I have to say, in re: the title, that "improper grounding due to removed RF shield" is indeed a cause of malfunction in one of Adrian's computers that I'd believe in.. :P
It always drives me crazy when he does that. I repair a lot of electronics from the late 80’s to mid 90’s, and quite a few circuit board designs use the rf shield to connect ground planes like this. It’s really not that unusual.
Hi Adrian, the separation of grounds was indeed for RF shielding but I suggest you solder only one "bridge" to the external ground to avoid "ground looping". You would notice this also by hearing strange noises through the speaker. It probably won't cause any noticeable difference but in the case of any surges a star ground is always a rule of thumb. Cheers and keep up the great vids!
So it helped in shielding? So only circuitry that needed shielding was put on that ground plane? But anything you plug in externally in that SCSCI port undoes that. Very confused on the purpose of dual ground planes. But I was thinking that likely you only needed to bridge it in one place.
Do you think he would get power supply switching noise and other artifacts? CRTs can generate a ton of noise too. IIRC, early 1990's macs developed a reputation for good quality audio production. I wonder if isolating analog audio ground this way contributed to that?
@@8BitRetroJournal As pragmax suggests the purpose is to reduce noise from the digital circuitry. Not sure if it's the exact right way of putting it, but the sharp edges of digital signals cause noise that can be picked up by the ground plane (I know, it's an oversimplification of the issue). Having a separate ground for the analogue parts of the circuit means that noise can't cause interference with the audio signals. So yeah, connecting an external device which shorts these two ground planes could also cause the digital noise to to be picked up by the audio circuitry, which may also be the case with Adrian's modifications. However it's a little more complicated than that, since you also need to take into account return current paths, and that's where having a star ground helps. Connecting an external device would also only connect the two ground planes at a single point, similar to a star ground, so that might not be so bad.
Haven't watched the whole video yet, so not sure if it has anything relevant to this, but: 1) if you're worried about bridging grounds partially via audio connections, bear in mind that sound connections commonly have capacitive isolation, and 2) "shield" grounds commonly have little to no noise on them in the first place because they aren't used for signaling at all.
Upon watching a bit further, I remembered something about SCSI: it uses bus termination, so there should be little to no noise reaching ground from the SCSI bus itself.
@@frankowalker4662 and now you’ve recapped a recap of a recap of recap which I have just recapped. Since this can go on forever I suggest we now put a cap on this conversion.😃
Commenting to raise this. It’s pretty common to keep analog and digital grounds separated except for some central star ground point. Think of a PC case, for example. The whole case is grounded to the PSU. The motherboard grounds to the case via the screw posts. Drives ground their chassis via metal to metal contact to the case. The Pro Audio Spectrum 16 sound card uses the ISA bracket for its audio ground. (I found this out when I removed the bracket after replacing old corroded jacks with new ones that were slightly shorter.)
@@nickwallette6201 I weep when i see modern ATX enclosure front panels connect audio ground and USB ground together. When you get interference from increased load, no surprise. The mainboards are remarkably careful about star-grounding the audio unit today.
Using the shield as the return path is not very clever though. It creates an enormous loop for the high impedance and noise sensitive audio signals. Audio frequencies are so low that they don't cause a problem for RFI so it's not necessarily a problem to involve the shield in the audio circuit. Until some other RF signal decides it would like to join in 😂
I have absolutely no knowledge on electronics, apart from the very theoretical basics I learned in High School. Your videos, though, made me learn so much more about retro technology and how it works in practice, that I managed to repair a C64 and a C64C. I still can’t believe it, this channel is a blessing!
Dunno but pretty sure he’d be the guy that every tech related workplace has who deliberately keeps really quiet about how much he really knows otherwise he’d be swamped with requests to fix shit (on his own time and with no extra duties pay ofc)
That RF shield, and digital/analog ground split was the best way to prevent EMI from exceeding the emission rules of the day. Split grounds also helped keep digital noise from getting into the sound signals. It works if done right. I use split grounds in my microprocessor projects when I use the A/D and D/A in the microcontroller. It keeps the analog signals separate from the digital signals. Apple probably had to work hard to meet the emissions requirements and that was their solution. Great video!! Thank you!!
Having a split ground plane to minimise interference between digital and analogue circuitry is standard practice. Using the RF shield to connect the two was probably just the cheapest/most convenient solution.
Every time you remove an RF shield, I struggle... Like, they actually perform a function (not just using them for grounding and this kind of weirdness), but also, they block RF emissions (which can screw with cell phone, Wi-Fi, and Bluetooth reception, among other things). Like even if it has no RF specific components inside, some of those old CPUs and ASICs can generate radio noise, and noisy power rails can act as antennas. It's part of why they have those RF shields in the first place...
Exactly! His I'll will towards shielding will probably bite back in much more subtle ways than just grounding issues. I guess I have to hope he lives out in the middle of nowhere
@Adrian It's interesting to hear you reflecting on these compact Mac's. My SE Superdrive was the computer that got me through university and I remember loading up Netscape & Z-Term and getting it online for the first time. It had a 68000 16Mz accelerator in it but was still unbearably slow. Even Maelstrom was sometimes my game of choice to distract me from the all those papers I wrote on that little Mac. A colour classic would have been my dream setup too, and thanks to your channel I have a Color Classic II on its way to me right now from a Japanese auction website. Who knows if it will work or even survive the journey, but all I know that I need to get it out of my system and find out. If I can eventually coax Maelstrom to work on it then that will bring some nice closure too. :)
Congratulations on figuring out the sound problem! It's blowing my mind, I'm sure I'd never figure it out if I had to debug this system. The destroyed shield made sure if you assemble the machine it'll never work on it's own.
I never owned a Macintosh classic until a few years ago (now have Classic, Classic 2, and SE/30). Really love the compactness and the HANDLE! First time seeing the Color Classic 2 in action; while I'm not a fan of the case changes, the colors do make a tremendous difference. Nice job on the restoration, glad you stuck it out until the end!
The RF shield revenge! Very surprise about the main cause of non working internal audio. Excellent serie, great job of restauration/troubleshooting for a wonderful machine at the end. I’m waiting for the next série now lol
That jelly donut shirt cracks me up every time... I only found your channel a few months ago, not big into Macs but I like that people restore these old computers, my elementary school had a few older units knocking around when they got the newer G3's. That startup "bong" always reminds me of sneaking off somewhere during lunch to play the Mac version of SimAnt or something on one of the older color Mac's they had around.
(13:49) It is quite common to have two separate grounds, an analogue ground and a digital ground. Typically these grounds would be joined together at a single common point (ie. Star Ground) in this case the RF shielding. I have repaired many switch mode power supplies that have an isolated ground, and yes, you must connect them together when the supply is on the bench, under test, or it will not work.
Ground path through the RF shield was used to avoid ground loops. Ground loops are created when having more than one path towards the ground. Although you might thing this would be good, in electronics, such a configuration creates an antenna. As a result of the loop, the circuit will pic up interference. The frequency of that interference is related to the length of the loop and you can refer to it as the characteristic frequency (or range of frequencies). In case this is within the operating range of your circuit, then you will be picking up noise. I would do what other people already suggested, only connect the two grounds in a single point.
Hi Adrian I really look forward to seeing your work. I have been teaching myself about electronics and the repair of. You are a great instructor and bring all your work down to a understandable level Thanks again.
I've done many designs where single point grounding was used on a PCB (star point) especially when using high speed ADCs, but it seems like bad design to rely on external mechanical ground connections for a circuit to work. Perhaps this was the only way to get the board to pass FCC testing - sometimes when a product fails a test you get desperate!
Some DLP projectors also use the shielding as an electrical connector on the power board. A loose screw would keep it from powering up. Specifically all the Viewsonic PJD models.
We had an old Mac when I was in elementary school. It eventually got raffled off, and I lost. I was devastated. The "rich kid" got it. I would have loved getting it, but I hope Ben got some good use out of it. Love your videos. Thanks for the updates.
I fully enjoyed your fault finding solution in the sound. So funny finding out the shielding was the problem. In the future removing the shielding may be the problem why things don't work . Lesson learned. Great watching look forward to future videos.
Adrian's channel is what convinced me to just enjoy retro computers vicariously - I bought one (Tandy Model 200) and it works, but it will be the one and only. Otherwise will stick with modern neo-retro computer recreations, FPGA MiSTer, and the like.
Using the RFI shield as the audio ground would probably have been done to help stop radiated emissions from a cable plugged into the audio jack. If the audio ground were connected to digital ground, any noise on the digital ground would be able to radiate from the ground conductor of the cable plugged into the audio jack.
RF shields are important. Example; my C128's RF shield doubles as on large heat sink! They are very important in Apple Color Classics. Concerning that Apple didn't use a ground cable!! Great video Adrian!
One reason it might be so slow is that JPEG uses quite a bit of trigonometry and it might be that the FCOS instruction isn't available with that CPU (I think it was added to the base CPU with the 68040) which would mean the math has to be handled via software emulation.
The split ground-plane was to isolate the external static risks from getting to the chips. Essentially, equivalent to suspenders+belt. My first mac was a Color Classic. One game that worked great on my CC was Mortal Pongbat.
Lol, those kind of ground failures can drive you completely crazy. And then the feeling after hours of searching when you solve the riddle and to find out how ridiculous this is. 😅. Well done and very entertaining video. thx.
Another example of a split ground relying on an RF shield for connectivity is the cartridge port on a ColecoVision game console. Pin 13 of the cartridge port replies on a mechanical connection to the lower RF shield to be electrically connected to the central ground.
I remember my older brother playing Maelstrom on his Mac, and I think he even talked to the developer of it about Mac programming in general. It's also available in many Linux repos.
I've also noticed in Atari 2600's (4 switch models) the RF shielding is important for keeping a good ground otherwise you get additional noise on the RF modulator.
I was amazed at the quality of the picture which wasn't what I was expecting at all, I thought it would look much more basic. We had some of these Macs in the school lab although I don't know exactly which model but definitely didn't have much ram and had the black and white screens which I think is why I was so impressed to see such a detailed image as I just think of the basic mono graphics the ones I used had,
Cool video adrian, excellent troubleshooting - you have a new subscriber :) i have a bunch of old macs - mine are all PowerPC machines but I love seeing these older classic macintoshes, i hope to get some soon.
26:50 I remember reading a copy of Games-X maybe around 1990, and being blown away by digitised photos from Gunship 2000 and similar games. It seemed like space technology, after years of low-colour / low-resolution life on the C64 and Atari ST.
Coleco did the same thing on one of the cartridge pins where the bottom RF shield provides the ground to the pin. If you don't have the bottom RF shield installed, the system and games still work but there is at least one home brew game release that uses that ground on the pin in order to operate properly. So there is at least one other system that did something similar.
I think it's because those exterior connector would touch the shield anyway, but if you connect the grounds on the other side you'll have a loop. Since you don't want loops to exist (as an external magnetic field can induce a current in a loop), they avoided connecting the groundplane to the shield on the digital board side.
Adrian, Apple did not split the analog ground plane from the digital ground plane to reduce EMI. Splitting the ground planes is a standard tactic to reduce noise in the analog circuitry for several reasons(*). When digital circuitry switches, especially CMOS, it will briefly draw a much larger "gulp" of current to switch voltages (needs to charge/discharge various capacitances). Because of ohm's law and the fact that real ground planes have nonzero impedance, this will actually produce a voltage spike on the parts of the ground plane the current flows through, which screws with all circuitry, but especially analog circuitry -- digital circuitry is somewhat protected by noise margins on logic levels. This is mitigated to a certain extent by bypass capacitors, which essentially spread the current spikes over a longer period of time -- the same amount of current will eventually need to flow in order to recharge the bypass caps. However, the current flow from digital circuitry still tends to produce noise on analog circuitry. Splitting the ground planes forces the return currents from the analog and digital circuitry to take completely different paths (**). Usually, the ground planes will be connected on the PCB, perhaps at a "star ground", but using the chassis or RF shielding is not terribly uncommon. * Note that this sort of design gets very complicated and there are disagreements about when to or if one should use split ground planes. However, my goal is to just explain why they were used here, not talk about what should be done for new designs. Please don't take this as advice on how to design PCBs. ** Note: modern sources advocate analyzing return paths instead of blindly splitting planes.
I wonder if simply installing a scsi terminator on the back would be a suitable solution? Especially for those who don't want to solder the ground planes together. Or just have the RF shield installed I guess.
Hi Adrian, love your channel. You mention different parts for your videos at the beginning. RUclips does a great job of suggesting the parts in order if it's part of the title. Something to consider if you can do that. I know you're probably bouncing around projects and other things so might be hard to keep track which part you're on :)
You can actually browse the web in a relatively modern way even on a classic mac like this. The trick is to get around the encryption which can be quite easily by connecting through a squid proxy that essentially performs a MITM attack and just strips the encryption. I spent a ... reasonable amount of time figuring it out, but I'm sure most of the people that would ever care to try wouldn't have much trouble with it.
I have a color classic that has NEVER booted up. After seeing you working with one, I took it off the shelf (mostly to see what motherboard it had, no Mystic board in there!) and I discovered it had NO RAM onboard. I need to find my "Box-o-Ram" and see if I have anything compatible in there. I'll let you know what I find out!
My Color Classic motherboard looks a bit different. For me I had the same audio issue as you BUT also a bad solder joint on the leg of the TI audio chip.
I did the VGA mod on one of my Mystic CC today. It was very easy and works perfectly. Sadly the Comm-slot and LC-PDS cannot be used simultaneously. I had to connect an AppleTalk to Ethernet adapter to get the CC on the net while having the Apple //e card installed.
Great vid.. Glad to see that Color Classic happy again... Personally, I think just for nostalgia reasons, the Color Classic isn't in my top tier. I like my Classic Macs in B/W. ;-) But great to see it running and with Ethernet! And that 040 was smoking for that era!!
Called it at the first mention of audio not working through speaker, but does work in headphone jack. Bad connection in the internal jack. There is a mechanical 'switch' that turns off the speakers when you insert a headphone jack. Most common issue with that kind of hardware. Troubleshooting 101: Follow Olcums Razor.
There was probably rf pickup on the ground plane on the motherboard. The sheild creates a faraday cage to box in the RF. That's why the external connectors are attached to the sheid ground.
Y’know, they might not have done it for RFI reasons, it might just have been to reduce the noise in the audio. A lot of modern motherboards have identical looking isolation around the audio circuitry, visible thru the solder mask as a gap in the ground plane. They boast that it stops it from picking up chirping noises from the data bus, but obviously those conduct thru the conductors rather than thru the air (hence, not necessarily for RFI). But I presume the modern motherboards get their ground from the brass standoffs or something, as they don’t have such an external shield.
Hi Adrian, long time viewer, I especially enjoy your troubleshooting and repair videos. In part #3 you 3D printed a new rear panel for your custom Color Classic. What brand is the 3D printer and would you recommend it ?
Hmm. I wonder if the second ground was to isolate the noise signal from "digital blipiness" coming through the speaker? Just a thought. Great content as always!
