Asante MacCon SE30 affecting video

[Elfen]

This is not a UE8  or related chips issue. Here's why. If it was a UE8 problem you would have bars everywhere on everything - icons, open windows, filenames, etc. This is what your screen would look like (excuse my POOR Photoshop Skills!)

See how it interferes with everything...

Instead, you got this. Again, Excuse my poor Photoshop Skills...

I had to go through the forum, and remembered this... You (and everyone here saying its a UE8 Error) should read it - all 2 pages of it!
https://68kmla.org/forums/index.php?/topic/23040-se30-jail-bars-llllll/

In thinking because it is not effecting other things but just the desktop, I come to the conclusion that it is an System/OS error/conflict with the Ethernet card. Why it is happening, I would not know without having it in front of me and giving it a once-over.

And it's not the Video RAM either - the 2- 4464 chips on the left side of the board near UE8. If it were your video screen would be really trashed.

All I can say is this:

1) Get a Clean Minimal System on a floppy that the SE\30 can boot from. (Do not have the ethernet card drivers on it.)

2) Disconnect the hard drive (the power cable, not the SCSI cable)

3) Boot with the floppy.

4) See how it boots with and without the ethernet card connected.

If you can, have a boot floppy for System 6 and System 7 to do this test with. I can bet that this is an OS issue and not a hardware issue.

 

[Elfen]

Just had a crazy idea...

When you do this test, when you have the ethernet card out, look at it's PDS Connector. There should be 3 rows of pins on it. I'm thinking - you might have a bent/broken/missing pin. If so, this could be the problem.

 

[TimHD]

Not crazy. I did check it, even took a picture with my iPhone 6, though I didn't upload the picture to the site because (i) it looked fine as I could site all the three rows of pins (female on PDS looked fine too) and (ii) was having such dramas with getting hi-res images on this site (how do you do it?). In any event, here it is... https://68kmla.org/forums/uploads/gallery/album_134/gallery_5459_134_1041542.jpg

Speaking of images, we're back at UE8...

Again, just to preface things, this board was recapped by Maccaps and passed all the test (was fine before it went in too, the recap was mostly preventative and the SE30 was chiming ok well before this 'fix', this was not a battery leakage victim, just old before recapping. Some Bourn filters were replaced, but otherwise, passed the burn in tests fine apparently.

So, here's some images of the UE8 on my SE30 board. 

Front: Large

Rear: Large

Direct Links:

https://68kmla.org/forums/uploads/gallery/album_134/gallery_5459_134_947058.jpg

https://68kmla.org/forums/uploads/gallery/album_134/gallery_5459_134_3215868.jpg

Welcome if anyone can see any issues...

 

[Elfen]

Good to hear about the Pins...

UE8 looks fine from the photo graph and it is not the problem. I'm not defending MadCap's work but he does great work in bringing dead and dying machines and he's even fixed a couple of my machines as well.

Did you read my post (last post on Page 1) before this first past on page 2? Here's the Link:

https://68kmla.org/forums/index.php?/topic/26506-asante-maccon-se30-affecting-video/?p=287650

Basically, if it is a UE8 issue, then there would be vertical bars every where and the pattern of the bars would depend on what data bits are effected on the VRAM. Your SE\30's problem is this is happening only on the desktop, not through the windows, menu bar or icons. A UE8 issue would have bars through the windows, menu bar and icons. You should sit down and read this thread entirely and you will see that it is not a UE8 issue:

https://68kmla.org/forums/index.php?/topic/23040-se30-jail-bars-llllll/

So this narrows it down to a System/OS error or conflict. System/OS errors or conflicts are hard to figure out. Once it is figured out, it would be easy to fix, but figuring out what it is - that's the issue. UE8 is just a DUMB Signal Latching Gate Array that grabs the bytes from the VRAM to display it on the screen. There is no intelligence or CPU Control in UE8 for it to discriminate where to put or not put bars through the windows, menu bar and icons. That takes software which UE8 has no control over. This is why if you have an issue with UE8, you will have bars all over the place and not just in the desktop only.

