EEVblog
Вставка
- Опубліковано 11 вер 2019
- Can you find the fault in the IBM PCjr wireless keyboard before Dave does?
A most excellent adventure in troubleshooting.
Forum: www.eevblog.com/forum/blog/ee...
#Repair #IBM #Vintage
Bitcoin Donations: 38y7DE8HEHNj8fGDtUr4PkCn9nWxiorvvy
Litecoin: ML7oQokTwB38bgzzjLDbRV97HKAHuwRfHA
Ethereum: 0x11AceA38DCA9DbFfB4F35f3F746af65F9dED28ce
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
AliExpress Affiliate: s.click.aliexpress.com/e/c2LRpe8g
Buy anything through that link and Dave gets a commission at no cost to you.
Stuff I recommend:
kit.com/EEVblog/
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.com/crypto-currency/
T-Shirts: teespring.com/stores/eevblog
Likecoin - Coins for Likes: likecoin.pro/@eevblog/dil9/hcq3 - Наука та технологія
"You don't want to get into reverse engineering it..."
Reverse engineers it, anyway.
Brian Sullivan But not much. Would have been educational to study the circuit even if working perfectly. This battery powered device from 35 years ago manages to run for a long time on a cheap battery. It has some well designed circuits to consume near zero power in standby yet wake up on any keypress without loosing the data from that first keypress. Lots of modern IoT devices fail to achieve this.
I love that you show all the futzing around, dead ends, and initial circuit misunderstandings. These are the lessons that give the new practitioners confidence that they too can attack these problems, and it's perfectly normal to follow a few fruitless leads before hitting the real problem. Well done Dave!
8:22 It's raining in Daves lab again ;)
Per Hansson, just spit it out...What are you trying to say?
@@v8snail ;-)
Super interesting. Love the fact that you went through the whole process. I don't like it when people fix stuff very quickly based on experience (although i do respect them for it) and then assume that you have the same knowledge they do. At some point in their lives they had to go through this exact same procedure and find out for themselves the 'hard' way.
This is what troubleshooting is about: fumbling through it, retracing your steps, getting lucky sometimes, but all in a methodical manner. Thanks for the videos!
Tantalum caps are the devil.
Had one die in my dynamic signal analyzer that kept the crt from working; it was a dead short. Now another one has failed as well, same symptoms.
I had one fail in my spectrum analyzer also.
Finally, I got a dead Flir A40 that had a blown n-channel fet that burnt some traces because of a shorted tantalum cap; got that back up and running too.
Great video, Dave. Reminded me of some of your older videos from a few years back. Always enjoy watching your videos regardless.
But Dave, you totally missed a chance to blow up a tantalum! THINK OF THE VIEWS!!! (j/k)
Over voltage on tantalums is an absolute party.
Capacitor fireworks display! Woo!
Think of the SMELL! Yuck, that burny metallic scent takes forever to dissipate.
@@Ziferten The smell is somewhat "tantalizing" to some people.
I'll show myself out.
@@Ziferten He's on an island continent, nearly on the other side of the world from me. If that smell propagates THAT far, I'll be genuinely impressed.
I have been tinkering with electronics for 30 years but you still always manage to teach me something. Thank you.
Great video Dave!
Teardown + repair = *thumb up* , great vid Dave! Thanks!
It's the tantalum. I hate those things. They're always failing on power supply boards.
they fail *everywhere*...i recently repaired a RIAA (turntable) preAmp..there had been a bunch of them on the raisl...and guess what, the PS kept failing, 78/7915 run into fail-save. I replaced them all with generic 105°C caps and here we go again, like new....sh!tty TT-Caps!! replace them...every time, everywhere!
@@tubical71 To be fair, they aren't ESR efficient unless you run them close to their rated voltages. But we all know what that means in the long run.
Years of dumpster diving I finally found a working i7 like you with your planted finds.
Thou shalt check Tantalum caps.
build Mr Carlsons capacitor tester :)
What's the deal with that? I've gather it's controversial in some way?
@@EEVblog He's managed to figure out how to test caps for leakage, and potentially predict a future failure. Whole thing done with batteries, instead of the old-school high voltage testers. I understand the plans are on Patreon, which might be the "controversial" bit. Video here: ua-cam.com/video/LhovRIM5xAo/v-deo.html
Not sure how fit it is for low voltage systems such as these.
@@EEVblog Maybe it is the analog witchcraft making it controversial :D
@@EEVblog i built one they are the ducks guts
i really enjoy these repair videos along with the soldering videos.
I once had repaired some over/under voltage protection device, and found that voltage on a simple voltage divider wasn't what it should have been and ended up replacing every component in that part of the circuit (two resistors and comparator) and the damn voltage was still wrong. As the last effort I removed a tantalum that was parallel with R2 of the voltage divider and it worked! After I told my coworker what happend he said that I should try putting it back in beacause it was bizzare that the cap was the fault, and after soldering it back in, it still worked...
Ended up reporting it as "ghosts issues" hahah
Learned so much about troubleshooting from your videos, thanks Dave!
It was not bizarre that the tantalum was throwing off the voltage divider -- same failure mode as in Dave's video. It was somewhat bizarre that reinstalling it didn't recreate the problem -- maybe the heat or lead-bending cured it!
Graham Wideman he said it was bizzare beacuse it was a passive component that died, from experience of repairing other devices we have. Its always active components
@@MrZombiekiller7777 slovenc?
B Bogataj Hrvat
Years ago I heard of somebody fixing a DNS failure by swapping a network patch cable. That's something that is physically incapable of fixing that problem, but did anyway. (Networks are weird sometimes.)
Great video, always nice to see troubleshooting and repair like this! :)
That's an unusual failure mode. Most tantalum cap faults I've found were short cct. Sometimes blowing the top of the caps off which makes it a lot easier to find.
Sounds like a related failure. There's not much that can go wrong if there's a 1M resistor between the cap and the PSU. Not sure why it's voltage dependent though.
That Fn that was lit when Alt-Fn was pressed in DEC VAX times :)
What a great reminder that I'm almost 50 y.o. :) Thank you!
Gotta admit, when I first saw this guy I was put off by his voice. Now though, I not only have a great respect, but actually find him refreshing and entertaining!! Obviously knowledgeable and passionate. It great you give your time for us. Much appreciated, and keep ‘em coming !
Ps. How about telling us a bit about yourself? You’ve got a lot of fans out there
he actually did...watch older videos and also EEVBlog 2 :)
Love these troubleshooting videos!
nice one Dave..glad you did the visual inspection first!...altho...them solder joints looked great for such an old bit of gear...
WacKEDmaN Made by professionals, not amateurs.
'just wanted to use the IC tester'. Yeah buddy, I believe you.
Great find Dave, well done!!! 10/10
love the repairs Dave, thanks!
Great video as always Dave!
I just love your repairing videos
Good troubleshooting video in the mixed domain and quite concise.
Loved this investigation!! now I know many on the bad cap effects! Thanks Dave!
I like watching you work!
I really enjoyed this video - Well done Dave!!
i really like when you do these repair videos!
Need to diagose kb problem on zx specy, this is a great walk thru for a novice like me. Ta for your work!
glad you got fixed i have issues with those caps before
That was some awesome troubleshooting. I know you said this could have been faster but I’d love to do this stuff. Awesome video. New Sub.
Great video!!! Greetings from France Laurent
Hat's off, Dave! So great seeing logic ICs and troubleshooting that found a real problem after all that weird behavior.
Mr. Carlson's Lab on UA-cam designed an advanced cap tester that will check for leakage like that
I am intrested in building it but the details are only avalible to his patrions.
@@TheEPROM9 for $2 a month it is worth it. I built the cap tester and it defiantly would have detected the leakage in that cap.
He has some different testing jigs. One is an old scope that shows an L or some figure for diode checking I think.
@@Mosfet510 It is a curve tracer or an octopus circuit. It is similar to a Huntron Tracker. It plots voltage and current on an old oscilloscope. I haven't built it yet, but it is a simple circuit. They have been around for decades and are kind of an insider secret for repair techs.
I do not even build anything of it but still think that *at least* 2$ each for Dave, Paul and Clive is a *must* for anyone learning about electronics!
Alright this was enjoyable, it took me back, it make me wanna go fix stuff again.
And the expensive tantalum cap was put there for better reliability
Tcll5850 Nope, smaller and more expensive than electrolytic cans back then. They were a luxury component only used when the special properties were needed. Also back then Tantalum wasn't a conflict mineral.
Ah the PCjr. Learned to program in BASIC on that puppy.
Always nice to see some proper component level troubleshooting rather than shooting in the dark. If I wanted to see hitting the common troublemakers by rote until it runs and wondering what the hell actually fixed it I'd just record my own bench.
Awesome fix!!
I love repair videos!
great video dave.thanks mate...
Great video, real edge of the seat stuff. More !!
Watch out! The alien on your ceiling started drooling! What's that goo from 8:21?
The main issue with that keyboard was the bad stabilizators. My BTC also works on a similar conductive rubber dome over PCB design (but with Cherry MX compatible sliders), but at least the wire stabilized keys worked very well after a good clean - which meant complete disassembly, and a bath in a warm, soapy water. The pin stabilizers needed some lubricant in the form of some universal grease (which I also applied to the wire stabilizers), the result was the best feeling keyboard I've ever owned. Probably some better Alps switch would outdo it in almost every way, but I couldn't try any of those.
Great video. Showing so much : )
I think Dave should make a video on how to interpret Australian sayings... I mean I sort of get them, but more explanation would be great. Dave would probably just insert more sayings into his explanations... Fun times to be had...
I remember the IBM PC jr. GOOD JOB!!!
People, if he started with the tantalum and replaced it directly he wouldn't have a video to post. I enjoyed it.
This is what you get Dave for hoping my next project doesn't work ;)
I have to comment about the traces on the board. I have a mixture of 'what where they thinking' and 'oh my god that's art' running through my head, in reference to all those trace paths. guess it was easier to manufacture if nothing was precise. Gotta appreciate the beauty in these things...
Lupine Dream Its amazing how well some of this old stuff performs with these one or two layer PCBs and no ground plane.
The ~400 kHz sounds like the carrier signal for the infrared, which is then modulated by the bit pattern of whatever key is pressed (turning on/off the carrier signal to the IR LED's as needed). Maybe you could do an episode about how IR remotes actually work (gated carrier on the TX, narrow bandpass on the RX), and why some early systems were prone to random receptions by just having sunlight shining on the receiver.
What a video.
Thanks!
that was an awesome troubleshoot! Tubular!
Dave, the methodical videos showing how you would test the components are great, make me wish I pursued EE instead of Network/Desktop engineering.
Thanks. Seems everyone just wanted me to replace the caps and be done :-/
Can't unsee the spit flying at 8:20
Kamagatza Bob's your auntie!
Great troubleshooting vid. Thanks Dave.
Hey Dave, in case nobody else pointed this out yet, the reason for a 0.8mm pcb is most likely because they used this thickness in other keyboards of the day that used buckling springs and capacitive sense. In those boards, the pcb had to be curved, hence the need for flex. I suspect they probably just went with whatever they were used to at the time, even if the flex was not needed here. The PCjr keyboards (both of them) were a bit of a disaster.
I had one of those as a kid. I did a lot of programming on it. I had the technical reference manual and did a lot of low-level programming on it. I built a controller to add a second floppy drive and I had an adapter to use a PC XT keyboard. I learned Turbo Pascal and assembly language on that machine. I wrote a lot of software on it.
The only keyboard I ended up repairing turned out to have a mechanical rather than an electronic fault.
One key kept on repeating, as though being held down. After taking the kb apart and examining the mechanics, I found the problem. The metal bracket which provided the spring return for the key was just slightly kinked. It was barely noticeable, but it was enough so that the bracket was pressing down on the conductive rubber mat and making it contact the PCB track so it appeared as though the key was being held down. I removed the bracket and smoothed out the kink and that was the repair.
@8:22, what just fell on his desk, drool, or something from the ceiling (right below the green power wire on the right)
Spittle from when he said "trace"
HAHAHAHAHA I NOTICED
It's snowing in Oz land :) or dandruff maybe ^^
It's from the Alien hanging on the ceiling...
More like this Dave
Hi! SYC is an Argentinian Company that sells electronic components. It as funny that you've downlaoded that ds. LOL
Hey, I buy my components in SYC! It's a small electronics shop in Buenos Aires.
8:22 liquid falls in right hand corner of bench
Educational and entertaining, too. Winner, Winner, prime rib dinner!
I approve, another bit of vintage computer hardware brought back from the dead. Tanterlums, transistors & diods are ususly the first things I check, caps, depends on when they were made or the type. 60s- valve stuff, early 90s to late 80s & mid 2000s just replace. If its outside those known bad periods then I tend to fire up & test.
Good video tracking down the problem & solving the real issue.
