• Please review our updated Terms and Rules here

Screwed up my IMS 8000

Chuck(G) said:
8" drives always use a 500Kbps data rate; no one that I'm aware of uses 300K or 250K, which are data rates for the smaller format media.
Click to expand...

Doesn't the TestFDC utility see the 8" drives as 1.2MB 5.25 drives as set in the BIOS? It runs tests at low density/data rate and then high density/data rate. It's testing the drive controllers ability to read/write at the full range of modes. Mine passes all the tests normally.

I do admit that I'm not all that well up on data rates etc for these drives. When adjusting the main settings in IMD and the settings it asks for to read/write/format a disk, I use what info the IMS8000 manual has about disk formatting plus a Wikipedia table that lists most if the 8" floppy format specifications.

So I should be setting IMD to 500kbs FM in all cases?
 
Last edited:
Thanks for clearing that one up, takes some of my guestimates away! I'll have to read up properly on the history and specifics of these drives. Herb's retrotechnology site has a lot of info - gets a bit too detailed at times though.
 
Like anything else, history is the best guide.

SA800 8" floppies were early on the scene and spec-ed for 500Khz clock and 360 RPM.

When 5.25" floppies were introduced, the medium couldn't handle the packing density of the 8" drives (think linear track length), so the spindle speed was lowered to 300 RPM and the data rate halved to 250KHz. 35 tracks (later 40) allowed a bit more "slop" in the positioning mechanism. At about the same time, 100 tpi (77 tracks) and 96 tpi (80 tracks) were introduced by reducing the track width.

1.2M floppies were originally introduced as an exact replacement for the 8" drives. Different media formulation and better heads did it. 96 tpi, 360 RPM and 80 tracks meant that you could drop a 1.2M drive in without changing signals. But these early drives were high-density only--no 360K/720K capability. Later, the drives could be switched electronically to work with the 5.25" low-density de-facto standard. However, the spindle still spun at 360 RPM, not 300. There were two answers to that problem. The first was to simply record "360K" media at a slightly higher (360/300) data rate. Sort of a Mohammed-to-the-mountain approach. The other was to make the drive spindle speed switchable between 300 and 360 RPM, so that only two (250KHz and 500KHz) data rates were needed.

Some 1.2M drives support either approach (e.g. Teac FD235GF), depending on jumpering.
 
FIXED IT! Nothing to do with faulty UARTS or drives, the resistor pack on the cpu board was the wrong way round! I must have desocketed, tested every IC on this machine 4 times in the last 2 weeks, even bought spares for the chips I can't test. 1.30am and finally I notice the writing on the resistor pack isn't upside down like its supposed to be!!

Anyway, back in business. It passes its memory tests now too which is a new one. Must have been that last 2114 that tested bad.

Now id like to know why a movcpm results in lots of nonsense characters and the system locking up. Works in every other way but I can't regen CP/M to a new disk.

Oh well, I'm just pleased the thing works again
:mrgreen::happy4:
 
Movcpm contains a complete copy of CP/M (as well as the BIOS). If for some reason, the CP/M copy in the movcpm.com file isn't the one you're using, the effect is the same as if you'd booted someone else's system.
 
I should probably have added earlier - movcpm used to work. The manual advises that the IMS system disk comes factory configured as a 32K system. It says to issue the movcpm ** command so it will build a system based on the max amount of memory it finds and then run the IMSGEN command (their version of sysgen) to take the memory resident system and copy it to the first disk boot sectors of a disk placed in the same drive (swapping the factory disk with a formatted disk you want to make bootable)


Previously, it would run the movcpm part ok most times but the sysgen would always fail with a disk error or parity error. Now all my RAM seems to be passing tests, perhaps this is new behaviour? If there is an incompatibility between the movcpm on the disk and my system, perhaps I should go through all my disks and try all the instances of movcpm I find (there are many) and see what works?


More to play with anyhow.
 
Good? Naa.

Remember - I have a fully socketed system, a chip tester - both help one to bypass the skilled method of using logic probes etc. With my setup its S100 repairs for dummies! Couldn't be easier and it still took two weeks to find an inverted component!

Lessons learned:

- Dont get fixated on a particular line of enquiry. My issues began at the same time i found I had an IC on the I/O board the wrong way round - immediately, despite finding no fault with all the testable chips, I assumed I had damaged one of the untestable UART/PPI/Buffer ICs or worse still, the 2708 EPROM.

- Photograph / Scan ALL S100 boards front and rear - there is then no doubt regarding jumper settings, IC positions / unpopulated socket positions etc

- My quick n dirty method of 'diagnosis' has its drawbacks, besides teaching me nothing about digital logic theory and circuitry. Repeatedly removing and replacing all the ICs increases wear and damage to the sockets, indroducing other harder to pippoint issues plus getting blase' about the whole affair and making silly mistakes.

On the upside - I have:

- bought more spare parts for my IMS8000 including a spare UART IC, 25 of one type of interface IC and 4 of another (they were dirt cheap) (i forget the types)
- Learned that i can flash an EEPROM with the IMS CP/M bootloader AND/Or the IMS TurboDOS bootloader and use it in place of the MM2708q with an adapter (JonB's)
- Learned about cutting traces and soldering in jumpers to change the factory default 2708 setup. I can now switch between the 2708 / 16 / 32 EPROM ICs with the jumpers I've added to the board. Interestingly, I also discovered I can use single supply voltage 2716 / 32 EPROMS. Some of the new jumpers set for triple or single rail power - triple is only for the 2708, the 16/32 use a 5V setup - so can i use TI2516s or Intels single rail 2716s?
- Learned all sorts of other stuff along the way - disaster brings the mind to the manual!

Thanks all for your assistance :)

PS - Superbrain? Having fun with S100 for now :mrgreen:


JonB said:
Well done mate.

You're getting good at this. Superbrain next?
Click to expand...
 
Last edited:
My new friend:

WP_20190411_08_43_04_Pro_LI.jpg

I've had it for years - scan the boards, scan all documentation.

Time to get it in the attic and use it!
 
That's a good point. You have put in proper joists haven't you? A typical loft floor (upstairs ceiling) has thinner (weaker, cheaper) joists as it only has to support the weight of the ceiling.
 
JonB said:
That's a good point. You have put in proper joists haven't you? A typical loft floor (upstairs ceiling) has thinner (weaker, cheaper) joists as it only has to support the weight of the ceiling.
Click to expand...

Nope - 148 years ago, the builders of my house decided to go with 3 x 2's I have since put a chipboard floor which added a lot of strength, it also hangs off the roof beams round the edge of the house. i tend to put heavy stuff over the middle wall area!
 
You must log in or register to reply here.
Back
Top