Just FYI - when you were looking at the scope trace of the audio signal, it was jumping up and down when you plugged in the external speakers. I have a theory on what caused that. Audio circuits are AC coupled, at least usually. It’s always nice when you have +/- voltage rails so the AC can be centered around 0V, but quite often it’s using a 5V DC supply. This means you have to center the AC waveform around some DC bias voltage to make sure the negative swing of the signal can go below the reference point. You would then use a capacitor to remove the DC bias and re-center the waveform around 0V. Hence, AC coupled. (Even if you have a -5V rail, you still need to AC couple to make sure you don’t have a ground potential difference. Moreover, in many cases the -5V rail is only used for amplification and the D/A is still DC biased. I haven’t looked at the Mac schematics though.) The “outside” or jack side leg of that cap needs to have a ground reference, since there’s no DC flow through the cap. Usually, the input stage of whatever you’re plugging into (like a powered speaker or a mixer, etc.) will have a resistor to ground. This sets the input impedance to something between 1K and 100K, often around 10-47K. Instrument and mic level inputs are higher (maybe 1M), balanced pro audio is 600R, but you get the idea. That R to ground pulls the AC coupled signal to ground, but isn’t strong enough to short the output, so you get a voltage swing around Gnd. So what you may have been seeing on the scope when you plugged in the audio cable is the external ground being applied to the AC coupled waveform, bringing the bias point down. Particularly if the ground reference was missing on the internal speaker circuit path.
well on most motherboards from pc's , at leasy pentium 2-4 era audio was also insolated grom digital ground they had only connections made by single 0 ohm resistor and had it's own power rail filtration caps all that was made to stop digital noise interference to analog aution which would produce high pitch noise hiss , but also audio interfering with bass sounds pulling ground up and making digital errors on the other side i belive in this case shield worked as low resistance resistor link that would stop crosstalk along with rf interference interesting bug non the less
![Macintosh IIsi Fix+Recap, AppleTalk stuff, NuBus and PDS cards and other Mac goodies [MAXI SIZE]](http://i.ytimg.com/vi/Dniqwec7_ko/mqdefault.jpg)
![Macintosh IIsi Fix+Recap, AppleTalk stuff, NuBus and PDS cards and other Mac goodies [MAXI SIZE]](/img/tr.png)
The ghosts of all those RF shields you’ve tossed out have had their revenge! ;)
I have to say, in re: the title, that "improper grounding due to removed RF shield" is indeed a cause of malfunction in one of Adrian's computers that I'd believe in.. :P
to be fair..this machine's RF shield was so rusty it HAD to go.
Fellow RF shield thrower here. Apple shenanigans 😂
It always drives me crazy when he does that. I repair a lot of electronics from the late 80’s to mid 90’s, and quite a few circuit board designs use the rf shield to connect ground planes like this. It’s really not that unusual.
Adrian, don't let this experience discourage you to toss out future RF shields!
This feels like nothing but divine retribution for serial RF-shield snubber Adrian.
When/if you ever open a merch store, the first T-shirts available should say, "It works! It freaking WORKS!"
❤️ when you say that.
Right up there with that classic film line, "It's alive, it's ALIVE!" (1931 Frankenstein - for those that don't watch B&W films)
"Hello, Chip Dippers!" "Welcome back, Cave Dwellers." "Hi, it's Jan Beta." "IT WORKS! IT FREAKING WORKS!" I know who has my favourite catchphrase 😄
Hi Adrian, the separation of grounds was indeed for RF shielding but I suggest you solder only one "bridge" to the external ground to avoid "ground looping". You would notice this also by hearing strange noises through the speaker. It probably won't cause any noticeable difference but in the case of any surges a star ground is always a rule of thumb. Cheers and keep up the great vids!
So it helped in shielding? So only circuitry that needed shielding was put on that ground plane? But anything you plug in externally in that SCSCI port undoes that. Very confused on the purpose of dual ground planes. But I was thinking that likely you only needed to bridge it in one place.
Do you think he would get power supply switching noise and other artifacts? CRTs can generate a ton of noise too. IIRC, early 1990's macs developed a reputation for good quality audio production. I wonder if isolating analog audio ground this way contributed to that?
@@8BitRetroJournal As pragmax suggests the purpose is to reduce noise from the digital circuitry. Not sure if it's the exact right way of putting it, but the sharp edges of digital signals cause noise that can be picked up by the ground plane (I know, it's an oversimplification of the issue). Having a separate ground for the analogue parts of the circuit means that noise can't cause interference with the audio signals. So yeah, connecting an external device which shorts these two ground planes could also cause the digital noise to to be picked up by the audio circuitry, which may also be the case with Adrian's modifications. However it's a little more complicated than that, since you also need to take into account return current paths, and that's where having a star ground helps. Connecting an external device would also only connect the two ground planes at a single point, similar to a star ground, so that might not be so bad.
Haven't watched the whole video yet, so not sure if it has anything relevant to this, but:
1) if you're worried about bridging grounds partially via audio connections, bear in mind that sound connections commonly have capacitive isolation, and
2) "shield" grounds commonly have little to no noise on them in the first place because they aren't used for signaling at all.
Upon watching a bit further, I remembered something about SCSI: it uses bus termination, so there should be little to no noise reaching ground from the SCSI bus itself.
