Pac0
Posts: 5
Joined: Sat Jun 09, 2018 3:24 pm

Voltage normalised (0x00000000)

Sat Jun 09, 2018 3:42 pm

This is my first post, and I have registered in order to ask for this message. I know other threads that people discused about this issue, but I would put my opinion about... because I think is not serious what I read...

I saw this message in the raspberri logs and I'm surprised because engineers said that only appears when the pi is powereed by a poor tension. I disagree with this opinion because I powered my pi with 5A - power supply and I continue receiving the same warning every minute. I raised the tension to 5.5V and I continue receiving warnings

Could be this is a marketing operation because someone linked an official power supply, but this is not the solution...
Other person said:
News at 10. Person ignoring smoke alarm killed in fire. (never EVER turn off a smoke detector)
well My smoke alarm is blinking but I don't see smoke not fire... could be the alarm have a malfunction???

Guys please take the people that have a raspberrry seriously and don't take us for fools.

klricks
Posts: 5899
Joined: Sat Jan 12, 2013 3:01 am
Location: Grants Pass, OR, USA
Contact: Website

Re: Voltage normalised (0x00000000)

Sat Jun 09, 2018 8:05 pm

The micro USB power cord is likely your issue. Too thin and/or too long and/or poor quality.
I use a 10A power supply and still got the on screen 'lightning bolt' low voltage indicator until I made my own micro USB cable.
Measure the voltage at GPIO pin 2 (5V) and GND (any metal port shield).
Unless specified otherwise my response is based on the latest and fully updated Raspbian Stretch w/ Desktop OS.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 20775
Joined: Sat Jul 30, 2011 7:41 pm

Re: Voltage normalised (0x00000000)

Sat Jun 09, 2018 9:21 pm

Pac0 wrote:
Sat Jun 09, 2018 3:42 pm
This is my first post, and I have registered in order to ask for this message. I know other threads that people discused about this issue, but I would put my opinion about... because I think is not serious what I read...

I saw this message in the raspberri logs and I'm surprised because engineers said that only appears when the pi is powereed by a poor tension. I disagree with this opinion because I powered my pi with 5A - power supply and I continue receiving the same warning every minute. I raised the tension to 5.5V and I continue receiving warnings

Could be this is a marketing operation because someone linked an official power supply, but this is not the solution...
Other person said:
News at 10. Person ignoring smoke alarm killed in fire. (never EVER turn off a smoke detector)
well My smoke alarm is blinking but I don't see smoke not fire... could be the alarm have a malfunction???

Guys please take the people that have a raspberrry seriously and don't take us for fools.
First post, and accuses the PRF of using this as an excuse to sell power supplies.

Wow. Just wow.

Your power cable is crap.

That message is the result of the power supply system on the B+ (a SMPS chip) detecting a low voltage on the board. Low enough that in some circumstances it might cause problems. That is a FACT, not a marketing exercise.

Why would you think we would be so desperate to sell power supplies when we sell 5 million Pi's a year?
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Please direct all questions to the forum, I do not do support via PM.

Milliways
Posts: 315
Joined: Fri Apr 25, 2014 12:18 am

Re: Voltage normalised (0x00000000)

Sun Jun 10, 2018 5:36 am

Pi users seem to be fixated on the current rating of their power supply (often with little or no understanding of WHAT current actually is).

This IS (or at least should be) the MAXIMUM current the PSU can safely deliver at the rated voltage. Of course MANY PSU don't actually perform to their stated rating.

You can run a Pi3B+ from a 1A supply - there is ABSOLUTELY NO POINT in using a supply rated > 2.5A (because the current is limited by a polyfuse) although this is OK.

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Sun Jun 10, 2018 3:36 pm

Milliways wrote:
Sun Jun 10, 2018 5:36 am
You can run a Pi3B+ from a 1A supply - there is ABSOLUTELY NO POINT in using a supply rated > 2.5A (because the current is limited by a polyfuse) although this is OK.
You are correct though a supply rated at over 2.5A tends to be more likely to give out the full 5V at 2.5A than one rated lower.

But even then it's not that simple. I use 5V 5A supplies which reliably deliver 5V at 5A, or whatever a Pi or anything else requires. Except they appear to have 'soft-start current limiting' which means they do not immediately deliver what they are capable of, causing under-voltage lightning bolts during booting.

