SE/30 various issues

[cb88]

I bought an SE/30 that if I let sit awhile or gently addjust the video board it will will boot up from floppy.

It was sold as not working but when I got it it booted up from SCSI to 7.1 and had MS word installed on it a few reboots/days later it slowly degraded to no SCSI device found and then to SimasiMac.

I have it partly taken appart so far.. SCSI drive is out and I tried connecting some of my other drives (a 2GB barracuda, a 4Gb quantum and a 143Gb SCA with an adapter for kicks) ended up with the same issue it could be that I am not using the patched drive setup? The original drive was a 40Mb model and it still sounds perfectly good (I hear it spin up and hear the heads move) when booting up the mac just can't see it.

Now I have gotten variations of SimasiMac.. from grey to horizontal bars. They don't always occur.. seem to be less likely if I turn it of a bit adjust the analog board and try again sometimes there is a delay between the video appearing and the system coming up. It also eventually locks up after awhile of running even if it does startup correctly.

I do have a new PRAM battery haven't put it in yet since I don't have the logic board out yet. from what I can see with the board still in there it looks clean with the exception of all the caps being hot glued. Also the transfomer at the top of the analog board has hot glue on it.. no glue on video board.

I've booted 6.0.8 and 7.1 from floppy... and can have my PC setup to modify some floppies for testing.

I don't have a way of grounding the CRT so don't want to do much till I take care of that? what would be the safest way to do that... I hear turn brightness up and jerk the mains out of the wall ... or make a grounding kit with a screwdriver and a the center pin of a power cable being sure it does not ohm between the other two pins as a cheap cable would.

 

[Macdrone]

Recap the logic board, the issues your talking about start there. Even if the logic board doesn't look a mess it is. I promise. If you count the threads youll see they are mostly bad caps over and over. Start with logic board, then to analog if needed. A lot of people here do repairs on their own or send it to people on the list. Im 50/50. I just cant see well enough to do small things so I send away for help.

 

[cb88]

Took the logic board out.. it looks quite clean as far as my untrained eye can tell... took out the battery had what appears to be a date on it GL/24/88 ... so its a miracle it hadn't leaked in the slightest. It boots up still... still no SCSI. Restarted it and it booted up fine to a couple different floppies still. So I haven't damaged it further at least

I did take the ROM out... it looks like its at stock SE/30 ROM. It looks like a have 4x 1Mb ram simms and 4 256k as that is the only combination which would add up to 5Mb right?

 

[uniserver]

You can take a picture of the main board and post it if you want.

IF it has not already been re-capped it needs it.

 

[nvdeynde]

Recap the logic board for start. All SE/30 logic boards need to be recapped even if the capacitors still look good, believe me they are not.

Also keep in mind that the SE/30 doesn't have a strong PSU so the drives you are connecting may be out of specs for the PSU and drawing too much power.

Termination can also be an issue with newer drives. Make sure all jumpers are set correctly on the HDD.

Try on of your hard drives in another compact mac so you're sure they are not the problem.

If after recapping the SCSI still doesn't work then you're probably gone for hours of testing traces and VIA's.

The cause are often broken VIA's, however I had one case where the SCSI chip was faulty.

Keep us posted

 

[cb88]

Indeed... I've read up on the issues in detail. I've already contacted uniserver to get 2 of my boards recapped. As I told him im waiting on the hobby money to reaccumulate since I just got a SparcStation 20 (dual sm81's, 3 pcimcia cards and 2 quad fast enthernet cards)

I'm aware the SCA drive is too much for it... it a rather hot drive note as bad as some though.

The drive original drive in it was only 40Mb I doubt it was pulling too much also tried a 2Gb barracuda... also a relatively hot drive and a cooler 4Gb quantum. I do wonder what the exact effect of the bad caps are.. I guess mainly most noise in the system so things act up.

 

[Macdrone]

Bad capacitors do not allow proper voltage to be maintained. It cannot be stable without them. Imagine its a crazy type of battery.

 

[cb88]

lol indeed... BSCpE and half of an EE degree lol ;-) so I should hope I know what a capactor is

I was refering to what exact effects it has in the system... hypothetically speaking that cause the problems and by noise I mean the signals degrade enough that they don't jive with the on off voltage of some transistors somewhere on occasion.. because it does work *sometimes* but there is something going on during the system startup that makes it at least more susceptible to these problems.... still happens while the system is running though eventually.

And trust me you still learn all about bypass capacitors in EE lab ... that and don't use the crappy breadboards in the kits they gave us!

 

[Macdrone]

I am sorry, I wasn't taking a shot at you there sorry.

 

[cb88]

No need ... after all I am the guy that ordered a DB25 to SCSI3 cable when he needed a SCSI-II cable.. X.x for my 2GB JAZ drive

Oh and in replacing various car door handles messed that up about 3 times... cursed mild dyslexia and confusing photos!

 

[Macdrone]

Well overall people who fix things rarely use directions, I do the same.

 

[cb88]

Not that either I have 2 manuals for the car one is hardback even they had photos of the opposite door handle for each one they were selling turned the wrong direction sigh... I did just order the *correct* cable I hope! I'll keep the other for a later date ...

 

[James1095]

Many of the capacitors serve as decoupling for various ICs on the board. When a logic line switches states, it is fighting parasitic capacitance of the PCB traces and whatever load that line is driving, as well as parasitic capacitance internal to the chips themselves. The effect is that the chips tend to draw small surges of current each time anything switches states, which obviously happens many many times a second. Without these decoupling capacitors, the resistance of the traces feeding power to the different parts will fluctuate with each of these surges and the parasitic inductance can cause unwanted resonance, leading to a great deal of random noise riding on the power rails. The capacitors serve as localized storage to absorb these spikes, "decoupling" them from other parts of the circuit. It can be thought of as similar to water towers located in various places around town. They help maintain a steady water pressure at the individual neighborhoods despite variations in demand by creating a local reservoir, isolating the area from pressure variations in the main line.

The other common application of those problematic SMT electrolytics is coupling and DC blocking in the audio amplifier circuit. The audio signal is passed through a capacitor, which passes AC but blocks DC, removing the DC offset and resulting in an AC audio signal. When these fail, the audio becomes soft or vanishes.

 

[uniserver]

The capacitors serve as localized storage to absorb these spikes, "decoupling" them from other parts of the circuit. It can be thought of as similar to water towers located in various places around town. They help maintain a steady water pressure at the individual neighborhoods despite variations in demand by creating a local reservoir, isolating the area from pressure variations in the main line.

That is awesome explanation!!! You should teach community college or something part time.

 

[cb88]

That really is a good explination even for those that know what decoupling capacitors are... you explain why they are there. And also why the audio fails ... that was puzzling me (I can still just barely hear mine with headphones lol).

 

[Mk.558]

All the more that points to bad caps near the C3 thru C6 area. Some important SCSI traces run near C8 thru C10.

...and yes, that was actually a pretty clear explanation of decoupling and blocking caps. :approve:

If you can't fix it right away, I would wash the board with 99% alcohol and suspend the board so the caps are pointing towards the floor, as to minimize cap damage, if any. (Not sure if that would help or not, but it's nice to think about.)