Help Fund the Amiga 4000 Replica Project!

[tbtorro]

Originally Posted by grelbfarlk

Why not just throw the cash at Hese instead? I mean I'd rather support work that's already done by an Amiga guy and when he releases his design people can "degrade" it back down to a Rev B if they want.

You can't just "degrade" a full, complex, digital computer motherboard from one revision to another. The Revision D board has considerable differences and routed entirely different than Rev B. It would be equivalent to a redesign, with all the work involved. Remember that the main goal of this project is to replace battery damaged boards. 99% of all boards are Rev B and not CR.

And forgive me but paying a professional firm, with production grade equipment and software, and several trained engineers who do this for a living seems a much safer bet for a good result, and completely accurate, and correctly made board than "throwing the cash" to one man hobby project (however competent, we have no idea if hese is a PCB layout professional). And his board is manually routed which is basically a different layout than the original, and it's unknown if it was done considering EMI elimination, noise, crosstalk, correct ground plane design, signal propagation etc.. There is much more to PCB layout than just connecting pads and vias, it's a whole science in itself.

There are modern PCB analysis techniques that guarantee fault free data, they'll run full test point analysis on the board to trace out all (or most) signals first, after which probably a combination of xray, and laminate separation (where they remove the substrate one by one to expose the next copper layer) for full mapping. This is a known automated process and the yield of these are close to 100% correct layout data, and the routing will not be done fully manual. So the result will be a near 100% replica of the original Rev B board, and wouldn't you agree that the original C= engineers knew much more about correct PCB design for their own hardware, than any "amiga guy" today (inlcuding myself)?

I would like to stress that by no means I want to belittle hese's work, which is wonderful, and by all means welcome to all of us in the community, all I'm saying, I'd love to see guaranteed professional grade boards. There is place for both Rev B and CR.

[tbtorro]

Originally Posted by grelbfarlk

Not to mention, I doubt Commodore auto-routed the board back in 1991, the original was probably hand routed.

The usual process is a combination of both. Partial autoroute, combined with manual design retouches at problematic areas.

[tbtorro]

Originally Posted by idrougge

Exactly how is "hand-traced" a bad thing? Is automatic routing supposed to be better?

Yes and no. But for example there are design rules you can set in the layout software to avoid problems like noise, crosstalk, bad signal propagation etc.. which are automatically taken into account during autorouting so human error can be avoided or minimized.

[grelbfarlk]

It just seems to me a needless duplication of effort if Hese's board already works and is done, that was my point and to support an Amiga developer.

Take any Classic Amiga hardware project in the last 10 years did they hire a professional design company to layout the PCB?
Not mostly, with some exceptions.
Does this bother anybody?
Well I guess so, now.

[tbtorro]

Originally Posted by grelbfarlk

It just seems to me a needless duplication of effort if Hese's board already works and is done, that was my point and to support an Amiga developer.

Take any Classic Amiga hardware project in the last 10 years did they hire a professional design company to layout the PCB?
Absolutely not, maybe with a couple exceptions.
Does this bother anybody?
Well I guess so, now.

I guess you can't please everyone. If after all the explanation it's still not clear to you what the benefits are then.. Whatever..

[grelbfarlk]

Originally Posted by tbtorro

I guess you can't please everyone. If after all the explanation it's still not clear to you what the benefits are then.. Whatever..

I am pleased that there will probably be two A4000 motherboard PCB replacements. Then there's Elenachen too. Even more pleased if any of the designs get enhanced with new stuff or improvements.

[Acill]

Tbtorro is correct about the dissection process here as well I donated my working A4000 motherboard to this process and its being used for the process. If you dont like the process I chose you have the option to disagree. I can guarantee the reason this hasent be done before is because of the cost. I am getting a significant discount on this project. The reason I went local is because of the ability to get it, be able to check in on the progress, and to have a US manufacturer with the ability to produce boards if a group buy gets requested as well as having the ability to get the board redesigned if also requested.. This pays a full time person dont forget, and delivery is estimated at 4 weeks.

In the end we will have two great options as a result. One newer style that a lot of people didnt get to own, including myself and one that is the most common and prone to failures being available to use as a 1:1 direct replacement for those of us that are doing repairs. I can tell you without a doubt I would rather use a new board and do a transplant to fix one then the time consuming fault finding I do now!

[Shadowfire]

Originally Posted by tbtorro

Yes and no. But for example there are design rules you can set in the layout software to avoid problems like noise, crosstalk, bad signal propagation etc.. which are automatically taken into account during autorouting so human error can be avoided or minimized.

In any professional package, using the manual point-to-point routing protocol will automatically generate traces compliant with your constraints, or at the very least, flag a DRC error if you violate them. It's not something you can miss unless you're so bad that you (a) didn't set up the constraints, (b) willfully ignored the errors, or (c) didn't bother running DRC. Unless they're using something which DOESN'T have constraints for capture/layout/routing, in which case, yeah, they probably shouldn't be doing a board this big with that kind of tool.

