rPi 5 announced

Here:

best news of all is that the initial production will be for individuals, not corporations.

3 Likes

And the bad news is that it needs a special PSU.

I know it looks like a USB-C port, they even waffle about USB PD support… but that just means you can plug it into a fast charger without blowing it up. It cannot negotiate any high power PD modes, there is no onboard power regulator, it is 5v max.

The PSU for this is a proprietary one delivering 5v, 5A. Referred to as a ‘custom’ mode in their specs.

Big pass.

1 Like

priced only $5 more than the 4GB rPi4B+ so I have a feeling there will be lots of initial testers helping to evaluate the product before the corporations are allowed onboard.

BTW, it will run with a 5V,3A PS but with that it will limit USB output power to 600mA and it’ll throttle the CPU back if you want to get more performance.

1 Like

That’s two releases in a row where they have flubbed the power supply. I think the original rPi 4 was worse with their confusion about how resistors work; not sure how you make that mistake and still otherwise successfully pull off a working board.

Wish they’d start doing USB-C right.

2 Likes

To add to that… it has two micro-hdmi sockets… a better option would be to replace one of them with a usb c port supporting passthrough video. Even better would be that, and also using a full size hdmi socket…

1 Like

Oh dear… no no no no no

  • That looks like extortion
  • if you do not buy their ‘faux’ USB-C PD charger with its ‘custom mode’ they will cripple your machine.

I wonder if this also happens if you power it via the gpio header.

I expect third-party adapters with cables to become available. It’s one of those “ok it’s technically allowed in the standard, but good luck finding hardware right now” things, but I’m sure this will change. I don’t think it’s proprietary in any way, just an unusual design choice. I don’t see any indication that it’s faux; I just think it’s unusual.

Maybe a more flexible PMIC just ran them over their target cost, and they decided that 5V @ 3A for most uses, and 5A if you want to not use a powered USB hub, was a reasonable trade-off.

There’s some good conversation in this thread:

2 Likes

I think it’s more like if you don’t provide it with enough power, they limit the output power the USB ports will provide which is supposed to be enough for IO like keyboards/mice so that the board keeps running.

All this USB-C dance-around sounds like there are licensing issues they don’t want to pay or they really need a senior EE designer to clean up the mess.

1 Like

Well, not supporting 9V or 12V means cheaper PMIC and probably fewer passives — likely also fewer inductors I think. And you can see the space on the board dedicated to power management.

But I’m still annoyed that even a cheap phone can take 9V and somehow they didn’t manage to find room on the board or in the budget for more flexible PD options.

I don’t think there would be any additional licensing costs to that — do you have a source for that suggestion?

1 Like

The source is just a random thought that a more complex interface chip would be another license fee.
But from what I gathered, they are likely already paying the $5,000 /yr fee to joiin USB IF or and that seems enough. Also read an Ankler thread discussing why USB-C PD system were so much more expensive and it noted chip count to implement is a few $$ more because of the complexities of USB PD. The added dollars to implement on a ~$15 assembly(25% of $60) are significant at retail.

They are already using USB-C PD, is my understanding. It’s just that they are only set to negotiate 5V; their regulator stage isn’t set up to handle 9V, 12V, or 20V. Therefore, they need 5A for all the power needs, and the supply has to be willing to negotiate supplying 5V at 5A (few do, it seems, even though I believe it’s covered in the spec) and the cable has to be tagged for 5A.

So they really pushed the costs of handling many voltages on the board over to the added costs in the power supply which is pushing more current at 5V. AND if there are not many 5V/5A USB-C power supplies that move also drive sales up of their PSU as Owen pointed out.

1 Like

Which, for the vast majority of loads that don’t need to supply lots of downstream USB current, could reasonably be expected to result in a lower mean total cost of acquisition. Many existing power supplies will be completely adequate for I suspect almost everyone…

I don’t think it’s necessary to assume malicious intent, even if I disagree with either an engineering or business decision here. Unlike last time we don’t even need to assume a rather naïve mistake. I can disagree with the choice but recognize that the choice could be rational and not malicious.

Edit: I got a little lost in all the noise about it. From the page in the OP:

So it’s a custom design supporting eight power domains of at least three voltages (5V, 3.3V, and 1.8V) and probably more, and they still didn’t include the ability to handle higher voltage supply. :sob: