STM CircleOS forum / Problem with regulator U17

baghirra · 2009-01-01 23:48:50

We have bought two Primer2 units. After some testing, loading toggle application and restoring original Primer2 firmware, the same problem appears on both units. Regulator U17 is too hot and

on first unit - out voltage is 2.5V and restoring original Primer2 firmware was successful,
on second unit - out voltage is cycling between 0V and 3.1V (probably thermal shutdown), restoring original firmware is impossible.

Any suggestions?

tybeede · 2009-01-02 05:48:48

I think the ground wire helps with power dissipation, you could cut the trace and green wire from the reg to ground with something a little heavier.

tybeede · 2009-01-02 06:07:26

It would be interesting to get some temperature and current readings from the reg. A thermistor and some thermal grease on the top of the chip might be helpful to determine any benefit from the larger ground path.

Francis · 2009-01-02 09:16:14

It's strange that the regulator problem appears on the two Primer you have. Until now, we see this problem only on a few units (2% of the production) and it was with U9, not U17...

Are you sure that you restore the right version of the firmware (it seems that most of the users restore the Primer1 firmware instead of Primer2')?

When the Primer1 firmware has been restored by mistake, the problem is that it shuts down the power supply when starting... The only way to workaround is to keep the button pressed for a while (until a new firmaware has been loaded).
The problem with the U9 regulator is not a thermal issue. The current is not so high. The main issue seems to be an overvoltage at the input due to the inductance of the thin wire..

For U17, I don't find it 'hot' on the Primers I have here... moreover, U17 is used only for the backlight and the output amplifier of the audio DAC. It should not interact on the loading of the firmware.

apasty · 2009-03-09 10:38:57

Hello,
I am new player with Stm32primer2, and after trying to upgrade to V3.6, system breakdown!
I thought it was upgrade problem and tried different version of upgrade, without succeed!
After investigation, I remarked that I could heard welcome message at start up (louder at start and quickly inaudible).
So the main problem, LCD backlight stay off.
About U17 is the problem because only 1.3V at output (against 3,1V at normal level).
So I am not sure on diagnostic! So much current is pumping on U17 or U17 broken down? I stay surprising me because that’s appended just after software update!
Do you have some advice on it.
Thank you.
Alain

Francis · 2009-03-09 13:15:04

There is no link between the upgrade and the regulators (the latest versions should be better because they reduce the overall consumption). But the problem is not within the current capabilities (the overall consumption is quite low, around 20% of the possible current for each regulator).

Now, we have seen (several customers too) that both regulators U9 and U17 are weak. Sometimes, they could be altered (for U17, backlight will toggle...) and sometimes they burn. The problems happen after touching the PCB with a finger. We discussed this issue with ST (with the team responsible for these regulators), and we didn't find any well identified explanation, but a set of suspicious things:
1. ST engineers recommend to use ceramic capacitors for C23 and C56 instead of tantalum. Their application note was not so clear (it says that tantalum is possible), but they now strongly recommend ceramic.
2. the routing of the board should be optimized, above all the traces for the feedback and for the input voltage.

It could be that the noise created by each regulator at its imput is added to the noise of the other regulator. The overall noise overstresses the input capacitors.

I made various tests, and after modifying #1 and #2, I was not able to burn a regulator anymore with my finger. These changes will be present on the next production... but, at the moment, the best solution consists in avoiding to touch the PCB (replacing the capacitors could be helpful too).

repzak · 2009-03-09 13:25:47

Hello,

It is possilbe to add a 100nF 1206 on the top of the tantalium capacitors, that will i do atleat...

Kasper

apasty · 2009-03-09 13:33:08

Hi Francis,

Thank you for your fast answer.
I will follow your advice about capacitors.
It not seems to me that U17 be burn!
So, PCB should be important investigating point, but I haven't found any documentation on it: do you have one?

Alain

Francis · 2009-03-09 14:23:06

The gerber files of the PCB are in the resource section.
For the future PCB (that will be indexed as 1.2), we rerouted the input voltage (as a V instead of having the same track that supply power to the first regulator then to the second). Moreover, we optimized the placement and the routing of all tracks around the regulators.
To Kasper: I didn't try this solution. I spent weeks with different tries but I didn't try this solution. I will ask ST about this and I will let you know.

# 1 2009-01-01 23:48:50 Problem with regulator U17

Problem with regulator U17

# 2 2009-01-02 05:48:48 Problem with regulator U17

Re: Problem with regulator U17

# 3 2009-01-02 06:07:26 Problem with regulator U17

Re: Problem with regulator U17

# 4 2009-01-02 09:16:14 Problem with regulator U17

Re: Problem with regulator U17

# 5 2009-03-09 10:38:57 Problem with regulator U17

Re: Problem with regulator U17

# 6 2009-03-09 13:15:04 Problem with regulator U17

Re: Problem with regulator U17

# 7 2009-03-09 13:25:47 Problem with regulator U17

Re: Problem with regulator U17

# 8 2009-03-09 13:33:08 Problem with regulator U17

Re: Problem with regulator U17

# 9 2009-03-09 14:23:06 Problem with regulator U17

Re: Problem with regulator U17

Board footer