I love this joke "Just to recap, I recapped the motherboard!!!"
He Re-capped the Re-cap. :)
@@frankowalker4662 and now you’ve recapped a recap of a recap of recap which I have just recapped. Since this can go on forever I suggest we now put a cap on this conversion.😃
@@Charlesb88 LOL.
This joke will scar me for the rest of my life, JUST TO RECAP.
You have to recap right? It's food, I mean for the motherboard; I guess 🤷
15:30 - Split ground planes are to prevent digital logic ground currents from contaminating the lower level analogue circuitry...
Commenting to raise this. It’s pretty common to keep analog and digital grounds separated except for some central star ground point.
Think of a PC case, for example. The whole case is grounded to the PSU. The motherboard grounds to the case via the screw posts. Drives ground their chassis via metal to metal contact to the case.
The Pro Audio Spectrum 16 sound card uses the ISA bracket for its audio ground. (I found this out when I removed the bracket after replacing old corroded jacks with new ones that were slightly shorter.)
@@nickwallette6201 I weep when i see modern ATX enclosure front panels connect audio ground and USB ground together. When you get interference from increased load, no surprise. The mainboards are remarkably careful about star-grounding the audio unit today.
Using the shield as the return path is not very clever though. It creates an enormous loop for the high impedance and noise sensitive audio signals.
Audio frequencies are so low that they don't cause a problem for RFI so it's not necessarily a problem to involve the shield in the audio circuit. Until some other RF signal decides it would like to join in 😂
That’s cool! I designed that back panel and posted it on 68kmla.
Since we're being clickbait-y :) "RF Engineers get revenge on Adrian with this ONE simple trick!"
"The SECRET TRICK that Apple does NOT WANT YOU TO KNOW!"
12:49 thank me later. ;D
This simple trick to repair ALL your old hardware the support doesn't want you to know.
I have absolutely no knowledge on electronics, apart from the very theoretical basics I learned in High School. Your videos, though, made me learn so much more about retro technology and how it works in practice, that I managed to repair a C64 and a C64C. I still can’t believe it, this channel is a blessing!
I wonder what Adrian does for a living? He is so knowledgeable and hardly ever seems to get completely stumped.
Dunno but pretty sure he’d be the guy that every tech related workplace has who deliberately keeps really quiet about how much he really knows otherwise he’d be swamped with requests to fix shit (on his own time and with no extra duties pay ofc)
I got a feeling that he’s an electrical engineer or graphic designer like his dad
Hasn’t he said before that he’s in IT? He just has really good troubleshooting skills.
Yeah, electrical engineer was my guess.
As the guy in SimCity 2000 says: "YOU CAN'T CUT BACK ON -FUNDING- RF SHIELDS! YOU WILL REGRET THIS!"
That RF shield, and digital/analog ground split was the best way to prevent EMI from exceeding the emission rules of the day. Split grounds also helped keep digital noise from getting into the sound signals. It works if done right. I use split grounds in my microprocessor projects when I use the A/D and D/A in the microcontroller. It keeps the analog signals separate from the digital signals.
Apple probably had to work hard to meet the emissions requirements and that was their solution.
Great video!! Thank you!!
Only stupidly simple problems take so long to diagnose. That’s something I’ve learnt over the decades
Just plug in a Zip drive, problem solved. ;D
Having a split ground plane to minimise interference between digital and analogue circuitry is standard practice.
Using the RF shield to connect the two was probably just the cheapest/most convenient solution.
Every time you remove an RF shield, I struggle... Like, they actually perform a function (not just using them for grounding and this kind of weirdness), but also, they block RF emissions (which can screw with cell phone, Wi-Fi, and Bluetooth reception, among other things). Like even if it has no RF specific components inside, some of those old CPUs and ASICs can generate radio noise, and noisy power rails can act as antennas. It's part of why they have those RF shields in the first place...
Exactly! His I'll will towards shielding will probably bite back in much more subtle ways than just grounding issues. I guess I have to hope he lives out in the middle of nowhere
Don't forget the other way around. Noise can screw with circuitry, especially analog circuitry, but by no means is digital circuitry immune.
@Adrian It's interesting to hear you reflecting on these compact Mac's. My SE Superdrive was the computer that got me through university and I remember loading up Netscape & Z-Term and getting it online for the first time. It had a 68000 16Mz accelerator in it but was still unbearably slow. Even Maelstrom was sometimes my game of choice to distract me from the all those papers I wrote on that little Mac. A colour classic would have been my dream setup too, and thanks to your channel I have a Color Classic II on its way to me right now from a Japanese auction website. Who knows if it will work or even survive the journey, but all I know that I need to get it out of my system and find out. If I can eventually coax Maelstrom to work on it then that will bring some nice closure too. :)
Did you get your Color Classic running?
How did it go?
I’m currently restoring my old Color Classic, and I’m looking forward to the day when I can say, “It works! It freaking works!”
Congratulations on figuring out the sound problem! It's blowing my mind, I'm sure I'd never figure it out if I had to debug this system. The destroyed shield made sure if you assemble the machine it'll never work on it's own.
"Just for a quick recap I took the board and recapped it" nice, nice.
I never owned a Macintosh classic until a few years ago (now have Classic, Classic 2, and SE/30). Really love the compactness and the HANDLE! First time seeing the Color Classic 2 in action; while I'm not a fan of the case changes, the colors do make a tremendous difference. Nice job on the restoration, glad you stuck it out until the end!
The RF shield revenge! Very surprise about the main cause of non working internal audio. Excellent serie, great job of restauration/troubleshooting for a wonderful machine at the end. I’m waiting for the next série now lol
"It Freakin, Freakin works", that's one happy Adrian, right there.
I think we need a T-shirt with this on
That jelly donut shirt cracks me up every time... I only found your channel a few months ago, not big into Macs but I like that people restore these old computers, my elementary school had a few older units knocking around when they got the newer G3's. That startup "bong" always reminds me of sneaking off somewhere during lunch to play the Mac version of SimAnt or something on one of the older color Mac's they had around.
(13:49) It is quite common to have two separate grounds, an analogue ground and a digital ground. Typically these grounds would be joined together at a single common point (ie. Star Ground) in this case the RF shielding. I have repaired many switch mode power supplies that have an isolated ground, and yes, you must connect them together when the supply is on the bench, under test, or it will not work.
WARNING: Do never ever remove a RF shield from a motherboard. Sometimes it's sounds bad. ;-)
Thank you for this another amazing video.
Ground path through the RF shield was used to avoid ground loops. Ground loops are created when having more than one path towards the ground. Although you might thing this would be good, in electronics, such a configuration creates an antenna. As a result of the loop, the circuit will pic up interference. The frequency of that interference is related to the length of the loop and you can refer to it as the characteristic frequency (or range of frequencies). In case this is within the operating range of your circuit, then you will be picking up noise.
I would do what other people already suggested, only connect the two grounds in a single point.
That was an excellent bit of troubleshooting Adrian. Great job!
Adrian, glad to see you finally fix the sound issue. Like others below I had a chortle when it was the RF Shielding :)
"You know what I do to RF shields...into the trash!"
It was only a matter of time before an RF shield got its revenge on you
Hi Adrian I really look forward to seeing your work. I have been teaching myself about electronics and the repair of. You are a great instructor and bring all your work down to a understandable level Thanks again.
Wow-THAT is CRAZY! So cool you figured that out!
I've done many designs where single point grounding was used on a PCB (star point) especially when using high speed ADCs, but it seems like bad design to rely on external mechanical ground connections for a circuit to work. Perhaps this was the only way to get the board to pass FCC testing - sometimes when a product fails a test you get desperate!
Bad design practice / 90s + apple engineering seem to go hand in hand
Some DLP projectors also use the shielding as an electrical connector on the power board. A loose screw would keep it from powering up. Specifically all the Viewsonic PJD models.
We had an old Mac when I was in elementary school. It eventually got raffled off, and I lost. I was devastated. The "rich kid" got it. I would have loved getting it, but I hope Ben got some good use out of it. Love your videos. Thanks for the updates.
So happy you figure this out Adrian !! Nice work !!
I fully enjoyed your fault finding solution in the sound. So funny finding out the shielding was the problem. In the future removing the shielding may be the problem why things don't work . Lesson learned. Great watching look forward to future videos.
Oh man, Maelstrom. *That* brings back memories.
(06:02) The audio is AC the 1 volt DC is the bias voltage... The 1-volt bias is NOT an A/C signal, it is very clearly DC.
Adrian's channel is what convinced me to just enjoy retro computers vicariously - I bought one (Tandy Model 200) and it works, but it will be the one and only. Otherwise will stick with modern neo-retro computer recreations, FPGA MiSTer, and the like.
I loved this series very much. Thank you. 😀
Well done in finding why the sound would not work and sound advice for anyone else having the same problem. 👍
Using the RFI shield as the audio ground would probably have been done to help stop radiated emissions from a cable plugged into the audio jack. If the audio ground were connected to digital ground, any noise on the digital ground would be able to radiate from the ground conductor of the cable plugged into the audio jack.
RF shields are important. Example; my C128's RF shield doubles as on large heat sink! They are very important in Apple Color Classics. Concerning that Apple didn't use a ground cable!! Great video Adrian!
One reason it might be so slow is that JPEG uses quite a bit of trigonometry and it might be that the FCOS instruction isn't available with that CPU (I think it was added to the base CPU with the 68040) which would mean the math has to be handled via software emulation.
The delay in opening the jpeg isn't because of the image compression, it's primarily the reduction to 256 colours. That is really processor intensive.
The split ground-plane was to isolate the external static risks from getting to the chips. Essentially, equivalent to suspenders+belt.
My first mac was a Color Classic.
One game that worked great on my CC was Mortal Pongbat.
Lol, those kind of ground failures can drive you completely crazy. And then the feeling after hours of searching when you solve the riddle and to find out how ridiculous this is. 😅. Well done and very entertaining video. thx.
Thanks for this Color Classic repair series, it was really fun to watch!
Another example of a split ground relying on an RF shield for connectivity is the cartridge port on a ColecoVision game console. Pin 13 of the cartridge port replies on a mechanical connection to the lower RF shield to be electrically connected to the central ground.
Hello from the future. This was a great fix of a great looking mac. Thanks
Awesome series and I loved it. Our first computer was a black and white mac around 1991
I remember my older brother playing Maelstrom on his Mac, and I think he even talked to the developer of it about Mac programming in general. It's also available in many Linux repos.
I've also noticed in Atari 2600's (4 switch models) the RF shielding is important for keeping a good ground otherwise you get additional noise on the RF modulator.
"Additional" No can do, meaning there is no headroom for that on them. heh - cool figuring it out.
Congrats on figuring out the sound issue!
I was amazed at the quality of the picture which wasn't what I was expecting at all, I thought it would look much more basic. We had some of these Macs in the school lab although I don't know exactly which model but definitely didn't have much ram and had the black and white screens which I think is why I was so impressed to see such a detailed image as I just think of the basic mono graphics the ones I used had,
Conclusion: RF shields may be useful! LOL! Love your content Adrian, thank you for the videos!
I think Dave on Eevblog discusses the split planes approache in one of his latest pcb review videos.
LOL. When you tossed that RF shield (as bad as it was) I was like "uh.. i think you might need that ! It's not just Rf shielding!"
Adore your tenacity. I would have given up on this long ago😁
Nice EEVBlog 121GW Multimeter. I've been trying to buy one but they are unobtanium now.
Cool video adrian, excellent troubleshooting - you have a new subscriber :) i have a bunch of old macs - mine are all PowerPC machines but I love seeing these older classic macintoshes, i hope to get some soon.
26:50 I remember reading a copy of Games-X maybe around 1990, and being blown away by digitised photos from Gunship 2000 and similar games. It seemed like space technology, after years of low-colour / low-resolution life on the C64 and Atari ST.
Coleco did the same thing on one of the cartridge pins where the bottom RF shield provides the ground to the pin. If you don't have the bottom RF shield installed, the system and games still work but there is at least one home brew game release that uses that ground on the pin in order to operate properly. So there is at least one other system that did something similar.
I think it's because those exterior connector would touch the shield anyway, but if you connect the grounds on the other side you'll have a loop. Since you don't want loops to exist (as an external magnetic field can induce a current in a loop), they avoided connecting the groundplane to the shield on the digital board side.
That is one awesome vintage MAC. Awesome job with all of the repairs as well!! The video looks awesome on my late 2009 macbook!
I can only imagine the reaction from your local retro buddies when they watch this video and see the RF shield karma! haha
Adrian, Apple did not split the analog ground plane from the digital ground plane to reduce EMI. Splitting the ground planes is a standard tactic to reduce noise in the analog circuitry for several reasons(*). When digital circuitry switches, especially CMOS, it will briefly draw a much larger "gulp" of current to switch voltages (needs to charge/discharge various capacitances). Because of ohm's law and the fact that real ground planes have nonzero impedance, this will actually produce a voltage spike on the parts of the ground plane the current flows through, which screws with all circuitry, but especially analog circuitry -- digital circuitry is somewhat protected by noise margins on logic levels. This is mitigated to a certain extent by bypass capacitors, which essentially spread the current spikes over a longer period of time -- the same amount of current will eventually need to flow in order to recharge the bypass caps. However, the current flow from digital circuitry still tends to produce noise on analog circuitry. Splitting the ground planes forces the return currents from the analog and digital circuitry to take completely different paths (**). Usually, the ground planes will be connected on the PCB, perhaps at a "star ground", but using the chassis or RF shielding is not terribly uncommon.
* Note that this sort of design gets very complicated and there are disagreements about when to or if one should use split ground planes. However, my goal is to just explain why they were used here, not talk about what should be done for new designs. Please don't take this as advice on how to design PCBs.
** Note: modern sources advocate analyzing return paths instead of blindly splitting planes.
I wonder if simply installing a scsi terminator on the back would be a suitable solution? Especially for those who don't want to solder the ground planes together. Or just have the RF shield installed I guess.
I've made that torte recipe, it's actually REALLY good. I highly recommend it.
So quickly?
@@katho8472 As if the site appeared at the same time as this video... LOL
@@ToTheGAMES Haha of course it existed longer but I supposed that "shadow" got aware of it through this video, as did I, being just a German noob :P
@@katho8472 Nah, I found that site years ago crawling the web with Netscape navigator back in the early 00s.
Hi Adrian, love your channel. You mention different parts for your videos at the beginning. RUclips does a great job of suggesting the parts in order if it's part of the title. Something to consider if you can do that. I know you're probably bouncing around projects and other things so might be hard to keep track which part you're on :)
You can actually browse the web in a relatively modern way even on a classic mac like this.
The trick is to get around the encryption which can be quite easily by connecting through a squid proxy that essentially performs a MITM attack and just strips the encryption.
I spent a ... reasonable amount of time figuring it out, but I'm sure most of the people that would ever care to try wouldn't have much trouble with it.
I have a color classic that has NEVER booted up. After seeing you working with one, I took it off the shelf (mostly to see what motherboard it had, no Mystic board in there!) and I discovered it had NO RAM onboard. I need to find my "Box-o-Ram" and see if I have anything compatible in there. I'll let you know what I find out!