Lets look at the SE\30 board. Right under the 68030 is the Video ROM (usually on a socket). Rock it up a bit on both sides with a screwdriver or  IC extractor and press it back down into place to break any corrosion that might be in the socket.

Any other problems? What kind of ROM SIMM is on the SE\30? A SE\30 (24bit dirty) ROM SIMM or an updated IIfx or IIsi (32bit clean) ROM SIMM? If it is a 32bit clean ROM SIMM, that might be the problem. Sure, it will give you 32bit access to RAM but it will also make your SE\30 think that it is something else (a IIfx or IIsi) and not an SE\30. This can produce issues with the System OS and hardware. "About this Mac" on the higher versions of System 7 should tell you what Mac it is on depending on the ROM's Gesault number. If you have a different ROM in your SE\30 board, it will show up as a different Mac on those versions of System 7. Any card expecting a SE\30 seeing that it is in another machine because of a ROM SIMM change will throw up issues, no matter how slight or crazy. If you need 32Bit Clean access to RAM, you can use the Mode32 Control Panel and Extension with the original SE\30 ROM SIMM instead of an updated ROM SIMM from a IIfx or IIsi.

I'm not denying that your SE\30 has a problem, for it does. But it is not hardware with UE8 or the VRAM. This is a Software/OS & Hardware issue conflict. The question is "why." And this is going to be a very hard problem to figure out.

 

[Bunsen]

(how do you do it?)

Or you can upload it as an attachment (use the full comment editor from the "More reply options" link below the reply box) and then (optionally) add it into the post.

 

[uniserver]

https://68kmla.org/forums/uploads/gallery/album_108/gallery_5459_108_1856616.jpg

this picture is only 1 of 2 things.   1. a trace or via that has opened up connecting the vram to the UE8 i/c

2. UE8 its self.

In the picture it looks like I replaced UE8 already... so this means that maybe the flexing of the board when inserting the PDS card or some flex from the pds card being installed is causing one line to disconnect from the Vram to UE8.

This can be checked by using a multimeter with the se.30 schematics and doing a continuity check pin to pin.

Might want to check those lines with the PDS card installed as you said this issue is only apparent when pds card is installed.

 

[techknight]

I dont think so. Look at the Mac OS Startup, and the extension. No bars. 

It sounds as if the expansion card is conflicting with the same slot-address as the video. 

 

[uniserver]

ah then we need to both give Elfen some credit....

because you are correct. and so is he.

Cudos Elfen !    

 

[apm]

Board flexing is a good thought. A good way to test that would be just to push down gently on the board without the card inserted and see if the problem reproduces itself. uniserver is right that a broken trace is more likely than a dead UE8 (which does look to be new). 

It is highly likely that either UE8 or the circuitry around it is involved. It will not be an OS or software issue because you see it even on the earliest power-on screen. Edit: Elfen and techknight are correct! I had misremembered the photo as being of pre-OS loading but it's not. You can still see some wrong pixels in the MacOS screen (see below) but it's not impossible it is just a software issue. The way to be sure is to boot to a flashing ? disk and see if the bars remain. Anyway, I've kept the rest I'd written before:

The behaviour you're observing is that every 8th pixel is black if the one immediately following it is also black. That has to have something to do with the shift register UE8, because the first pixel of the pair is bit 0 in one group of 8 pixels, then the next pixel to the right is bit 7 in the following group of 8 pixels. Bit 0 comes from UC7, bit 7 comes from UC6, but the value of one bit affects the other. The only way that could happen is where they come together, i.e. shift register UE8 which converts them into a serial data stream.

Now it may not be as simple as a broken trace from the VRAM (then you would see the jail bars all the way down the screen) but there are lots of other, subtler things that can go wrong. For example, there might be a problem with the SREGLO signal coming into pin 15 which controls when UE8 loads each group of 8 bits. If that connection has gone flaky, or if there is something wrong in the PALs which generate it, then the shift register might load data at the wrong time compared to when the VRAM wants to send it out, and that could produce unexpected effects. Alternatively, maybe the pullup resistor on the CLR and SER signals (pins 9 and 1) has failed and that signal is just floating. I'm not sure exactly but I would start checking continuity with the schematic, which you should be able to find on this forum somewhere. Look for open circuits and also possibly shorts between pins.

Some particular traces to check (by no means exhaustive):

UE8 pin 2 <--> UC7 pin 2 (VID0)

UE8 pin 14 <--> UC6 pin 23 (VID7)

UE8 pin 15 <--> UG7 pin 14 (SREGLO)

UE8 pin 9 <--> UE8 pin 1 and RP8 pin 8 (PU2) <--- this goes somewhere else too on a schematic page I don't have

UG7 pin 15 <--> UC6 pin 1 and UC7 pin 1 (SC) <--- should read 22 ohms because of R23

 

[Paralel]

I agree with the Elfen, I could easily be some strange OS issue. Stranger OS issues have happened before.

 

[Elfen]

ah then we need to both give Elfen some credit....

because you are correct. and so is he.

Cudos Elfen !  

Thank you! Now I can walk through the silicon gates proudly now that the Mac Gods have smiled on me!

Board flexing is a good thought. A good way to test that would be just to push down gently on the board without the card inserted and see if the problem reproduces itself. uniserver is right that a broken trace is more likely than a dead UE8 (which does look to be new). 

It is highly likely that either UE8 or the circuitry around it is involved....

I would love to agree with this except for 1 major factor prevents me from doing so: The SE\30's display is fine without the card as stated on the very first post:

Anyone know why my Asante Maccon SE30 ethernet card is causing this striping to appear on the boot screen? (there's also the occasional weak interference to the screen image too).

Jumpers on card set to default, SE30 boots fine without the ethernet card installed.

If this was a UE8 or associated chips problem, then it would be happening with and without the card.

But something TimHD stated there - a weak interference on the video screen. That suggests to me that the PSU is border line - giving out just enough power to keep things going. Sounds like it needs a recapping. I would suggest the Ethernet Net card to be recapped too. Have you checked the 5V output from the exterior floppy port? This way the 5V from there has travelled and shared with the rest of the system and should still be around 5V (4.9V at a minimum). Anything less than the minimum, the PSU needs to be adjusted or recapped. I forget what pin is 5V on the floppy port but that should be easy to find with the experts here.

Lets try something crazy - boot with the floppy drive disconnected so there is less of an electrical strain on the PSU. If it comes up OK without the floppy drive, then its the PSU being borderline and needs to be adjusted and recapped.

I worry though that this is still an System/OS & Hardware issue.

 

[Elfen]

I agree with the Elfen, I could easily be some strange OS issue. Stranger OS issues have happened before.

For a while I thought this was Mac Virus doing something crazy like this but I never heard of any of the 12 or so Mac 68K viruses or the 65 virus variants with such behavior. The only two suspects I can think off is mVIR and WDEF, but off hand they don't do this. And of all of them, only 3 or 4 can sense that it is on a network to infect connected volumes which mVIR and WDEF does not do.

 

[TimHD]

Hey, thanks everyone. Am feeling I am asking a lot but it's one of those 'it seems so close' that I don't want to give up on it. It's also totally baffling (at least I think it's not a simple stupid mistake that I'd be embarrassed about!). I'll try a few of those ideas over the weekend and report back (to give you guys a little break from my questions too!!). Rest assured it's appreciated at my end. Thanks again all

 

[techknight]

ah then we need to both give Elfen some credit....

because you are correct. and so is he.

Cudos Elfen !  

Never paid attention to his posts. Most of them are too long, and I hate reading long posts. Keep it short and sweet.

anyways. There are 2 sets of jumpers on the right of your card. Change them around, thats your slot-address. 

 

[Paralel]

IDK if its necessarily a virus, it could easily just be damage to the system files that can't be detected by the built-in checks that are used to detect system corruption.

 

[uniserver]

haha...  don't worry Elfin, I think you are pretty great buddy.

Charles

 

[ktkm]

Any luck on this? I have almost the exact same problem, with the same card as mentioned, but with horizontal flickering lines that show up after approximately 30 minutes from booting!

 

[Jinnai]

Three year necropost...

 

[ktkm]

Three year necropost...

I thought they would have solved it by now!?  ;-)