Milliways
Posts: 315
Joined: Fri Apr 25, 2014 12:18 am

Re: Voltage normalised (0x00000000)

Mon Jun 11, 2018 12:01 am

hippy wrote:
Sun Jun 10, 2018 3:36 pm
Milliways wrote:
Sun Jun 10, 2018 5:36 am
You can run a Pi3B+ from a 1A supply - there is ABSOLUTELY NO POINT in using a supply rated > 2.5A (because the current is limited by a polyfuse) although this is OK.
You are correct though a supply rated at over 2.5A tends to be more likely to give out the full 5V at 2.5A than one rated lower.

But even then it's not that simple. I use 5V 5A supplies which reliably deliver 5V at 5A, or whatever a Pi or anything else requires. Except they appear to have 'soft-start current limiting' which means they do not immediately deliver what they are capable of, causing under-voltage lightning bolts during booting.
That is WHY most computers use the kind of voltage detector chip used in the Pi3 to hold the processor in reset until the voltage is OK - and we are talking mS not seconds.

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Mon Jun 11, 2018 11:03 am

Milliways wrote:
Mon Jun 11, 2018 12:01 am
hippy wrote:
Sun Jun 10, 2018 3:36 pm
I use 5V 5A supplies which reliably deliver 5V at 5A, or whatever a Pi or anything else requires. Except they appear to have 'soft-start current limiting' which means they do not immediately deliver what they are capable of, causing under-voltage lightning bolts during booting.
That is WHY most computers use the kind of voltage detector chip used in the Pi3 to hold the processor in reset until the voltage is OK - and we are talking mS not seconds.
Unfortunately it seems there is no timed hold-in-reset for the Pi, or, if there is, it's not long enough for my PSU.

The PSU puts out 5V, the reset controller /power manager sees that, lets the Pi come out of reset, the current draw increases, the PSU limits that, voltage drops, lightning bolt appears, and this continues until the PSU is eventually happy providing the full current demanded.

I'm not criticising the PI here; just describing the behaviour I see, noting that simply having a PSU with a high current capability doesn't make everything issue free.

I was surprised at the PSU behaviour, surprised it seems to have such a long soft-start period. I have a few and they all behave the same way. I am assuming that, although the low-voltage lightning bolts appear, the voltage is high enough to let the SoC and on-board stuff work as expected which is why I don't have problems. I might measure it one day.

It has only been an issue with my Pi 3B, is fine on everything else, and only occurs during booting the Pi.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 20775
Joined: Sat Jul 30, 2011 7:41 pm

Re: Voltage normalised (0x00000000)

Mon Jun 11, 2018 11:38 am

hippy wrote:
Mon Jun 11, 2018 11:03 am
Milliways wrote:
Mon Jun 11, 2018 12:01 am
hippy wrote:
Sun Jun 10, 2018 3:36 pm
I use 5V 5A supplies which reliably deliver 5V at 5A, or whatever a Pi or anything else requires. Except they appear to have 'soft-start current limiting' which means they do not immediately deliver what they are capable of, causing under-voltage lightning bolts during booting.
That is WHY most computers use the kind of voltage detector chip used in the Pi3 to hold the processor in reset until the voltage is OK - and we are talking mS not seconds.
Unfortunately it seems there is no timed hold-in-reset for the Pi, or, if there is, it's not long enough for my PSU.

The PSU puts out 5V, the reset controller /power manager sees that, lets the Pi come out of reset, the current draw increases, the PSU limits that, voltage drops, lightning bolt appears, and this continues until the PSU is eventually happy providing the full current demanded.

I'm not criticising the PI here; just describing the behaviour I see, noting that simply having a PSU with a high current capability doesn't make everything issue free.

I was surprised at the PSU behaviour, surprised it seems to have such a long soft-start period. I have a few and they all behave the same way. I am assuming that, although the low-voltage lightning bolts appear, the voltage is high enough to let the SoC and on-board stuff work as expected which is why I don't have problems. I might measure it one day.

It has only been an issue with my Pi 3B, is fine on everything else, and only occurs during booting the Pi.
Note the Pi also takes a bit more current during startup, as it's starting up all the peripherals.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Please direct all questions to the forum, I do not do support via PM.

