Mac 512k chassis ground issue?

[d_a_parker]

Hey all,

I recently acquired a Mac 512k in great cosmetic condition, but it does not boot. It beeps once when powered on, but has no picture on the display. I'll give more diagnostic history below, in case it's relevant, but first... I'm reading 22.3V AC on the chassis ground when plugged in but not powered on, and 28.3V AC when powered on. This is measured with the negative probe on signal ground and positive probe on the chassis ground. The only components on this board which bridge between the circuits and chassis ground are 3 RIFA safety caps, which I replaced, but it made no difference. Is this normal for these Macs?

Some history on this...

This is a Mac 512k with the HyperDrive upgrade. The original symptoms were a loud clicking sound from the speaker, and no picture on the CRT. After some troubleshooting, I traced the clicking issue to the power supply board for the Hyper Drive and removed the entire kit (PS board, HDD, drive cage, SCSI board, and EMI filter). The clicking is now gone (sort of, more on that in a bit), and the Mac now beeps when powered on but there is still no picture on the screen. Brightness control, voltages, and signals in the video circuit all seem to be OK, but it's like the CRT itself is unpowered. There is a faint glow at the tip of the neck of the tube, but no other signs of life. A spectrum analyzer on my phone shows nothing coming from the CRT when powered on.

Voltages and signals between the logic board and video board seem fine. Signals and voltages on the logic board itself also seem fine.

The Mac makes the clicking sound instead of a beep when no floppy drive is connected. If I connected either the original internal drive, or an external one, it solves this problem. Is this also normal?

Thanks!

 

[d_a_parker]

UPDATE: There's still AC voltage in the chassis ground plane, but I narrowed it down a bit more and have to make some corrections...
1. There is ~13V AC when plugged in but powered off, and ~17V AC when powered on. The measured voltages are higher (22V-28V) when my meter is plugged into its charger, but these lower voltages have been measured and verified with 3 different DMMs. A scope reveals a 60Hz AC sine wave when voltage is present in this ground area.
2. This is not measured between signal ground and chassis ground. I thought the metal bracket attached along the back of the logic board was signal ground, but it's not. The only true signal ground seems to be on J7 pins 7 and 9.

So my questions are: is it correct that I see two distinct ground planes in this Mac 512k, which do not share continuity? And is it normal for the analog board's chassis ground to show AC voltage and a 60Hz sine wave?

I'm trying to figure out if I'm on the right path here for finding the fault in this Mac, or chasing a non-issue.

Thanks!

 

[Hugo Holden]

If you could post the schematics this whole issue should be easy to solve.

Unfortunately many appliance designs of all kind made in the USA and elsewhere for the USA market they did not have earthed commons or an earthed chassis.

In many cases the chassis and often the common power supply line, assumed an "indeterminate voltage" which floated somewhere between phase and neutral.

It is a really bad idea (unless a double insulated device with no connections to anything else), in that if the apparatus, whatever it might be, gets inter-connected with another one by a cable, then there are, at the moment of the connection, high discharge currents as the two commons in the two different items of equipmen assume the same potential. This is a good way to destroy interface IC's, such as the line driver IC's used for serial (RS232) interfaces.

In the ideal world, all appliances that can connect to others with interfaces, should have a three wire line cord and a grounded chassis and the secondary side of the DC power supplies not floating, but with he common (negative ) connected to ground too. This way there are no issues even hot plugging the equipment.

The Y capacitors that connect both phase and neutral to the chassis are "supposed" to be special quality parts, that don't have a short circuit failure mode. But that is like believing that the Moon is hollow or that crop circles were made by Aliens. They do short out at times. In the case that the chassis is not earthed this can raise the chassis potential to live(phase) and be dangerous. In many cases these capacitors are now more of a liability than an asset, and the line power supply now is so polluted with interference from every Man and his Dog's switch-mode power supply in every cheap appliance they have bought with a fake CE logo, that the Y capacitors are now pointless.

If you don't earth your chassis and negative voltage supply lines with a three wire line cord, you may as well cut out the Y capacitors, it is safer. If you earth the chassis, all will be good regardless of the presence of good or bad Y capacitors.

In any case, its an individual problem, and a lot of guesswork until the schematic is posted and then a good solution can be suggested.

 

[Al Kossow]

the ac input circuit
the whole description of the power supply is in the URL

is the ac issue because of not using an electrician's voltmeter?

 

[d_a_parker]

If you could post the schematics this whole issue should be easy to solve.

The schematic I'm using is found here: https://bitsavers.org/pdf/apple/mac/schematic/EarlyMacSchematics.pdf

The unit is properly grounded (I have verified that the receptacle connects to earth ground). The original RIFA safety capacitors at C33, C37, and C38 were not blown and seemed to be ok, but I still replaced them with new Kemets from Mouser (P/N PME271M610MR30 and PME271Y447MR04).

Would an electrician's DMM would make a difference in this case? I've never heard of someone needing one to fix a Mac before. I'm not 100% sure this grounding issue is even incorrect or a problem. I'm just trying to find out if I'm looking in the right place, since all voltages and signals on both boards check out fine, yet there is no life in the CRT.

Thanks!

 

[d_a_parker]

the ac input circuit
the whole description of the power supply is in the URL

Thanks for the link! I was using that document earlier in the process, for both the circuit diagrams and the connector pinouts, and then I couldn't find it again when I wanted to download it. Thank you!

 

[andy]

I think your chasing a red herring. The ground pin on the power socket should have a wire that's screwed to the metal chassis. You can confirm that by measuring the resistance from the ground pin to the chassis. The ground of the logic board is connected to the chassis by the metal brackets either side of the rear jacks. If those aren't making good contact with the chassis (as they might not be when the case isn't screwed together), you might see a phantom voltage on the logic board's ground. This shouldn't cause any functional problem, and will go away when it's fully assembled.

I would start by re-soldering all of the heavy connections on the analog board. Particularly, the ones on the plug in sockets tend to crack.

 

[d_a_parker]

I think your chasing a red herring. The ground pin on the power socket should have a wire that's screwed to the metal chassis. You can confirm that by measuring the resistance from the ground pin to the chassis. The ground of the logic board is connected to the chassis by the metal brackets either side of the rear jacks. If those aren't making good contact with the chassis (as they might not be when the case isn't screwed together), you might see a phantom voltage on the logic board's ground. This shouldn't cause any functional problem, and will go away when it's fully assembled.

I would start by re-soldering all of the heavy connections on the analog board. Particularly, the ones on the plug in sockets tend to crack.

Thanks, Andy. I failed to mention before that I did reflow many of the analog board connections already, based on advice I found in various other posts about this issue. Most people with this issue at least seem to have power to the screen, though (i.e., it glows more brightly when they crank up the brightness, or they get the no-disk icon). In my case, the CRT is acting like it has no power, yet it does. It may turn out to be a dead CRT, but I was suspicious that a broken ground may be the issue instead. That's seeming less likely, though.

 

[Hugo Holden]

It is good the chassis is properly grounded. You need to identify why there was around 17V AC between the signal ground and chassis ground. The exact reason, by looking to see why and where the signal ground (normally the supply common) was not connected to the chassis. It can be either with a separate wire link or when a pcb screws down to a chassis as Andy said. If you just start re-soldering random places and the problem appears to go away, you won't be 100% sure have fixed it and the problem could reoccur later. If you cannot work out how the original grounding was achieved it would be better to add a ground link. CRT's themselves are usually pretty reliable, unless air gets in or the heater goes open. Mostly when a VDU doesn't operate its a fault in the CRT's support circuitry.

 

[cjs]

Shouldn't you be able to determine if the voltage between the grounds is a phantom voltage or not by connecting the two grounds with, say, a 1 MΩ resistor? (Or would doing such a thing be a bad idea. My knowledge of electronics is limited.)

 

[d_a_parker]

As usual, it was a foolish mistake on my part... When I removed the logic board so I could better probe things, I removed those screws and thus disconnected it from the chassis. I assumed that the ground wire on J7 kept them connected, but it does not. That's only for the logic ground, not the actual electrical ground. As soon as I reconnected the logic board's ground plane to the chassis with an alligator clip, the phantom voltages disappeared and my readings (both DMM and scope) got much better.

The display on this Mac still doesn't work, but it's probably not a grounding issue, and I'll start a different thread for that.

Thanks all!