Mac II with no Nubus slots working - can another Mac II owner measure some voltages/resistances?

[JohnR]

Hi, I've been wanting a Mac II for years and have finally purchased one described as "dead". I'm making good progress with it and it's no longer "dead". The logic board has been cleaned, recapped and had the batteries replaced. The power supply still works and has good voltages. The soft power system is working fine. The machine chimes and boots into an operating system however no image ever appears on the screen. I've tried 5 different video cards in all slots (3x Toby, 1x Mac Display card 8*24, 1x Lapis) and no luck. All cards work fine in my IIsi. The Mac appears on the LocalTalk network but I'm unable to connect to it with Timbuktu as it thinks there's no monitor plugged in. This leads me to believe that no Nubus slots are working. It's the later revision with the Nuchip controller.

I've checked voltages at every pin on the Nubus connectors with the machine running and compared the outputs to my IIsi Nubus adaptor card. (the only other Nubus machine I have). Pins A5,A24,A25,A28,A30,C5,C24,C25 and C28 all measure at approx. 5 volts. The same pins on the IIsi measure at 3.2 volts, which happens to be the same voltage as many other pins on both machines.

With the machines switched off I measured the resistances of all those suspect pins to ground and they all measured approx 200 ohms on the Mac II vs. 180 ohms on the IIsi. All the suspect pins are connected to chip RP7 on the logic board, which also measures 200ohms between pins 1-15 and the 5v pin 16.

Would anyone here who happens to own a Nuchip equipped Mac II be able to do a few quick measurements for me? I'd love to know:

Voltage of pin A5 to ground (while running) - my Mac II reads 5v, my IIsi reads 3.2v
Voltage of pin A4 to ground (while running) - my Mac II reads 3.2v, my IIsi reads 3.2v
Resistance of pin A5 to ground (not running) - my Mac II reads 200 ohms, my IIsi reads 180 ohms
Resistance of pin A4 to ground (not running) - my Mac II reads 180 ohms, my IIsi reads 180 ohms
Resistance between pins 15 and 16 on chip RP7 - my Mac II reads 200 ohms

Thanks so much in advance
John


Love it!

(this pic prior to recapping)

 

[Melkhior]

Pins A5,A24,A25,A28,A30,C5,C24,C25 and C28 all measure at approx. 5 volts. The same pins on the IIsi measure at 3.2 volts, which happens to be the same voltage as many other pins on both machines.

During the measurements, was there a NuBus card somewhere in the system or did you do 'empty system' measurements? Also, how long after boot did you take measurements - similar for both or not? The Slot Manager probing the slot will cause activity early during boot, and then there should be no more activity if there's no card afterwards, I believe.

All the pins you list carry signals, so they should be either 5V (pulled up) if nothing is going on, or toggling between 5V (unasserted) and 0V/GND (asserted) during actual bus cycles. 3.2V looks like an averaged-out value during bus activity - though 3.2V is enough to record a '1' at TTL logic level (2.0V at destination, 2.4V required at source, TTL is very tolerant...), so it could be a quite low but legal no-activity value as well. A logic analyzer (you don't need a fast one for 10 MHz NuBus) or an oscilloscope would probably be more useful than a pure voltmeter.

A24, A25, C24, C25 are the ARB[0-3] (arbitration) signals. I'm surprised they could be asserted in your IIsi, as they should not need assertion unless there's a bus-mastering device in the system other than the NuChip itself [1]. A30 is RQST, and indicates a bus master is requiring the bus - which explains why ARB are also active, though that's weird.

A5/C5 are the TM[01] signals, A28 is ACK, C28 is START. Those are needed for all bus cycles, so should have activity whenever there's a bus cycle. They should rest at 5V when there's no device.

Honestly I would expect 5V (or close) from all those signals with no devices, so it's weird to have 3.2V on the working IIsi... it's probably not anormal for the Mac II, assuming no bus activity occurred during measurement.

It's probably difficult to diagnose the issue for the Mac II without at least a logic analyzer. You need to check the clock pin for activity first (10 MHz, 25% duty cycle), and then observe the START, ACK and TM signals to see if there's some probing going on during boot, along with some of the address signals, the upper address bytes should be $Fs with 's' the slot number (negative logic, so 'b00000xxx, the upper bit of 's' is always set as it's between $9 and $E). Without device, the bus cycle will time-out. If there's 'reasonable' probing, then you can try to see what happens when the device is plugged. Normal bus cycle should assert ACK much faster. If that works, then some of the NuBus functionalities are working and you might need to check the data lines - if one of them is broken (traces or buffers), then the ROM probe could fail. You may also need to check the address lines to make sure none of them are broken, either (otherwise the ROM access would be at a wrong address, also failing).

Good luck!

[1] Arbitration is used when more than one master requires the bus. But the current bus owner can start a transaction without arbitration provided no other master is requesting the bus, so when there's only one master arbitration is not needed.

 

[JohnR]

Thanks so much for your very thorough reply!

Those measurements were taken with and without a card in a nubus slot. Voltages were measured before, during and after boot. There wasn't any noticeable change in the readings (keeping in mind I'm only using a multimeter). Resistance was checked with the computer off.

That being said I could really get away with just knowing the resistance between pin A4 and ground, and pin A5 and ground on a NuBus connector in a working late revision Mac II while switched off. If the resistances are similar to mine (175-180 ohms on pin A4, 200-210 ohms on pin A5) then I'll assume that's normal and proceed to purchase a suitable oscilloscope or logic analyser to probe deeper (and learn how to use one lol).

 

[David Cook]

I have a Macintosh II I'll be working on this evening. If someone else hasn't replied by then, I'll perform the measurements and post them.

 

[JohnR]

Thank you David!

 

[David Cook]

On David's Macintosh II with memory but otherwise just the motherboard (no NuBus cards or floppy or hard drive):

Voltage of pin A5 to ground (while running) - David's Mac II reads 5.16V
Voltage of pin A4 to ground (while running) - David's Mac II reads 3.27V
Resistance of pin A5 to ground (not running) - David's Mac II reads 213 ohms
Resistance of pin A4 to ground (not running) - David's Mac II reads 177 ohms
Resistance between pins 15 and 16 on chip RP7 - David's Mac II reads 201 ohms

To be clear, you're talking about A4 & A5 on the Nubus slot connector, right? Not the CPU address lines.

A4 goes through resistor pack 8.
A5 goes through resistor pack 7

 

[JohnR]

Thanks David, that is a huge help. It definitely seems like I’ve been barking up the wrong tree in the wrong forest .
I’ll study Melkhior’s suggestions now and dig deeper. I’ll post my findings

 

[David Cook]

Please post any updates or diagnostic methods you find helpful. I have a IIx with the same symptoms (computer runs but nubus doesn't work) that I plan on fixing one day. The Nuchip on that gets *hot*, so I suspect it is dead. I have purchased a socket to help test via part substitution. However, unless someone has a Nuchip laying around in a parts bin, I'm afraid I'm going to need to sacrifice another board.

 

[pezter22]

I am certainly interested in your repair too. I have a similar issue with my Mac IIx motherboard with no nubus. I don't get a chime (sound chip issues), but my CPU gets warm and it seems to boot, but no video. My nubus video card is also showing activity as can be seen by my thermal camera image. It's been stumping me for a few months.

I will definitely recheck the traces mentioned here. Thanks for giving me the incentive to look again. Good luck with your repair.

 

[JohnR]

No solutions yet but I’ve been painstakingly beeping out all the data and address lines between the chips, so far so good. Next step will be checking every pin on every chip one by one and hopefully finding something. Borrowed a thermal camera from work and had a look.. the only thing that stood out to me was the Nuchip is not warming up at all, not sure if that’s normal

 

[JohnR]

Success!!! After many many hours of studying the Bomarc schematics and checking continuity of a seemingly endless list of pins and connections I found the connection between UC7 pin 6 and RP4 pin 14 was incomplete. I quickly ran a bodge wire for testing and it freaking works!! What a great reward after spending all this time!! I found the trace next to chip UC7 and it's still connected to RP4, so the break must be underneath UC7. If my repair skills improve I may remove it one day and fix it properly, but for now I'll run with my ugly yellow oversized wire haha.

Thanks again so much for all your input



Good little SE30 serving files and internet over LocalTalk to the Mac II

 

[Melkhior]

Success!!!

Excellent Congrats on your new Mac ii!

(...) I found the connection between UC7 pin 6 and RP4 pin 14 was incomplete

Ouch, if I read that schematics correctly that would disconnect A5 (buffer to slot connection) - that would definitely prevent all the slots from working :-(

 

[pezter22]

Congrats!

 

[JC8080]

It's always great to see someone do a repair like this!

 

[Sideburn]

The best (and addicting) part of repairing this stuff is the “it freaking works” moment.

 

[JohnR]

Oh yes it was a great feeling. The machine is reassembled now and has been running solidly. I've been experimenting with multiple nubus video cards and discovering some don't play so nice with each other lol. Also have an Ethernet card on the way for it

 

[Phipli]

Oh yes it was a great feeling. The machine is reassembled now and has been running solidly. I've been experimenting with multiple nubus video cards and discovering some don't play so nice with each other lol. Also have an Ethernet card on the way for it

If you run out of slots, you can always get an eMachines combo ethernet and video card