Hi Adrian, I greatly enjoy all of your videos. I'd love to see you build and MSX2 Omega kit.
Good show. Nice print job on that cover too. Thanks.
NEC did the same thing with the original white PC Engine. The RF shield on the bottom used to connect ground to the exp port.
My Color Classic motherboard looks a bit different. For me I had the same audio issue as you BUT also a bad solder joint on the leg of the TI audio chip.
I did the VGA mod on one of my Mystic CC today. It was very easy and works perfectly. Sadly the Comm-slot and LC-PDS cannot be used simultaneously. I had to connect an AppleTalk to Ethernet adapter to get the CC on the net while having the Apple //e card installed.
Adrian got 0wn3d by RF shield. I love this sweet irony ^^
Great vid.. Glad to see that Color Classic happy again... Personally, I think just for nostalgia reasons, the Color Classic isn't in my top tier. I like my Classic Macs in B/W. ;-) But great to see it running and with Ethernet! And that 040 was smoking for that era!!
Woah! Maelstrom! I completely forgot about that game. We used to play in middle school.
My Analog Ground/Digital Ground radar was blipping as soon as you were jostling the plug in the socket. Weird SCSI hooked into 'Analog' ground.
Called it at the first mention of audio not working through speaker, but does work in headphone jack. Bad connection in the internal jack. There is a mechanical 'switch' that turns off the speakers when you insert a headphone jack. Most common issue with that kind of hardware. Troubleshooting 101: Follow Olcums Razor.
There was probably rf pickup on the ground plane on the motherboard. The sheild creates a faraday cage to box in the RF. That's why the external connectors are attached to the sheid ground.
Y’know, they might not have done it for RFI reasons, it might just have been to reduce the noise in the audio.
A lot of modern motherboards have identical looking isolation around the audio circuitry, visible thru the solder mask as a gap in the ground plane. They boast that it stops it from picking up chirping noises from the data bus, but obviously those conduct thru the conductors rather than thru the air (hence, not necessarily for RFI).
But I presume the modern motherboards get their ground from the brass standoffs or something, as they don’t have such an external shield.
I liked this series! Great to see it working now imo
thanks for all the video,s adrian superb work as always .
I *had* Maelstrom on my Quadra 610! It was one of my favorites.
Very well done Adrian!!
Hi Adrian, long time viewer, I especially enjoy your troubleshooting and repair videos.
In part #3 you 3D printed a new rear panel for your custom Color Classic.
What brand is the 3D printer and would you recommend it ?
Awesome troubleshooting.
Hmm. I wonder if the second ground was to isolate the noise signal from "digital blipiness" coming through the speaker? Just a thought. Great content as always!
My Color Classic has the same sound issues as yours. Even the audio chip gets hot. Sad that your fix didn’t work for the original logic board.
wonderful work.. thank you so much for the video...
Nice Adrian. I’m addicted to every one of your videos. I need to get an oscilloscope and learn how to read them from a signal standpoint.
Just FYI - when you were looking at the scope trace of the audio signal, it was jumping up and down when you plugged in the external speakers. I have a theory on what caused that.
Audio circuits are AC coupled, at least usually. It’s always nice when you have +/- voltage rails so the AC can be centered around 0V, but quite often it’s using a 5V DC supply. This means you have to center the AC waveform around some DC bias voltage to make sure the negative swing of the signal can go below the reference point.
You would then use a capacitor to remove the DC bias and re-center the waveform around 0V. Hence, AC coupled.
(Even if you have a -5V rail, you still need to AC couple to make sure you don’t have a ground potential difference. Moreover, in many cases the -5V rail is only used for amplification and the D/A is still DC biased. I haven’t looked at the Mac schematics though.)
The “outside” or jack side leg of that cap needs to have a ground reference, since there’s no DC flow through the cap. Usually, the input stage of whatever you’re plugging into (like a powered speaker or a mixer, etc.) will have a resistor to ground. This sets the input impedance to something between 1K and 100K, often around 10-47K. Instrument and mic level inputs are higher (maybe 1M), balanced pro audio is 600R, but you get the idea. That R to ground pulls the AC coupled signal to ground, but isn’t strong enough to short the output, so you get a voltage swing around Gnd.
So what you may have been seeing on the scope when you plugged in the audio cable is the external ground being applied to the AC coupled waveform, bringing the bias point down. Particularly if the ground reference was missing on the internal speaker circuit path.
Clip away the back section off the shield and put it back in the case?
Audio and power connectors get worn also they usually switch as well which can be a problem audio and voltage ground are usually separate depends
well on most motherboards from pc's , at leasy pentium 2-4 era audio was also insolated grom digital ground they had only connections made by single 0 ohm resistor and had it's own power rail filtration caps
all that was made to stop digital noise interference to analog aution which would produce high pitch noise hiss , but also audio interfering with bass sounds pulling ground up and making digital errors on the other side
i belive in this case shield worked as low resistance resistor link that would stop crosstalk along with rf interference
interesting bug non the less
NCR tower used a 68040 to run Unix and 68010s for TTY terminal handling a very popular system.