Testing 2021 new Official Micro USB Power Supply.

[RPIgoodSBC]

This is a test of the 2021 new Official Micro USB Power Supply using a consumption meter:

000mA ... 5.23V
300mA ... 5.19V
600mA ... 5.15V
900mA ... 5.10V
1.20A ... 5.06V
1.50A ... 5.02V
1.80A ... 4.97V
2.10A ... 4.93V
2.40A ... 4.89V
2.70A ... 4.84V
2.90A ... 4.81V
2.98A ... 4.80V (shutdown)

This is another excellent Power Supply with very good performance.

Product page and data:
https://www.raspberrypi.com/products/mi ... er-supply/
https://datasheets.raspberrypi.com/powe ... -brief.pdf

Hardware used:
https://es.aliexpress.com/item/33022448123.html

Related ... Testing official USB-C Power Supply rpi4 ...
viewtopic.php?f=63&t=243942

NOTE: Another test more exact is at page 2 of this thread.

[FTrevorGowen]

This is a test of the 2021 new Official Micro USB Power Supply using a consumption meter:

000mA ... 5.23V
300mA ... 5.19V
600mA ... 5.15V
900mA ... 5.10V
1.20A ... 5.06V
1.50A ... 5.02V
1.80A ... 4.97V
2.10A ... 4.93V
2.40A ... 4.89V
2.70A ... 4.84V
2.90A ... 4.81V
2.98A ... 4.80V (shutdown)

This is another excellent Power Supply with very good performance.

Product page and data:
https://www.raspberrypi.com/products/mi ... er-supply/
https://datasheets.raspberrypi.com/powe ... -brief.pdf

Hardware used:
https://es.aliexpress.com/item/33022448123.html

Related ... Testing official USB-C Power Supply rpi4 ...
viewtopic.php?f=63&t=243942

@RPIgoodSBC are you happy (again) for me to add your results to my webpages, with due credit & reference to this post. I've performed a comparison set of measurements to its predecessor, but, as yet, not updated my external webages.
Trev.

[ame]

Australia/New Zealand model still has that stupid circular flange on it that nobody has been able to explain.

[Gavinmc42]

Australia/New Zealand model still has that stupid circular flange on it that nobody has been able to explain.

Sidecutters

[ame]

Australia/New Zealand model still has that stupid circular flange on it that nobody has been able to explain.

Sidecutters

But no-one has stated what they're for. They aren't present on any other version, despite the fact that the body of the PSU is identical.

Oh, and the cable exits upwards when plugged in to an NZ outlet. It looks awful.

[pi-anazazi]

Who needs a power supply for a computer that you can't buy (for the comming months)?

[jamesh]

Who needs a power supply for a computer that you can't buy (for the comming months)?

We are on track to make just as many Pi's this year as last. I am sure you are aware of the supply chain issues worldwide - its not just us. If we could make more , don't you think we would?

I believe that STM32's are currently on a two year lead time?

[jamesh]

Australia/New Zealand model still has that stupid circular flange on it that nobody has been able to explain.

Sidecutters

But no-one has stated what they're for. They aren't present on any other version, despite the fact that the body of the PSU is identical.

Oh, and the cable exits upwards when plugged in to an NZ outlet. It looks awful.

ISTR this is due to the same PS being used in NZ/Aus and also some Pacific rim countries, and a few of those require the flanges for safety reasons.

[ame]

Sidecutters

But no-one has stated what they're for. They aren't present on any other version, despite the fact that the body of the PSU is identical.

Oh, and the cable exits upwards when plugged in to an NZ outlet. It looks awful.

ISTR this is due to the same PS being used in NZ/Aus and also some Pacific rim countries, and a few of those require the flanges for safety reasons.

Last time I asked, you said you didn't know. You said nobody knew. I even emailed the manufacturer. They didn't know.

The cable exits upwards because China. It's not actually designed for Australia/New Zealand, just that our power prongs are the same as China, but upside-down (because we're in the Southern Hemisphere).

