I appreciate including the failures. A lot of retro tech channels make it look like this old hardware just springs to life and cooperates 100% of the time, but the reality is that a lot of this hobby is trying, reading, trying again, changing some settings, trying again, replacing some parts, trying again, and even then sometimes still not getting it working.
i'm 44, and also missed being in the game at the right time in the 90's, but 20y ago I've bought some Sun hardware from friends : two Ultra30 + one Ultra1 + one SPARCstation 4 + 2 SunBlade 100 (those 2 last ones came from a national french lab, INRIA), all of this works flawlessly yet. Honestly I think it worths the experience to invest a few bucks in such old piece of hardware even on auction and have fun playing with it. Of course it's not that fast compared to modern hardware but it's nice and rock solid. Can support Solaris or NetBSD or FreeBSD for some. When you look at how powersupplies were enginereed you understand they designed all of these very seriously. One day i'll buy some SGI station, just to play with Irix on its original dedicated hardware. This is not a matter of nostalgia, as we didn't know the era when it appeared on the market (would love to play with a Next Cube too), but it keeps all of this alive and maintained, but it's more like collecting old cars, from time to time you just enjoy the ride 😎.
@clabretro , I found from experience that the Sun PC on-a-card suffered from the same issue as the old GoldenGate Amiga Bridge Boards, with the BIOS chips going bad after not being powered on for a while. Depending on the make and model, I found that the chips, needed to be re-flashed, or if, the were a weird flash-mask fromtheybrid, replaced. Also, for the Sun PC on a card, the power delivery to the BIOS Roms can be a bit unstable as the Caps age, and the caps that deliver power to the ROM are also part of the chain that powers the CPU and RAM, so if the ROM gets fried, the RAM and CPU may be damaged too. I advise testing the circuits for continuity and the voltages and amperage being in spec.
I remember those Athlon XP CPUs being able to fry themselves by heat. They had no built in protection from overheating so booting one without proper cooling they died from heat pretty quick. Hope you get that SunPCI III working!
I think Athlon XP had that protection built in, not sure if that applies to all the XP models. Athlon Thunderbirds and Durons would fry themselfes to death for sure because those didn't have any built in protection at all.
Only use plastic picks to pry plastic parts! Using metal to pry always damages plastic. Also HP keeps the service manuals online even for 20 year old laptops, so if you can't figure out how to open them, use the service manual. By the way doesn't that Sun x86 card have a VGA port? See if you get any output there?
Regarding the Sun Ray 270, that error indeed looks a lot like one of those errors on standard monitors. The 270 has a VGA input, maybe it error is due to it being left on monitor mode. Maybe giving it a VGA signal and pressing the input button would fix it?
@@clabretro It's also worth trying to connect the VGA output of the board to an external monitor, to see if anything comes up to prove the board operational
Those Sun Ray 2 thin clients brings back memories. When I went to uni in 2011 we still used those, and they hung around until 2014 iirc. Cant say that system was amazing, but it was still pretty impressive seeing that it was still serving the entire university well enough to not be a nuisance until the day they were replaced.
wonderful Video ! i was watching it as i was personally fumbling with the ILO 4 on my (new to me) server ! glad to see it isnt just me getting a bunch of failures :) great video this week keep it up!
Thank you very much for re-uploading a video with Sun Microsystems content.👏👏 I hope you manage to figure out the problem with the PCI card and the AMD Athlon processor.💪 At least you were able to fix the 2FS thin client.😃 Very curious that it belonged to Oracle Japan itself.🤨
Not seen alom in ages. Making me feel old. I bought a t1k a few years ago in case I ever wanted to relive 3am on-site ufs fsck. The loneliest feeling in the world.
Cool shirt. I visited Connections Museum earlier this year and long story short I can't wait to get back to Seattle so I can visit again! There is just so much to see.
Believe it or not I haven't actually been, just wanted to support the museum, been enjoying their UA-cam channel. I grew up in Seattle so I make it back there fairly often, it's on my list. Glad to hear it was good!
I may be totally off, but is that ZIF socket one where you drop the chip in and then slide it down like... 1/2 a mm to engage the pins? There was also one you needed to use a screwdriver blade in the little ears to engage but it's hard to see if this might be that.
Yeah the laptop was like that, the SunPCI is *similar* but without an actual sliding mechanism. I did try to make sure it had good contact, but I'm still suspicious of it.
Man, that's too bad about the Sun card, I'm really interested in seeing how it works. Hopefully there's another one out there you can use to get up and going. Thanks for the video!
2000's and earlier laptops are such a pain to take apart. I tried to take my 2004 HP apart and even after removing about 87 screws and a dozen panels I couldn't even get to the DC barrel jack. I think one of the steps involved removing the display and it's associated cabling and I just NOPE'd out of that endeavour. Honestly surprised the disassembly instructions didn't involve reducing the components down to their base chemical elements. "Step 4,371: Harvest all the copper from the mainboard before continuing to next step." Thankfully in the last 10 years or so, laptop makers realized that their machines didn't have to be built like a 5 layer lasagna and now almost everything can be reached after removing a handful of screws and prying off a bottom plate.
Modern laptops (particularly consumer model HP ones) often have critical screws under rubber feet which are impossible to put back. Older laptops have more screws but are also seemingly better built overall.
Seeing the inside of that 2100 makes me appreciate the interiors of laptops right now. I'm a dell field tech and the inside of that makes the insides of current laptops look empty. I remember taking apart my HP Pavillion dv9700 back in 2010, still don't like taking apart laptops, never liked the plastic clips keeping the covers in place. Unfortunately most of my jobs are laptops even though I specialize in server repairs.
SunRay 270: There is a VGA input on that board, those thin clients obviously have a monitor function in addition to the thin client. Check if there is a way to switch the input back to the thin client.
Yeah I forgot to film trying that, it doesn't respond to input button changes. I also tried (after the video, after some viewers had the good suggestion) to try that with a VGA input actually connected and also no luck.
I bought a couple of sun ultra ac 5's off craigslist one time just to check them out. Spent a week messing with them. Was when I was really into linux. Got debian going on one. Then I spent a week troubleshooting things as the thing wouldn't boot and I had to mess with the eprom. Turned out one of the paired rams was bad. Thing absolutely would not boot without two good paired ram sticks. Once I got it going, I couldn't really do much with it. The tech was so far behind the times.. it was a fun experiment, and got to learn a bit about eprom. But, eventually gave them away to a sun collector.
Great video, my fan was fragged in my III board, I don't suppose you tested the pinout of the fan, I wanted to make sure red was +5v and Orange is Ground
@@clabretroThanks for the comment! My fan was bad, took care of that but ran it without a cooler or heatsink which was unwise - so I had to order another CPU ebay which will hopefully work.
For the SunPCI III Pro, try cleaning the PCI connector fingers. They look badly oxidized in spots. I was waiting for you to do this and disappointed that it was never addressed, but I know from experience that sometimes it's the most simple things that get overlooked when troubleshooting.
They'd been cleaned shortly before the video and I actually cleaned them really well after the video to have another try, no luck. They might look worse on camera than they actually are; I sure was hoping it was that though.
I swear laptops are designed by phycopaths. Why do they make them so bad from a service/disassembly point of view? I'm aware this is one of the better ones out there! Anyway appreciate all the SunRay stuff, I used to use them at university and have a bit of a soft spot for them.
Planned obsolescence started back in the 50's with automobiles, it's gotten way worse, now subscriptions expire and a device won't work if you don't renew, That's why I only buy open source. It's really sad to have watched all this happen in the computer industry, I started using them in the early 80's, by llate 80's and early 90's it exploded and you could buy tons of parts and just build a pc no proprietary crap, had to be compatible but that's not the same. And everything was made to repair or upgrade
I have a 2021 ASUS gaming laptop and if its CPU was in a socket, I could probably have it out in about 4 and a half minutes. Of course, now that access has improved greatly, so much more is soldered to the board.
I have an old Toshiba satellite & it was the worst. Only broke one speaker connector from age & had a few screws left. Just to apply fresh arctic silver and clean the fan.
Have you locked the cpu in the socket? Otherwise it has no electrical contact! It looks like a sliding mechanism that you can push with a flat hed screw driver up and down to shift the whole frame with the cpu up and down.
Yeah - the laptop has an actual sliding locking mechanism, but the SunPCI doesn't. I think the chip is meant to be slid sideways against/into the connectors. I did try to make sure it had the best contact I could, though I'm still suspicious of that.
I still can't get over how cool those SunRay workstations are! I know Windows has Terminal Services that works KINDA similar, but I gained a good deal of experience with that working for H&R Block about 13 years ago, and that experience was VERY poor. Now that I'm thinking about it, I'd bet you could do something similar to the sunray with linux and some of those Chromebox linux conversions. Think I should look into that, actually! Would be really nice to have a terminal in my garage.
Howdy! Sun expert here who used a lot of these new. Part of your problem booting the Sun PCi III? You're doing it in a V240. It wasn't actually supported in those, and these cards are very picky. The Sun PCi cards (original, II, and III) are EXTREMELY architecture and firmware specific. The Sun PCi III will just not work in a V240. The PCI bus doesn't support it. I can see just from the ALOM that it's not detecting the card at all - there's no pci@2 devices. The card itself is probably fine though - by the III, they were actually surprisingly reliable for Sun equipment. Also, the only known major customer for SunRays, is Sun. The machines were so incapable and ridiculously licensed that they couldn't give the hardware away. (They tried. We said no. Multiple times.) And I never once heard a Sun employee speak even slightly favorably of them.
Hey! Interesting, I read in the 3.1 product notes that the III supports the v240: docs.oracle.com/cd/E19614-01/817-2968-11/817-2968-11.pdf (though I have a PCi III Plus here). I did see it post once but never again, which lead me to thinking it is an issue on the card.
@@clabretro might be possible, but I'd wager there's a stack of caveats including a minimum OBP version. It's been over a decade (obviously) but 'failed to respond. flags are 0' is to the best of my recollection, a failure in reading the PCI registers and NOT the card itself. The card itself is about 99% self-contained. So I suspect what's going on here is that even if the V240 gets VenID and DevID, it's unable to figure out the state of the card from the PCI configuration space. Especially since the BDF looks incorrect and it's failing to read the MAC address. Doing a `set-defaults` (will set ALL defaults including passwords!) in OBP might help, but I think the version of OBP is just unhappy. However, thinking on it, I also wonder if the battery on the SunPCI might not be the fault. Yes, there IS a battery - it's part of the RTC, and of course, 'not user replaceable!' I believe it's between the SODIMM and the large QLC marked 'BELLEROPHON' (next to the USB/FW riser connector.) /var/preserve/sunpcilogs might have something useful too.
Just a thought, are the sun thinclient "faceplates" interchangeable? maybe you can source a non working but nice looking version for very cheap and replace the broken faceplate? If this dosent work - I think it was superglue and baking soda making it even stronger and you could sand and repaint it :D
Have you tried using the parts from the pci card in other machines to test the ram and cpu? For the 2nd system, it has both vga in and out, do either do anything?
testing those parts (probably in that laptop, after I fix it) is next on the list. yeah I did try the input button on the 270 (forgot to film it) and it didn't make a difference. but one viewer suggested doing that with an actual vga input going in, so I'll try that next
@@clabretro what about the vga out? I’m curious if either the broken or working system also output simultaneous vga at the same time as internal video?
yeah I *did* try the VGA out of the 270 (and the PCI card for that matter) but no luck. forgot to put all that in the video too, of course!@@JohnKiniston
On Video is not visible is you connected the Athlon CPUs into the socket. Did you slide the CPU in the socket? (it's not enough to just drop them in the socket, sliding is important and not easy)
Yeah the laptop has the typical slider to engage the CPU, and I did that. But the SunPCI doesn't... I did push the chip to engage the pins as best I could (it seems to be the same mechanism, just without the thin sliding layer of plastic typical sockets have, if that makes sense). So I'm still a bit skeptical of the connection the SunPCI is making.
The one on the laptop does but the SunPCI board doesn't... but it does feel loose in there. I made sure to push the chip tight in the direction of the socket pins before tightening the fan down, but I'm still a bit suspicious of it.
I think those All-In-One thinclients are literly just modified off the shelf monitors with regular off the shelf monitor controllers. That error about the cable not detected was likely just coming from the monitor controller, just like a Dell monitor would show if no input was detected, TLDR; monitor and monitor controller is getting power and funtioning but the thinclient computer compent istelf is not sending a display signal to the controller for the monitor.
Is there a connector for a PC speaker? I had an issue on my old Slot 1 board where the BIOS had a slight corruption causing the 'CPU failure' alert to go off, fortunately in my case I was still able to boot into DOS and reflash it which fixed it. I got the fix from one of CuriousMarc's videos. Is it also possible to pass through a PCI card? Maybe one of those diag cards could give an opcode if the CPU is still executing code, but not getting far enough to initialise communication with the Sun.
@@clabretro, I found from experience that the Sun PC on-a-card suffered from the same issue as the old GoldenGate Amiga Bridge Boards, with the BIOS chips going bad after not being powered on for a while. Depending on the make and model, I found that the chips, needed to be re-flashed, or if, the were a weird flash-mask fromtheybrid, replaced. Also, for the Sun PC on a card, the power delivery to the BIOS Roms can be a bit unstable as the Caps age, and the caps that deliver power to the ROM are also part of the chain that powers the CPU and RAM, so if the ROM gets fried, the RAM and CPU may be damaged too. I advise testing the circuits for continuity and the voltages and amperage being in spec.
Have you actually measured the rail voltages on the X86 board? Also, if its regular PCI-64, can't you put it in a donor X86 machine without fear of damaging the host.
I haven't yet but I'm thinking the same direction... get it in a machine where it's convenient to probe around with the multimeter and see what it's up to. It's not a total failure... there is a status LED on the back of the card that consistently blinks amber (which means "potential problem" according to the documentation), forgot to put that in the video. Will be an interesting one to figure out.
"It's not the hardware, it's not the hardware, it's not the hardware.... OK... It's the hardware...." heartbreaking. The highs and lows of old hardware.
You have to push the cpu sideways with a flathead screwdriver from one of the slots next to it inside the cpu socket! Otherwise the pins don't make contact. There are 2 slots next to the cpu, directly opposite to each other. When you look at them you'll notice that one of them is nearer to the cpu than the other. This is the one you need to lever with a screwdriver against the cpu! The other one is to remove the cpu from this position. Try it and give me a note whether it worked, I'm 99% sure that's the problem!
so, you're totally right - I did try that a few times but it still didn't work. I think it's part of the puzzle though, that socket is pretty low quality and some of the pins are a little suspect. I can never get the CPU back in as firm as it was when I first took it out either. I think there's still a good chance there are bad caps or other components at play here, but I might actually just replace the CPU socket entirely when I get some better desoldering gear!
These socket-types are a bit clunky. You need a well fitting screwdriver, maybe you can wedge another one between the cpu and the first one to get more leverage. Had those in laptops of that era. Changing sockets is a pain, even with good equipment. I did that several times. Best way seems to be hot air, but with a big nozzle, high airflow and matching temperature (
weird, it didn't get held for review or anything, I don't see it on my side either. I did see it before it disappeared though just hadn't responded yet! I'll try that jumper to see if I can get a boot screen
@@clabretro YT probably thinks I'm a bot, because I keep editing my comments multiple times. For some reasons, my thoughts start to flow only after hitting the Submit button.. I'm weird like that.
The message on the 270 was generated by the screen itself. It's the same "no signal" message you get on a normal monitor when the VGA is disconnected, or not at the right specification. That really rules out any screen problem. I'd start with checking power on the "main" board onwards. And literally as I finish this comment, you've just said what I've written - almost word for word!! I'm just a mind reader. Ignore me!! 😂
strange, right? I had no idea until I filmed this video, I had thought those slots were clips holding the case together or something. The Sun Ray 2s also have the same thing. This is all I've found about it so far: www.infoworld.com/article/2656281/sun-rolls-out-new-sun-ray-thin-clients.html Apparently they were thinking admins could store device config on the sims?
@@clabretro ooh you found *something* about the sims, i went looking and couldn't find anything - at first i thought they would use the sim as some sort of out of band managment... for some reason, lol. Interesting, I wonder if anyone has actually ever used the sims in the sunrays!
I did try hitting the input button (forgot to film it) but it had no effect. another viewer suggested actually hooking a vga input up to see what it does which I'll try next
I have a broken USIII 750 module for my (what was) Dual CPU Sun Blade 1000 which is now unfortunately single CPU! Like you mine worked once!!! and never since! Hard to find these modules cheaply if at all!
The Compaq sounds like it's having a small power delivery issue....like a power rail is missing, Defo a component lever repair of sorts but definitely fixable with the correct knowledge
I havr a sunray 270 with the exact same fault. Ive tried external video input, also tried the external input with no result either. Did you ever get a fix?
Are there other avenues into the X86 board? Like can it be powered in this slot with its own monitor and keyboard so you can see what is happening to it? Does it have any PCI/ISA expansion; perhaps something you can shoehorn a diag board with 7 segment LEDs for boot codes? I don't really know how much special sauce they've put into the BIOS or any other subsystem on this board but it could be interesting. If you do not have the ability, perhaps a colab with another youtuber might be in order. Ask them to flash a new BIOS to a chip for you, if you are suspecting that to be an issue. There might be a website out there with this BIOS saved for posterity. Something I would try in addition to recapping, is a BIOS reset or clear CMOS jumper if it has one.
Yeah all great ideas. The back VGA has no output, unfortunately. While it's a PCI board itself, I don't think I have access to whatever internal PCI bus it's using somehow for the x86 machine it hosts. There is what appears to be a JTAG header on it, so that could be interesting. I think I need to do a little voltage probing while it's in the machine to see if there is a basic power failure going on. It does have a diagnostic LED on the back which blinks amber at all times, the manual says this "may indicate a problem," which we obviously have haha.
That on screen error message on the sunray 270 looks like a generic monitor message. I would expect that the display is a commodity monitor and if it doesn’t see a signal it just displays that message. I would try it without the USB board connected to the monitor. I wouldn’t be surprised if the same message shows up on screen.
