Peter's site blog

Please read ! 
If NO IMAGES will be shown, use instead of


I have prepared a new blog with wordpress at !!!

Thank you.
A word about 'Spinrite' ... a repeatedly echoed myth from many forum users 
Saturday, April 29, 2017, 01:00 PM
Posted by Administrator
Spinrite is a hard drive diagnostic software, and was first published in 1987. The last version 6.0 was published in 2004. Before I go on with my explanations, remember, Spinrite should be necessary only for MFM and RLL drives, modern IDE drives have their own defect management.

Last time I recognized my IBM Portable PC has at least 110KB clusters marked as bad. But I didn't expirienced problems so far (read or write errors), so I was happy. But even 110KB are space which you could need, so I decided to "refresh" the 20MB MFM hard drive with Spinrite II. After 4 hours only 15% were processed, so I decided to stop it with pressing ESC, which is NOT a hard break, Spinrite finishes it's last action (last processed cluster).
Now I tried to boot again and guess what, NOW I had read errors. Also, the drive seeks desparately for a readable sector often. CHKDSK results in 250KB bad clusters.

This was surprising for me first, because BEFORE I started Spinrite, there was no noticeable error !
Also, I got many defective directory entries, which I could NOT REPAIR with CHKDSK /F.

But meanwhile I understood what happened. For such an old MFM hard drive, this is a torture or at least stress. So instead of doing something good, I tried to kill my MFM hard drive.
A full Spinrite run take many hours, and can't be compared to a normal load of software usually used for such a computer.

So the best idea is NOT USING SPINRITE SNAKE OIL, instead, try to backup what you want first, start low level formatting by using the controllers BIOS (start DEBUG, enter "g=C800:5"), then boot with a working DOS floppy disk, use FDISK and FORMAT C:/S, and restore the saved files back. Last but not least, as already stated, the processing time of Spinrite is far too high - it can take days, not hours. A low level format by using DEBUG take minutes only but refreshs the sectors (re-magnetize them), too.

If in doubt about the drive health, you may use a DOS 6.2 (6.22) boot disk with SCANDISK, to detect unreadable clusters. I recommend NOT TO USE Spinrite, because your old drive will hate you for it. Believe me.

SCANDISK should be called with the parameter /SURFACE and for monochrome screens with /MONO, too.
add comment ( 263 views )   |  permalink   |  related link   |   ( 3 / 1928 )
Got an IBM Portable PC 5155. Really slow, but it's a nice engineering piece. 
Sunday, April 23, 2017, 09:00 PM
Posted by Administrator
This Portable PC is heavy, but it's an IBM PC/XT in a case with a build in amber CRT screen.
It's factory configuration only includes 256KB RAM, an IBM PC floppy controller and a CGA graphics card. The internal screen is feed by an FBAS (Composite) video signal (and not with 9-pin connector).

An original 64-256KB memory card was installed, which results in 512KB in total.
But I managed to add interface cards, and I replaced the original memory card with a Mycomp MEMS memory card, which can be populated with gigantic 2048KB RAM, and this card can offer conventional *and* expanded memory (at the same time).
In the screenshot, you can see the result after it was booted (IBM DOS 5.02 is used).

Believe me, repairing or just changing the cards, DIP switches or memory chips on the mainboard is *NO* fun. Inside the case it's really (over)crowded. But on the other side, in terms of mechanical quality, this PC seems to be very good quality.

Take a look at the 'related link' below, if you are interested in further details.
add comment ( 179 views )   |  permalink   |  related link   |   ( 3 / 1956 )
Amstrad PC1640 on steroids ? Not really, but still a surprise... 
Monday, March 13, 2017, 12:30 AM
Posted by Administrator
Recently I got some problems related with an Amstrad PC1640, the floppy drive was not working properly (formatting and read errors, also when reading disks, formatted on the Amstrad, but on a different PC). So I decided to replace the floppy drive, which is in fact a very unusual, belt driven OEM Alps drive (DFD222A01, Amstrad part no 40046).
The first surprise while opening the PC1640 - I discovered a small additional pcb which was connected with many cables to the area on the mainboard the crystal is located.
There was also a 3-way switch to change the frequency from 4 MHz to ~10 Mhz.
The additonal board and the switch is marked green in the picture below.

Also, I discovered (never used it) an Intel 8087 (marked light blue in the picture) !
This might be useful when using one of the first AUTOCAD versions.

Regarding my wish to replace the floppy drive - this was a desaster. I was reading >John Elliott's Amstrad hardware compatibility page< and hoped that my TEAC FD-54B can replace the original drive.

