Mystery: rMBP and replacement rMBP repeatedly bricked (likely fried)-geniuses no help

[marekkurlmann]

Hello!

I have experienced a very strange issue with a new rMBP that has confounded Apple techs and geniuses. Apple has given me a replacement machine and replaced the logic board, and each time, within a few days, it's bricked. Apple has offered no advice except "we'll fix it the next time it happens." I turn to you––hopefully more interested in solving this issue than Apple––for advice.

Many thanks for any input. Here is what happened:

Death #1
On Jan. 2, I purchased a new late-2013 rMBP. About 10 days later, the machine mysteriously restarted in the middle of the night and would not boot. All I got was a black screen; no software or BIOS loaded. The screen was illuminated and the Apple logo lit up, but I couldn't boot to an external drive or reset the PRAM. The power light on the MagSafe connector indicated that the SMC reset was working, however.

I took the machine to the Apple Store and a genius confirmed the above. He sent the machine to a repair depot, which recommended replacing the logic board and display. At this point, the Apple Store simply gave me a brand new machine.

Death #2
I used the new machine for about a week and the same thing happened again––the machine restarted and would not boot. Black screen, no BIOS, etc. I took the machine to the Apple Store and the geniuses were stumped. I asked what could cause this to happen and was told that it would be shipped off to techs at a repair depot who would investigate. There was no investigation. They replaced the logic board and display and returned the computer to me without any advice

Death #3
Two days after receiving the repaired computer, the computer bites the dust again. An Apple genius offered that a USB device may be creating a power surge and that I should stop using all USB devices with the computer until I find out which device is defective. But this seemed an amateurish diagnosis at best: Can USB devices really fry a logic board? Aren't USB connections limited in the power they can transmit?

Possible causes
Power adapter: After the second frying, I volunteered a piece of information that was dismissed by the geniuses: I am using a 2006-era 85W MagSafe power supply with a MagSafe 2 adapter. While Apple says this is safe, two Apple techs advised me that this could be responsible for power issues damaging the logic board. (They told me this after the third death.)

External devices Can an external USB device destroy a motherboard? I have been using all of the external devices connected to the destroyed rMBPs with previous-generation MBPs for years with no issues whatsoever. [CLARIFICATION: The previous MBP was a mid-2012 model with USB 3 ports; these peripherals have all played nice with USB 3 ports for about 20 months.] The peripherals include:

- 3 powered USB hubs
- Multiple hard drives
- DVD writer
- DAC
- Printer
- Scanner

Software On both machines that died, I had migrated from the same backup via Migration Assistant. Is it conceivable that a software issue could prevent a machine from booting as I have described?

Thanks again for any advice!

 

[Intell]

A USB device can fry a logicboard. It's more likely to happen on a USB 3 port than the older 1 or 2 port. I suggest discontinuing use if the old MagSafe. While it is safe to do, that MagSafe could be defective or nearing the end of its life.

 

[jav6454]

Why use that old MagSafe adapter? Although higher spec'd power adapters do only provide what is needed to Macs that consume less (say 65W), it is very possible that the power adapter doesn't play nice with the MagSafe 2 port conversion and instead is providing full 85W.

This in turn will fry your computer. Use the adapter that came with your Mac. USB devices can fry your mother board, but only if you force draw more power than what the USB port is designed to deliver.

 

[marekkurlmann]

Why use that old MagSafe adapter?

I'm not going to use it any longer. I didn't even consider it could be the source of the problem because the geniuses had originally dismissed it out of hand.

USB devices can fry your mother board, but only if you force draw more power than what the USB port is designed to deliver.

What exactly do you mean by "force draw?" Rigging the USB device to draw more power than it's designed to draw?

 

[jav6454]

What exactly do you mean by "force draw?" Rigging the USB device to draw more power than it's designed to draw?

Exactly. Some non-compliant devices might do this.

 

[marekkurlmann]

Exactly. Some non-compliant devices might do this.

The device directly plugged into the computer is a single powered USB hub. All devices (including other powered hubs) are plugged into this single hub. Would a defective/non-conforming device still be able to "force draw" excess power from behind the hub?

Thanks for your help, jav!

 

[jav6454]

In theory a compliant hub should divide the amount of power available. However, cheap ones will most likely cause damage. It happened to me once.

 

[Swampus]

Were all of those USB devices hooked up at the time of each failure? What kind of printer?

Is everything plugged into the same circuit (same breaker or fuse for all outlets involved)? What else is on that circuit? Anything with a compressor motor (air conditioner or refrigerator)?

----------

...Some non-compliant devices might do this.

This is true. I'd add that it's possible to buy hubs that claim to be compliant even though they're not. This is especially true in the $10 range of self-powered hubs on eBay & such.

