Canon G2010 turns off in seconds after pressing power button

[PeterBJ]

I opened a power supply from a Canon MP600. The reservoir capacitor is 150µF/400V, and no bleeder resistor is connected in parallel. Here the reservoir capacitor can be seen:

BTW the power supply has not been connected to the mains for several months, and I verified with a DMM that there was no voltage on then capacitor.
I found a YouTube video about a similar power supply, It is in German but notice that at 5:28 the voltage over the reservoir capacitor is still 220 VDC, and the mains was disconnected from the power supply at the beginning of the video. I recommend to wait at least 10 minutes after disconnecting the mains before opening the box.
I noticed in the second video linked to by stratman that the power supply looked very similar to the type K30321with an output of 24 VDC @ 1,0 A. At 0:16 you can see that the type number is K30377 and the output is 24 VDC @ 0.75 A.

I wonder if the two power supplies behave in the same way? So can they be tested in the same way? Here is a sketch of the K30321 with the output connector:

The pins marked + and - are the output, it is 8 VDC in standby mode and 24 VDC in operating mode. With no voltage applied to the control pin marked C the output is 8V. With +3 V on pin C with the - pin as ground the output is 24 V. Maybe this could be used to test the C2010 power supply? You can get the 3 V control signal from a lithium coin cell or a couple of AAAs in a battery box.

 

[stratman]

notice that at 5:28 the voltage over the reservoir capacitor is still 220 VDC

Thanks for your explanation. What was causing the bleed off of the capacitor - simply time not connected to a current or his testing of the voltage?

 

[Artur5]

Certainly the PSU of the MP600 looks way more powerful than the average Canon PSU I dismantled in the past. Right now only I have at hand a K30233 from my late IP3000.
As you can see in the photo, the input electrolytic capacitor is only 22uFarad and there’s not even a small heatsink anywhere.
Maybe power supplies of multifunction machines need to give more wattage than plain printers because it seems to be general consensus that the first generation of Pixmas were machines very well built.

Anyway, I tested that one right now and as soon as you disconnect the cable from the mains, the tension between the pins of that 22uF first capacitor goes from 330V DC to 24V in a small fraction of second and then it sinks down very slowly. One minute later it’s still 19V.
The K30233 is rated for an output of 24V-1.2A DC and should be used only on 220-240VAC mains, while I recall that other Canon PSUs had universal inputs ( 100-240 VAC ). Maybe because of that, the internal design is less sophisticated.

 

[The Hat]

I tested that one right now and as soon as you disconnect the cable from the mains, the tension between the pins of that 22uF first capacitor goes from 330V DC to 24V in a small fraction of second and then it sinks down very slowly. One minute later it’s still 19V.

What the differences between power supply’s is telling us, is to leave the printer turned off for at least 5 minutes to allow the power to drain out before turning back on to insure a clean boot, Question:- how much power remains in the logic board if the printer remains plugged in..

 

[PeterBJ]

Thanks for your explanation. What was causing the bleed off of the capacitor - simply time not connected to a current or his testing of the voltage?

Electrolytic capacitors are not perfect, they have a leakage resistance that will discharge the capacitor slowly or quickly.

Most modern digital multimeters have an input resistance of 10 MOhm. Measuring the voltage over the capacitor with that instrument will not speed up discharge very much. Some less expensive instruments have an input resistance of 1 MOhm. Using this instrument might speed up the discharge. See this about capacitor discharge and time constants.

There are many kinds of switch mode power supplies, and they do not all behave in the same way. I think that in the circuit examined by @Artur5 most of the charge in the capacitor is used up by the voltage converter until the voltage becomes too low and the converter stops. This would explain the quick discharge to a low voltage and then the slow discharge only determined by the capacity and leakage resistance of the reservoir capacitor.

I still think that you should wait 5 or better 10 minutes before opening a power supply, you don't know how quickly the reservoir capacitor will discharge to a safe voltage.

I also think Artur5 is right in that Canon printer power supplies rarely fail. Instead failing to power on is caused by a protection circuitry reacting to an overcurrent. This overcurrent is most likely caused by a failed printhead or logic board.

 

[Artur5]

@The Hat, do you mean the printer powered off but connected to the power outlet ?.
According to the service manual of the Pixma IP3000, the averege consumption when the machine is working would be 15W. On standby 0.5W and powered off 0.25W. Given the small size of the capacitors of the PSU, I don't think that it would take more than a few seconds to discharge them when you remove the power cable, if the PSU is connected to the printer. It's more difficult to know how long it will take to discharge totally the more significant components in the logic board, such as the volatile memory. Your guess is as good (or better) than mine. We can ask Canon, can't we ?

 

[stratman]

you don't know how quickly the reservoir capacitor will discharge to a safe voltage.

From your article,

"The time constant, τ is found using the formula T = R*C in seconds."

If you know 1T then multiply by 5 to get 5T and then at that time the Capacitor is considered discharged.

If they taught this in my Physics 101 class I sure do not remember it!

 

[stratman]

Your guess is as good (or better) than mine.

I'd say give it a try and touch it BUT I like you Artur5.

DO NOT TRY IT!

 

[The Hat]

DO NOT TRY IT!

 

[Artur5]

Thanks for your concern guys, much appreciated. .. but why shouldn't I ask Canon ?

Worst thing they could do is to ignore my request, or do you expect that they would retaliate in a more sinister way ?