Milliways
Posts: 315
Joined: Fri Apr 25, 2014 12:18 am

Re: Voltage normalised (0x00000000)

Mon Jun 11, 2018 12:05 pm

hippy wrote:
Mon Jun 11, 2018 11:03 am
Milliways wrote:
Mon Jun 11, 2018 12:01 am
hippy wrote:
Sun Jun 10, 2018 3:36 pm
I use 5V 5A supplies which reliably deliver 5V at 5A, or whatever a Pi or anything else requires. Except they appear to have 'soft-start current limiting' which means they do not immediately deliver what they are capable of, causing under-voltage lightning bolts during booting.
That is WHY most computers use the kind of voltage detector chip used in the Pi3 to hold the processor in reset until the voltage is OK - and we are talking mS not seconds.
Unfortunately it seems there is no timed hold-in-reset for the Pi, or, if there is, it's not long enough for my PSU.

The PSU puts out 5V, the reset controller /power manager sees that, lets the Pi come out of reset, the current draw increases, the PSU limits that, voltage drops, lightning bolt appears, and this continues until the PSU is eventually happy providing the full current demanded.

I'm not criticising the PI here; just describing the behaviour I see, noting that simply having a PSU with a high current capability doesn't make everything issue free.

I was surprised at the PSU behaviour, surprised it seems to have such a long soft-start period. I have a few and they all behave the same way. I am assuming that, although the low-voltage lightning bolts appear, the voltage is high enough to let the SoC and on-board stuff work as expected which is why I don't have problems. I might measure it one day.

It has only been an issue with my Pi 3B, is fine on everything else, and only occurs during booting the Pi.
What you are saying is your PSU is inadequate.

I have stated before (but received lots of negative comments) this is a design fault of the Pi.
The PSU may be adequate, but if you are going to use a 1M lead to connect 5V to it voltage drop is a problem. The resistance to keep voltage drop within limits is ridiculously low. You COULD use hefty cables, but these are rare.

In the 1970s and 1980s I used to design computer systems and the 5V power was always an issue. Many systems in the day "solved" this issue with on-board regulators feeding higher voltages to the board.

Modern computer systems either have on-board batteries and feed a higher voltage (~16-19V) which reduces current, and allows reasonable voltage drop OR have switch mode PSU in the case - connected by a fist full of cables.

Given that we can't change the Pi design the solution is to use decent short cables. I do this (and don't use cheap $10 plug packs).

NOTE the Pi itself DOES NOT NEED 5V; NOTHING on the board requires this - they will run down to ~4V (I haven't tested Pi3B+ yet), so even though the low voltage warning is triggered will work (this may not apply to USB peripherals - but I use a powered hub for high power peripherals).

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Mon Jun 11, 2018 5:20 pm

Milliways wrote:
Mon Jun 11, 2018 12:05 pm
What you are saying is your PSU is inadequate.
No; I am not. My PSU drops below the low-level voltage threshold the Pi uses but that does not compromise operation of the Pi. So I do not consider it inadequate even if the Foundation consider it a critical issue which must be rectified, because ...
Milliways wrote:
Mon Jun 11, 2018 12:05 pm
NOTE the Pi itself DOES NOT NEED 5V; NOTHING on the board requires this - they will run down to ~4V (I haven't tested Pi3B+ yet), so even though the low voltage warning is triggered will work (this may not apply to USB peripherals - but I use a powered hub for high power peripherals)
It is worth noting that 5V does go to the BCM2835 on the original B and perhaps other boards.

User avatar
Imperf3kt
Posts: 1386
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: Voltage normalised (0x00000000)

Mon Jun 11, 2018 9:39 pm

A little something about 5v '5A' USB power supplies / chargers.
I have one that is capable of maintaining [email protected], yet when connected to the Pi via a very expensive '2.4A' rated cable, I found it still inadequate.
Know what the problem was?
The USB port itself.

The PSU indeed delivers [email protected]~5A and the USB cable indeed carries enough current to supply my Pi, but the rectangular bit you plug it into, was limiting current flow.

Its like having a dam with 30 million liters of water, a garden hose large enough to stand up in, but a connection the size of a pin prick.
Stop plugging your fan directly into the GPIO 5v
https://www.electronics-tutorials.ws/power/transient-suppression.html

Pac0
Posts: 5
Joined: Sat Jun 09, 2018 3:24 pm

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 11:50 am

jamesh wrote:
Sat Jun 09, 2018 9:21 pm
Pac0 wrote:
Sat Jun 09, 2018 3:42 pm
This is my first post, and I have registered in order to ask for this message. I know other threads that people discused about this issue, but I would put my opinion about... because I think is not serious what I read...

