Micron Xceed Color 30 fail

[Bolle]

Nice error chasing going on right there.

Good to see the schematics were kind of useful and I didn’t waste all those hours buzzing them out. :tongue:

I would go with jumpering the signal until you run across a replacement part.

There still are switching ICs but you would have to look at datasheets on how they handle video (or higher frequency for the matter) signals.

I used some switching ICs to switch analog audio signals and they were fine for that... gotta look up what chip I used exactly and what frequency range they support. Also totally not sure if they would end up to be pin compatible with the Maxim chip.

 

[Crutch]

Oh yes, I was remiss in not saying BIG thanks for the schematics ... I would have gotten nowhere without them.

Yeah I looked around but it looks like all the new analog switching ICs on Digikey/Mouser now use another package (not DIP-14) or have entirely different pinouts due to supporting more channels (plus to your point not obvious if they'd work for these video signals).  However I found someone in China selling the exact Maxim IC for $5 plus shipping on eBay.  I just ordered one, hope it's legit otherwise I'm following your advice and jumpering the signal.  Thanks again!!

 

[unity]

Im just thrown this out there. Have you inspected the back of the Micron board? I have a couple Micron boards and one would not work. Tried all sorts of stuff until I noticed a through hole lead was bent and pressing on another causing a short. The leads on both my boards should have been cut shorter, many are way too long.

 

[Crutch]

Thanks unity. I checked that (would have been good to have an explanation for the gradual failure of the board!) but the leads look OK, if slightly on the long side. 

 

[Crutch]

Many thanks once again @Bolle for all your help above. Today thanks to your assistance I replaced the Maxim video switch IHC5341CPD with a replacement IC I found on eBay, shipped from China but looks legit. After soldering it up, my Micron Xceed Color30 with grayscale adapter instantly works 100% perfectly. Attached is a photo of my repaired card, I had replaced U12 (Maxim chip)  and (earlier, because it was overheating) U1 (yeah you can tell I didn’t have a very steady hand when I surface-mounted my replacement IC at U1, but all the contacts are good and secure). 

I feel pretty lucky to have a working grayscale adapter that didn’t cost me $2000. Now for an exciting game of Tetris Max ... 

 

[JDW]

@Crutch I too have a Micron Xceed Color30 PDS video card and the official grayscale adapter.  I've noticed a horizontal "twitch" on my internal CRT (with the screen with decreasing and increasing very slightly (a couple millimeters only), yet quickly and somewhat noticeably).  I am thinking if could be the electrolytic capacitors on the grayscale yoke board that need to be replaced because when removed the Xceed video card and grayscale card and put the stock yoke board back in, I don't see the twitch anymore.  Have you noticed this horizontal twitch on your grayscale setup?

Another question...

To properly discharge the CRT, you are suppose to put the ground lead of the discharge tool to the ground lug in the upper left corner of the CRT and NEVER to connect it to the metal chassis, as that could zap the motherboard.  However, the Micron Xceed grayscale adapter splices the ground wire that leads to the CRT upper-right corner ground lug to the chassis ground!  In other words, the wire harness coming from the back of the Grayscale adapter connects chassis ground to the CRT upper-right ground lug.  I actually discharged the CRT a couple times without thinking.  I don't see any ill effects, but perhaps the bleeder resistor in my machine had already discharged the CRT before I manually tried it.  I would think it potentially could be a problem for the logic board.  Have you discharged your CRT with the motherboard still connected?  I'm really curious why they connect that CRT ground lug to the chassis ground.  It's a bit unsettling.

Thanks.

 

[unity]

The ground system in all connected as one.

 

[Crutch]

@JDW I haven’t noticed the jitter you mentioned but I will look closely tonight. I also thought the ground system was all connected as one - indeed I just checked and the original Pina volume actually recommended discharging to the chassis near the motherboard (photo, p. 12). (The second Pina, covering the Classic and SE, suggested using the screw near the CRT—but didn’t say why.)

 

[JDW]

@Crutch  thank you for checking the screen jitter. I look forward to hearing your feedback on that.  My jitter is not persistent but rather comes and goes. 

Even the official Apple service manuals for the SE/30 mention specifically to discharge to the ground lug of the CRT in the upper left corner and not to the chassis.  

