IIgs RAM Expansion Preventing Boot?

[pb3623]

So in between working on our A/UX imaging project, trying to work through some Q950 issues and (on the side) my new SGI O2... I have a IIgs that I got in trade about six months ago that’s pretty much sat the whole time. 

• It’s a ROM 01 unit.
  
• MicroDrive CF card w/ 512 MB CF
  
• ByteBoosters 4 MB RAM expansion card
  
• serial card
  
• I clipped the PRAM battery and got the alligator clip battery pack to replace it
  
• I bought an upgraded PSU from ReActiveMicro and that got installed last week
  
• I bought the Spanish HD15-to-SCART cable, a SCART-to-HDMI upscaler and an HDMI-DVI cable that doesn’t appear to pass the signal
  

Main issue is this: with the RAM expansion card installed (yes, it’s pointed in the correct direction), there’s a high-pitch, almost CRT whine from the PSU and the screen is filled with inverse @@@@@@@@@@@@ (like it does briefly on power-up before POST). Without the card, it boots so in my mind, it’s the card or the (original) caps on the logic board.

I remember it starting up when I first got it, enough to see it worked (connected to composite).

There’s only apparently 256KB on the board so I’d love to see if the RAM card can be salvaged!  Thanks!

 

[cruff]

Shorted capacitor on the memory board dragging down the power supply voltage?

 

[nglevin]

I did have some trouble with that ByteBoosters card, enough that I wrote to its maker back when I bought it.

The GGLabs one is much better. Met its creator at a uh, Maker Faire in Oakland back when I was still in the Bay. He was showing off an SMB1 demo on his IIgs along with some work in progress peripherals, good stuff

As for that memory card bit! https://gglabs.us/node/1935 explains what it is, I believe he's still selling stock on the eBays. 

Solid work, solid dude.  I believe he still works at the big green company in his not-spare time? Just, gosh, wish I had his hardware know-how...

 

[pb3623]

Shorted capacitor on the memory board dragging down the power supply voltage?

I'll need to check tonight! No, I hadn't recapped it yet and hadn't considered that a high priority since the RAM card did work at one point, and when it ceased to work I immediately blamed the stock PSU.

Any one in particular I should start looking at, on the off change it's a 47uF that I have as spares?

I did have some trouble with that ByteBoosters card, enough that I wrote to its maker back when I bought it.

The GGLabs one is much better. Met its creator at a uh, Maker Faire in Oakland back when I was still in the Bay. He was showing off an SMB1 demo on his IIgs along with some work in progress peripherals, good stuff

As for that memory card bit! https://gglabs.us/node/1935 explains what it is, I believe he's still selling stock on the eBays. 

Solid work, solid dude.  I believe he still works at the big green company in his not-spare time? Just, gosh, wish I had his hardware know-how...

Big green = Nvidia? It's been a while since I heard it called that (when I was in a PC build phase)

I've no problem giving him my business if I can narrow down root causes and show it's the ByteBoosters card, just having spent about $200 on parts in the last month. Makes sense about the caps, though - more RAM equals more voltage. I don't even get a lit LED.

Should I be turning my multimeter on the expansion slot?

 

[Gorgonops]

As for that memory card bit! https://gglabs.us/node/1935 explains what it is, I believe he's still selling stock on the eBays. 

That's the memory card I have for my IIgs, it seems to do the needful. I snagged it when it was on sale for $40, which was a nice bonus.

I'm guessing you don't have the original card to try in its place, do you? I find myself... pretty skeptical, that the ByteBooster's card would draw enough power to be the straw that breaks the camel's back even if the motherboard capacitors are on the moldy side. (it's implemented with modern, low-power RAM and should suck a lot less than the original card, which sported 32 41256s.) Does it work if you pull the CF or serial card instead?

(Also, just curious: why do you have a serial card in a IIgs? The onboard ports do everything a Super Serial card can do.)

I had something like that happen with my IIgs, IE, the power supply whining and the screen freezing, when I plugged in a bad floppy drive that had the interface board shorting against the metal drive frame. It makes me wonder if you might have a short on that RAM card or in the slot connector.

 

[Unknown_K]

What do you guys use to test IIgs RAM cards anyway?

 

[pb3623]

As another data point, I ran the self-test (Ctrl-Open-Closed-Reset or whatever equivalent in my AEK II) and got the following:

(SYSTEM BAD    09010001)

which i I believe can be traced to (among other things) insufficient voltage.this after I took apart and cleaned the board (and I have that upgraded PSU). 

@Gorgonops I have the serial card because it came with it. I have two FDs. The seller told me he prefers to connect one drove to each port, rather than daisy-chain. Guess that’s not true!

 

[NJRoadfan]

The error is related to the Apple Desktop Bus chip. Its likely the RAM card has a dead short in it somewhere. A chirping power supply with no activity on the machine is a clear indication of this. Monitor the power rails and see if they drop below their rated values.

 

[pb3623]

The error is related to the Apple Desktop Bus chip. Its likely the RAM card has a dead short in it somewhere. A chirping power supply with no activity on the machine is a clear indication of this. Monitor the power rails and see if they drop below their rated values.

Would a video help? Just to make sure I'm describing the sound correctly as a chirp?

What would be the test points on the board and card that I would need to probe? There isn't any visual short - no whiskers on the solder joints, no errant solder paste, no loose traces. I can tell you the front of the RAM card (particularly the jumpers you're not support to do anything with - that were used for factory production) gets real hot so I'd guess you're right and one of the ICs is overheating? 

 

[Gorgonops]

The card getting really hot is a *bad* sign.

I can't find documentation one way or the other, but I'm guessing that square IC near the jumpers is a CPLD that provides all the decode logic, etc. (You have this one, right?) I wonder if it's possible the programming for the CPLD was erased/damaged somehow so it's now effectively internally shorted. If that's the case I suspect you're stuck trying to contact the maker and see if they can fix/reprogram it.

 

[cruff]

Any one in particular I should start looking at, on the off change it's a 47uF that I have as spares?

I have the ByteBoosters RAMGS-4MB/XLP version, there are five surface mount capacitors on it, so it is unlikely the 47 uF ones you have will fit.  If it is getting hot near the jumpers, and if your board is similar to mine, then C1, C4 or C5 could be suspect, as well as ICs X3 or X4.  You might get away with unsoldering one of those caps at a time and testing on the off chance one of them is the culprit.

 

[david__schmidt]

How sure are you that the card is oriented correctly?  I know you said it is, but... how sure are you really?  The cards that fit into the memory slot in a GS are traditionally facing "backwards" as far as their chip orientations go.

The power supply whine is a dead giveaway as a dead short, as many have said.  No need for a video there - that's the design of the power supply (specially, the "crowbar circuit") in action.  It senses the short, and shuts itself down and repeats.  We perceive that cycling as it passes through the audible range.  Heat is another great indication of a short.

 

[Gorgonops]

One nice thing about the gglabs card, I guess, is because of its shape it'd be obvious if it were in backwards. (Wouldn't be able to close the case.,)

 

[pb3623]

@david__schmidt I'm positive the card's correctly oriented. There's an arrow that points to the front of the card/IIgs case.

@cruff My comment on the 47uF's  were related to the logic board. I agree using them on a card of this size would be overkill, electrically and physically!  

All,

I took some close-up pictures of the upper-right quadrant of the card, particularly where the heat builds up.

I took note of component D1, which appears to be small capacitor (is it a diode?) but it almost appears chipped. (i.e., the top of the diode isn't flat). I tried to capture it as best I could.

All other components appear visually intact. Perhaps the card just went bad. I'd rather try and fix it (w/ or w/o the Byteboosters' help) but, if that doesn't pan out, at least I'm confident the board itself (and the new PSU) appear to be good. 

https://imgur.com/a/I5k9zTc

 

[lisa2]

D1 looks like a surface mount LED to me, does is light up?

 

[pb3623]

Yeah, that explains it - boy, do I feel silly. I should have caught that when I typed diode. But, in my defense, no - the LED doesn't light up so it wasn't obvious to me what this way, besides looking very similar to the cap next to it.

 

[cruff]

D1 is a red LED that does light up.  And yes, there is a silk screened arrow marking the correct orientation of the card (arrow towards the front).

 

[cruff]

C4 and C5 measure 6.8 nF in circuit with my Smart Tweezers.

C1, C2 and C3 measure about 300 nF, which means they may be three 100 nF caps in parallel.

IC X3 is marked as a LD33C, which is probably an adjustable voltage regulator, but I'm not turning up a datasheet based on that marking code right at the moment.