X-Mass help please

[protek]

Does the X-Mass need to be initalized otherwise before trying to initialize it in Symbos?

I think I found a probable cause to my B: drive acting up. The 5v line on the molex connector was loose. I re-crimped the wire.

[TotO]

No, it is not a LBA problem as all the X-MASS work in LBA using BONNYDOS and ACMEDOS (wip).
It is related to the problem that I have pointed to you 3/4 weeks ago about SymbOS fail to properly handle some "modern" flash drives.

As I have said, for testing the drive SymbOS send the "READ NATIVE MAX ADDRESS" command (0xF8).
This command is optional according the ATA standard. It is why, some device are not able to answer it.

In our case, if the DOM return busy (0x80), it is skipped and work properly.
If the DOM return error (0x51), it look not supported by SymbOS.

So, before sending an optional command, you have to know if the device support it.
(that don't mean the device not support LBA if it not support the command... Yes, it look stupid)

EDIT: I have not SyX here to know how to, but gerard does... 

[gerald]

The mandatory IDENTIFY DEVICE command (0xEC) will tell you all you need:
- word 19 : bit 9 is set to indicate LBA mode,
- word  60-61 contains the number of addressable sector (LBA24)
- word 83 : bit 10 is set to indicated LBA48 support
- words 100-103 contain the number of addressable lba in LBA48

Not the latest ATA spec, but could be useful : http://www.t13.org/documents/uploadeddocuments/project/d1410r3b-ata-atapi-6.pdf

[Prodatron]

Ok I will just remove the detection, as its probably senseless anyway. Should be able to provide an update today.

PS: Bonnydos is using  CHS

[TotO]

It's true. I have updated my post.

[protek]

Thanks for all the support guys!  Looking forward for the update.

And thank you also TotO and others for wonderful pieces of hardware!

[TFM]

The mandatory IDENTIFY DEVICE command (0xEC) will tell you all you need:
- word 19 : bit 9 is set to indicate LBA mode,
- word 83 : bit 10 is set to indicated LBA48 support

That's what I use for my IDE ROM (thanks to you pointing that out) and yesterdays tests work like a charm. 

[netmercer]

Hello,
I have a problem with wonderful X-Mass and my extern power supply.

If the X-Mass and MotherX4 are connected to the CPC's expansions port and powered by the CPC all work fine.
But if I connect an external power supply to MotherX4 and switch on the external PSU first and then the CPC, I get into trouble. The X-Mass returns 0x51 in status register (0xFD0F) and 0x04 in error register (0xFD09). I don't know how to clear this error. (I tried three different PSUs, same result)

Perhaps somebody can explain... 

[TFM]

How much Ampere does your 5 V PSU have?

Error &04 tells unknown command. So it looks like there is some bit storm when you switch it on.

My external 3 A 5 V PSU works fine.

Can you power it all up and then make a Reset with the button on the X4? What does FD0F and error register tell in this case?

[TotO]

Sound realy strange. 
Check that your external PSU do not provide a too high voltage. (5V ±10%)

[Bryce]

Any chance that you are mixing a linear PSU and a switch-mode PSU together to power one device??

Also, remeber, that by turning on the external PSU you are actually powering the CPC through the expansion port (if the MX4 doesn't have a diode to avoid this). This could put some ICs into states that they shouldn't be in and also messes up the reset circuitry of the CPC).

Bryce.

[TotO]

As you know, the MotherX4 use 2 Schottky diodes to avoid to power it back, up to 3A current.

[Bryce]

As you know, the MotherX4 use 2 Schottky diodes to avoid to power it back, up to 3A current.

I thought it did, but I couldn't remember for sure and I don't have one here.

Bryce.

[netmercer]

Thank you for your answers.
I'm using the GT65 for powering the CPC and a 5V 1A PSU for powering the MotherX4.
The output voltage of the external PSU is about 5.1V. I tried already different PSU's, one of them was a linear working PSU, the others switch-mode.
I already pushed the reset button on MotherX4 in order to clear the error. But the status register responds always 0x51.
I also tried a soft reset without success (writing 0x0E to 0xFD06 device control register).
Any other ideas ? 

[Bryce]

If the CPC is being powered by a GT65 you definitely shouldn't connect a switch-mode PSU to the MX4.

Bryce.

[Prodatron]

A switch-mode PSU will damage the M4X expansions?

[CraigsBar]

If the CPC is being powered by a GT65 you definitely shouldn't connect a switch-mode PSU to the MX4.

Bryce.

so my mx4 should not be powered by the atx bench power supply and the plus by a cm14 then? Should I be worried?

[Bryce]

A switch-mode PSU will damage the M4X expansions?