Because some service technicians were accidentally grounding to the metal frame of the SE/30 instead of that CRT ground lug, they were zapping their motherboards. And for that reason Apple came out with a revised discharge tool that had a resistor inside — the thinking being that the resistor would perhaps attenuate the voltage spike that was killing the motherboards in cases where the CRT was accidentally discharged through the chassis instead of through that CRT ground lug. And that is the entire basis for my concern about the Grayscale adapter shorting the CRT ground lug directly to the chassis.  They obviously did it for a reason, but I want to know what that reason is in light of the CRT discharge implications of presents.  None of the official Micron documentation that I can find mentions any dangers associated with CRT discharge grayscale board attached.  But that doesn’t mean there isn’t a danger. 

 

[Crutch]

Thanks @JDW. This is interesting to me - how does the CRT grounding lug at top right get to ground (I just looked at mine and still have no idea)?  The chassis is connected to the grounding screw of the PSU which is in turn connected to the grounding plug in the wall outlet (or, if you're working on the machine, should ideally be properly grounded through a clip that's grounded to a wall outlet).  If the chassis is grounded to the wall outlet, is it really possible to fry the motherboard when discharging the CRT to the chassis?  (I could see this being a danger if the chassis isn't grounded to anything, like if a technician is just working on it loose on a tabletop.)

 

[JDW]

You are NOT supposed to discharge the CRT with the power cord connected to the wall.  This is mainly a safety precaution because if you discharge foolishly and somehow flip the PSU switch with the cord connected, you're in a world of hurt!  So always disconnect that power cord before you discharge in the normal way, or just use Professor Tom Lee's trick (Stanford University school of engineering, and author of the Classic Mac Repair Notes) and turn up the display to full brightness while powered on and yank the power cord, which discharges it quite nicely.

READ THIS about why NEVER to discharge to the chassis ground.

The ground lug in the upper left corner of the CRT attaches to the metal belt that wraps completely around the rim of the CRT and is not directly connected to chassis ground.  Not all grounds must be the same ground.  How then does the CRT get discharged?  Well, think about a capacitor.  You can discharge it without connecting its ground lead to ground, right?  Right.  You need only short the leads together and you get the spark (never try that on really high capacitance values though or the cap might blow up -- 1000uF or below is best), and the spark indicates the capacitor discharged, just like the spark or pop you sometimes hear at the CRT cap when you discharge the CRT.

So because that CRT ground lug is NOT connected to chassis ground (and in turn, NOT connected to earth ground via the 3-prong power cord), there is no voltage spike that would travel through chassis ground and potentially zap the motherboard.  (READ THIS, CRT Safety - 31)  HOWEVER, the Micron Xceed Grayscale adapter affixes to the CRT Yoke and it's single ground pin splices two thin black wires (why so thin relative to the stock wire, I do not know), one wire leading to the CRT ground lug in the upper left and the other wire leading directly to chassis ground, which goes against everything I've heard about what you should or shouldn't do in a compact Mac.

To be honest, I've been a bit afraid to run my Grayscale setup without the chassis ground connected (which is an important test, I think) simply because Micron did that for some reason.  I doubt disconnecting that one ground would blow up the yoke board because the other black wire still leads to the upper left CRT ground lug, but due to the rarity of these boards, I've been too chicken to try it.  Perhaps someone braver than I will step up to the task.  So for now I can only wonder and take care to remove the PDS card and completely disconnect and remove the motherboard BEFORE I discharge with the grayscale adapter attached.  It worries me.

I look forward to hearing your test results about CRT jitter.  Again, my eyes are seeing only HORIZONTAL expansion and contraction, and that movement is only a couple millimeters so it isn't hugely noticeable, but it is noticeable and is why I wanted to post my experience in hopes of getting feedback from my fellow Micron Xceed Grayscale users.

Thanks.

P.S. Below are the JPG schematic of the Grayscale adapter compared with the Apple adapter and also a PDF of the patent application.  Note though that schematic markings on the patent don't perfectly match the JPG schematic.  The JPEG schematic is correct. The 4 caps in yellow are the electrolytic capacitors that probably should be replaced.

View attachment Grayscale_video_conversion_system.pdf

 

[Crutch]

Thanks for that explanation - and just to be clear so nobody thinks I am crazy, I was not suggesting fiddling around with an SE/30’s innards while it was plugged into the wall!  I was referring to grounding mat setups that involve grounding the mat to a wall outlet, with presumably the chassis sitting on the mat, but I realize now I was very unclear.

I just used my SE/30 for half an hour and saw no sign of screen jitter — if I observe any, I’ll let you know.

By the way, while repairing my Micron kit recently I checked the electrolytic caps you marked with an in-circuit ESR cap tester and all seemed OK. Not sure how definitive that is however.

 

[JDW]

No, you were clear, but because there will no doubt be a self-professed CRT expert who will cross this thread and start telling me I don't know what I'm talking about, I put that piece of information in my previous post.  Seriously, I've read such posts in this very forum by people with decades of CRT experience AND who also have had no major issues discharging compact Mac CRTs and they feel that plugging the cord into the wall is the right thing to do because of the earth ground, when in fact it isn't.  So I was trying to drive that point home so there won't be future debate on that.  

Please let me know if your screen jitters when you have time to check it.  Thanks!

 

[Bolle]

Disconnecting the wiring harness from the neck board will disconnect chassis ground from the tube.

Just unplug that one connector before discharging the CRT and you will be fine.

I personally only discharge the CRT when I have to remove the anode cap. The logicboard and any PDS cards that were in the machine have already been removed from the machine at that point anyways.

 

[JDW]

@Bolle

Thank you for the advice, but can you please tell us WHY Micron spliced the ground of the Grayscale Adapter Yoke PCB with one thin wire leading to the CRT Ground Lug (upper left corner screw of CRT) and the other wire leading to the same chassis ground screw as the PSU?  Keep in mind the Apple Yoke board does NOT do that, using only 1 thick wire from the yoke board to the Ground Lug in the upper left of the CRT.  I really want to know why Micro did that.  (See schematics and other technical info from my earlier post.)

BTW, I discharge the CRT when I need to remove the Grayscale yoke board to swap it with the Apple stock CRT yoke board.  It would be unwise to do such a swap without first discharging the CRT.

 

[Bolle]

It would be unwise to do such a swap without first discharging the CRT.

There shouldn't be any serious HV exposed on the back of the CRT so you don't really need to do that every time you swap the neck board. (I don't )

Not sure why they did connect those two grounds in the first place though. How is this done on other later compact Macs? CC, LC5xx? Do they share one ground for everything?

Gotta take a look at my spare Apple CRT neck board to see what's going on.

 

[JDW]

@Bolle

Since you have no fear, fire up your SE/30, then power it off and unplug the cord.  Rub your fingers across the Yoke board while still attached and see if you get a shock.  If no shock, then you're absolutely right about there being no need to discharge the CRT!    

I don't know about the Color Classic or even the Classic because I don't own any compact Mac newer than the SE/30.  But again, my grayscale adapter question is two-fold:

1. Why did Micron use such THIN WIRES to run from the Grayscale Yoke board to the Ground Lug and to the Chassis Ground?  (The stock Yoke Board's Ground Lug wire is very thick!)

2. Why did Micron care to connect the Ground Lug ground wire to the Chassis Ground in light of the fact Apple did not do that and Apple makes it very clear in their own technical documentation to discharge CRTs ONLY to the Ground LUG and never to the chassis.  Indeed, it was because technicians failed to heed that advice that Apple had to come out with a revised discharge tool with internal resistor to limit the damage that would occur to the motherboard when discharging the CRT -- apparently it would fry motherboards.

 

[Bolle]

Just checked... the CRT ground frame is connected to chassis ground with the original neck board in place as well. If you look at the original neck board you can see that there is only one big ground plane that connects all grounds together.

They just didn't splice it together (because there is nothing to splice as all wires are soldered in) so it's not that obvious as on the Micron board.

Splicing that wire into the harness was probably cheaper than having it soldered to the neck board.

 

[JDW]

@Bolle

Thank you for investigating that.  But that opens up the even bigger question of why Apple, in their official documentation, says specifically to discharge ONLY to the Ground Lug in the upper left cornet of the CRT and NOT to the chassis ground?  Why did connecting the old discharge tool (or a screw driver with wire attached) to the chassis ground fry motherboards?  Why then did Apple create the new CRT discharge tool with resistor to lessen the voltage spike in the event technicians turned a blind eye to Apple's advice and grounded their discharge tool to the chassis anyway (instead of the ground lug)?

The only thing I can think if us that there would be more resistivity in between the ground lug and the motherboard ground than if you would to use the chassis ground instead.  More resistivity would lessen the spike that ultimately killed motherboards.  Would you agree?

 

[Bolle]

If you think of the CRT as a big capacitor then the ground lug is the shortest path just like in your "just touch both legs of the cap"-example. That's what you want to do when discharging.

When connecting to the chassis ground the remaining HV has to find it's way to the CRT ground (through the analog board to the neck board and possibly through the logicboard as well if that's the shortest path depending on where on the chassis you connected the tool).

As soon as you are using a tool without a resistor to prevent sparks it is possible that HV arcs over somewhere on its long way. If you have the chassis connected to ground on a power outlet that could help as the HV doesn't have to travel all the way back to the tube but can just go down the drain :tongue: