Labolatory power supply fails to power up BeagleBone Black

[bremenpl]

Hello there,
I have a very strange problem with powering up BeagleBone Black (rev C).... When i try to power it up from some cheap AC adapter it works fine, but when I connect to to my labolatory power supply the power LED on board is lid for a second and then doesnt power up the MCU and turns off intead. I have connected both supplys to osciloscope and they are both stable, the laboratory one even more. Why is the power controller on the BeagleBone Black refusing to power up the MCU when powered from lab supply? I have no idea what is this about. I would aprichiate any help.

[Gerald Coley]

Have you tried leaving the power supply on and plug in the cable to the board or are you using the power switch on the power supply?

Gerald

On Friday, November 14, 2014, bremenpl <brem...@gmail.com> wrote:

Hello there,
I have a very strange problem with powering up BeagleBone Black (rev C).... When i try to power it up from some cheap AC adapter it works fine, but when I connect to to my labolatory power supply the power LED on board is lid for a second and then doesnt power up the MCU and turns off intead. I have connected both supplys to osciloscope and they are both stable, the laboratory one even more. Why is the power controller on the BeagleBone Black refusing to power up the MCU when powered from lab supply? I have no idea what is this about. I would aprichiate any help.

--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Gerald

 

ger...@beagleboard.org

http://beagleboard.org/

http://circuitco.com/support/

[Bremenpl]

I just plug the power and the board normally should turn on but it doesnt. I then press the pwr btn but the result is the same as if i plugged the power supply in.

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/xRPC_39ju9Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

[Bremenpl]

I read you post again:
I am plugging a cable that has power already, not using the switch.

Dnia 14 listopada 2014 13:06:11 Gerald Coley <ger...@beagleboard.org> napisał(a):

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/xRPC_39ju9Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

[evilwulfie]

As Gerald has said before, check the ramp up of the power supply.
Some of the lab supplys have slow ramp ups.
Yes they are very stable and VERY good regulation but
the BBB design requires a fast ramp up supply.

On 11/14/2014 3:03 AM, bremenpl wrote:

Hello there,
I have a very strange problem with powering up BeagleBone Black (rev C).... When i try to power it up from some cheap AC adapter it works fine, but when I connect to to my labolatory power supply the power LED on board is lid for a second and then doesnt power up the MCU and turns off intead. I have connected both supplys to osciloscope and they are both stable, the laboratory one even more. Why is the power controller on the BeagleBone Black refusing to power up the MCU when powered from lab supply? I have no idea what is this about. I would aprichiate any help.

[Bremenpl]

What is the exact circuit that requires a fast ramp?

W dniu 2014-11-14 o 14:48, evilwulfie pisze:

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/xRPC_39ju9Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

-- 
Bremenpl

[Gerald Coley]

The power management IC, TPS65217C.

Gerald

[Bremenpl]

Is there a workaround for this? I thought that pre charged capacitors could help but in my application power to the BeagleBone Black is aplied in the same time as to the rest of the circuit.

W dniu 2014-11-14 o 15:19, Gerald Coley pisze:

[Gerald Coley]

We are working on one if we can collect enough information to confirm the issue. it appears to be a power supply issue that creates a condition that the TPS65217C does not like. If you use a good power supply, the issue is not there.

Gerald

[Bremenpl]

I dont agree here, and this is the weird part. A cheap 5V wall plug adapter works great and a certified meanwell supply does not. Can it be a case related to linear/ switching power supply?

[Gerald Coley]

Not likely. As I said we need more boards that fail and the power supply that is used with them. What I have seen is that it tries to switch to the battery, that is not there, for whatever reason. Theory is it sees a voltage drop due to the inrush limit on certain power supplies. I have fixed this on a few boards by connecting the TS signal hard to ground using the battery pins on the board.

What I see when the board "does not power up" is the PMIC cycling on and off. You can look at the 5V output rail of the PMIC for this condition.

Gerald

[Bremenpl]

I understand, thank you for the info. I will make more attempts at monday with another bbb board. For now I have a 10k resistor on the TS pin because I intend to use the battery to. If the voltage is measured on the resistor ill swap it to lower value one. I hope ill get it working using meanwell power supply.

[Gerald Coley]

Another test here would be to connect the battery and see if after it switches to the battery if it switches back to the DC input.

Gerald

[Bremenpl]

How can I know either it switched without taking the battery out?

[Gerald Coley]

The output voltage level is different. When the DC is contented it is supposed to go to the DC input and go into charge mode for the battery. Measure the output voltage of the PMIC. If it is 5V, then it is DC in. If it is 3.7V +/- then it is running on the battery.

Gerald

[Bremenpl]

Ill get it tested, thank you. If this worked it would solve my problem.

[Gerald Coley]

And mine.

Gerald

[Bremenpl]

A question regarding battery though- if i leave the bbb under battery supply, but turn off the mcu with pwr button, what will be the current drawn by TPS chip? Wont it dry the battery through the night? Lets say its a single li-ion cell 2200mAh

[Graham]

I have a similar problem, using a "B&K Precision 1550" Lab Bench Power Supply.
When you turn on the Output of the supply, the BBB Power Light blinks once and
the BeagleBone Black does not boot.

If you put an oscilloscope on the output of the supply and watch it turn on, then
the power supply overshoots up to 6 Volts, before returning to the 5 Volt setting.

It turns out that the Power Management IC used in the Beaglebone Black will
do an over-voltage shutdown at 6 Volts for protection. So, that is why it does
one-blink then shuts-off.

So, not only do you need a certain rise time, which the B&K power supply
meets, but  you can not have any large overshoot.  No good power supply
should have this much overshoot, but the B&K does, so it has problems
running a BBB.

--- Graham

==

[John Syn]

From: evilwulfie <evilw...@gmail.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Friday, November 14, 2014 at 5:48 AM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: Re: [beagleboard] Labolatory power supply fails to power up BeagleBone Black

As Gerald has said before, check the ramp up of the power supply.
Some of the lab supplys have slow ramp ups.
Yes they are very stable and VERY good regulation but
the BBB design requires a fast ramp up supply.

It could also be that he has current limit set too low. Set the current limit to 2A or more. It would be helpful if we could see an oscilloscope capture of the startup waveform. 

Regards,

John

[John Syn]

From: Graham <gra...@flex-radio.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Friday, November 14, 2014 at 8:03 AM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: [beagleboard] Re: Labolatory power supply fails to power up BeagleBone Black

I have a similar problem, using a "B&K Precision 1550" Lab Bench Power Supply.
When you turn on the Output of the supply, the BBB Power Light blinks once and
the BeagleBone Black does not boot.

If you put an oscilloscope on the output of the supply and watch it turn on, then
the power supply overshoots up to 6 Volts, before returning to the 5 Volt setting.

That is just crazy. I have never seen a bench power supply do this. Perhaps you should contact B&K, because there must be something wrong with the voltage regulation on your power supply. 

Regards,

John

It turns out that the Power Management IC used in the Beaglebone Black will
do an over-voltage shutdown at 6 Volts for protection. So, that is why it does
one-blink then shuts-off.

So, not only do you need a certain rise time, which the B&K power supply
meets, but  you can not have any large overshoot.  No good power supply
should have this much overshoot, but the B&K does, so it has problems
running a BBB.

--- Graham

==

On Friday, November 14, 2014 4:03:37 AM UTC-6, bremenpl wrote:

Hello there,
I have a very strange problem with powering up BeagleBone Black (rev C).... When i try to power it up from some cheap AC adapter it works fine, but when I connect to to my labolatory power supply the power LED on board is lid for a second and then doesnt power up the MCU and turns off intead. I have connected both supplys to osciloscope and they are both stable, the laboratory one even more. Why is the power controller on the BeagleBone Black refusing to power up the MCU when powered from lab supply? I have no idea what is this about. I would aprichiate any help.

--

[Gerald Coley]

I have seen it on one board. Funny thing is it does not always do it.

Gerald

[Graham Haddock]

Gerald:

Well, in the case of the B&K 1550, set to 5.0 Volts, and any reasonable
current limit, if the power supply output is turned on,

and the power supply cap in the Beaglebone Black is fully discharged,

then the BBB never starts. Just one blink of the power LED then nothing.

On the oscilloscope, voltage rises from zero to 6 Volts in one ms, then

decays back to 5.0 Volts over 100 ms. or so. Probably trips the overvoltage
detect in the BBB power chip, which must be latching.

