• Please review our updated Terms and Rules here

SE/30, wobbly horizontal display, worsened by HDD activity

T

trondl

Experienced Member
Joined
Dec 9, 2019
Messages
57
Location
Oslo, Norway
Hi from Norway and thanks for your time!
To verify some parts from my other SE in service (SE vcfed thread), I thought my SE/30 would be a perfect candidate.
Not so easy. This became a catch 22.

First of, the SE/30 has previously been recapped and thoroughly washed (5 years ago?), corroded solder vias and the nearby chips with dull colour has been reflowed just now.
All is in theory good except for when previously testing the SCSI drive with Lido 7 i get "OSErr= 5 bad scsi command: phase error".
The disk is "fully" accessible though according to the sector reads, apart from trying to format it. This is incredible in it self as this drive was opened to remove gooey rubber years ago.
This I hope I fixed by the recent reflowing of the SCSI controller (dull solder).

Now, the HDD is another problem for later.
What I can see during boot is an intermittent horizontal "collapse" on both sides about a couple of "pixels" in width.
This is even more present when testing said HDD in Lido 7.

My immediate thought was ripple from a shared supply line between horizontal drive and HDD.
I see that there are two 12V lines from the PSU, one for the "Disk" and one for "Sweep" which I guess is the CRT.
First line of attack was the analog board.
Solder inspected for cracks, caps verifiable in circuit verified and unverifiable had one leg freed and verified with ESR tester.
All within reasonable spec.
Next attack was the PSU it self.
All solder points verified, caps on secondary verified-ish out of circuit but measure in the lower uF side of within spec.

I've read that a wobbly horizontal may be a failing HOT, but since this is directly affected by HDD activity, I have my doubts.

Any recommended lines of attack I should consider?
Don't have all the caps for the PSU or analog board in stock atm, but I fear the PSU caps have to be replaced as they're all "weaker".
 
I have seen this, recap the analog board and it should stop. When it was recapped they may have only reworked the motherboard. Almost every Mac from this era has capacitor issues.
 
The capacitors in the power supply are notorious for leaking. Additionally, they sometimes have the horrible yellow glue that absorbs moisture and becomes conductive. If the capacitors are those horrid brown crapstain Nichicon caps or ELNA, they need to go. You generally don't have to replace the large main line capacitor though, they rarely go bad. Did you inspect for wetness under the capacitors? You should also check the base of the capacitor where the legs come out. You'd see clearish drops of electrolyte or green corrosion in the cups where the leg poke out.

I had a customer's Mac SE several years ago get a case of the death wobbles before the BU406 went thermonuclear and caught on fire.

I40ogYZ.jpg

A0xpgxp.jpg
 
Thanks for input!
Clarification, the horizontal "collapse" or jitter, is about even across, not "noisy" as in the horror image just posted.
Looks just like in this video:
In that clip, the HOT is the sinner, but in my case worsened by the HDD is what get me confused.
The HOT shares what with the HDD? 5V or 12V?
My guts say the 12V line but not too much knowledge on CRT circuits yet.
The caps I think may be worth investigating on the analog board is: C11, C25, C13, C14.
My initial attack will be to investigate these closer, especially C14 (4700/16)
Red herring warning: could the case fan cause these oscillations?
Are HOTs easily tested with a transistor tester, or are they "weird"?

Both SE and SE/30 have the Sony CR-44M PSUs.
SE/30 PSU secondary has blue Rubycons and two larger black ones I cannot find the maker of in circuit.
No brown "cement" (I've seen the horror before in old Japanese synthesizers. Always removed) at the base, just the white "chewing gum" on top kind.
No wetness found, except for the two smaller caps (470uF) close to the ceramic resistor.
The wetness was leftover flux melted by desoldering (80% sure).
The q-tip was light brown as in flux, and smelled like 80s flux, not fish (the smell of the Toshiba T1000 laptop with very leaky brown Nichicons is still fresh in mind).
Again, they're all "in spec" ESR and capacitance wise, albeit lower half.
 
Just did a "full" inspection of the digital board. Found no direct shorts between the smd chips.
Also did what I should do immediately when recapping, lift the Molex socket as it is near the caps.
Corroded pads indeed.
The FDD and HDD is starting up, so 5V and 12V should be somewhat present.
Should the CRT generate a raster without the digital board fully initialized?
The tiny crystal near the main square crystal, is it vital to initialization or just the RTC clock?
It was a bit corroded.
 
I don't think the analog board starts without getting a video signal from the logic board.

The CRT has its own +12v rail called +12VSweep independent from the +12v rail the rest of the system uses, at least in my SE with the Sony power supply.

If you end up recapping the analog board, DO NOT replace the horizontal output capacitor with a normal capacitor. It's in the picture above, the largest cap between the coils and above the yoke connector. This is a special extremely low ESR capacitor that isn't made anymore. If you replace it with a normal capacitor, it will explode and potentially take out other parts of the analog board. These capacitors don't often go bad, so it is probably fine. If it is bad, you'll need to use a microwave film capacitor. These are a bit annoying to fit because they're larger, but they work.

I don't think that you could find any small faults with a BU406 with a simple transistor tester, you'd probably need a curve tracer and an oscilloscope and another known good BU406 to compare with. It'd be less expensive to just replace the BU406, as that test equipment is hundreds of dollars.

But it sounds like you have more damage on the logic board that needs to be cleaned up. If the board doesn't boot at all anymore, you likely have broken traces. Worst case, you can buy a new logic board and use the current one as a parts donor.
 
Likely broken traces yes. Too many dull vias and pads I had too touch up. Will have to dig for a while now.
Preliminary symptoms in current state:

BU406: 0.0Vdrop between Base and Emitter both ways in circuit (pin1 and 3).
Smoking gun? Shorted to ground? A failure here shouldn't prevent a boot?
Considering it was already "wavy", reflowing/heat may have finished it off.

Reset button causes a click sound.
80% sure there is no CRT glow at base (not dark enough room?).
Verified 5V and 12-ishV at HDD yesterday.
HDD wouldn't spin up first time today. Quick power off/on x amount of times started it.
Fan spin.
Did a quick check on PSU input on analog board: looks like 5, 12 and -12 is present and stable.
 
Short addition: I had forgotten to solder the lifted leg on C25 (between "12V Source" and ground) on the analog board.
Unsure if this could have caused BU406 to fail or other faults inbound.
 
Pretty sure the BU406 shouldn't have a B-E short, it should be replaced. I would not power the analog board up anymore with that transistor failed, very bad things will happen.

I would also get rid of that hard drive. If the motor can't spin up, it may pull heavy current and drag down the power rail. It can also cause the motor driver on the hard drive control board to blow up. If it's a Miniscribe drive, the heads may be getting stuck to the platter and causing stiction.
 
Did a quick desolder of the BU406 before trying anything else.
So it appears that it isn't shorted, something else is...
Transistor removed, something is still shorted between pin 1 and 3.
 
You must log in or register to reply here.
Back
Top