• Please review our updated Terms and Rules here

MTI MLV11M Controller with MFM

gnupublic

Experienced Member
Joined
Jan 23, 2019
Messages
198
Location
Germany, Berlin
Hello,
the summer is heating the town, it's too hot for the minis in the flat, so I tried something with mico PDP11/73.

My PDP11/73 came with an MTI MLV11M controller together with a Rodime RO204 disk as bootdisk. The rodime had problems from the beginning on with booting, but when it came to temperature after 20 minutes, it did. I was able to save the RT11 on it in 2019.
Meanwhile the Rodime has passed away after a longer time period of getting more and more errors. So I wanted to replace that disk against another MFM disk.
I did some googling, but did not find any informations nor documents for the controller how to do it. Just 2 things:
- someone was (long time ago) searching for a file (rl02dg.sav) for this controller. Nobody replied.
- And somebody said that this controller would emulate multiple RL02 for the system.

On the RT11 of the Rodime, I found the file rl02dg.sys. So I started this file with "r rl02dg.sys" and was happy this is the program for testing and formating disks with this controller.
So I did. The controller only formats, if ECC is disabled and I took some time to find out the jumper for ECC. There is now way to configure the disk in any way, you can just format it. When this is done, you can do an ".INIT/BADBLOCK DL0:" in RT11 and use this disk. It's 10MB big like expected for RL02. So with a 20MB MFM disk I expect 2 RL02 disks. DL0: and DL1:.

But if Itry to access DL1:, the System hangs. No matter if I do an INIT or DIR. I tried with different disks up to 40MB, but always the same. When accessing DL2: or DL3: I get an error message, but DL1: always hang.

Disks I used: ST-225, ST-125, NEC 3142

MTI_MLV11M.jpg

Does someone here has further informations for this controller? Any hints?

Have fun,
Volker
 
I wish that I knew more as according to my notes I have one of these boards in my stash. AFAICS from the markings in your photo my board looks to be the same as yours. Will be following along with fingers crossed that you figure it out :-}. Did the original Rodime RO204 setup support DL1:? Was that part of your backup?

In the meantime could you please post a copy of rl02dg.sys? No other Rodime or MTI information on your backup?
 
I have the same controllers, but no software or documentation. I can try to play around with them, it would not be nice to get this program first - rl02dg.sys, since I have not found anything on the board yet. As a last resort, I can even disassemble it and see what it does.
 
I cannot remember if I ever tryed DL1: on the Rodime. Sorry.
I will attach the file later this day when I'm home, or maybe tomorrow. (Family event today).
How to attach? As RX50 diskimage?
 
I'm guessing the problem is the controller is built only for the Rodime drive. Which has 8 heads and 320 cylinders. The ST225 has 4 heads and more cylinders.

My guess is the controller puts "DL0:" on heads 1-4 and "DL1:" on heads 5-8. Since the Seagate doesn't have 5-8 it's not going to do anything right when accessed.
 
That's the reason I tried the NEC 3142 also. STASON shows it as 664/8/17. Same behaviour.

Do you think that a third party company builds a controller just for one drive?
 
Does someone here has further informations for this controller? Any hints?

Have fun,
Volker


Volker,

i do have some MTI controllers also and there is also a little stack of documentation i did not look into up to now.
So there may be a chance that i will find something over the next weekend.
If yes i can post some information...

What about setting up an mfm emulator with the rodime config ?

Best regards :)
 
Volker,

i do have some MTI controllers also and there is also a little stack of documentation i did not look into up to now.
So there may be a chance that i will find something over the next weekend.
that would be helpful
I just noticed I have nothing on the MLV11 under http://bitsavers.org/pdf/microTechnology

also I notice hack42 listed having a manual, but a simple search isn't turning anything up on the internet archive where they've been dumping their scans

cleaned up and un-bloated pdf of the diagnostic manual on bitsavers now
someone may want to ping them about scanning the hardware manual
 
Last edited:
Hello, here are the three most interesting drivers I found on the disk. I extracted the files with putr using copy/bin, from the backup file I did from the Rodime.

I made an empty RX50 floppyimage, did an INIT in simh and "cp a copy/bin" to save the files on the floppy image.
There is the rl02dg.sys file that belongs to the MTI MLV11M controller also I found some other drivers for DQ615 and DQ614, that might be of interest for others.
The RT11 ist nothing special, but full of letters and text with names and addresses. Not for common interest.

Hopefully I did it right and the files are useable.

Thank you Al for your prompt reaction!

Have fun,
Volker
 

Attachments

  • Archiv.zip
    67.8 KB · Views: 6
That's the reason I tried the NEC 3142 also. STASON shows it as 664/8/17. Same behaviour.

Do you think that a third party company builds a controller just for one drive?
No but they may load ROMs specific to the drive. Remember MFM is stupid and doesn't abstract head/cyl/sec like IDE does. The fact that it's emulating an RL02 also points to static mappings.

I have and use an MIT ESDI card in my 11/83: It's extremely fast, does MCSP with burst mode DMA, and you hook up a serial port to some of the pins to program the drive specifics. Maybe there is a serial port on your board somewhere?
 
The sizing was done by simh. I just attached a nonexisting file "at rx0 rx50_drv.dsk", did an "init dx0" and "copy/bin" the files. I can mount that file in simh and run the program, so the binary should be OK.

According to the documentation from Al, there is a dialog in the program, I misunderstood.
There is a question for the number of devices (1-4), then following by the question if the device is attached, followed by the question if it is used.
This happens as often as you answerd the number of devices.

That let me think it should be the number of physical devices. But it is meant as the number of logical RL02 devices.

With the NEC 3142 (8 heads) I was able to answer 2 and then get DL0: and DL1: working. I am able to do an INIT on both.
But this is not possible with ST-225 or ST-125. With this disks (4 heads) I can only manage to get one RL02 working. And when accessing the second, the system hangs.

So I would guess any RL02 needs at least 4 heads? And the controller is expecting drives with 8 heads. The Rodime has 8 heads. Maybe the parameters for the Rodime are in the the ROM and the NEC is the same in heads, sectors and has more cylinders. So it works like the rodime with some unused cylinders.
So again, 50% loss of usable space.
 
Ok, so my theory was right. Makes sense.

The problem isn't that RL02's have 4 heads (they don't of course), it's that the mapping of RL02 commands to the MFM drive itself follows specific hard coded rules in the controller. This looks like a simple one relatively speaking so it would need to do quick translations. Since they had a drive with 8 heads and a very low track count (it sounds like it is from the RD50 vintage) they went with mapping the first two heads+all cylinders to "RL02 unit 0 head 0" and the second two heads+all cylinders to "RL02 unit 0 head 1". Then the next 4 heads did the same thing for the "second" RL02.

Remember that RL02's, RL01's, RK05's, and even RM02/03/05/80 drives used addressing explicitly by head, sector, and track. Later controllers would use "go to block xxxx" which is oddly enough more like how a DF32 worked (where you specified a sector "address" and the controller figured out which head and where). So anything that uses that format is probably going to be a simple controller that you would upgrade by changing the ROMs.

Enjoy the system!
 
Back
Top