No, but mixing two types (The GT65 has a linear PSU) isn't a good idea. It won't damage anything, but it can cause serious stability problems. Ideally you should use one powerful linear PSU to power both the CPC and the MX4 with a separate connector directly to the MX4 so that the MX4 current doesn't have to pass through the CPC expansion connector. I use a Spectrum +3 PSU and an Amiga PSU with adapters.

Bryce.

[pelrun]

Powering the same rail with two regulators simultaneously doesn't work well, it doesn't matter whether they're linear or switched. They will each have a slightly different idea of what voltage the rail should be at, and will fight each other.


I disagree with Bryce a bit on whether switched psu's are a bad idea - certainly the cheap ones aren't so great, but decent ones with a high switching frequency and proper output filtering (capacitor plus ferrite ring/bead) don't give me any problems. I'm using a BEC from my rc plane to drive the 5v rail quite successfully.

[protek]

I made some progress today. With Prodatron's new build the Symbos mass storage applet was able to initialise X-Mass. 

[IanS]

Powering the same rail with two regulators simultaneously doesn't work well, it doesn't matter whether they're linear or switched. They will each have a slightly different idea of what voltage the rail should be at, and will fight each other.

The supply to the cards on an MX4 (or at least the one I have v1.1) if fed from a Bat85 Schottky diode. There are two diodes feeding the supply from the edge connector or the external input to the supply pins of the cards on the MX4.

(Bat85 seems a little underpowered, only having a max current rating of 200mA, not quite sure why the silk-screen has 5V/2A". I'd suggest removing the diodes and replacing the one from the external input with a link if I was using the external input.)

[TotO]

Strange, because when I have ordered them on Farnell, the website description shown 3A. 
So I decided to print 2A to be sure to not excess that. Yes, You should remove it to avoid problems.
(nobody get this issue since 2 years on over 100 boards dispatched... I'm embarrassed)

[Bryce]

I disagree with Bryce a bit on whether switched psu's are a bad idea - certainly the cheap ones aren't so great, but decent ones with a high switching frequency and proper output filtering (capacitor plus ferrite ring/bead) don't give me any problems. I'm using a BEC from my rc plane to drive the 5v rail quite successfully.

I fully agree, but how often do you actually see a high quality SMPS? The market is flooded with cheap and nasty SPMSs that are just one big noise source. This is what most people are buying when the get an SMPS.

Bryce.

[netmercer]

Sure, there are a lot of lousy switch-mode PSU's for sale, which could cause trouble.
But in my special case (X-Mass hanging up in error status during switch on, X-Mass powered first, CPC powered second) I don't agree with the theory, that switch-mode PSU is causal for this problem.
I did two additional tests, which I repeated several times. In both cases the X-Mass was powered by a switch-mode PSU (5V, 1A).

1.) I powered the CPC first and afterwards I powered the X-Mass. All work fine !!!
2.) I connected CPC and X-Mass to a switchable power socket and powered them simutaneously. All work fine !!!

IMHO the chronology is the cause, but I don't understand why.

This minute I'm testing only with X-Mass connected to MotherX4 and therefore a possible current limit of 200mA by the Schottky diode shouldn't be a problem. But in future I want to connect further equipment, like RRB, RS232, MegaFlash etc. ...

I'm astonished that I read in this forum about people, which run a fully equipped MotherX4 without any problems with their external PSU. Perhaps some boards are armed with a "super"-diode....

[CraigsBar]

Sure, there are a lot of lousy switch-mode PSU's for sale, which could cause trouble.
But in my special case (X-Mass hanging up in error status during switch on, X-Mass powered first, CPC powered second) I don't agree with the theory, that switch-mode PSU is causal for this problem.
I did two additional tests, which I repeated several times. In both cases the X-Mass was powered by a switch-mode PSU (5V, 1A).

1.) I powered the CPC first and afterwards I powered the X-Mass. All work fine !!!
2.) I connected CPC and X-Mass to a switchable power socket and powered them simutaneously. All work fine !!!

IMHO the chronology is the cause, but I don't understand why.

This minute I'm testing only with X-Mass connected to MotherX4 and therefore a possible current limit of 200mA by the Schottky diode shouldn't be a problem. But in future I want to connect further equipment, like RRB, RS232, MegaFlash etc. ...

I'm astonished that I read in this forum about people, which run a fully equipped MotherX4 without any problems with their external PSU. Perhaps some boards are armed with a "super"-diode....

mine has all 4 slots occupied, from front to back....

X-mem
X-mass
PlayCity
Mini booster

It works perfectly when power either by an iomega zip drive 5v 2a brick, or a pc atx supply.

I think you have a super diode :p