If you cycle the output off, wait about three seconds, then on, and the caps
in the BBB have dropped enough to reset the BBB power chip, but not

decayed all the way back to zero, then the over shoot voltage is less when
you turn it back on, and the BBB will generally start on this second try.

P.O.S. power supply.

--- Graham

==

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/xRPC_39ju9Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.

[Bremenpl]

I have tested the board again and heres what I have come to:
- Using meanwell 5V ac adapter works when the 10k resistor is NOT soldered to battery header. If theres resistor soldered, but there is no battery attached it stops working and the power led blinks all the time. It really seems that the power controller is trying to switch to the battery that is not there. After placing the battery everything works correct. I only home it wont get discharged to fast when the board is off.

W dniu 2014-11-14 o 16:58, Gerald Coley pisze:

[Gerald Coley]

Replace the resistor with a wire and just short it to ground.

Gerald

[Bremenpl]

But will trhis allow battery to work properly?

W dniu 2014-11-17 o 11:48, Gerald Coley pisze:

[Bremenpl]

I have observed an interesting thing now. I have replaced the resistor as you adviced. Now I can indeed use the BeagleBone Black without battery. Now when I run on a regular power supply and the battery is in and charging, when i remove the main power supply it seems like BeagleBone Black is performing a safe shutdown (im not sure im not connected to serial debug), because it doesnt automatically go off but it takes its time turn off user leds and power led. Is this made on purpose? Why didnt he act like this when there was 10k resistor instead of 0R?

W dniu 2014-11-17 o 11:48, Gerald Coley pisze:

Replace the resistor with a wire and just short it to ground.

-- 
Bremenpl

[Bremenpl]

And also an important thing- I cannot turn on BeagleBone Black on battery only now.

W dniu 2014-11-17 o 11:48, Gerald Coley pisze:

Replace the resistor with a wire and just short it to ground.

-- 
Bremenpl

[Gerald Coley]

When you remove it yes. In the meantime I am hoping this stops people from blowing up processors. If battery is needed it can be removed.

[Gerald Coley]

I can accept that. This woul be only for non battery operation.

[Bremenpl]

This functionality is actually very good for me. I was intending to make a system that prevents user from turning off BeagleBone Black without proper shutdown. I thought i will have to implement this in code, but since the BeagleBone Black is shutting down automatically after the main supply is down (running on the battery in that time) i dont have to do this on my own. Please correct me if I am wrong in understanding this.

W dniu 2014-11-17 o 12:44, Gerald Coley pisze:

[Gerald Coley]

As long as there is nothing that can cause the PMIC to shutdown before it has completed the full power up cycle we are OK. It is the sitting down before that process has completed is what concerns me.

[Bremenpl]

So you are saying that if I remove the main power supply while
BeagleBone Black is turning on (like 7 seconds?) the safe shutdown wont
occur?

I have checked some scenarios:
- Turn on BeagleBone Black with battery in and power supply. Then remove
power supply before BeagleBone Black fully inits. If I do that the board
wont turn off and I have to do it manualy after full init. I gues thats
ok, I can check for power right after I turn on my program.

- Second scenario is turning the board on with battery in and usb cable
with power in. In this case BeagleBone Black keeps resseting after 2-3
secs. Not sure whats happening here.

W dniu 2014-11-17 o 12:58, Gerald Coley pisze:

> As long as there is nothing that can cause the PMIC to shutdown before
> it has completed the full power up cycle we are OK. It is the sitting
> down before that process has completed is what concerns me.
>
> Gerald
>

--
Bremenpl

[Gerald Coley]

That all depends on the state of the current SW configuration. My focus is shutdown before SW even wakes up and has it's first cup of coffee, before SW can make anything happen.

You first scenario is after SW has already run via the uBoot and done what it needs to do.

Per the datasheet there may be issues detecting the battery when DC and USB are connected. Check the datasheet for more information to see how it affects you.

Gerald

--
Bremenpl

--
For more options, visit http://beagleboard.org/discuss
--- You received this message because you are subscribed to the Google Groups "BeagleBoard" group.

To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Bremenpl]

My software is supposed to be some userspace software that is turned on via script when the board is fully on. Also in the end application I dont connect power pin on ish slave connection.

You received this message because you are subscribed to a topic in the Google Groups "BeagleBoard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beagleboard/xRPC_39ju9Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beagleboard...@googlegroups.com.