[pi-anazazi]

Who needs a power supply for a computer that you can't buy (for the comming months)?

We are on track to make just as many Pi's this year as last. I am sure you are aware of the supply chain issues worldwide - its not just us. If we could make more , don't you think we would?

I believe that STM32's are currently on a two year lead time?

I know, not really the right thread for this, but I'm working hard for months now to understand, how come? You build a constant flow (OK, raspis it's more batches, I guess) of hardware (cars, computers, you name it) and have to have contracts for supplies. Why in the world is there all of a sudden a delay in the range of months or years in the supply? All needed for producing vaccines?

Some car manufacturers with long-term supply (hoarding?) seem to cope better with the problems.

From your perspective: Who/what failed here? I really want to understand what's going on here...

[rpdom]

I know, not really the right thread for this, but I'm working hard for months now to understand, how come? You build a constant flow (OK, raspis it's more batches, I guess) of hardware (cars, computers, you name it) and have to have contracts for supplies. Why in the world is there all of a sudden a delay in the range of months or years in the supply? All needed for producing vaccines?

Some car manufacturers with long-term supply (hoarding?) seem to cope better with the problems.

From your perspective: Who/what failed here? I really want to understand what's going on here...

Did you not follow the news for that last year or so?

This https://www.bbc.co.uk/news/technology-55936011 gives some insight an there are many others.

[jamesh]

Who needs a power supply for a computer that you can't buy (for the comming months)?

We are on track to make just as many Pi's this year as last. I am sure you are aware of the supply chain issues worldwide - its not just us. If we could make more , don't you think we would?

I believe that STM32's are currently on a two year lead time?

I know, not really the right thread for this, but I'm working hard for months now to understand, how come? You build a constant flow (OK, raspis it's more batches, I guess) of hardware (cars, computers, you name it) and have to have contracts for supplies. Why in the world is there all of a sudden a delay in the range of months or years in the supply? All needed for producing vaccines?

Some car manufacturers with long-term supply (hoarding?) seem to cope better with the problems.

From your perspective: Who/what failed here? I really want to understand what's going on here...

Our production is constant flow - we make 7M or so devices a year, the production line is never idle (well, maybe overnight). They might be made in batches of particular product, but the total flow is continuous.

The problem for the entire Si industry, is that overnight, lead times for parts went from Y weeks to Y * 2, or even bigger multipliers.

So for example (and these are made up numbers, and not at all Raspberry Pi specific - again, this is industry wide), lets say you have a part on a 10wk lead time, not uncommon, usually much longer. If, overnight and completely out of the blue, that lead time changes to 20wks, you have a gap in supply of 10wks. You can use stock to try and keep things going, but eventually you run out of that as well.

So, why did the lead times change so much? Don't know, but it is a combination of many factors. Covid is one of them, high demand is another. I don't think vaccine production has anything to do with it though.

[pi-anazazi]

We are on track to make just as many Pi's this year as last. I am sure you are aware of the supply chain issues worldwide - its not just us. If we could make more , don't you think we would?

I believe that STM32's are currently on a two year lead time?

I know, not really the right thread for this, but I'm working hard for months now to understand, how come? You build a constant flow (OK, raspis it's more batches, I guess) of hardware (cars, computers, you name it) and have to have contracts for supplies. Why in the world is there all of a sudden a delay in the range of months or years in the supply? All needed for producing vaccines?

Some car manufacturers with long-term supply (hoarding?) seem to cope better with the problems.

From your perspective: Who/what failed here? I really want to understand what's going on here...

Our production is constant flow - we make 7M or so devices a year, the production line is never idle (well, maybe overnight). They might be made in batches of particular product, but the total flow is continuous.

The problem for the entire Si industry, is that overnight, lead times for parts went from Y weeks to Y * 2, or even bigger multipliers.

So for example (and these are made up numbers, and not at all Raspberry Pi specific - again, this is industry wide), lets say you have a part on a 10wk lead time, not uncommon, usually much longer. If, overnight and completely out of the blue, that lead time changes to 20wks, you have a gap in supply of 10wks. You can use stock to try and keep things going, but eventually you run out of that as well.

So, why did the lead times change so much? Don't know, but it is a combination of many factors. Covid is one of them, high demand is another. I don't think vaccine production has anything to do with it though.

If it was my bussiness I would ask my supplier for GOOD reasons. Apparently they all have the same "problem", whatever it is. But I don't see any good reason. Besides chemicals supplies and plastic ware of high standards (needed for PCR testing and manufacturing of vaccines). I still don't understand what's going on in Taiwan and South Korea. Prices for air fright are nowadays in the same order of magnitude as container shipping, so no excuse...

[jdb]

From your perspective: Who/what failed here? I really want to understand what's going on here...

Remember when all the toilet roll ran out? Or more recently in the UK where there was a petrol shortage as soon as someone said "there's a petrol shortage"?

Exactly the same thing. Demand shocks in inelastic production chains cause shortages. Chip manufacture has a hard cap, and that cap is determined by the number of ten-billion-dollar fabs there are in the world.

[jamesh]

I know, not really the right thread for this, but I'm working hard for months now to understand, how come? You build a constant flow (OK, raspis it's more batches, I guess) of hardware (cars, computers, you name it) and have to have contracts for supplies. Why in the world is there all of a sudden a delay in the range of months or years in the supply? All needed for producing vaccines?

Some car manufacturers with long-term supply (hoarding?) seem to cope better with the problems.

From your perspective: Who/what failed here? I really want to understand what's going on here...

Our production is constant flow - we make 7M or so devices a year, the production line is never idle (well, maybe overnight). They might be made in batches of particular product, but the total flow is continuous.

The problem for the entire Si industry, is that overnight, lead times for parts went from Y weeks to Y * 2, or even bigger multipliers.

So for example (and these are made up numbers, and not at all Raspberry Pi specific - again, this is industry wide), lets say you have a part on a 10wk lead time, not uncommon, usually much longer. If, overnight and completely out of the blue, that lead time changes to 20wks, you have a gap in supply of 10wks. You can use stock to try and keep things going, but eventually you run out of that as well.

So, why did the lead times change so much? Don't know, but it is a combination of many factors. Covid is one of them, high demand is another. I don't think vaccine production has anything to do with it though.

If it was my bussiness I would ask my supplier for GOOD reasons. Apparently they all have the same "problem", whatever it is. But I don't see any good reason. Besides chemicals supplies and plastic ware of high standards (needed for PCR testing and manufacturing of vaccines). I still don't understand what's going on in Taiwan and South Korea. Prices for air fright are nowadays in the same order of magnitude as container shipping, so no excuse...

You can ask as much as you like. When you have a supplier of something, if they cannot get it to you, the reason are irrelevant even thought I am sure there are very good reasons. If they are a single supplier (and most silicon vendors are), then customers have no option but to wait. Redesign around a different component could work, as long as you can get that other component, but it's not cheap and again, takes time.

[FTrevorGowen]

Getting back on the O.P.'s topic - here's a plot of @RPIgoodSBC's data together with mine:

OfficialRasPi_uUSB_PSU2021.jpg (153.72 KiB) Viewed 1729 times

(2018 comparison is here: http://www.cpmspectrepi.uk/raspberry_pi ... 28Dec.2018 .29 )
Trev.

[pi-anazazi]

From your perspective: Who/what failed here? I really want to understand what's going on here...

Remember when all the toilet roll ran out? Or more recently in the UK where there was a petrol shortage as soon as someone said "there's a petrol shortage"?

Exactly the same thing. Demand shocks in inelastic production chains cause shortages. Chip manufacture has a hard cap, and that cap is determined by the number of ten-billion-dollar fabs there are in the world.

Toilet paper went somewhere on stock (private households). Who has the stock of chips in your example?

Petrol: there were not enough drivers to drive all the trucks due to covid quarantine and some other bold, mildly clever political moves. All explainable. But where do all the chips go? The fabs are still there and producing, afaik...

[RPIgoodSBC]

This is a test of the 2021 new Official Micro USB Power Supply using a consumption meter:

000mA ... 5.23V
300mA ... 5.19V
600mA ... 5.15V
900mA ... 5.10V
1.20A ... 5.06V
1.50A ... 5.02V
1.80A ... 4.97V
2.10A ... 4.93V
2.40A ... 4.89V
2.70A ... 4.84V
2.90A ... 4.81V
2.98A ... 4.80V (shutdown)

This is another excellent Power Supply with very good performance.

Product page and data:
https://www.raspberrypi.com/products/mi ... er-supply/
https://datasheets.raspberrypi.com/powe ... -brief.pdf

Hardware used:
https://es.aliexpress.com/item/33022448123.html

Related ... Testing official USB-C Power Supply rpi4 ...
viewtopic.php?f=63&t=243942

@RPIgoodSBC are you happy (again) for me to add your results to my webpages, with due credit & reference to this post. I've performed a comparison set of measurements to its predecessor, but, as yet, not updated my external webages.
Trev.

Yes of course. You can use this data freely as you like.
A greeting.

[bjtheone]

Our production is constant flow - we make 7M or so devices a year, the production line is never idle (well, maybe overnight). They might be made in batches of particular product, but the total flow is continuous.

The problem for the entire Si industry, is that overnight, lead times for parts went from Y weeks to Y * 2, or even bigger multipliers.

So for example (and these are made up numbers, and not at all Raspberry Pi specific - again, this is industry wide), lets say you have a part on a 10wk lead time, not uncommon, usually much longer. If, overnight and completely out of the blue, that lead time changes to 20wks, you have a gap in supply of 10wks. You can use stock to try and keep things going, but eventually you run out of that as well.

So, why did the lead times change so much? Don't know, but it is a combination of many factors. Covid is one of them, high demand is another. I don't think vaccine production has anything to do with it though.

If it was my bussiness I would ask my supplier for GOOD reasons. Apparently they all have the same "problem", whatever it is. But I don't see any good reason. Besides chemicals supplies and plastic ware of high standards (needed for PCR testing and manufacturing of vaccines). I still don't understand what's going on in Taiwan and South Korea. Prices for air fright are nowadays in the same order of magnitude as container shipping, so no excuse...

You can ask as much as you like. When you have a supplier of something, if they cannot get it to you, the reason are irrelevant even thought I am sure there are very good reasons. If they are a single supplier (and most silicon vendors are), then customers have no option but to wait. Redesign around a different component could work, as long as you can get that other component, but it's not cheap and again, takes time.

With regard to chip suppliers, fabs were impacted by Covid (slowdowns and shutdowns). First you need to understand that fab time is typically sold/scheduled 12-18 months (or longer) out and they are trying to sell 100% of capacity. This means any major event produces massive ripples. On top of this, there was two significant reactions from buyers; 1) to try and increase buys to protect their supply chain, and 2) offer higher prices to try and buy other folks fab time. Some fabs are reacting to market demands and moving production to higher margin devices. It is ugly out there.

There is also the question of how does the fab deal with the lost production time. Do they ripple and impact everyone downstream or do they just bail on the one scheduled run, and try and go back to delivering to contract, or a combination of both. Really sucks if your stuff was in the production run impacted.

I would say Y*4 right now.

[cjketle]

It is possible your measurements of voltage droop are a bit on the pessimistic side - to use the charge measuring device identified you presumably used extra leads and connectors -something like

RPi PSU -> usb-c -> extra lead(?) -> USB-A-> Voltage Measurement Device -> USB-A -> load

IME USB-A connectors are often not happy with high current loads and might (with any extra lead) contribute to the measured PSU droop at the higher currents. The odd 0.1-0.2V drop is critical which at 2A means less than 0.1 ohm might be significant - and would not be present when used to power the Pi directly.

Chris K

[FTrevorGowen]

It is possible your measurements of voltage droop are a bit on the pessimistic side - to use the charge measuring device identified you presumably used extra leads and connectors -something like

RPi PSU -> usb-c -> extra lead(?) -> USB-A-> Voltage Measurement Device -> USB-A -> load

IME USB-A connectors are often not happy with high current loads and might (with any extra lead) contribute to the measured PSU droop at the higher currents. The odd 0.1-0.2V drop is critical which at 2A means less than 0.1 ohm might be significant - and would not be present when used to power the Pi directly.

Chris K

Have you looked at my tests over many years (since the "first Pi") - @RPIgoodSBC's setup for a previous test is there too (and the link to a previous, similar thread): http://www.cpmspectrepi.uk/raspberry_pi ... plies.html
In general the main "source" of "voltage droop" is the length of any connecting cable the effective resistance of which can be determined from the slope and is > than any connector resistance. The real problem with any "easily unplugged" connectors carrying significant current is the damage caused by "micro-arcs" when they are unplugged. The official power supplies have used leads with much lower internal resistance (ie. thicker wires) than, historically, many "charging leads" supplied for use with 'phone chargers - a charger does not need to supply >4.65 V at its maximum current, it just needs to be > than the battery's (back) voltage its charging. See http://www.cpmspectrepi.uk/raspberry_pi ... .2FCharger

Trev.

[cjketle]

Agreed, the correct way to do it is to eliminate any extra connectors and lead lengths that could skew the results. The pointer to the Alix measurement device shows a strictly USB-A to USB-A tool though (see pic) unlike the original which showed a device with a USB-C input socket as well. This would need extra connectors/adapters for the USB-C input from the RPi PSU that could be a source of error.

Chris K

[RPIgoodSBC]

Micro USB connector from Raspberry Pi power supply to ---> micro USB conector load HD35 and AT35 connected to the male usb HD35 from the female connector.

[FTrevorGowen]

Agreed, the correct way to do it is to eliminate any extra connectors and lead lengths that could skew the results. The pointer to the Alix measurement device shows a strictly USB-A to USB-A tool though (see pic) unlike the original which showed a device with a USB-C input socket as well. This would need extra connectors/adapters for the USB-C input from the RPi PSU that could be a source of error.

Chris K

In theory yes. In practice the main error is that such devices (I've used several types, see below**) is that they are uncalibrated. Hence the effective cable (+ any other) resistance is calculated from the slope and the nominal current available at 4.65V (the low voltage warning trigger threshold) determined from that. The slope resistance will, to some extent, take account of the "sense resistor" if there is one. A Hall effect sensor could be used but, again need to be calibrated. In practice any current drawn by a Pi will be dynamic and hence there is also a "time constant" associated with the trigger warning. Ideally one should also use an oscilloscope to observe the supplies dynamic behaviour. Unfortunately having retired from >40yrs of R&D, just before the "first pi", I no longer have access to such and hence used basic, static, methods****.

Trev.

** see http://www.cpmspectrepi.uk/raspberry_pi ... octor.html
http://www.cpmspectrepi.uk/raspberry_pi ... ector.html
http://www.cpmspectrepi.uk/raspberry_pi ... istor.html
and http://www.cpmspectrepi.uk/raspberry_pi ... sters.html

**** see http://www.cpmspectrepi.uk/raspberry_pi ... t.2FMethod
and http://www.cpmspectrepi.uk/raspberry_pi ... Test_Board

[cjketle]

If you are using the HD35, and not the AT35 (other than as a voltmeter) pointed to by the top post, then yes, my comments do not apply. I think we are all agreeing - it is a good quality power supply

Chris K