Amstrad PC1512 Hard Disk

[Gordonjj]

Hi folks,

I’m trying to resurrect an Amstrad PC1512 that’s been in the attic for about 25 years. I’ve fixed a corroded leg on a ROM, removed the extra 128k RAM chips to get past a memory parity error and I can boot DOS (original
System disks).

Anyway, the MiniScribe MFM HDD has sat down but I have an ST225 and an ST125 and both the original 8 bit controller and another I have seem to support both drives - both are WD controllers and I can LLF both drives using debug and g=c800:5, and subsequently format with /s and I can read data back from the drives. All looks good.

What I can’t do is get it to boot from the drives though - with the original controller (can’t find a part number yet...) it’s just that nothing happens, with the other (an LCS-6210) when there’s no boot disk in A:, it starts the preformat utility where i expect it to boot (without running c800:5 manually). It’s almost as if the controller thinks the disk is not formatted and tries to start again preboot, but it does it with both ST125 and ST225 (they have the same geometry, which is also the same as the MiniScribe M8425).

Can anyone think of anything obvious I’m missing here...? I don’t think there’s any setup for the XT to tell it to boot from HDD so it should just work but I’m struggling to think of what to try next.

Thanks In advance!
Gordon

 

[Malc]

,,,Anyway, the MiniScribe MFM HDD has sat down but I have an ST225 and an ST125 and both the original 8 bit controller and another I have seem to support both drives - both are WD controllers and I can LLF both drives using debug and g=c800:5, and subsequently format with /s and I can read data back from the drives. All looks good.

What I can’t do is get it to boot from the drives though

Sometimes ( Not always ) running " FDISK /MBR " from a DOS boot floppy fixes this.

- with the original controller (can’t find a part number yet...) it’s just that nothing happens, with the other (an LCS-6210) when there’s no boot disk in A:, it starts the preformat utility where i expect it to boot (without running c800:5 manually). It’s almost as if the controller thinks the disk is not formatted...

It's been a long time since i owned a 1512 and 1640 but IIRC the "Preformat" utility does not Partition and format the drive ready for installing DOS. Once you have run the Preformat with NO errors you need to run FDISK and partition the drive and after a reboot format the drive and then install DOS.

 

[Gordonjj]

Sometimes ( Not always ) running " FDISK /MBR " from a DOS boot floppy fixes this.

Thanks - had to be worth a try - not sure this option was in the DOS 3.2 as it didn't seem to do anything and same result after booting.

It's been a long time since i owned a 1512 and 1640 but IIRC the "Preformat" utility does not Partition and format the drive ready for installing DOS. Once you have run the Preformat with NO errors you need to run FDISK and partition the drive and after a reboot format the drive and then install DOS.

Absolutely correct - I did the LLF then formatted using /s to copy basic boot over - command.com is there, it just infuriatingly doesn't boot!

To confirm, the drive is an ST-225 and the controller is an LCS-6210C. The manual for the controller has the ST-225 as supported and the card is in the configuration as set out in the manual.

System boots from A: (5.25" floppy) and I run the preformat:



Next, I run fdisk, select 1 and make the whole disk available for DOS / active and reboot as required after fdisk.

I then run DOS format:



Then, after reboot with no disk in A:, I'm dropped back into the preformat routine running in the ROM on the LCS-6210C...



Odd, isn't it?!?

 

[Malc]

Dunno what to suggest then, Too many years have gone by since i had one of those.

 

[Tharg]

I have a 1512 with the same problem- dead Miniscribe HD- this one the 8438 model. The PC boots fine with the system floppy, and works fine otherwise. It's just the HD/controller card that is dead.

Is it possible to swap out the controller card to see if this is the problem? Can the controller card be checked and fixed?

 

[mrwebber35]

Same MiniScribe 8425 same DTC-5051CR card I suppose. The MiniScribe is the Winchester with ST412 Interface that the card doesn't automatically recognize I think. The card has the BIOS chip BXDO7 which is for ST412 interface but not the 21.4MB formatted drive with 615 cylinders and 4 heads. It has to be "free formatted which requires the memory for the BIOS to set for 16 bit not 8 bit. This card doesn't have the right jumper to set for the larger manually set format G=C800:5 for the slightly different drive. Other DTC-5150CRH cards have a "W3" 2 pin jumper that my card doesn't have. That's where I'm at now. Anyone know what the hell I'm saying? I'm not too sure. My guess is the Amstrad PC1512SD came preprogrammed and the HD came pre formatted. Nobody was aware that the card was intended as an interface and not set up for a total reformat.

 

[modem7]

Same MiniScribe 8425 same DTC-5051CR card I suppose. The MiniScribe is the Winchester with ST412 Interface that the card doesn't automatically recognize I think.

Correct. The 'ST-506/ST-412 Interface' does not provide a mechanism for the controller to interrogate the drive for drive information (make, model, cylinder count, head count, etc.)

... It has to be "free formatted" ...

More commonly known as 'dynamic' formatting. See [here] for XT-class HDD controller 'categories'.

Anyone know what the hell I'm saying? I'm not too sure.

Certain models of DTC controllers are 'multi-geometry' controllers. They do not support dynamic formatting.

Certain models of DTC controllers are hybrids. The user has the option is using either of two modes:
- (Multi-geometry) Set the controller's drive geometry jumpers to match the geometry of the attached drive, then perform a low-level format; OR
- (Dynamic) Set the controller's drive geometry jumpers to the 'free format' setting, then perform a low-level format

Same MiniScribe 8425 same DTC-5051CR card I suppose ... The card has the BIOS chip BXDO7 which is for ST412 interface but not the 21.4MB formatted drive with 615 cylinders and 4 heads.

Online, I see that the MiniScribe 8425 has a geometry of 615/4 (cylinders/heads). No write precompensation (WPC).

In a hobby environment, WPC can be ignored. Even 'back in the day', some manufacturers may have sometimes ignored it (like as described in 'variation #1' at [here]).

Information on your DTC 5150CR is at [here]. Strangely, no mention of the W6 geometry jumpers. At the bottom of that page are the drive geometries supported by the DXD06 BIOS and the BXD07 BIOS.
Although there are no direct geometry matches to the MiniScribe 8425, there is a geometry that is very close: 612/4 (cylinders/heads)
Selecting that geometry would result in the last three cylinders of the MiniScribe 8425 not being used, which is about half a percent of capacity.
I don't think that many people would be concerned about that half a percent, even manufacturers.

 

[mrwebber35]

Yup, the illusive W6... the search continues.

 

[modem7]

Yup, the illusive W6... the search continues.

The W6 jumper-to-drive-table-entry is quite possibly the same as W6 in the DTC 5150CRH (note the 'H').
See the final page of the 5150CRH Installation Guide at [here].

If so, the jumper-to-drive-table-entry is:

Jumper positions 3,4, 7, and 8 are for the first drive.
Jumper positions 1,2, 5, and 6 are for the second drive (if any).

TABLE |  Jumper Jumper   Jumper Jumper
ENTRY |   3      4        7      8
------+--------------------------------------
0     |  ON     ON       ON     ON     (1111)
1     |  OFF    ON       ON     ON     (0111)
2     |  ON     OFF      ON     ON     (1011)
3     |  OFF    OFF      ON     ON     (0011)
4     |  ON     ON       OFF    ON     (1101)
5     |  OFF    ON       OFF    ON     (0101)
6     |  ON     OFF      OFF    ON     (1001)
7     |  OFF    OFF      OFF    ON     (0001)
8     |  ON     ON       ON     OFF    (1110)
9     |  OFF    ON       ON     OFF    (0110)
10    |  ON     OFF      ON     OFF    (1010)
11    |  OFF    OFF      ON     OFF    (0010)
12    |  ON     ON       OFF    OFF    (1100)
13    |  OFF    ON       OFF    OFF    (0100)
14    |  ON     OFF      OFF    OFF    (1000)
15    |  OFF    OFF      OFF    OFF    (0000)

Which geometry corresponds to which table entry will depend on what is in the card's BIOS ROM.

The DTC 5150CRH Installation Guide shows what applies to the CXD03A BIOS, although listed are drive models rather than geometries. Of course, with a drive model, the geometry can be worked out.

There should be a list somewhere for the BXD07 BIOS.

Note that in a lot of cases (not all), you can see the drive table information for a controller (XT class controller) via the ROMLIST switch of SpeedStor software. Example at [here].

 

[modem7]

There should be a list somewhere for the BXD07 BIOS.

Doh! I actually pointed to it in the final paragraph of post #7.

( Curse this old age. And now I've forgotten the URL for my web site. I know that it has something to do with my avatar. Perhaps rectal-thermometers-r-us.com ? )

 

[mrwebber35]

The W6 jumper-to-drive-table-entry is quite possibly the same as W6 in the DTC 5150CRH (note the 'H').
See the final page of the 5150CRH Installation Guide at [here].

If so, the jumper-to-drive-table-entry is:

Jumper positions 3,4, 7, and 8 are for the first drive.
Jumper positions 1,2, 5, and 6 are for the second drive (if any).

TABLE |  Jumper Jumper   Jumper Jumper
ENTRY |   3      4        7      8
------+--------------------------------------
0     |  ON     ON       ON     ON     (1111)
1     |  OFF    ON       ON     ON     (0111)
2     |  ON     OFF      ON     ON     (1011)
3     |  OFF    OFF      ON     ON     (0011)
4     |  ON     ON       OFF    ON     (1101)
5     |  OFF    ON       OFF    ON     (0101)
6     |  ON     OFF      OFF    ON     (1001)
7     |  OFF    OFF      OFF    ON     (0001)
8     |  ON     ON       ON     OFF    (1110)
9     |  OFF    ON       ON     OFF    (0110)
10    |  ON     OFF      ON     OFF    (1010)
11    |  OFF    OFF      ON     OFF    (0010)
12    |  ON     ON       OFF    OFF    (1100)
13    |  OFF    ON       OFF    OFF    (0100)
14    |  ON     OFF      OFF    OFF    (1000)
15    |  OFF    OFF      OFF    OFF    (0000)

Which geometry corresponds to which table entry will depend on what is in the card's BIOS ROM.

The DTC 5150CRH Installation Guide shows what applies to the CXD03A BIOS, although listed are drive models rather than geometries. Of course, with a drive model, the geometry can be worked out.

There should be a list somewhere for the BXD07 BIOS.

Note that in a lot of cases (not all), you can see the drive table information for a controller (XT class controller) via the ROMLIST switch of SpeedStor software. Example at [her

Thank you for the spec sheet. Best I've seen so far.

My card came with W6 jumpers set ON pins 1-2... 1,2,3,4,6,8. Only 5 and 7 were off.
This is exclusive for the BXDO7 8 bit BIOS. If this is wrong I have no way of knowing.
Cannot set the card to the "FREE FORMAT" because other DTC-5150CRH info says it needs the 16 bit ROM BIOS set by W1.
I tried W1 setting for 16 bit and the floppy stops working on next boot. At least putting it back is still letting it work.
Must need the ROM memory chip replaced or re-flashed I don't know. Hate to do that wrong.

Question: the jumper setting say that they are set so you can use a different DMA channel than what an IBM PC/XT uses.
What exactly is DACK3, DREQ3 factory settings? The other setting is DACK1 or DREQ1. Obviously factory is for the channel used by the IBM PC/XT but
<<< is the Amstrad PC1512 an XT that is different than the IBM? >>>

I tried SSTOR and it find the Microscribe 8425 615 cylinders etc and you'd think all is good.... but then it says it's for DOS 3.30 and higher. Amstrad used DOS GEM 2.0. FDISK for GEM is different than DOS 6.22., GEM can not recognize the partition info. SSTOR says there is an older version than the 6.06 version that is available on the web. Does anyone have the older version?

Incidentally SSTOR found no errors on the drive before it wrote a whole new partition and formatted it. GEM couldn't format it though. Now SSTOR can't read a single track. The rabbit just keeps getting farther down the hole. I run faster and faster but...

The Amstrad PC1512 owners manual says it is a 512K/640K system unit with 8MHZ 8086 processor and a 20MB HD. The manual is exclusively Amstrad hardware. No further information is provided in the owner’s manual.

Lots is unique to the Amstrad, the mouse is different and can only be used on the Amstrad; same with the monitor and keyboard. All have PCBA’s that have AMSTRAD inscribed on them. That means designed by Amstrad for Amstrad. There are Amstrad FDD’s and only they are mentioned in owner’s manual. HD is sketchy since the PC did not come with an Amstrad HD. That is 3rd party. I don't have the owners manual for the MicroScribe 8425XT. I know the channel jumper is set for DS0, which must be correct. There are jumpers like J13 and I know that is for self-diagnosis when powered on. It can do that going back and forth methodically over the disk for hours without LED error flashes. Why no errors if the disks cannot be read or tested? I don't know what the other jumpers do like J15, J16, J18, J12.
I took Computer Science at the college level for 3 semesters back in 1985. I'm not to sure but they must have really wanted these things to work but just didn't have the circuitry or software to simplify it for us poor simple IT folk. Nothing was plug and play obviously. Now they have self-driving parking cars. Ya, the past was harder for the regular folk. Now we are all geniuses. I hear astrophysicists need analog computers to find the mysteries of the universe here in 2023. I think we know why.

 

[modem7]

Lots is unique to the Amstrad, the mouse is different and can only be used on the Amstrad; same with the monitor and keyboard. All have PCBA’s that have AMSTRAD inscribed on them. That means designed by Amstrad for Amstrad.

Note that in some cases, computer companies found it easier/cheaper to licence designs from other manufacturers. For example, take a look at the card comparison at [here].

I tried W1 setting for 16 bit and the floppy stops working on next boot. At least putting it back is still letting it work.
Must need the ROM memory chip replaced or re-flashed I don't know. Hate to do that wrong.

Per [here], W1 on the 5150CR has nothing to do with bit size. It will be the ROM type/size.
Setting 8K = EPROM type 2764, or 27C64 (i.e. 8 KB in size)
Setting 16K = EPROM type 27128, or 27C128 (i.e. 16 KB in size)

As shown at [here], the BXD07 is in a 2764 EPROM (or PROM equivalent), and so leave the W1 jumper set to the 8K setting.

With W1, what DTC were doing is allowing for the fact that down the track, they may have decided to offer an upgraded ROM for the card, one where the code grew to the point where an 8 KB sized PROM/EEPROM was no longer large enough.

Question: the jumper setting say that they are set so you can use a different DMA channel than what an IBM PC/XT uses.
What exactly is DACK3, DREQ3 factory settings?

DACK3 and DRQ3 are the two pins in the ISA slot that DMA channel 3 uses.

The other setting is DACK1 or DREQ1. Obviously factory is for the channel used by the IBM PC/XT but ...

The IBM supplied hard disk controller (made for IBM by Xebec) in the IBM XT used DMA channel 3. Other manufacturers of XT-class hard disk controllers followed suit.

... <<< is the Amstrad PC1512 an XT that is different than the IBM? >>>

Hopefully, someone familiar with the PC1512 will know if the PC1512 uses DMA channel 3 for something else.

 

[modem7]

I tried SSTOR and it find the Microscribe 8425 615 cylinders etc ...

No, it did not 'find' the drive. What it did was ask the 5150CR controller for the geometry of the attached drive. Your 5150CR will have determined (at power-on time) the geometry by examination of the W6 jumpers.

... and you'd think all is good.... but then it says it's for DOS 3.30 and higher. Amstrad used DOS GEM 2.0. FDISK for GEM is different than DOS 6.22., GEM can not recognize the partition info.

I am unfamiliar with Amstrad computers. There is a service manual for the PC1512 at [here], and paragraph 11 on page 25 suggests that one can choose between GEM and DOS.

SSTOR says there is an older version than the 6.06 version that is available on the web. Does anyone have the older version?

See [here].

Incidentally SSTOR found no errors on the drive before it wrote a whole new partition and formatted it. GEM couldn't format it though. Now SSTOR can't read a single track.

Normally, one doesn't use SpeedStor to partition then high-level format a hard drive.

Normally, one would use SpeedStor to do a low-level format (SpeedStor uses the term 'initialize') of the drive, then use FDISK to create a DOS partition, then use FORMAT to high-level format the drive (partition actually).

 

[JohnElliott]

I tried SSTOR and it find the Microscribe 8425 615 cylinders etc and you'd think all is good.... but then it says it's for DOS 3.30 and higher. Amstrad used DOS GEM 2.0. FDISK for GEM is different than DOS 6.22., GEM can not recognize the partition info. SSTOR says there is an older version than the 6.06 version that is available on the web. Does anyone have the older version?

There is no FDISK for GEM. GEM runs on top of DOS (or DOS Plus), and the FDISK on the PC1512 boot disk is the one for MSDOS 3.2.

FDISK /MBR was introduced in MSDOS 5, so won't be available on the PC1512 unless you use a more recent version of MSDOS.

There is a service manual for the PC1512 at [here], and paragraph 11 on page 25 suggests that one can choose between GEM and DOS.

Not quite. There is the possibility of a dual boot between MSDOS and DOS Plus - GEM isn't an operating system in that sense. The dual boot is installed by FIXLDR.COM.