But you can't replace that floppy drive for two reasons:
a) the metal cage for the floppy drive does not have the correct "holes" at rear...
b) even if you take a saw and make the holes bigger, the new floppy drive still does NOT work

I know the jumper settings has to be changed from (IBM PC compatible) DS1 to DS0 (because of the twisted cables of the IBM PC), but the Amstrad ALWAYS SAYS the floppy controller has problems with the drive. Also, the motor light was always on, which points to DRIVE SELECT problems. But DS1 doesn't work too.
Also, rotating the cable connector ends up also in drive light always on problems but the floppy controller do not show up the error message.
So I decided to check again the original Amstrad floppy drive, turning the motor (with the belt), and also checked the drive head movement. I didn't recognized any problem, so I exchanged the non-working TEAC FD-54B with the Alps OEM drive (the Amstrad one) back.
After reassembling, it suddenly worked without errors. Strange.
add comment ( 187 views )   |  permalink   |  related link   |   ( 3 / 1909 )
Transferring binary files to a DOS PC without anything else but DOS ? 
Monday, February 27, 2017, 08:02 PM
Posted by Administrator
I've a Toshiba T5200 laptop, which has a working harddrive with an installed DOS but a faulty floppy disk drive. There is a way to connect a floppy drive to a parallel port by using the correct adapter, but I do not own this adapter.
So I thought it must be an other way existing to transfer files, like using MS-DOS KERMIT.
Unfortunately transferring binary files over a cross-over cable without any file protocol is NOT possible because of non ASCII characters with a special meaning.
If there is nothing else but DOS existing, you can use MODE COM1 and CTTY at least.
MODE COM1 with additional parameters like the BAUD rate sets the serial port.
CTTY COM1 will redirect the command line to the serial port then.
But how to convert the binary file(s) to an easily approach without any effort for me ?
I remembered DEBUG - you can type in the data of a binary file just by using the 'e' command, and then set then name with the 'n' command. Also, you have to set the file size of it by using the CX register.
I wrote a small C program to accomplish this mission, the program just converts any file smaller than 64KB to a text file, which can be used for a recreation of a binary file just by redirecting the generated text file as input for the DEBUG command.

(btw. redirection does not work with DOSBOX, but on every real or virtual PC with an installed DOS)

Feel free to download the binary and the source of the file by clicking on "related link".

P.S.: You can even print the text file to paper, so you will be able to get it up and running even if all transfer options are gone or if all floppy disks of the world will fail ;-)
add comment ( 166 views )   |  permalink   |  related link   |   ( 3 / 1982 )
What to do if Floppy Drives don't read floppy disks ... 
Friday, January 6, 2017, 04:00 PM
Posted by Administrator
Sometimes you can't read floppy disks in your computer's floppy drive.
It can be related with the floppy disk media itself.
But the drive can be misaligned or faulty, too.
If you suspect the drive to be defect, you can replace it (if the drive is still available or can be taken out from an other (old) computer).
You can try to clean the r/w heads with special cleaning diskettes, or may be with cotton buds and pure alcohol. But this helps only in rare cases.

At least you should narrow down what is not working, and if it helps, do something (see below).

So there is a checklist I created:
a) Do your drive have a "belt" instead of a more modern direct drive ?
b) If so (with belt), check the belt. Does it have still a tension ?

c) if you want to use the drive and it has all power connectors correctly connected, does the LED light up if you try to access the drive (regardless of successfully reading a floppy) ?
d) if the LED does not give any sign, are you sure you connected the data line correctly (beware of the old IBM PC drive select crossover cables) ?
e) if the LED lights up but you can't read any sector, did you tried a different floppy disk or may be also a different data cable ?
f) if the LED lights up during access but even with different floppy medias reading doesn't work, did you try to format (if you can boot up with a different drive) a floppy with that problematic drive ?
g) if formatting a floppy media works, but you can't read this formatted floppy disk on other computers, did you're using the right media (DD versus HD floppy disks) ?
h) is the formatted floppy readable but only in the same drive ?

If yes is the answer for a), but b) can be answered with no, replace the belt with a similar, but new belt.

For drives without belt or with a belt with enough tension:
If you answer c) and d) with yes, continue with the subsequent points e) to h), otherwise you are lost and have to replace the drive.
If f) was successful, but g) and h) was answered with yes, you can try to re-align the drive, it sounds that the drive is misaligned. Just take another already formatted and proofed to be good floppy disk, insert this floppy disk, and try to change the position of the r/w head of the floppy very CAREFULLY. During this procedure, try to read the disk continuosly.
For 3.5" inch floppy disk drives, this will work sooner or later, I've successfully resurrected such drives a few times already.

The shown TEAC 3.5" floppy drive is an exception, because the logic board, see red arrows, is already also a mount for the drive head mechanics.

Another example is this drive from Sony, MPF72C-1, used for Thinkpads as an internal drive:

To move the head, just untighten CAREFULLY the one torx screw which holds the plastic head frame. If you move the head, do it in very small steps. One millimeter is very much.

Real "alignment floppy disks" are to be used with additional eletronic equipment like oscilloscopes. You will not have success just with such a media only.

If you're curious what possibilities exists for reading floppy drives, visit the excellent page from, see "related link" below. Try also >Herb's Floppy Drive info collection<.
add comment ( 185 views )   |  permalink   |  related link   |   ( 3 / 1973 )

<<First <Back | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | Next> Last>>