Yeah the trick is the display board gets its power through the main CPU board. It's something I'd like to try though, figure out how to power just that board properly to see if it does the same thing.
Please try to use a plastic guitar pick or spudger to pry off clips and if that doesn’t work try a thin, wide metal thing… I think your ifixit kit came with some. It pains me to see you use a small screwdriver.
Yeah couldn't find it, I actually have another kit on the way. The whole time I was thinking "hmm I should really wait for that to arrive" haha. No harm done though.
I wish I lived near you, I have the experience and equipment to do some more in-depth repairs, including surface mount chip swapping, etc. If you do want me to take a look at anything, I can share my mailing address on Patreon or something and I'd be more than willing to do the work gratis.
Also, you really should pick up an iFixIt toolkit. They come with very high quality screwdrivers and bits, and lots of plastic spudgers for opening plastic parts like that without causing damage. They are extremely useful!
Is the Athlon XP still loose in the socket? If it is loose the CPU has no electrical contact inside the socket! After inserting the CPU it has to be pushed towards the PCI connector to secure and connect it in the socket! Insert a flathead screwdriver in the little groove on the socket above the CPU and then twist the screwdriver to push the CPU sideways towards the PCI connector. It takes some force because the pins of the CPU have to be pushed in the "connection groove" in the socket. Take a close look at the socket with a magnifying lens if you don't know what I mean. I'm sorry but I can't explain it in a better way because I'm German and English isn't my native language.
Very good point. I did actually push the CPU towards the connector like you say, but I forgot to get it on video! The laptop has the sliding style socket, but the SunPCI needs to be pushed like you're describing. It still doesn't sit very tight, so I'm a little suspicious of it still.
@@clabretro In my experience this type of socket needs "unreasonable amounts of force" to connect the CPU properly. Every time I applied reasonable amounts of force the machines acted exactly like the Dell & the SunPCI! I call this kind of state "braindead". If a Socket A-CPU dies because of overheating (cooler not properly mounted - Socket A-CPUs don't have overheating protection because there is no thermal sensor inside the Die) they usually have a brown spot on the underside (the side where the pins are) in the middle spot that is free of pins.
Heh I know, couldn't find mine. I actually have more on the way, the whole time I was thinking "hmm I should have really waited for those to show up." Anyway, no harm done haha.
For monitor it is most likely the inverter that provides power to the backlight failing ,it provides just enough power for it to turn it up and than it fails
I appreciate including the failures. A lot of retro tech channels make it look like this old hardware just springs to life and cooperates 100% of the time, but the reality is that a lot of this hobby is trying, reading, trying again, changing some settings, trying again, replacing some parts, trying again, and even then sometimes still not getting it working.
oh yeah, lots and lots of trying stuff. thanks for watching!
I love that the Linksys stack is growing. I hope it exceeds the highest peak of your home.
35:06 lol. It me.
More Sun Microsystems stuff I like instantly. Being in my early 20s part of me wish I was born 15 years earlier looking at all this amazing gear.
i'm 44, and also missed being in the game at the right time in the 90's, but 20y ago I've bought some Sun hardware from friends : two Ultra30 + one Ultra1 + one SPARCstation 4 + 2 SunBlade 100 (those 2 last ones came from a national french lab, INRIA), all of this works flawlessly yet. Honestly I think it worths the experience to invest a few bucks in such old piece of hardware even on auction and have fun playing with it. Of course it's not that fast compared to modern hardware but it's nice and rock solid. Can support Solaris or NetBSD or FreeBSD for some. When you look at how powersupplies were enginereed you understand they designed all of these very seriously. One day i'll buy some SGI station, just to play with Irix on its original dedicated hardware. This is not a matter of nostalgia, as we didn't know the era when it appeared on the market (would love to play with a Next Cube too), but it keeps all of this alive and maintained, but it's more like collecting old cars, from time to time you just enjoy the ride 😎.
@clabretro , I found from experience that the Sun PC on-a-card suffered from the same issue as the old GoldenGate Amiga Bridge Boards, with the BIOS chips going bad after not being powered on for a while. Depending on the make and model, I found that the chips, needed to be re-flashed, or if, the were a weird flash-mask fromtheybrid, replaced. Also, for the Sun PC on a card, the power delivery to the BIOS Roms can be a bit unstable as the Caps age, and the caps that deliver power to the ROM are also part of the chain that powers the CPU and RAM, so if the ROM gets fried, the RAM and CPU may be damaged too. I advise testing the circuits for continuity and the voltages and amperage being in spec.
I remember those Athlon XP CPUs being able to fry themselves by heat. They had no built in protection from overheating so booting one without proper cooling they died from heat pretty quick. Hope you get that SunPCI III working!
I think Athlon XP had that protection built in, not sure if that applies to all the XP models. Athlon Thunderbirds and Durons would fry themselfes to death for sure because those didn't have any built in protection at all.
I love this! Like others have said, seeing what doesn’t work out is often just as entertaining as when it does! Fun to see how you approach problems
Thanks! Glad you liked it.
Only use plastic picks to pry plastic parts! Using metal to pry always damages plastic. Also HP keeps the service manuals online even for 20 year old laptops, so if you can't figure out how to open them, use the service manual. By the way doesn't that Sun x86 card have a VGA port? See if you get any output there?
Yeah couldn't find my spudger, new kit on the way :)
It does have a VGA port and unfortunately no output from it.
hp recently took my pc off the website 😢
@@realstealthninja which laptop is that?
@@RedShift5 not a laptop but a desktop with a a laptop motherboard, HP P2-1366il
Regarding the Sun Ray 270, that error indeed looks a lot like one of those errors on standard monitors. The 270 has a VGA input, maybe it error is due to it being left on monitor mode. Maybe giving it a VGA signal and pressing the input button would fix it?
interesting... I had tried the input button, but not while giving it an actual input. I'll give that a shot!
