Print

[anotherLISAguy]

You guys did such a great job in pointing out differences in the TwoPort card earlier, that I thought you could help with this curiosity. 

Was reviewing some eBay images online and noticed two different types of Lisa 2 motherboards (not XL). While I understand different revisions may have different layouts and electronics, these have or don't have a series of resistors! 

The first image is a Lisa 2 motherboard with resistors installed, the second image is an earlier revision motherboard without resistors installed!
This strikes me odd since the PCB clearly identifies locations for electronics where none exists in the second image. 
It seems a bit more dramatic than the clipped R47 mod and I don't recall any 'that motherboard doesn't work with X' info.

Is there some functional difference between these motherboards?

TIA

[sigma7]

Those are damping resistors on the parallel port to reduce the signals ringing, which affects the reliability of the ProFile.

The original Lisa 1 Motherboard design doesn't include those resistors; the holes used for the retrofit are vias and so not actually intended for installation of any parts; they were just the most convenient place to install the resistors.

The 2/10 Motherboard design includes the resistors, as does the Sapient version of the Lisa 1 Motherboard.

[ried]

I have always wondered about this, and it's nice to finally know the answer. The left-most resistor has been added on one of my spare boards, too.

Photos for reference:




And one with no resistors at all:

[anotherLISAguy]

Those are damping resistors on the parallel port to reduce the signals ringing, which affects the reliability of the ProFile.

The original Lisa 1 Motherboard design doesn't include those resistors; the holes used for the retrofit are vias and so not actually intended for installation of any parts; they were just the most convenient place to install the resistors.

The 2/10 Motherboard design includes the resistors, as does the Sapient version of the Lisa 1 Motherboard.

AH - so the initial L1 design had hole placement in place as a potential modification (for dampening), and later L2 builds (not L2/10/XL) added them to ensure dampening - or do I have that backward - either way, both motherboards are functionally equivalent except for the latter having dampening added to that port

Thanks for explaining that curiosity!

[stepleton]

AH - so the initial L1 design had hole placement in place as a potential modification (for dampening)

Not exactly... Apple got lucky. The holes were already there, but they weren't for components: they were vias, ways of connecting a circuit trace on one side of the board to a trace on the other side. But a component leg fit inside these vias, and they happened to be in just the right places, so they could stick in resistors without having to come up with a new board design. This kind of low-price modification is sometimes called a "bodge".

[anotherLISAguy]

AH - so the initial L1 design had hole placement in place as a potential modification (for dampening)

Not exactly... Apple got lucky. The holes were already there, but they weren't for components: they were vias, ways of connecting a circuit trace on one side of the board to a trace on the other side. But a component leg fit inside these vias, and they happened to be in just the right places, so they could stick in resistors without having to come up with a new board design. This kind of low-price modification is sometimes called a "bodge".

Got it - the vias were in place to connect across layers, and leveraging them with resistors to dampen the signal ringing was a bonus 'bodge'.
So what was the effect of signal ringing in ProFiles - extra read/write errors or actual digital corruption?

Thanks...cuz' you know inquiring minds

[AlexTheCat123]

So what was the effect of signal ringing in ProFiles - extra read/write errors or actual digital corruption?

Interestingly enough, I've never noticed any ProFile issues on boards that are missing the resistors. ArduinoFile prints all of the ProFile commands that it receives to the serial console, and I always watch them fly by whenever I'm running the Lisa off one of my ArduinoFiles since it's really fun to watch. But in all the time I've spent watching it, I haven't once seen it report any errors, other than some timing stuff that I've since fixed and that was completely unrelated to signal integrity.

So I wonder how much of a difference the resistors actually make? I guess they have to be somewhat important, given that Apple put in the time and money to add them, but I don't really see the neccessity!

[blusnowkitty]

I have a real ProFile disk that behaves strangely depending on what port it's connected to during POST. If I have the drive connected to the built-in port, it will work correctly. If I have the drive connected to a parallel card port, the ProFile gets confused and tries to seek the heads past end of disk during the boot device identification phase of POST. I wonder if there's some connection between resistors and no resistors? I'll have to check my backplane later.

Or I think someone here said that the parallel card VIAs were clocked a little bit faster than the built-in port...

[anotherLISAguy]

So I wonder how much of a difference the resistors actually make? I guess they have to be somewhat important, given that Apple put in the time and money to add them, but I don't really see the neccessity!

It may have been some earlier stress testing that flagged the issue, the L1/2 today probably would never reach those levels by collector/hobbyist.

[AlexTheCat123]

I have a real ProFile disk that behaves strangely depending on what port it's connected to during POST. If I have the drive connected to the built-in port, it will work correctly. If I have the drive connected to a parallel card port, the ProFile gets confused and tries to seek the heads past end of disk during the boot device identification phase of POST. I wonder if there's some connection between resistors and no resistors? I'll have to check my backplane later.

Strange! I wonder how it's doing that? If you try to get a ProFile to access a block that's out of range (0 to 25FF on a 5MB ProFile), the ProFile should refuse to do the seek and set some status bits to tell you that the block is out of range. So it's really odd that yours is actually trying to seek to it. I've never seen that happen before!

I would be inclined to think that it's probably got nothing to do with the resistors though, and more to do with the differences between the parallel interface on the card versus the I/O board.

It may have been some earlier stress testing that flagged the issue, the L1/2 today probably would never reach those levels by collector/hobbyist.

Good point. Maybe it's just a precaution to cover some rare edge case that they occasionally saw during testing.