Of additional concern would be upstream current. This would be in clear violation (section 7.2.1), yet I've observed (and measured) it on more than one occasion with cheap self-powered hubs. Some A$$h@le paper pusher thought these would be a really good deal until boards started failing.

From section 7.2.1 (emphasis added):

• Self-powered hubs: Power for the internal functions and downstream facing ports does not come from
VBUS. However, the USB interface of the hub may draw up to one unit load from VBUS on its upstream
facing port to allow the interface to function when the remainder of the hub is powered down. Hubs that
obtain operating power externally (from the USB) must supply five unit loads to each port. Battery-
powered hubs may supply either one or five unit loads per port.

No device shall supply (source) current on VBUS at its upstream facing port at any time. From VBUS on its
upstream facing port, a device may only draw (sink) current. They may not provide power to the pull-up resistor
on D+/D- unless VBUS is present (see Section 7.1.5). When VBUS is removed, the device must remove power
from the D+/D- pull-up resistor within 10 seconds. On power-up, a device needs to ensure that its upstream
facing port is not driving the bus, so that the device is able to receive the reset signaling. Devices must also
ensure that the maximum operating current drawn by a device is one unit load, until configured. Any device that
draws power from the bus must be able to detect lack of activity on the bus, enter the Suspend state, and reduce
its current consumption from VBUS (refer to Section 7.2.3 and Section 9.2.5.1).

 

[marekkurlmann]

Thanks, Swampus! Here is some more information:

Were all of those USB devices hooked up at the time of each failure?

Yes, they were. Only the DAC (high-end Audioengine model) and DVD writer (Samsung) were pulling bus power. All others––hard drives (Rosewill enclosures), printer (Samsung), scanner (HP)––had their own AC power supplies. I include the brands so you can see that I'm using mainstream peripherals. No $3 fly-by-nights from eBay.

What kind of printer?

A five-year old Samsung laser printer (with its own AC power supply, of course)

Is everything plugged into the same circuit (same breaker or fuse for all outlets involved)? What else is on that circuit? Anything with a compressor motor (air conditioner or refrigerator)?

Yes. It's all on the same circuit. Nothing with a compressed motor on the circuit, only computer equipment and lights. I'm using three separate surge protectors for all the equipment, if that's at all relevant.

I'd add that it's possible to buy hubs that claim to be compliant even though they're not. This is especially true in the $10 range of self-powered hubs on eBay & such.

These aren't bargain-basement hubs. They're Belkin USB 2.0 hubs, about 4-5 years old, and a hub built into my Dell display. I've used these exact hubs with at least half a dozen Macbooks without problem.

 

[Chatran]

Try not to use powered usb hub if possible (or dont use the ac adapter provided). These always create problems even in desktop.

 

[simonsi]

Four options for the cause here that I see.

1. The infinitesimally small chance that you had two logic boards fail in the same unusual way.
2. The power adapter
3. The powered USB hub
4. One of the USB devices (also requires the hub to pass-thru the upstream voltage fault)

I'd disregard 1.

For 2 & 3 I gladly junk BOTH and not be interested in finding out which was the cause.

4 is the difficult one, I have no data on how likely that is and diagnosis by trial-and-error will get real expensive, real quick. options here are:
a) Russian Roulette (connect one at a time until you fry the board again),
b) Voltage test each device. (are there voltage surge protectors for un-powered USB, I'm not even sure if that is possible???)
c) Junk ALL USB peripherals and start again.

Risk here is that Apple refuse a 3rd repair and ask you to front up due to the likelihood of it being an external issue.

 

[CausticPuppy]

What exactly do you mean by "force draw?" Rigging the USB device to draw more power than it's designed to draw?

Yes, or even shorting it out. Since the USB can provide power, if you short it out, it'll fry. There's no fuse to blow. So beware if you plug in some cheap chinese knockoff thingy into your USB port.

 

[Swampus]

A five-year old Samsung laser printer (with its own AC power supply, of course)

Yes. It's all on the same circuit. Nothing with a compressed motor on the circuit, only computer equipment and lights. I'm using three separate surge protectors for all the equipment, if that's at all relevant.

Laser printers have been known to wreak havoc on occasion. Some can be as bad as a window AC unit when it comes to inrush and ground potential rise. (FWIW: I think this is a little extreme, but Portland General Electric recommends that a computer and all of its peripherals be plugged into the same circuit except for a laser printer, which they suggest be on a different circuit with no physical connection to the computer other than fiberoptic).

These aren't bargain-basement hubs. They're Belkin USB 2.0 hubs, about 4-5 years old, and a hub built into my Dell display. I've used these exact hubs with at least half a dozen Macbooks without problem.