@@clabretro It's also worth trying to connect the VGA output of the board to an external monitor, to see if anything comes up to prove the board operational
Those Sun Ray 2 thin clients brings back memories. When I went to uni in 2011 we still used those, and they hung around until 2014 iirc. Cant say that system was amazing, but it was still pretty impressive seeing that it was still serving the entire university well enough to not be a nuisance until the day they were replaced.
wonderful Video ! i was watching it as i was personally fumbling with the ILO 4 on my (new to me) server ! glad to see it isnt just me getting a bunch of failures :) great video this week keep it up!
thank you!
Thank you very much for re-uploading a video with Sun Microsystems content.👏👏 I hope you manage to figure out the problem with the PCI card and the AMD Athlon processor.💪
At least you were able to fix the 2FS thin client.😃 Very curious that it belonged to Oracle Japan itself.🤨
thanks for watching!
you could try taking out the BIOS chips on the SunPCI and putting them back to clear any bad connections in the sockets
good idea
Curious what would happen if you went from two head to one head. I'd imagine it'd just move the windows over or can you just not access them?
yeah good question, I haven't actually tried it yet. I'll let you know when I do
Not seen alom in ages. Making me feel old. I bought a t1k a few years ago in case I ever wanted to relive 3am on-site ufs fsck. The loneliest feeling in the world.
Cool shirt. I visited Connections Museum earlier this year and long story short I can't wait to get back to Seattle so I can visit again! There is just so much to see.
Believe it or not I haven't actually been, just wanted to support the museum, been enjoying their UA-cam channel. I grew up in Seattle so I make it back there fairly often, it's on my list. Glad to hear it was good!
These things are so cool. Well done bud.
From the video the pci connector on that board looks corroded a bit. Could be lighting though. Maybe a polish with ipa or deoxit?
Good call, I had cleaned it with IPA but that was weeks ago, so I think I'll clean it again
I may be totally off, but is that ZIF socket one where you drop the chip in and then slide it down like... 1/2 a mm to engage the pins? There was also one you needed to use a screwdriver blade in the little ears to engage but it's hard to see if this might be that.
Yeah the laptop was like that, the SunPCI is *similar* but without an actual sliding mechanism. I did try to make sure it had good contact, but I'm still suspicious of it.
@@clabretro sad trombone :/
I have spread this video out over multiple days just so it can feel longer haha. Love the sun content.
😆 thanks for watching!
Man, that's too bad about the Sun card, I'm really interested in seeing how it works. Hopefully there's another one out there you can use to get up and going. Thanks for the video!
Yeah I'll get it working one way or another. Thanks for watching!
@@clabretro Slim chance, but maybe both of the cpus are bad? Maybe there is a way to get the post codes out of the card?
it definitely could be, I'll try another CPU if the re-cap doesn't work
2000's and earlier laptops are such a pain to take apart. I tried to take my 2004 HP apart and even after removing about 87 screws and a dozen panels I couldn't even get to the DC barrel jack. I think one of the steps involved removing the display and it's associated cabling and I just NOPE'd out of that endeavour. Honestly surprised the disassembly instructions didn't involve reducing the components down to their base chemical elements. "Step 4,371: Harvest all the copper from the mainboard before continuing to next step."
Thankfully in the last 10 years or so, laptop makers realized that their machines didn't have to be built like a 5 layer lasagna and now almost everything can be reached after removing a handful of screws and prying off a bottom plate.
Modern laptops (particularly consumer model HP ones) often have critical screws under rubber feet which are impossible to put back. Older laptops have more screws but are also seemingly better built overall.
21:12 VESA mount, usually for attaching to monitor arm.
Seeing the inside of that 2100 makes me appreciate the interiors of laptops right now. I'm a dell field tech and the inside of that makes the insides of current laptops look empty. I remember taking apart my HP Pavillion dv9700 back in 2010, still don't like taking apart laptops, never liked the plastic clips keeping the covers in place. Unfortunately most of my jobs are laptops even though I specialize in server repairs.
Yeah they're super hard to work on. Haven't been inside a newer one but glad to hear it's a little better haha
07:20 I wonder if the 2800+ would work in the SunPCi.
I bet it would, but I'm thinking it's not the CPU at this point
Man. Always excited to see a brand new video dude. Great content
Thanks as always!
SunRay 270: There is a VGA input on that board, those thin clients obviously have a monitor function in addition to the thin client. Check if there is a way to switch the input back to the thin client.
Yeah I forgot to film trying that, it doesn't respond to input button changes. I also tried (after the video, after some viewers had the good suggestion) to try that with a VGA input actually connected and also no luck.
I bought a couple of sun ultra ac 5's off craigslist one time just to check them out. Spent a week messing with them. Was when I was really into linux. Got debian going on one. Then I spent a week troubleshooting things as the thing wouldn't boot and I had to mess with the eprom. Turned out one of the paired rams was bad. Thing absolutely would not boot without two good paired ram sticks. Once I got it going, I couldn't really do much with it. The tech was so far behind the times.. it was a fun experiment, and got to learn a bit about eprom. But, eventually gave them away to a sun collector.
Great video, my fan was fragged in my III board, I don't suppose you tested the pinout of the fan, I wanted to make sure red was +5v and Orange is Ground
I can't remember if I get it on video anywhere but the fan does work, so probably caps or that CPU unfortunately!
@@clabretroThanks for the comment! My fan was bad, took care of that but ran it without a cooler or heatsink which was unwise - so I had to order another CPU ebay which will hopefully work.
For the SunPCI III Pro, try cleaning the PCI connector fingers. They look badly oxidized in spots. I was waiting for you to do this and disappointed that it was never addressed, but I know from experience that sometimes it's the most simple things that get overlooked when troubleshooting.
They'd been cleaned shortly before the video and I actually cleaned them really well after the video to have another try, no luck. They might look worse on camera than they actually are; I sure was hoping it was that though.
I've got a Sun Ultra 60 that was once owned by SAAB automobile. That's kinda cool too.
