New to me GRIDcase 3 Plus

[0xDEADBEEF]

BTW this guy did good research on 26pin JVC MFM
https://knm.org.uk/blog/2017/04/the-jvc-26-pin-hard-disk-interface-part-1/
https://knm.org.uk/blog/2017/04/the-jvc-26-pin-interface-part-2/

The JVC 26-pin Interface – Part 3 « MattInX's Miscellaneous Musings

knm.org.uk

 

[Gerry71]

It is not XT 8bit IDE. It is 26 pin JVC MFM. I vot it working with DREM.info MFM emulator after about a week of debugging, rework and 6 firmware updates. Oleksandr now has settings and HDD image with DOS for the next person who buys it . I will be making a post about this adventure later.

Thanks for clarifying.
May I ask if the Drem.info replaced only the JVC HDD or is it connected directly to the Grid motherboard?

 

[0xDEADBEEF]

DREM replaces only the HDD, it has JVC 26 pin connector, it is just MFM drive emulator. So you do need the controller.

 

[Gerry71]

OK, here is my brain dump of gridcase 3p related stuff

MFM controller + DREM3 PCB happy together.
View attachment 1287862

bottom of MFM controller

View attachment 1287863
Component side. It uses common WDC chip
View attachment 1287864

The controller is connected to motherboard with an IDC cable, 34 pin as I counted. I believe it carries partially decoded ISA power.

I have mapped 26 pin floppy connector (next to power supply). In 3P it is directly routed to external DB25 floppy connector using flex cable, basically 1-1 pin mapping. In base gridcase 3 I believe it is directly connected to internal floppy and routed to external floppy too. PDF attached.

I have also attached STL file for 3d printing enclosure to hold together MFM controller + DREM + external floppy connector. (hdd plate no fdd v10.zip). I tried to sandwich it all together with bare gotek board but there is not enough vertical space ;(

Also, HDD image with DREM config (gridcase3p.zip)

And low lever format (llf13.zip) I used to format DREM.

The MFM controller is unique to GRiD. Maybe if you map the connector you could connect some ISA device, there but if it is already partially decoded you may be out of luck. but maybe if it is just IO/data + power and some other ISA signals maybe you can solder someting like XT IDE board there with compact flash?

Also, random JVC drives will probably not work, I think you need specifically JD-S3812M0R0, it has special protective plastic underneath to insulate from controller PCB and possibly other stuff for GRiD.

I think your best course of action is to burn 4 128KB EPROMs or 27C010 equivalent flash chips with GRiD DOS 3.3 and anything you want permanently available and add a GOTEK in place of FDD. unless you find broken gridcase and salvage the hdd controller.

Also, how is your plasma? I have two 3Ps and both have effed up displays with missing lines and some weird pattern spreading on screen when scrolling and so on ;(

Hi,
just back after a busy period. Wanted to say thank you for providing all his precious info!
I'll be back to you regarding the plasma display: I ordered some chips I removed 20 years ago because they got burnt at that time.

 

[0xDEADBEEF]

Both my GRiDcases 3P have similar screen problems with messed up vertical sections and missing horizontal lines. I think the horizontal lines may be caused by bad cable though. Vertical smudging problems are either some controller chip problems or something. And small vertical line problems may be even in glass substrate.

Also, 3P available right now on eBay has the same display corruption (last pic) so it seems this is a common problem.

 

[0xDEADBEEF]

I took a pic of the display back.
The panel is PD640G200BA-104A and i believe it is made by NEC.
There are some similar panels PD640G200BA-BC panels on ebay for $600ish but that is very expensive and would probably require parts transplant. If the panel itself is bad we are looking at soldering like 2 thousand wires

 

[NeXT]

Wait, so on the plasma models the glass envelope is connected to the driver sustain board by soldered points and not ultrasonically welded to ribbons?

 

[0xDEADBEEF]

15XX 640x400 plasma uses flex ribbons
3P 640x200 plasma uses from what i can tell 640*2 + 200*2 metal wires between panel and PCB. As you can see on the picture they are through hole, because why not? the wires look like steel, quite stiff, they are covered with adhesive cloth tape to protect them

 

[MrImprovement]

Both my GRiDcases 3P have similar screen problems with messed up vertical sections and missing horizontal lines. I think the horizontal lines may be caused by bad cable though. Vertical smudging problems are either some controller chip problems or something. And small vertical line problems may be even in glass substrate.

Also, 3P available right now on eBay has the same display corruption (last pic) so it seems this is a common problem.

I am not an expert but from what little I know the good news is that the panel itself looks to be in good shape! However either the cable that connects to the display or some part of the display controller needs to be looked at. Thankfully the panel is good as that is the part that can not be easily repaired.

 

[0xDEADBEEF]

Not sure, the pattern changes somewhat when I press slightly close to edges of the panel, maybe it affects the pcb underneath?
I know for sure some of the missing vertical lines are because of the cable, they reappear if i move around display cable with the computer open. So I definitely will test or try to make new display cable as well.

 

[NeXT]

Finally had a chance to bring the system in and work on the screen in an ESD-free environment.
So the panel is indeed soldered to the board, however it has a wide cable attaching it, so pressing on the screen would of never located the bad connections in this case.

If you look closely in that photo at the closest Y sustain chip you might notice what isn't right. A number of the legs have developed cold joints and are just floating in the breeze. I'm almost willing to bet that this is our missing horizontal lines, but I forgot my AC adapter and will need to repair and test it tomorrow.

 

[0xDEADBEEF]

All right, this is interesting observation. If just reflowing it fixes stuff it would be awesome. I am more concerned about messed up vertical columns though.
Also, I think sometimes there is an issue with ever 2nd horizontal line caused by the cable from the motherboard.

 

[NeXT]

Went under the iron today. To my surprise while inspecting the rest of the board I found a lot of places where both the wave and surface-mount soldering was not great. Many of the wires running to the panel have voids at their solder joints.

It's making me wonder how many other panels where I've seen missing rows or corrupt columns are because of soldering issues. Anyways, after carefully reflowing the pins I was happy to see all my horizontal lines came back and I now have a completely healthy plasma panel.

 

[Stanton]

If true (soldering issues), hard to believe that (what I thought was) a high-quality product had such wide-spread manufacturing issues! While your (nice) equipment setup makes it easier, it's not like this isn't something that could be done to all those "bad" screens out there (although disassembly isn't that easy either).

 

[0xDEADBEEF]

What does make you think this was a high-quality product? It was expensive because of expensive components and being unique in the market. Also, since these were first laptops they were making up a lot of things as they went. But they made a lot of silly engineering decisions in my opinion.
The right hinge base is a disaster, the support for display catches inside the display is a joke, cooling and airflow is nonexistent, blocking of cartridge power supply pin in 100MB models (is it because of heat or power reasons?) is just downright silly. SIPPs sockets are so close you cannot fully populate them unless you find SIPPs with blobs instead of chips.
It is surprisingly serviceable though compared to some stuff from 90s.
And these things still work for tens of years, solder does not get better with age, especially with thermal cycling (this display does get warm, and guess what, no vents anywhere).

 

[NeXT]

GRiD has a MASSIVE misconception that they were rugged military-grade products, partially because GRiD UK's entire market segment today is that.

I mean, you got the magnesium alloy case, you have the high visibility gas plasma electroluminescent display options, you have external peripherals using GPIB/IEEE-488 rather than conventional buses, you have movie and television appearances and you have styling. You pick your GRiD up and drop it from the height of a table, it will break. Case, screen and hard drive. Poof, there goes a giant wad of cash. Almost like it's not remotely ruggedized?
Seriously, look at the features and interfaces on a retail COMPASS or GRiDcase, then look at later or even period correct mil-spec equipment. Basic stuff like screen padding, spec compliant ports, dust boots and debris ingress protection on the keyboard. The GRiD can't hold a candle to it until you start looking at the very special custom models they built.
Bubble memory? GPIB? Their own custom OS and applications? In retrospect they quickly switched to MS-DOS and floppy drives once they realized this was not going to make them a leader against competing portable PC's

In the 80's that I can tell, GRiD's market was a visually high-end laptop with gimmicks that was aimed mostly at executives who waned a VERY nice looking laptop on their desk. People who looked at the Compaq Portable and then saw this, for only $7000. GRiD however accepted custom orders so if you wanted additional functionality, interfaces or even TEMPEST compliance, it was available but it was not part of the main offering but that small slice of the pie is what gets referred to the most.
They are indeed very nice machines asthetically but like 0xDEADBEEF said, they have major design issues.

 

[0xDEADBEEF]

I followed @NeXT footsteps and resoldered all driver chips on the back of two of my failing GRiDcase 3P screens. Both were cured. There is still little bit of weirdness showing up but the difference is incredible.

before



after