Raspberry Pi B+ (V1.2) Dead

If your Raspberry Pi B+ (V1.2) is dead and shows no signs of life, no display, and will not boot, then this article might help you diagnose the problem and make repairs simple. If you have done something silly and applied more than five volts to your Pi B+, or perhaps applied an excessively large voltage in reverse polarity, then chances are your Pi is deceased, and the red power LED will not light. If you are lucky, the damage might be isolated at the power supply circuit and you might be able to fix it. However, you will require skills in soldering SMD components and working with WDFN-12L packages.

If you are unable to repair the board, then as they would say in the Monty Python show, your Pi is deceased, it has departed, it is a late Pi.

The Raspberry Pi B+ (V1.2) uses a step-down DC-DC converter integrated circuit (IC) U3 with the markings C2=CG UOZ. This IC is PAM2306AYPKE, which is used in battery-powered portable applications allowing the batteries to last longer. It is a much more efficient solution than the voltage regulators found in the earlier incarnations of the Raspberry Pi.

The PAM2306AYPKE is more than just a fixed voltage regulator IC, it is an intelligent power management system that actively reduces the current to the Pi circuits if it is not doing much. This enables the batteries to last much longer. This is one of my favourite chips by DIODES INCORPORATED and I use it at every opportunity.

This simplified schematic shows the pins responsible for the inputs and outputs. For detail, see the Raspberry Pi B+ circuit diagram.

This IC has dual programmable circuits and can therefore provide two independent and separate regulated outputs. Consequently, it also has two separate voltage inputs Vin1 (pin 7) and Vin2 (pin 1). As you can see, this IC therefore generates a 3.3 V rail at pin 8, and 1.8 V rail at pin 2. Both of these voltage rails should be present if your Pi is to work again.

The maximum input voltage this IC can accept is 5.5 V, and if you do something silly and feed more voltage to it, then you will destroy it. The IC has some internal protection against short circuits of the output rails. It also has thermal shutdown protection as well, and if the die temperature reaches +150 °C, then it automatically resets and remains locked in that condition until the temperature drops to +120 °C. As you can see, there are many safeguards, however they are not Brit proof and therefore, I anticipate many boards will die in the future.

If you notice the PAM2306AYPKE IC heating up, then you need to determine its temperature to see if it is over +150 °C. Look for any shorts to the 3.3 V and 1.8 V rails. If there is none, then chances are there has been damage to the IC and may require replacing.

Diagnostics

Diagnosis will require measuring voltages because we need to get these facts correct if we are to make a correct prognosis. We measure voltages with respect to the ground/earth, which on our board is the negative polarity or an earth point where the black lead of your digital meter would connect. You then use the red probe to measure voltages at various pins on the IC.

The first step is to look for a dot marking on the IC because it identifies pin 1. Once you know this, you can easily count the pins to the ones you are interested in. In our case, we need to measure the voltages on the pins that generate 3.3 V and 1.8 V. We should also check if there is indeed 5 V entering the IC, because if there is none, then we need to focus on the input stage, power socket, and the adapter.

In this diagram, the orientation and pinout of the voltage regulator IC PAM2306AYPKE (U3) is exactly as you would see it on the board as shown on the header of this page. The 5 V input to the IC is by the red pins, whilst the yellow pins are the regulated output voltages.

If your digital meter does not detect the 5 V input voltage as indicated in the diagram, then you need to focus your diagnostic in earlier stages of this circuit. This will be in the following pages of this multi-part article. However, let us assume for the time being that there is indeed 5 V entering the board and IC U3. If any one or both of the regulated output voltages are abnormally low, high, or absent, then your Raspberry will appear dead, and you will need to troubleshoot further. Unfortunately, U3 is a complicated integrated circuit, and it would be impossible to predict all the different modes of failure. I would need to buy hundreds of boards for testing to get accurate information. One possible type of failure would be if you were to detect extremely high voltages on either pin 2 or pin 8. For example, if these voltages were in the region of 4.5 V or thereabouts, then the IC may be short internally resulting in the input voltage (+5 V) appearing on the output pins. Abnormally high voltages can also affect the reset circuitry, which is usually active high, causing the processor to remain permanently frozen in a reset condition.

Another type of failure is when pin 2 and pin 8 read abnormally low voltages. In this case, it could be the regulator IC U3 itself producing low voltages due to internal damage, or it could be another IC connected to that voltage rail that may be short and loading that rail down. In this case, one would need to isolate the pin concerned from the voltage rail to determine which is at fault. To isolate the pins, one simple way is to remove the SMD inductor coils L1/L2 which are in-line to the rails. Alternatively, if you have some SMD soldering skills you may wish to isolate just the pins concerned.

Once you have isolated the pin, if it begins producing the correct voltage, then you know for sure that the regulator IC U3 is fine and the problem is with another IC that is loading that rail down. However, if the voltage remains abnormally low on the isolated pin, then you know that U3 itself is damaged.

I understand that there are newer versions of this board (PI 3B V1.2), however if it uses the same voltage regulator IC, then you will have to use similar principles outlined in this article. Currently, I am no longer working on the Pi.

This Article Continues...