I swear laptops are designed by phycopaths. Why do they make them so bad from a service/disassembly point of view? I'm aware this is one of the better ones out there! Anyway appreciate all the SunRay stuff, I used to use them at university and have a bit of a soft spot for them.
agreed, I know I was complaining but this one wasn't actually too bad haha
Planned obsolescence started back in the 50's with automobiles, it's gotten way worse, now subscriptions expire and a device won't work if you don't renew, That's why I only buy open source. It's really sad to have watched all this happen in the computer industry, I started using them in the early 80's, by llate 80's and early 90's it exploded and you could buy tons of parts and just build a pc no proprietary crap, had to be compatible but that's not the same. And everything was made to repair or upgrade
I have a 2021 ASUS gaming laptop and if its CPU was in a socket, I could probably have it out in about 4 and a half minutes. Of course, now that access has improved greatly, so much more is soldered to the board.
I have an old Toshiba satellite & it was the worst. Only broke one speaker connector from age & had a few screws left. Just to apply fresh arctic silver and clean the fan.
Have you locked the cpu in the socket? Otherwise it has no electrical contact! It looks like a sliding mechanism that you can push with a flat hed screw driver up and down to shift the whole frame with the cpu up and down.
Same questions.
Yeah - the laptop has an actual sliding locking mechanism, but the SunPCI doesn't. I think the chip is meant to be slid sideways against/into the connectors. I did try to make sure it had the best contact I could, though I'm still suspicious of that.
Maybe it needs a thermal pad then? To press it down firmly into the socket.@@clabretro
Not a bad idea, that might help
Wow, lots of memories seeing the compaq presario 2100.
I still can't get over how cool those SunRay workstations are! I know Windows has Terminal Services that works KINDA similar, but I gained a good deal of experience with that working for H&R Block about 13 years ago, and that experience was VERY poor. Now that I'm thinking about it, I'd bet you could do something similar to the sunray with linux and some of those Chromebox linux conversions. Think I should look into that, actually! Would be really nice to have a terminal in my garage.
Howdy! Sun expert here who used a lot of these new. Part of your problem booting the Sun PCi III? You're doing it in a V240. It wasn't actually supported in those, and these cards are very picky. The Sun PCi cards (original, II, and III) are EXTREMELY architecture and firmware specific. The Sun PCi III will just not work in a V240. The PCI bus doesn't support it. I can see just from the ALOM that it's not detecting the card at all - there's no pci@2 devices. The card itself is probably fine though - by the III, they were actually surprisingly reliable for Sun equipment.
Also, the only known major customer for SunRays, is Sun. The machines were so incapable and ridiculously licensed that they couldn't give the hardware away. (They tried. We said no. Multiple times.) And I never once heard a Sun employee speak even slightly favorably of them.
Hey! Interesting, I read in the 3.1 product notes that the III supports the v240: docs.oracle.com/cd/E19614-01/817-2968-11/817-2968-11.pdf (though I have a PCi III Plus here). I did see it post once but never again, which lead me to thinking it is an issue on the card.
@@clabretro might be possible, but I'd wager there's a stack of caveats including a minimum OBP version. It's been over a decade (obviously) but 'failed to respond. flags are 0' is to the best of my recollection, a failure in reading the PCI registers and NOT the card itself. The card itself is about 99% self-contained.
So I suspect what's going on here is that even if the V240 gets VenID and DevID, it's unable to figure out the state of the card from the PCI configuration space. Especially since the BDF looks incorrect and it's failing to read the MAC address. Doing a `set-defaults` (will set ALL defaults including passwords!) in OBP might help, but I think the version of OBP is just unhappy.
However, thinking on it, I also wonder if the battery on the SunPCI might not be the fault. Yes, there IS a battery - it's part of the RTC, and of course, 'not user replaceable!' I believe it's between the SODIMM and the large QLC marked 'BELLEROPHON' (next to the USB/FW riser connector.) /var/preserve/sunpcilogs might have something useful too.
The pci connection on that sun card looked tarnished. I would take a pencil eraser to both sides of it until it shines and try again.
not a bad idea, I had cleaned it several weeks ago with IPA but I'll give it a more thorough look now
Do more of these!!! I love repair vids!!
That’s something I’ve never noticed about those linksys routers, they stack rather well 😂
Might want to consider investing or building a dim bulb test socket.
Yeah been thinking about that, probably a good idea
Yet another great video. Thanks!
thanks!
Maybe you can look for some sort of serial debug on the x86 card. If one is on there, it might output something.
Yeah it might come to that!
Just a thought, are the sun thinclient "faceplates" interchangeable? maybe you can source a non working but nice looking version for very cheap and replace the broken faceplate? If this dosent work - I think it was superglue and baking soda making it even stronger and you could sand and repaint it :D
Oh yeah I'm sure I could find one in good outer physical condition. In the meantime I think the fix is looking pretty good, just that small crack!
Have you tried using the parts from the pci card in other machines to test the ram and cpu? For the 2nd system, it has both vga in and out, do either do anything?
testing those parts (probably in that laptop, after I fix it) is next on the list. yeah I did try the input button on the 270 (forgot to film it) and it didn't make a difference. but one viewer suggested doing that with an actual vga input going in, so I'll try that next
@@clabretro what about the vga out? I’m curious if either the broken or working system also output simultaneous vga at the same time as internal video?
yeah I *did* try the VGA out of the 270 (and the PCI card for that matter) but no luck. forgot to put all that in the video too, of course!@@JohnKiniston
On Video is not visible is you connected the Athlon CPUs into the socket. Did you slide the CPU in the socket? (it's not enough to just drop them in the socket, sliding is important and not easy)
Yeah the laptop has the typical slider to engage the CPU, and I did that. But the SunPCI doesn't... I did push the chip to engage the pins as best I could (it seems to be the same mechanism, just without the thin sliding layer of plastic typical sockets have, if that makes sense). So I'm still a bit skeptical of the connection the SunPCI is making.
...does that CPU socket move left or right to lock/unlock? if it does, that might be your issue.
The one on the laptop does but the SunPCI board doesn't... but it does feel loose in there. I made sure to push the chip tight in the direction of the socket pins before tightening the fan down, but I'm still a bit suspicious of it.
That was my first laptop! I miss that Compaq, wish I'd held onto it.
nice!
I think those All-In-One thinclients are literly just modified off the shelf monitors with regular off the shelf monitor controllers. That error about the cable not detected was likely just coming from the monitor controller, just like a Dell monitor would show if no input was detected, TLDR; monitor and monitor controller is getting power and funtioning but the thinclient computer compent istelf is not sending a display signal to the controller for the monitor.
Maybe try the VGA input with another computer to see if that even works.
Yeah that's what I'm thinking, all the video components operating just fine.
Is there a connector for a PC speaker? I had an issue on my old Slot 1 board where the BIOS had a slight corruption causing the 'CPU failure' alert to go off, fortunately in my case I was still able to boot into DOS and reflash it which fixed it. I got the fix from one of CuriousMarc's videos.
Is it also possible to pass through a PCI card? Maybe one of those diag cards could give an opcode if the CPU is still executing code, but not getting far enough to initialise communication with the Sun.
No PC speaker connector on the SunPCI that I can find, but yeah... I wonder if I have some BIOS chip issues here.
@@clabretro, I found from experience that the Sun PC on-a-card suffered from the same issue as the old GoldenGate Amiga Bridge Boards, with the BIOS chips going bad after not being powered on for a while. Depending on the make and model, I found that the chips, needed to be re-flashed, or if, the were a weird flash-mask fromtheybrid, replaced. Also, for the Sun PC on a card, the power delivery to the BIOS Roms can be a bit unstable as the Caps age, and the caps that deliver power to the ROM are also part of the chain that powers the CPU and RAM, so if the ROM gets fried, the RAM and CPU may be damaged too. I advise testing the circuits for continuity and the voltages and amperage being in spec.
Did you try hooking a monitor to the vga port on the x86 board? It might show errors to get you going in the right direction. Thanks for the video.
I actually did notice the linksys stack getting bigger, great bit
Do a vrm check , see if any of those mosfetts are ok,
good idea!
Have you tried it on Solaris 9? Maybe the Errors are because the solaris version is too new.
I did try it on Solaris 8 with the same outcome
Have you actually measured the rail voltages on the X86 board? Also, if its regular PCI-64, can't you put it in a donor X86 machine without fear of damaging the host.
I haven't yet but I'm thinking the same direction... get it in a machine where it's convenient to probe around with the multimeter and see what it's up to.
It's not a total failure... there is a status LED on the back of the card that consistently blinks amber (which means "potential problem" according to the documentation), forgot to put that in the video. Will be an interesting one to figure out.
"It's not the hardware, it's not the hardware, it's not the hardware.... OK... It's the hardware...." heartbreaking. The highs and lows of old hardware.
Funny thing with the 270, did you try pressing the input button before taking it apart? It's got a VGA input so you can use it as a monitor
If this is it I'm gonna my rolling laughing
I did, it wasn't that haha. I forgot to film trying that out
@@clabretro That's fair, it was worth a shot to ask! Lol
You have to push the cpu sideways with a flathead screwdriver from one of the slots next to it inside the cpu socket! Otherwise the pins don't make contact. There are 2 slots next to the cpu, directly opposite to each other. When you look at them you'll notice that one of them is nearer to the cpu than the other. This is the one you need to lever with a screwdriver against the cpu! The other one is to remove the cpu from this position. Try it and give me a note whether it worked, I'm 99% sure that's the problem!
so, you're totally right - I did try that a few times but it still didn't work. I think it's part of the puzzle though, that socket is pretty low quality and some of the pins are a little suspect. I can never get the CPU back in as firm as it was when I first took it out either.
I think there's still a good chance there are bad caps or other components at play here, but I might actually just replace the CPU socket entirely when I get some better desoldering gear!
These socket-types are a bit clunky. You need a well fitting screwdriver, maybe you can wedge another one between the cpu and the first one to get more leverage.
Had those in laptops of that era.
Changing sockets is a pain, even with good equipment. I did that several times. Best way seems to be hot air, but with a big nozzle, high airflow and matching temperature (
I don't know what's wrong with UA-cam again.. my comments seem to be disappearing..
weird, it didn't get held for review or anything, I don't see it on my side either. I did see it before it disappeared though just hadn't responded yet! I'll try that jumper to see if I can get a boot screen
@@clabretro YT probably thinks I'm a bot, because I keep editing my comments multiple times. For some reasons, my thoughts start to flow only after hitting the Submit button.. I'm weird like that.
The message on the 270 was generated by the screen itself. It's the same "no signal" message you get on a normal monitor when the VGA is disconnected, or not at the right specification. That really rules out any screen problem. I'd start with checking power on the "main" board onwards. And literally as I finish this comment, you've just said what I've written - almost word for word!! I'm just a mind reader. Ignore me!! 😂
great minds!
also wait what, the thin client has a sim slot??
strange, right? I had no idea until I filmed this video, I had thought those slots were clips holding the case together or something. The Sun Ray 2s also have the same thing. This is all I've found about it so far: www.infoworld.com/article/2656281/sun-rolls-out-new-sun-ray-thin-clients.html
Apparently they were thinking admins could store device config on the sims?
@@clabretro ooh you found *something* about the sims, i went looking and couldn't find anything - at first i thought they would use the sim as some sort of out of band managment... for some reason, lol.
Interesting, I wonder if anyone has actually ever used the sims in the sunrays!
wait that card is meant for a back plain right? i thought they would not work inside a computer? also have the same sunfire v240 great machine.
The SunPCI is just a PCI card, so it goes in those slots. Yeah I love the v240!
what about the VGA on it?@@clabretro
The sun 270 looks like you can use it as a monitor with VGA input? Does that work, is there a way to change inouts from external to the embedded pc?
I did try hitting the input button (forgot to film it) but it had no effect. another viewer suggested actually hooking a vga input up to see what it does which I'll try next
I wonder if you could mod the Sunray to be PoE enabled
probably, I'll have to check the voltage and current requirements. that'd be a sweet project
I have a broken USIII 750 module for my (what was) Dual CPU Sun Blade 1000 which is now unfortunately single CPU! Like you mine worked once!!! and never since! Hard to find these modules cheaply if at all!
The Compaq sounds like it's having a small power delivery issue....like a power rail is missing,
Defo a component lever repair of sorts but definitely fixable with the correct knowledge
yeah also a chance I didn't seat a ribbon cable properly too, I'll update when I get back to it!
What about the other RAM slot on the X86 board? The video looks like you just used that same slot with different RAM sticks?