[IanP]

Modern boards are a science with their GHz signals, low voltage logic with thousands very fine traces and vias across many layers requiring professional tools. The a A4000 motherboards are 25 year old technology probably built with tools far more primitive than what's available to the hobbyist today.

[tbtorro]

Originally Posted by Shadowfire

In any professional package, using the manual point-to-point routing protocol will automatically generate traces compliant with your constraints, or at the very least, flag a DRC error if you violate them. It's not something you can miss unless you're so bad that you (a) didn't set up the constraints, (b) willfully ignored the errors, or (c) didn't bother running DRC. Unless they're using something which DOESN'T have constraints for capture/layout/routing, in which case, yeah, they probably shouldn't be doing a board this big with that kind of tool.

Exactly, and that's why I rather trust a pro firm the have proper rule checks in place than a hobby project.

[Acill]

Amiga on the Lake has generously agreed to help us get this last bit of funding a bit easier for those that want to use PayPal for a donation. If you would like to donate but only have PayPal to use you can donate here and they will in return donate on your behalf to the campaign. Thanks so much to Aaron Smith for helping to offer this!

The donation link is here: http://amigaonthelake.com/aotl-donations/

[Acill]

Good news! I will be going to the shop on Wednesday to make the initial payment and have work started! The goal is to have the board work done and complete in 4 to 6 weeks. The goal is to recreate the entire design. They will provide everything they create, including the PADS and Altium design databases.

The donate page can be found here: https://www.gofundme.com/amiga-4000-...ith-schematics

Also dont forget you can have Amiga on the Lake donate for you if you want to use paypal and dont have a card supported here. http://amigaonthelake.com/aotl-donations/

[alexh]

Originally Posted by tbtorro

Remember that the main goal of this project is to replace battery damaged boards. 99% of all boards are Rev B and not CR.

Understood. The Rev D boards were designed to be partially populated so couldn't a Rev D board work as well as a Rev B board?

Originally Posted by tbtorro

I would like to stress that by no means I want to belittle hese's work, which is wonderful, and by all means welcome to all of us in the community, all I'm saying, I'd love to see guaranteed professional grade boards.

Understood. Could it not be cheaper to get them to validate hese's work? If he open-sourced it? Ran all the checks you mentioned?

Good luck to the project. I don't want a board but I would donate a bit of money to help preserve the data.

[SnkBitten]

Almost there....

Only $453 left to go

[Acill]

Today is a amazing update day for this project! I am happy to say that I put in the remaining balance myself to get things moving at the PCB design house. They have the original board to use, all the required information and the most important thing, they got paid! What this means is things have STARTED!!!!!!

The delivery items to us are even better now thanks to the significant speed at which we paid them, and the extra work I did to supply as much as possible to help with the creation of everything. So what will we get you may ask? Well EVERYTHING!!! The project databases in Altium, PADs and hold on......EAGLE!!!! All Gerber's, new matching schematics that go with the new layout done in tandem, BOM sheets, manufacturing guides, and things I missed I am sure. So this will allow us to not only get NEW BOARDS, but once I ahve it all organized and hosted someone can take these design files and improve on it or whatever you need!

I want to thank everyone that helped make this possible. It has been a lot of stress, work and back and forth communication with the design house, but I think in the end it will be an amazing achievement! This is a real first at this level for ANY Amiga community project and you should all be very proud! I know I am.

[grelbfarlk]

Congratulations, this will be very helpful to the community.

Now do something about replacement cases.

How about just finding a commonly available in great quantities, ATX case with a motherboard tray that will fit, then making a whole bunch of replacement trays?

[amigasith]

This is awesome, Acill

I was already following very closely what hese did with his A4000+ Alice motherboard, but what I was really waiting for was a Rev B motherboard replica. Hats off to this great project Now I can't wait to get mine

[Mr.Flibble]

Nice! Very happy to hear about Eagle being included now as it opens up porting to KiCAD in the future too

[Acill]

Originally Posted by grelbfarlk

Congratulations, this will be very helpful to the community.

Now do something about replacement cases.

How about just finding a commonly available in great quantities, ATX case with a motherboard tray that will fit, then making a whole bunch of replacement trays?

I believe we have a line of people waiting for this to get completed that are going to start converting the design over to ATX form factor. It will open the door to exactly this sort of thing though. As far as finding case, its easy to get a tray made if you want to spend the money to get one laser cut.

[Mr.Flibble]

Some other ease of use changes will be nice to make too, like converting the board to take a standard ATX power connector over the old AT style one for example, or fixing the design issues with the audio circuit