Weird Rom Board

ge_WeirdRomBoardDeutsche Version dieser Seite

Some time ago I bought a bunch of plain vanilla Newton stuff. Thought I...

The ROM board described here caught my eye as soon as I opened the package. Its dimensions are exactly the same as those of the ROM boards in the Newton 2000 / 2100 and eMate models, and it has the same connector.

Click on the images to get a higher resolution image in a new browser window.

Weird ROM board bottomRight under the copyright notice we find the text "Q FLASH BOARD". “Q” was the code-name for the 2000 / 2100 units. Well, well, well...
As fast as my trembling hands allowed, I ripped one of my Newton 2000s apart and plugged the board in.
The good news were that said Newton didn’t explode when I switched it on. No smoke, no excessive heat, no flames, no smell, nothing. Unfortunately, no signs of life, either. For as long as the board was installed, the Newton stubbornly feigned death. As soon as the standard ROM board was plugged back in, the Newton powered up just fine, obviously unaware of the recent experiment.

Standard 2100 ROM board bottomIn comparison to the weird ROM board the standard ROM board of the Newton 2000 and 2100 has no chips on this side.

Weird ROM board topMy weird ROM board shown from the other side. “EVT”, as shown in the bottom line of the white sticker, stands for “Engineering Validation Test”, which is the first of a sequence of new-product development stages. EVT units are test units used to find and fix bugs, tweak the design etc.
Typically the following stages would be “DVT” (Design Validation Test) and “PVT” (Production Validation Test).
This side is populated by three chips labeled
E28F016SA., which are 16 Mbit Flash ROM chips made by Intel. Together with their brother on the opposite side they provide the board with a total of 64 Mbit or 8 MByte. As the standard ROM board in the 2000 / 2100 models has a memory size of 8 MByte, too, there is no chance to increase the available amount of ROM by using this board.

Standard 2100 ROM board topThe standard ROM board has two chips made by Sharp on this side. They are labeled LHME5BT3 and LHME5BT4. I have spent the better part of three hours out there in the web trying to find a datasheet for these beasts, but in vain. Most likely they are mask ROM devices whose customer-specific content is programmed at the end of the manufacturing process. This would also explain why the final version of the operating system lives in only two chips as opposed to the developer board’s four. Mask ROM devices are much higher integrated than Flash ROMs.
Someone told me that this ROM  board looks basically like an enlarged Sharp copier ROM board. If you have one of these copier boards (dead or working) or documentation thereof that you would be willing to donate for scientific history research, or if you can point me to a datasheet for the LHME5BT3 and LHME5BT4 beasts, please
drop me a line.

eMate Screen with weird ROM boardWell, what would the life of a real Newton fan who is said to have marginal hardware knowledge be without the occasional challenge. If this board couldn’t be persuaded to power up a Newton, maybe it would be more accommodating when plugged into an eMate. Which is exactly what I did, with the result shown in the picture.
What we see here looks suspiciously like the first of the two warnings one gets after a hard reset, rotated 90°.
I can "tap" both buttons, but this requires placing the stylus nowhere near where the buttons actually are. I discovered this by drawing horizontal and vertical lines in a grid-like pattern until the buttons reacted. The buttons in both message boxes are highlighted simultaneously, which is not surprising because they only look like two message boxes, but are in fact only one.
If I tap the bottom button (which is probably the "OK" button), two new message boxes appear that strongly resemble the well-known second warning one encounters during a hard reset. Tapping "OK" again restarts the eMate, which then shows the first message again. No matter what and in what order I tap, the eMate always get back to this message, it never finishes booting the operating system.
Both the image and the behavior allow some conclusions, though. Obviously enough of the operating system is loaded to allow the eMate to sense and react to pen input. The display works, at least kind of, which also requires a fair amount of the operating system to be loaded.

Time to do some thinking. Hmmm...

Why did this board come up in an eMate, but not in a normal 2000? The fact that the 2000’s code name is printed on it suggests that it was intended for this model, and there is a lot of further evidence. Since the screen content is rotated 90° on an eMate, wich always powers up in landscape mode after a hard reset, this ROM board must have been developed for a device that, like the 2000, powers up in portrait mode.
The eMate does not react to the keyboard. Not surprising if the ROM board was designed for a keyboard-less machine like the 2000. Neither does it chime after the hard reset, which would also make sense for a board that wasn’t to be used in an eMate (whose volume is controlled by a a mechanical slider), but in a device with software volume control like the 2000.

A possible explanation is that the 2000’s original design used the same RISC CPU as the eMate: The ARM710. Later the 2000 was redesigned for the SA-110 CPU . This assumption is backed up by what Mike Culbert, a former Newton hardware developer, once said in an interview:

“The MessagePad 2000 design was already under way using an ARM710 microprocessor. However, once we had StrongARM silicon in hand, it became clear it was a viable alternative."

In conclusion it seems likely that my weird ROM board came from an early (note the (C) 1995 date on the board) EVT prototype of the Newton MessagePad 2000. A prototype that had a motherboard based on the ARM710 CPU. It seems also likely that this prototype had a screen resolution different from the 2000’s 320 x 480 pixels. As there are two message boxes where there should be only one, with almost identical display content in both halves of the screen, it can be assumed that the operating system in this ROM board expects half the horizontal screen resolution of an eMate. This would be 240 x 320 pixels, which happens to be the screen resolution of the Newton 110, 120 and 130 models.
This would also explain why the screen content looks somewhat blurred. Maybe lines 1, 3, 5... are used for the left and lines 2, 4, 6... are used for the right half.
But who knows. Maybe it's  not a resolution problem at all, but the mere fact that the display control electronics of the eMate is not being correctly initialized by the ROM.

I would gladly try out some some other experiments with this board, but I am currently at a loss as to which kind of experiments this could be. If you have an idea what I could try, please
let me know.

Dead links? Questions? Anything unclear? Any syntax or grammatical errors in this description? Feel free to tell me about it. Yes, really. Don’t be polite, be helpful. If you aren’t, how am I supposed to improve my English?

Nothing like that? This page really helped you? Wow! What a perfect reason to sign my guest book...

[Home] [Site Map] [Software] [Hardware] [Books] [Stories] [This And That] [FAQ] [Cry For Help] [E-Mail] [Guest Book] [Legal Info]