I saw this message in the raspberri logs and I'm surprised because engineers said that only appears when the pi is powereed by a poor tension. I disagree with this opinion because I powered my pi with 5A - power supply and I continue receiving the same warning every minute. I raised the tension to 5.5V and I continue receiving warnings

Could be this is a marketing operation because someone linked an official power supply, but this is not the solution...
Other person said:
News at 10. Person ignoring smoke alarm killed in fire. (never EVER turn off a smoke detector)
well My smoke alarm is blinking but I don't see smoke not fire... could be the alarm have a malfunction???

Guys please take the people that have a raspberrry seriously and don't take us for fools.
First post, and accuses the PRF of using this as an excuse to sell power supplies.

Wow. Just wow.

Your power cable is crap.

That message is the result of the power supply system on the B+ (a SMPS chip) detecting a low voltage on the board. Low enough that in some circumstances it might cause problems. That is a FACT, not a marketing exercise.

Why would you think we would be so desperate to sell power supplies when we sell 5 million Pi's a year?
Well my power cable is short, is a crap, could be, but if usb cables are crap, why you don't power de raspberry by a standard connector. Could be users don't will find that kinf of issues...

B.Goode
Posts: 6544
Joined: Mon Sep 01, 2014 4:03 pm
Location: UK

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 12:03 pm

why you don't power de raspberry by a standard connector.
Can you provide a clear link to documentation that describes the "standard connector" that you think should be used?

(Noting that no simple physical connection can mandate or enforce the characteristics of the signal it is required to carry. For example consider the range of services that can be provided by using a 'standard' utp cable with an rj45 connection... )

Pac0
Posts: 5
Joined: Sat Jun 09, 2018 3:24 pm

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 12:06 pm

hippy wrote:
Sun Jun 10, 2018 3:36 pm
Milliways wrote:
Sun Jun 10, 2018 5:36 am
You can run a Pi3B+ from a 1A supply - there is ABSOLUTELY NO POINT in using a supply rated > 2.5A (because the current is limited by a polyfuse) although this is OK.
You are correct though a supply rated at over 2.5A tends to be more likely to give out the full 5V at 2.5A than one rated lower.

But even then it's not that simple. I use 5V 5A supplies which reliably deliver 5V at 5A, or whatever a Pi or anything else requires. Except they appear to have 'soft-start current limiting' which means they do not immediately deliver what they are capable of, causing under-voltage lightning bolts during booting.
Yes I saw this messages too. In other hand if you connect a USB external drive, you need the 2.5A or more. My pi is a control-center for a PV system, it has an HDD, with Mysql database, and several periphereals in i2c bus. And yes after shorten the usb "crap" cable I've increased the Vcc in order to get 5V at pin 2. But even so I receive the message at boot time...

Pac0
Posts: 5
Joined: Sat Jun 09, 2018 3:24 pm

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 12:14 pm

B.Goode wrote:
Tue Jun 12, 2018 12:03 pm
why you don't power de raspberry by a standard connector.
Can you provide a clear link to documentation that describes the "standard connector" that you think should be used?

(Noting that no simple physical connection can mandate or enforce the characteristics of the signal it is required to carry. For example consider the range of services that can be provided by using a 'standard' utp cable with an rj45 connection... )
The documentation is of public domain...
I was talking about jack-type connector the standard in the industry of CC equipment (until someone decided that USB connectors could be used)
the d¡famous wikipedia has page for that:
https://en.wikipedia.org/wiki/DC_connector
clear enough?

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 12:19 pm

Imperf3kt wrote:
Mon Jun 11, 2018 9:39 pm
Its like having a dam with 30 million liters of water, a garden hose large enough to stand up in, but a connection the size of a pin prick.
I wouldn't say pin prick, but the connection could be a limiting factor, if not the cable.

I have asked before what the current carrying rating of the micro-USB connectors are for power, whether they are actually specified for carrying the 2.5A a Pi may use, but have never received an answer.

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 11033
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 12:21 pm

Pac0 wrote:
Tue Jun 12, 2018 11:50 am
Well my power cable is short, is a crap, could be, but if usb cables are crap, why you don't power de raspberry by a standard connector. Could be users don't will find that kinf of issues...
WHY?
Because its simply the best solution, if you want a power supply that directly delivers the 5V, +-5% that among other things USB devices need.

Other "connector solutions" are simply worse, they have much more contact resistance than the gold plated microUSB connectors have, for example "barrel connectors" (I suppose you mean those when talking about a "standard connector") are much-much worse, the only reason they are so popular is because they are normally used in applications where contact resistance doesn't matter. If you input 9V DC, for a DC regulator, it doesn't matter if the barrel connector "looses" 0.5V. But that would be disastrous if you use it to directly power USB devices.

I hear this error a lot, people think that because barrel connectors a popular they can be used in this application, but no they can NOT!
They are normally tin-plated, with low contact pressure, meaning they have high contact resistance, (the heat the contact resistance produces can be easily dissipated because of the bulk of the metal, so they can still have a high amperage rating, but they still introduce voltage drops, and sometimes become hot), they are nor properly standardized (they can be AC or DC, and any voltage and both polarities, and different internal pin sizes etc etc), and they move around so they tend to introduce short contact breaks, which will crash a computer system (unless you put massive amounts of bulk capacitors behind them, and even then they tend to introduce system crashes). So they are S**T connectors for this application. If they were used power problems would go through the roof!

Its not hard to find an USB power supply, that can deliver 5V +_ 5% while delivering 2.5A, there are plenty of those, it more or less becoming a standard.
Its also not difficult to get an USB to microUSB cable that doesn't drop more than 0.1 Volt while 2.5A runs through it, if you simply don't buy them from cheap Chinese supplier, who sell you the cheapest made junk possible because you won't notice if it takes ten minutes longer to charge your phone with it, as that is what these are typically used for. They use metaled plastic-foil inside instead of proper 22AWG or thicker copper wire. Or as Dougie Lawson here calls it they use "wet string". :mrgreen:

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 12:44 pm

hippy wrote:
Tue Jun 12, 2018 12:19 pm
I have asked before what the current carrying rating of the micro-USB connectors are for power, whether they are actually specified for carrying the 2.5A a Pi may use, but have never received an answer.
When I had previously asked I must have picked the only published circuit diagram (B+) which did not include any part number for the connector so I could not check for myself :roll:

Circuit diagrams for the Pi 3B+, 3B, 2B, and A+ show the power-in connector as being "10103594-0001LF". All the specification data I can find on that suggests a maximum 1.8A rating, eg -

https://www.alliedelec.com/m/d/ee78f0da ... 333fdb.pdf

It would seem the connector itself is not specified to carry the current which the recommended PSU could supply, the current which could be drawn through that connector :o

Perhaps someone from RPT could clarify what the situation is, confirm the maximum rating of the "10103594-0001LF" micro-USB connector used ?

Pac0
Posts: 5
Joined: Sat Jun 09, 2018 3:24 pm

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 1:00 pm

its not hard to find an USB power supply, that can deliver 5V +_ 5% while delivering 2.5A, there are plenty of those, it more or less becoming a standard.
Its also not difficult to get an USB to microUSB cable that doesn't drop more than 0.1 Volt while 2.5A runs through it, if you simply don't buy them from cheap Chinese supplier, who sell you the cheapest made junk possible because you won't notice if it takes ten minutes longer to charge your phone with it, as that is what these are typically used for. They use metaled plastic-foil inside instead of proper 22AWG or thicker copper wire. Or as Dougie Lawson here calls it they use "wet string". :mrgreen:
I'm reticent, but I agree your exposition of the facts...
Now I'm powering the pi with 5.3v enough to arrive 5v to pin 2. This agree with your comment.
I'm waiting for a best cable a braided one, but you know, today most of this cables (best ones or bad ones) are manufactured in china or similar.
The problem can solves it self using a span (5") of cable...

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 20775
Joined: Sat Jul 30, 2011 7:41 pm

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 1:23 pm

hippy wrote:
Tue Jun 12, 2018 12:44 pm
hippy wrote:
Tue Jun 12, 2018 12:19 pm
I have asked before what the current carrying rating of the micro-USB connectors are for power, whether they are actually specified for carrying the 2.5A a Pi may use, but have never received an answer.
When I had previously asked I must have picked the only published circuit diagram (B+) which did not include any part number for the connector so I could not check for myself :roll:

Circuit diagrams for the Pi 3B+, 3B, 2B, and A+ show the power-in connector as being "10103594-0001LF". All the specification data I can find on that suggests a maximum 1.8A rating, eg -

https://www.alliedelec.com/m/d/ee78f0da ... 333fdb.pdf

It would seem the connector itself is not specified to carry the current which the recommended PSU could supply, the current which could be drawn through that connector :o

Perhaps someone from RPT could clarify what the situation is, confirm the maximum rating of the "10103594-0001LF" micro-USB connector used ?
And yet.....

Here I am, running a 1.5A external USB drive attached to a 3B+ running from the Raspberry Pi power supply, doing long term ethernet stability testing. I see no warnings, nothing has blown up.