Dave,
I see what you mean! I did a search for the schematic for the keyboard. I found a good "manual" on the Jr but although there WERE schematics for the main board, the power supply, the modem, the monitor, etc. there was NO schematic for the keyboard!!! There was a fairly decent block diagram for it, but no schematic. In case you want to view it here is the link:
file:///C:/Users/user/Downloads/PCjr_Technical_Reference_Nov83.pdf
Also, another of your viewers suggested that you build Mr. Carlson's Capacitor tester....I TOTALLY AGREE !!!! The thing is awesome!!! I checked his UA-cam site and it seems that the information for that device is limited to Patrons ONLY, so you would have to become a patron to get it, but it is WELL WORTH IT !!! His non patron site is listed as "Mr. Carlson's Lab" and you can find it here:
ua-cam.com/users/MrCarlsonsLabvideos
I may have found a link to the regular site that is about the tester:
ua-cam.com/video/LhovRIM5xAo/v-deo.html
Check that out and see if it is what we were talking about.
The link ...
@@0x8badf00d looks like the path to my hard drive was listed instead of the link. Try this one:
ftp://bitsavers.informatik.uni-stuttgart.de/pdf/ibm/pc/pc_jr/PCjr_Technical_Reference_Nov83.pdf
I think it's not just the leaky tantalum cap and the high-value resistor, but that in combination with the ultra-low input drain CMOS logic. If you had that much leakage current with a TTL setup, it would probably not be a problem. That would probably be what Dave meant about it being fine in another circuit design. (It would make for absolutely sucky battery life to use TTL on a battery powered keyboard, of course. That's why they didn't.)
Interesting to see the thought process of an engineer doing repair sometimes. In some regards its very different from a repair tech. Though I assume some of Dave's process was for the added educational video content as well.
It's very easy to remove the yellowing if you want to do so
Cool, classic.
I hate tantalum caps and avoid them like the plague. Even new ones can fail on you after a short time without warning and failure modes are totally unpredictable. I have had ones that started leaking, some failed shorting a rail and others failed open circuit. Add to this the shittiest conditions of mining the resources needed to make these and you end up with the worst components imaginable.
Reverse engineering a product must be interesting. Even though this is pretty basic product, was interesting to watch.
Nice debug! I'm a pretty patient guy, sometimes, but at one point the window might have looked good for a quick fix for jr. A tantalum!? lol jeez.
I think allot here will have had the joy of repairing old vintage computers. Watching you do your magic on that jr keyboard was amazing to watch. :)
The only thing I don't understand is, was that resistor there as a way of coping with tantaal failure mode? I've had the polarized buggers catch fire on me more than once. :/ Or had it another function?
BUT YOU NEED TO REPLACE THE CAPACITORS /s
hahaha, this is classic. Awesome circuit with the wake up circuit.
You must have some kind of enhanced PCJr keyboard, the one I knew was a chicklet keyboard, most horrendous.
Awesome repair..process of elimination
At about 8:22 (go to 8:20 and press play), what is it that falls from the sky? Bottom right hand corner just under the green wire...
Hi Dave, how to make an active ps/2 to USB converter, suitable for old IBM model m keyboards ? like the blue box
Seems I'll have to check the tantalums on one of my not booting laptops.
Will an RJ11 to USB adapter cable allow you to use this keyboard on a modern computer? Thanks!
Thanks Dave ! I did really enjoy this vid. What was that IC testing rig you used?
It's in the video.
@@EEVblog I tried finding it by your statements in the vid but the web site came up as an Asian Casino site. If I'm understanding your Aussie speak it is www.zgecu.com. The TO8662+ doesn't return anything in a search. I'm sure I must be misunderstanding something. Sorry Dave
@@dpyles9396 Try here: www.ebay.com.au/itm/TL866II-PLUS-High-Speed-USB-Universal-Mini-Programmer-10pc-Adapter-ICSP-SPI-Kit/223577232280?hash=item340e3dd398&_trkparms=ispr%3D1&enc=AQAEAAADMKvsXIZtBqdkfsZsMtzFbFsbX3WcW5fmB%2Fx7ZbaZTyexa37kSnlYLTvSLmiaSZNbV55CEYA5sBIKyF25t0uqo6FICU2BR4q29uRY914ZyJFysQkGQKzRVu7MFQ5QIQRYKOparbxIfDSmF%2BL96Eldc5QEeEEr%2FpqCFOOhp7sUwGfTgBPMP%2Bd2flpbBTYW9%2Fn23HJRjuixNIxRqNt0h%2Fc9KqY8cgOeHw32LlflXt%2FrV09lcZxFFEyTCY5pOAMobEJTWH9imO5ncgUAVdd%2BUrUZEBu73e%2BMJ3Lu9kFzC3ZHuZI2DfRXAxja7hqUS9HlyjLQB73AT%2F6G%2FHkG%2BedMRH9JjjErNLOsGSGeD5X%2BK7TTf8lW7pEfY0IcxJFiTx9RgH%2B2rQUjmH2XbUfJm1esRGNoZO%2FGT1zwTJJfwY2N9geNn7SzlRzwUpQtUmDQ7%2FJnUkc0kT0PQnO8%2FdHt43MPYQMJc2HgWi2uqZfpxo%2FiHTx4GMAUycYnebODuT5CwzrYFMRspfKQpFmEyuH3uYAkQOnh%2Foc1jeMEl%2FNek%2FOIvI1nj92a2vzekQIpErgSlLAF7eZlF8iMqtSMtHY%2FE9VcaCHBSbW%2FgSfVsuWP2dKca9XbKDFTLapHQ0CyOaN7c8%2BsFxD0vf71QOFMrHEIBbnJ4m8xu6QONbWApWodosua0B4PEStTei4BtEu3tFx4x0NLlht3qhoxqTx%2BlJTiA47j7UMRuS1K9gBeywigy0WthklRwltjCmSrJaLejHc45Cl02M21xLywc6GQTGRrncj0u9LzViHwq0QX1Sqv8IIE%2BqnRLt1aN3TtQOKnVyd%2BMJSSDN0RLPN0eotUSE0I2EiKJQHy45WkKAucsImWjUU4j5WBQQ9t8DACVtLIKsHacByUw9%2F1g1kWZm9zesUzSWliSJM4zQs14%2BUR%2FmPiZTRAiuCe9B7GO%2B0PtOgXxNaKuYFn54yRpUb02SejXsuTZyl4h3ync9W0pYBslgVvvzB6uL7ckCy%2F8VrfKZDrqMpQO4Mc3T6LSjBf%2By1MG6LBD%2FwWZfmENIWZez76Lt5f63LYvf8QHp68V5O0K%2FKV11Hw6TqQiwCuJA%3D%3D&checksum=22357723228077783c4819fc414d94215350925120f0&frcectupt=true
@@dpyles9396 Sorry, probably a bad example below. They are available much cheaper than this, less than 20 AU and well worth the money.
What dropped onto the bench at 8:21?
I don't know if anyone has pointed this out, but while that keyboard may seem low quality compared to other IBM keyboards from the era (and it is, tbh), it's actually the really nice keyboard for the PCjr. The standard one was a chiclet keyboard. Apple makes nothing but horrible keyboards now, but they're still not as bad as chiclet keyboards.
when i need to look up old TTL/CMOS chips datasheets the 1st couple google hits are ususally pdf from old, and lovely, TI datasheet scans converted into .pdf......wired that your search resuslts are so different....
TubiCal The TI website is pretty good also, usually my first stop.
"With no cap. May not actually be the cap, may be something else. But the cap is making is worse." Which is why I went into software :)
Chris Boyce Do you regret it? Capacitor failures are so much simpler than debugging intermittent failures in complex production code.
Hi Dave, these contacts looks good but thay can be gold. Search on youtube electroplating gold contacts, and plese tell us what you think about procedure. Maybe some review. Thank you!
when i first heard the problem i thought: there's gotta be some kind of latching circuit that's switching off the IRLEDs but not switching over to allow the wired setup to work, which relies on that circuit to know when to unlatch the first circuit to switch the IRLEDs on... and i was thinking: bitrot? stuck latch or flip-flop?
Is it possible that the computer once connected to the keyboard sends a signal to the infrared receiver waiting for a signal from the infrared transmitter ? I've seen those keyboards a long time ago, but never experimented with one. Seems like a weird idea anyway.
There is no ground plane or shielding. I wonder if a modern office environment with switch mode power supplies, etc causing something to float.
Syc is an Argentinian electronic store! I buy chips there
Love Keanu in the thumbnail.
Party on dude
8:21 nice loogie
Rawr!