Yeah good call out, I did actually try the other slot off camera without luck, forgot to include that in the video.
That Linksys stack is pretty, in a mid-2000s tech way.
Part of me yearns for the day when laptops (and cellphones too) had user removable and replaceable batteries.
Nice watch!
Surface mounted electrolytic caps are time bombs
Those yellow capacitors on the PCI-card are suspect to me. Tantalum caps...
I havr a sunray 270 with the exact same fault. Ive tried external video input, also tried the external input with no result either. Did you ever get a fix?
No fix yet, but I haven't looked any further. I'm fairly certain it's something on the main board itself rather than the video circuitry.
Are there other avenues into the X86 board? Like can it be powered in this slot with its own monitor and keyboard so you can see what is happening to it? Does it have any PCI/ISA expansion; perhaps something you can shoehorn a diag board with 7 segment LEDs for boot codes? I don't really know how much special sauce they've put into the BIOS or any other subsystem on this board but it could be interesting. If you do not have the ability, perhaps a colab with another youtuber might be in order. Ask them to flash a new BIOS to a chip for you, if you are suspecting that to be an issue. There might be a website out there with this BIOS saved for posterity.
Something I would try in addition to recapping, is a BIOS reset or clear CMOS jumper if it has one.
Yeah all great ideas. The back VGA has no output, unfortunately. While it's a PCI board itself, I don't think I have access to whatever internal PCI bus it's using somehow for the x86 machine it hosts. There is what appears to be a JTAG header on it, so that could be interesting. I think I need to do a little voltage probing while it's in the machine to see if there is a basic power failure going on.
It does have a diagnostic LED on the back which blinks amber at all times, the manual says this "may indicate a problem," which we obviously have haha.
Nice shirt!
That on screen error message on the sunray 270 looks like a generic monitor message.
I would expect that the display is a commodity monitor and if it doesn’t see a signal it just displays that message.
I would try it without the USB board connected to the monitor. I wouldn’t be surprised if the same message shows up on screen.
Yeah the trick is the display board gets its power through the main CPU board. It's something I'd like to try though, figure out how to power just that board properly to see if it does the same thing.
How do u configure Java cards to connect?
I like the shirt! good video!
Thanks for watching!
What happened to part 2 of the 3650
haven't tried it yet!
I want that shirt you have on. Where can I get one?
cottonbureau.com/people/connections-museum?hierarchicalMenu%5Bcategories__lvl0%5D%5B0%5D=Apparel
why would this SunPCi board be so picky about CPU?
Please try to use a plastic guitar pick or spudger to pry off clips and if that doesn’t work try a thin, wide metal thing… I think your ifixit kit came with some. It pains me to see you use a small screwdriver.
Heh I know, I couldn't find it. I have another kit on the way though, no harm done in the mean time.
P.S. treat yourself to some old guitar plectrums, they make great prying tools.
Yeaaaah couldn't find my plastic smudger. Nothing got hurt, luckily. I've got another kit on the way now though haha.
@@clabretro the plectrums last a lot longer. I still have my first 😆
Better using plastic to pry plastic. If you have the iFixit kit, you have plastic pry too and plastic plectrum.
Yeah couldn't find it, I actually have another kit on the way. The whole time I was thinking "hmm I should really wait for that to arrive" haha. No harm done though.
I'm getting ESD itch in the first three minutes.
if someone adds that PCI support arm to a modern case that still has an optical drive bay it might divorce me from my enthoo pro m lol
thats awesome
Wooohooooooo!!🎉🎉
At least this might needed a fix.
I wish I lived near you, I have the experience and equipment to do some more in-depth repairs, including surface mount chip swapping, etc. If you do want me to take a look at anything, I can share my mailing address on Patreon or something and I'd be more than willing to do the work gratis.
Also, you really should pick up an iFixIt toolkit. They come with very high quality screwdrivers and bits, and lots of plastic spudgers for opening plastic parts like that without causing damage. They are extremely useful!
so weird seeing desktop sized CPUS in a laptop
Nice vid, but I think you’re gonna have to get further into electronics to keep these alive!
agreed haha
Is the Athlon XP still loose in the socket?
If it is loose the CPU has no electrical contact inside the socket!
After inserting the CPU it has to be pushed towards the PCI connector to secure and connect it in the socket! Insert a flathead screwdriver in the little groove on the socket above the CPU and then twist the screwdriver to push the CPU sideways towards the PCI connector. It takes some force because the pins of the CPU have to be pushed in the "connection groove" in the socket.
Take a close look at the socket with a magnifying lens if you don't know what I mean.
I'm sorry but I can't explain it in a better way because I'm German and English isn't my native language.
Very good point. I did actually push the CPU towards the connector like you say, but I forgot to get it on video! The laptop has the sliding style socket, but the SunPCI needs to be pushed like you're describing. It still doesn't sit very tight, so I'm a little suspicious of it still.
@@clabretro In my experience this type of socket needs "unreasonable amounts of force" to connect the CPU properly. Every time I applied reasonable amounts of force the machines acted exactly like the Dell & the SunPCI! I call this kind of state "braindead".
If a Socket A-CPU dies because of overheating (cooler not properly mounted - Socket A-CPUs don't have overheating protection because there is no thermal sensor inside the Die) they usually have a brown spot on the underside (the side where the pins are) in the middle spot that is free of pins.
No meter? No scope? Let me guess, software geek? 😂
hey, how'd you know? 😂 this might be my excuse to finally get a scope.
Please buy a plastic spudger. You're moving this from asmr to horror show
Heh I know, couldn't find mine. I actually have more on the way, the whole time I was thinking "hmm I should have really waited for those to show up." Anyway, no harm done haha.
lol, I'm screaming inside- always see if you can get the part working you're going to be extracting first. Otherwise you're just wasting your time lol
yeah 😆
Invest in some non marring pry tools.
heh yeah... I couldn't find my spudger. didn't break anything though, I have a new set as well.
and seriously, get yourself some spudgers
Couldn't find mine, bought another kit 😂. No harm done though.
crusty
For monitor it is most likely the inverter that provides power to the backlight failing ,it provides just enough power for it to turn it up and than it fails