not at work so I don’t have access to these figures right now, but maybe Rick or David could share the speed test figures they have? |
Pitteloud Stéphane (8213) 25 posts |
could be interesting, yes. If this is confirmed, any possibilitie to trade off and go on the better model instead, the Arcin32 ? |
Rick Murray (539) 13851 posts |
That might be because of subtle differences in CF cards vs harddiscs. Have you had any luck using a normal harddisc on the internal IDE, and the CF on the IDE podule? You might find a CF card that won’t work on the internal bus does work on an extension card.
Let’s just speak the truth – the internal IDE is really miserably poor. Like I said earlier, I think it is stuck in PIO mode so essentially talks to the media in the slowest manner possible. Pretty much anything is going to be an improvement on that…
…including a bunch of latches and logic gates pretending to be an IDE interface. Yup, getting the processor to do most of the work itself is faster than the internal interface.
On all podules, your primary limitation is the bus speed. It’s 8MHz synchronous tops. Or slower (I think the regular access speed is something like 2MHz?). This means you’ll see about 6MB/sec maximum on a 32 bit podule. Older 16 bit podules would manage about half that, and it is roughly in keeping with the speeds that I can remember from my older machines – given a big wimpslot, a Morley cached SCSI podule managed around 1.5MB/sec from tape streamer to harddisc (which means it was about 3MB/sec because it was writing to a SCSI drive on the same bus). My favourite harddisc interface is the 16 bit Simtec IDE; never gave me any problem and god knows I’ve hooked all manner of drives to it in my time…but note that I never tried it with a CF card; I meant to, but then the 32 bit world came along and my RiscPC…well…it’s probably full of battery fuzz nowadays. :-( |
Pages: 1 2