I agree with simonsi that a blanket approach makes sense. I think I'd at least replace that top level hub. Maybe with a wireless?

...(are there voltage surge protectors for un-powered USB, I'm not even sure if that is possible???)
...

Yeah, you can get inline supressors. They were probably more common back in the RS232 days. I used to have to maintain physical connections between PCs in two buildings on different power supplies. Pretty easy to fry a serial card in that situation without inline protection.

But maybe a better option here would be to replace that top level hub with a wireless hub. That way, there would be nothing physically connected to the MBP, only a wireless dongle, no potential for upstream weirdness in case we're overlooking something in relation to the hubs. Seems really handy too. Move all around the house and still have all of your devices "connected".

To recap what we have so far:

1. Discontinue use of old MagSafe and use new one with grounded extension.
2. Verify that wall outlets are properly grounded.
3. Replace top level USB hub with wireless model.

 

[thekb]

Move. Your house is about to explode from defective wiring.

 

[marekkurlmann]

Thanks again, Swampus. You seem to understand the electrical principles at work, so I have two questions for you:

1) Assume a USB peripheral is connected to a powered hub. If that peripheral sends a power surge in the direction of the computer, will the hub send it on to the computer, or somehow "absorb" the shock ? (I ask because I have no way of testing each peripheral for defects, but can easily replace the hubs.)

2) You write

Verify that wall outlets are properly grounded.

Will using a surge protector offer any protection against improper grounding? (Seems like an obvious answer [no], but I have little knowledge of the principles at work.)

Thanks.

 

[TheIguana]

2) You write


Will using a surge protector offer any protection against improper grounding? (Seems like an obvious answer [no], but I have little knowledge of the principles at work.)

Thanks.

Nope, if your grounding at the outlet is off the surge protector has nowhere to send the excess voltage. Most of the electricity gets sent to ground in the case of a surge, the actual surge protector element will absorb a portion of the excess electricity and dissipate it as heat but the bulk of the electricity is shorted to ground.

 

[alphaod]

Well if you're old 2006 era power supply hasn't been replaced yet, it's time to get it replaced. Even though it technically shouldn't be an issue, it can cause an issue.

Then see if you still have this issue. Also try using your computer in other places.

 

[simonsi]

Then see if you still have this issue.

Hmmm - risky strategy....

 

[Swampus]

...Assume a USB peripheral is connected to a powered hub. If that peripheral sends a power surge in the direction of the computer, will the hub send it on to the computer, or somehow "absorb" the shock ? (I ask because I have no way of testing each peripheral for defects, but can easily replace the hubs.)...

If that top level hub is working correctly, it should block upstream bus current from downstream devices. But that's not the same thing as preventing a surge. Some do provide some level of basic protection, but I wouldn't count on most consumer level models (name brand or not) to prevent a surge. It would be possible for a diode to fail allowing current in both directions. Even in common surge protectors, components can fail without warning you. I guess the most common is an outlet strip that has a metal-oxide varistor added. Unless the strip has an indicator that failure has occurred, you'll never know that the MOV has failed (until you do).

I wonder, though, about the ground wire in the USB interface--Especially in relation to the printer. If you're running the MBP on the charger without the grounded adapter (or with bad ground), it creates a virtual ground (two resistors are used to split the voltage and reconnect it in series so that the midpoint becomes the virtual ground). It's possible to measure half of line voltage from case to ground (low amperage, not dangerous, most people don't notice) when running it this way. I assume that your laser printer would have a protective earth ground? And a grounded USB cable connecting the two. As the printer cycles on, inrush current would cause neutral to drop creating even more difference between those ground potentials? It's a situation that I would try to avoid, but you can't be the first person to hook their MacBook Pro to a laser printer while running from the two prong adapter. It seems like there must be something else going on here.

Will using a surge protector offer any protection against improper grounding?

Assuming that the outlets are otherwise wired correctly, a "mode 1" device will dump a surge from line to neutral (neutral should still be bonded to ground at the point of entry to your home). Even if you have a ground wire to your outlets, I'd still be inclined to use a mode 1 device for interconnected equipment.

Where do you live, anyway? What is your line voltage? Do you know when your house was built? Do you have other reasons to be suspicious of the wiring? Do you happen to have a multimeter (Volt-Ohm meter)?

Also, what model Samsung laser printer? What model hubs? I'll see if I can find some schematics or at least some specific details about the interface chipsets and components. Do you have a wireless router that could accommodate that printer?

 

[benguild]

I don't know if this has been said already, but it could be that the USB hub connected directly to the computer is busted or has water damage, and is shorting power directly into the machine.