Use appropriate equipment, everything is fine.

We have no intention of changing the USB connector, it works fine.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Please direct all questions to the forum, I do not do support via PM.

B.Goode
Posts: 6544
Joined: Mon Sep 01, 2014 4:03 pm
Location: UK

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 1:25 pm

Pac0 wrote:
Tue Jun 12, 2018 12:14 pm

I was talking about jack-type connector the standard in the industry of CC equipment (until someone decided that USB connectors could be used)
the d¡famous wikipedia has page for that:
https://en.wikipedia.org/wiki/DC_connector
clear enough?
Quoting from that Wikipedia reference:
There are a wide variety of sizes and designs for these power connectors, and many appear quite similar to each other yet are not quite mechanically or electrically compatible.
Yes, clear enough. I could not have put it more clearly.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 20775
Joined: Sat Jul 30, 2011 7:41 pm

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 1:28 pm

Pac0 wrote:
Tue Jun 12, 2018 1:00 pm
its not hard to find an USB power supply, that can deliver 5V +_ 5% while delivering 2.5A, there are plenty of those, it more or less becoming a standard.
Its also not difficult to get an USB to microUSB cable that doesn't drop more than 0.1 Volt while 2.5A runs through it, if you simply don't buy them from cheap Chinese supplier, who sell you the cheapest made junk possible because you won't notice if it takes ten minutes longer to charge your phone with it, as that is what these are typically used for. They use metaled plastic-foil inside instead of proper 22AWG or thicker copper wire. Or as Dougie Lawson here calls it they use "wet string". :mrgreen:
I'm reticent, but I agree your exposition of the facts...
Now I'm powering the pi with 5.3v enough to arrive 5v to pin 2. This agree with your comment.
I'm waiting for a best cable a braided one, but you know, today most of this cables (best ones or bad ones) are manufactured in china or similar.
The problem can solves it self using a span (5") of cable...
Another bit of kit on my desk is a bench power supply. I set that to 5.15v, and with a soldered up cable of about 50cm, everything works as it should. I can vary the voltage and I only get the lightning icon at about 4.7v on the power supply. So if you are having to go up to 5.3v to maintain a decent supply at the device, then the cable isn't very good at all.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Please direct all questions to the forum, I do not do support via PM.

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 3:56 pm

Pac0 wrote:
Tue Jun 12, 2018 11:50 am
why you don't power de raspberry by a standard connector.
Micro-USB has become the standard for 5V connections. The main problem is that cables are not always designed for the currents which the Pi and other high-current devices use. That's a cable issue, not an issue with the connector chosen.

I disagree about barrel connectors not being appropriate for use because there is an awful lot of consumer and professional equipment out there which continues to use those in high current applications which one would expect they would not do if they were not suitable, were as appalling as some suggest. Every monitor I have with an external power brick uses 12V, 2A or 3A, with a barrel connector and I have never had any problems.

But barrel connectors are far from being a single standard and it doesn't mean they would be appropriate for a Pi. While they might have been a preferred option for some before micro-USB became the standard that's not the case now.

In the case of 5V consumer goods anyway. There are different standards in other fields, notably for high current battery connectors. But power supplies and cables using those connectors are nowhere near as ubiquitous as USB cables.

User avatar
Imperf3kt
Posts: 1386
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: Voltage normalised (0x00000000)

Tue Jun 12, 2018 10:47 pm

Perhaps the next Pi should include pads we can solder to so as to completely avoid the micro USB port, while still retaining the polyfuse protection.
Stop plugging your fan directly into the GPIO 5v
https://www.electronics-tutorials.ws/power/transient-suppression.html

hippy
Posts: 3927
Joined: Fri Sep 09, 2011 10:34 pm
Location: UK

Re: Voltage normalised (0x00000000)

Wed Jun 13, 2018 9:43 am

Imperf3kt wrote:
Tue Jun 12, 2018 10:47 pm
Perhaps the next Pi should include pads we can solder to so as to completely avoid the micro USB port, while still retaining the polyfuse protection.
I am told it's insulting to the intelligence of the very bright folk behind the Pi to suggest such things though I agree with you that it could be desirable for some.

I would guess, seeing as I am sure that has been suggested in the past and obviously discounted because we don't have those, that it has been discounted because it's not something many people need, adds cost, would take real estate to provide holes to solder to, any such soldering invalidates the Pi's warranty, and, apart from lack of poly fuse, the GPIO power pins can be used, and a poly fuse can be added externally to the board.

Return to “Raspbian”