amiga-news.de - AmigaOne Board: Status Report by Martin Schueler

AmigaOne Board: Status Report by Martin Schueler
As promised on the BAS2001 (see First report) Martin Schueler has sent us a summary of the lecture about the AmigaOne board and technical details.

Status

The lecture mainly was similar to the one already lectured on the WoA, but of course contained updated information regarding the time schedule. Important about this is that the beta test will close middle of September, which at the same time will mark the starting shot for the first line.

Moreover the new revision of the AmigaONE boards - plugged into an Eyetech tower - was shown. One was able to tangible ("forcibly") convince oneself of how sophisticated and solid even the mechanical solution was done, and that it is by no means some unstable kind of tinkering solution. But we deliberately used a board only quipped with PCI/AGP sockets, to prevent our electrostatic sensible testing boards from any possible damage.

Following the status report from which I took the answers on respective questions, on the show:

During production the first revision formed short cuts on several points of which some were irreparable. It was not possible to obviate the latter during production without knowing the problem in advance.

For the new revision we had changed the respective points, but also several other points, in a way that reduced production requirements. Furthermore we corrected mechanical reference measurements, which prevented the first revision from being precisely assembled inside the casing.

Also the non-technical problems were fixed and the testings efficiently proceed on schedule since June 8th. It is especially joyful that all of the testings (contrary to common expectations due to complexity) succeeded on the first go, which was finally a result of their carful preparation, too.

All significant testings on board level are already finished. The following testing work is limited to verification of the logic within WestBridge and NorthBridge and it's interaction with the standard components.

This is to talk about ca. 10% of those testings are done. But virtually this means that already 50% of the problematic and time critical points are covered. Nevertheless those succeeded on the first go. So schedule risks are more and more decreasing. Furthermore 100% of the problematic points are functional of which malfunction would mean total outage, in worst case.

I also added three pictures:

picture_1.jpg
(640 x 480)

picture_2.jpg
(640 x 480)

picture_3.jpg
(640 x 480)

The first one shows the rev2 board (equipped with PCI/AGP only) as it was presented on the show in Rotterdam. But the other one shows one of the equipped rev2 boards as these were used for the testings. All of the testings succeeded on the first go using the board in the picture. Third picture: The slots in this picture are simply put on, cause I think it looks much nicer if it is fully equipped :-).

For the rest the board will be tested as stand-alone for a very long time. The connection to the A1200 will be done at the very end.

Martin Schueler
(ps)

[News message: 25. Jun. 2001, 18:21] [Comments: 0]
[Send via e-mail] [Print version] [ASCII version]

< Next message

Prior message >