 

[marekkurlmann]

Swampus, thanks for such a detailed answer. I'll update you with what I've done, provide some clarification, and answer your questions:

Where do you live, anyway? What is your line voltage? Do you know when your house was built? Do you have other reasons to be suspicious of the wiring?

I am in Los Angeles; standard North American 110-120 line voltage; building built circa 1955; no reason to be suspicious of wiring.

Do you happen to have a multimeter (Volt-Ohm meter)?

No, but I did pick up an outlet tester (http://amzn.com/B002LZTKIA). None of the outlets showed any problems. The test showed that the grounding on the outlets was fine.

Also, what model Samsung laser printer?

Samsung CLP-315 (http://www.samsung.com/us/computer/printers/CLP-315/XAA)

What model hubs?

2 Belkin USB 2.0 hubs (http://belkin.com/us/p/P-F5U237)
1 hub integrated into a Dell 2408WFP monitor (PDF user guide; product info page no longer on Dell's website

Note: I have have replaced these hubs with two Plugable USB 3.0 hubs (http://plugable.com/products/usb3-hub7-81x).

Do you have a wireless router that could accommodate that printer?

I don't, but I could move the printer to an outlet on the other side of my office. Whether that is another circuit is something I don't know.

I assume that your laser printer would have a protective earth ground?

Yes. It's a three-prong plug.

And a grounded USB cable connecting the two.

Do standard USB cables have ground wires? Are standard USB hubs grounded? I am using standard USB cables and hubs.

...you can't be the first person to hook their MacBook Pro to a laser printer while running from the two prong adapter.

I'm using a 2006-era 85W Apple power adapter with MagSafe connector with the Apple extension cord with a grounded plug. The adapter itself came with a MacBook Pro bought in Europe (110-240v), but the extension cord is American (standard American three-prong plug). I've used this adapter in the U.S. for many years without issue.

One peripheral that hasn't been discussed is the external display (the Dell 2408WFP monitor), which is connected via an official Apple DVI-to-mini DisplayPort adapter. Could a display fry a motherboard?

If you've got any further advice, I'm looking forward to hearing it. Even after replacing the hubs and the power supply, I still have not connected my MacBook Pro to USB devices and my external display for fear of what may happen.

Thanks again for all the helpful input!

 

[Swampus]

I am in Los Angeles; standard North American 110-120 line voltage; building built circa 1955; no reason to be suspicious of wiring.


No, but I did pick up an outlet tester (http://amzn.com/B002LZTKIA). None of the outlets showed any problems. The test showed that the grounding on the outlets was fine.

That type of tester will not reveal a bootleg ground or a reversed-polarity bootleg ground. The tester will show normal in both of these cases. Both would be very dangerous. Personally, I think there is cause have suspicion of the wiring for anything in the United States built before 1970. Some outlets might appear to function fine for years for some things, but will not provide a safe environment for interconnected electronics. It would be possible for an entire home or office to be done correctly except for one or two outlets. That's actually quite common. Such outlets, of course, are also potentially deadly to humans as well as to electronics.

Read this: http://ecmweb.com/contractor/failures-outlet-testing-exposed

Were you at this same location with the same equipment on the same outlets with the previous model MacBooks? Have you been using the grounded extension all along?

Do standard USB cables have ground wires? Are standard USB hubs grounded? I am using standard USB cables and hubs.

Yup. Ground loops over USB quite common. Most go unnoticed, but can cause major headaches for audio folks. Can be prevented with isolators (expensive). These connections, in themselves, are not usually the cause of hardware failure for things on the same circuit, but they can allow a path for other hardware problems or wiring issues to cause failure.

I'm using a 2006-era 85W Apple power adapter with MagSafe connector with the Apple extension cord with a grounded plug. The adapter itself came with a MacBook Pro bought in Europe (110-240v), but the extension cord is American (standard American three-prong plug). I've used this adapter in the U.S. for many years without issue.

So you've been using the grounded extension all along?

Given the age of your building and the shortcuts (hacks) that were sometimes taken in renovations, it might be wise to stop here for the moment. I think we need a licensed electrician to verify that what we're dealing with is safe. Once we determine that it is safe for you, then we can troubleshoot to make it safe for your electronics.

 

[marekkurlmann]

Were you at this same location with the same equipment on the same outlets with the previous model MacBooks?

I've used only one previous MacBook at this location. I've used this same equipment with half a dozen MacBooks.

So you've been using the grounded extension all along?

Yes!

Given the age of your building and the shortcuts (hacks) that were sometimes taken in renovations, it might be wise to stop here for the moment. I think we need a licensed electrician to verify that what we're dealing with is safe. Once we determine that it is safe for you, then we can troubleshoot to make it safe for your electronics.

Do you think that this may be overreacting a bit? What signs, if any, are there that something is wrong with the wiring?