-
Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.
Strange Mac SE startup anomaly
- Thread starter spaceinvader12
- Start date
mattislind
Member
I agree that 5.2V should not be able to trigger the the crowbar but the clicking sound is very typical for an emergency shutdown in the PSU imho.
The source might for example be over current or over voltage. In the Sony PSU both the 12V and 5V outputs together control the emergency shutdown. A quick glance at the schematic give that above 6.2V at the 5V or 15V on the 12 V would trigger the emergency shutdown. A reason for false triggering could be excessive ripple due to bad filtering capacitors. Check ripple on the 12V and 5V before the filter inductors.
Put dummy loads on the 5 and 12 V outputs and temporarily disconnect the zeners for the overvoltage system. Meassure ripple both before and after the filtering inductor.
Since the overvoltage system is measuring both voltages simultaneously uneven loading of the PSU can give this type of problems. I have seen this previously in s MicroVAX 2000 which need to have a dummy load if the hard disk is removed.
As far as I see it a capacitor stay reformed for quite some time. I find it unlikely that a capacitor heating up is creating this kind of repetitive cycle.
The source might for example be over current or over voltage. In the Sony PSU both the 12V and 5V outputs together control the emergency shutdown. A quick glance at the schematic give that above 6.2V at the 5V or 15V on the 12 V would trigger the emergency shutdown. A reason for false triggering could be excessive ripple due to bad filtering capacitors. Check ripple on the 12V and 5V before the filter inductors.
Put dummy loads on the 5 and 12 V outputs and temporarily disconnect the zeners for the overvoltage system. Meassure ripple both before and after the filtering inductor.
Since the overvoltage system is measuring both voltages simultaneously uneven loading of the PSU can give this type of problems. I have seen this previously in s MicroVAX 2000 which need to have a dummy load if the hard disk is removed.
As far as I see it a capacitor stay reformed for quite some time. I find it unlikely that a capacitor heating up is creating this kind of repetitive cycle.
techknight
Well-known member
But here is the thing.
if the capacitor leaks, or even dries out from usage beyond its lifespan, the capacitance lowers, and the ESR rises. This allows spikes to get through as the capacitors are no longer filtering/suppressing them. Sure a dummy load is important, but thats not the case here.
So, when the power supply fires up, or even when its running, you can get a few microsecond spike which would normally be filtered, to pass through. just long enough to trip out a detection circuit.
As a capacitor with high ESR used in a demanding circuit sees power, the current flowing through its high internal resistance will heat the capacitor, and when a capacitor heats up, the ESR starts to drop a bit, and the capacitance actually raises a bit. Temporarily "reforming" a capacitor.
I know this effect, and I have seen it 1000000000000000000s of times. Especially with mid 2000s LCD monitors and TVs. Eventually the capacitor completely fails, and the effect doesnt take place without a hair dryer to help aid things along.
Back in my days as a service bench tech, I used to snuff out those types of capacitors all the time with a heat gun/hairdryer trick. The amount of Panasonic projection TVs, and Hitachi, mitsubishi sets that did this to me...
The worst ones were the 1st generation Mitsubishi DLP sets. Dear gosh the amount of bad capacitors in those. 100s of them. all SMD and some through hole.
if the capacitor leaks, or even dries out from usage beyond its lifespan, the capacitance lowers, and the ESR rises. This allows spikes to get through as the capacitors are no longer filtering/suppressing them. Sure a dummy load is important, but thats not the case here.
So, when the power supply fires up, or even when its running, you can get a few microsecond spike which would normally be filtered, to pass through. just long enough to trip out a detection circuit.
As a capacitor with high ESR used in a demanding circuit sees power, the current flowing through its high internal resistance will heat the capacitor, and when a capacitor heats up, the ESR starts to drop a bit, and the capacitance actually raises a bit. Temporarily "reforming" a capacitor.
I know this effect, and I have seen it 1000000000000000000s of times. Especially with mid 2000s LCD monitors and TVs. Eventually the capacitor completely fails, and the effect doesnt take place without a hair dryer to help aid things along.
Back in my days as a service bench tech, I used to snuff out those types of capacitors all the time with a heat gun/hairdryer trick. The amount of Panasonic projection TVs, and Hitachi, mitsubishi sets that did this to me...
The worst ones were the 1st generation Mitsubishi DLP sets. Dear gosh the amount of bad capacitors in those. 100s of them. all SMD and some through hole.
Last edited by a moderator:
techknight
Well-known member
Yep this is true. People that restore Vacuum Tube TV sets, and Radios already know this. as early as the 80s, and definitely now most of the capacitors if not all are leaky/bad.That being said though, there will be a day in the future when all of this stuff goes bad. Not yet, but eventually it will happen.
Every once in awhile you get the lucky ones where the capacitors are actually still ok, or are good enough to operate properly. But its rare.
Modern SMD electrolytics are the worst because of the rubber they used, they crack/shrink and it causes the leaks we see. Same thing with a few of the standard through hole stuff.
Throughout my previous years as a service tech, I have learned that the expected lifespan and quality of an electronic device is inversely proportional to the number of electrolytic capacitors they have.
Oh, and the wonderful blunders of the engineers sticking fluid filled capacitors directly near a heat producing component/heatsink. Knowing the heat will eventually dry out the cap and kill the device. Infocus was notorious for this in some of the early to mid 2000s DLP projectors.
Last edited by a moderator:
