Raspberry Pi

[Captain Video]

good to know

only use an official Raspberry Pi 4 power cable, check

[The Latinist]

Is that short something you can break? The

No, they've combined the two leads on the PCB and run them into the same 5.1kΩ resistor. I don't see how you could fix that.

More details: https://medium.com/@leung.benson/how-to-design-a-proper-usb-c-power-sink-hint-not-the-way-raspberry-pi-4-did-it-f470d7a5910

Instead of trying to come up with some clever circuit, hardware designers should simply copy the figure from the USB-C Spec exactly. The Figure 4–9 I posted above isn’t simply a rough guideline of one way of making a USB-C receptacle. It’s actually normative, meaning mandatory, required by the spec in order to call your system a compliant USB-C power sink. Just copy it.

good to know

only use an official Raspberry Pi 4 power cable, check

You should be able to use most 3A phone chargers, which are usually not e-marked.  But, yes, the official charger is guaranteed to work.

[Captain Video]

I just checked and Microcenter finally has a bunch of the 1gb units in stock today. If they are still there tomorrow I can pick one up.  I wish they had some 4s.

[brilligtove]

Is that short something you can break? The

No, they've combined the two leads on the PCB and run them into the same 5.1kΩ resistor. I don't see how you could fix that.

More details: https://medium.com/@leung.benson/how-to-design-a-proper-usb-c-power-sink-hint-not-the-way-raspberry-pi-4-did-it-f470d7a5910

Instead of trying to come up with some clever circuit, hardware designers should simply copy the figure from the USB-C Spec exactly. The Figure 4–9 I posted above isn’t simply a rough guideline of one way of making a USB-C receptacle. It’s actually normative, meaning mandatory, required by the spec in order to call your system a compliant USB-C power sink. Just copy it.

good to know

only use an official Raspberry Pi 4 power cable, check

You should be able to use most 3A phone chargers, which are usually not e-marked.  But, yes, the official charger is guaranteed to work.

Ah. Not a blob of solder at the pins, then.

[The Latinist]

Yeah, the traces are printed that way.

It appears they were just trying to save money by eliminating  what they thought was a superfluous resistor. To be fair, part of their core mission is making the board affordable; but I heard the component only costs about 1/20¢, which means they’d save about $500 on a run of 1 million boards. Now they’re looking at bad PR and the cost of a redesign.  Penny wise and pound foolish, as they say.

I’m actually a bit more concerned about the heat issues; the new processor seems to be generating a lot of heat (and drawing significantly more power) under load. I don’t mind putting a heat sink on one, but I’m not willing to put a fan on it.  Perhaps for some use cases it could be under-clocked.

[BAWRFRS]

I’m actually a bit more concerned about the heat issues; the new processor seems to be generating a lot of heat (and drawing significantly more power) under load. I don’t mind putting a heat sink on one, but I’m not willing to put a fan on it.  Perhaps for some use cases it could be under-clocked.

I share your concern. I too would prefer to just add a heat sink, as I did for my 3B model (it came with two, in a Canakit, uh, kit). That said, I don't run my 3B with the top on its case. I keep it open during use.

I'm content to wait out the re-emergence of USB Boot and any other stuff that isn't quite operational out of the box. But the heat thing is not (much) solved by software/firmware updates. There is one that'll give you 3-5C reduction but if you're already at 81C, that's not quite enough to avoid throttling, I think!

FWIW, based on my reading of the many experiments people have done with cooling their Pi 4, it appears that a fan is the most efficacious solution, by far. A (typical) heat sink buys you some time, but does not solve the overheating (to the point of throttling) issue, especially in an enclosed case. You have to carry that heat away somehow, and a fan is the most reliable way to do it.

Other suggestions have been to simply not use a case, to have it exposed to the air, preferably on edge so that top and bottom are both exposed, etc. But that's a most inelegant solution, and makes me feel that at some point, they're chasing desktop performance that requires desktop-style cooling. Now, why can a smartphone or tablet be all enclosed and yet not overheat? IDK. I'm pretty sure smartphones aren't cooled by fans. Is the whole smartphone enclosure a heat sink? Is a heat sink that large sufficient, as opposed to those tiny ones Canakit provided me?

My (aging, thick) laptop has some tiny fans inside and they certainly blow out some heat. IDK how today's super-thin laptops vent heat; I suspect the case is a giant heat sink. But I can't help but wonder if the RPF is at a design crossroads. Something will have to give ... price or speed. They shouldn't need such speed to teach programming, but to the extend that any of that programming has to interface with a browser, modern web browsing is absurdly resource intensive. And with that speed/power comes the need to vent heat: fans or heat sink, liquid cooling, etc.

[mindme]

Building out my own retropie bartop.

[BAWRFRS]

Chris Barnatt at Explaining Computers with another great video ... in this one, he comes down squarely on the necessity of active cooling for the Pi.



A couple of caveats:

It seems to me that his tests demonstrate the necessity of active cooling to keep temperatures well below throttling level. However, it does not appear he demonstrated the necessity of active cooling as far as performance goes. Even a small heatsink, coupled with no Pi 4 case, ran the prime number task about as fast as with an actively cooling fan. It just did so at a much hotter (82C) temperature than the Pi did with a fan + heatsink (55C).

This leads to another issue - he didn't test the performance under the simple condition of no heatsink and no case. Might have it done the task as quickly, and at around 82C, simply by removing the case?

That said, the ability to run at lower temps could reasonably imply longer service life for the Pi.

For me, I'd not likely run the Pi 4 without some sort of enclosure. It would scratch up my wooden desktop surfaces and possibly get too hot to touch, unless I could figure out a way to mount it cleverly. I suspect a well-vented, fan-enabled case will prove a necessity.

[Captain Video]

Yea if i have to put fans on them some sound guy will see it and then they will suddenly hear it (inside joke) No fans.  Im thinking a big heat sink and a case designed to allow airflow or a metal case with a heatsink built in. I will already need to design something to hold two of those mini HDMI plugs in place with strain relief. 

I still haven found time to drop by microcenter. 

[The Latinist]

My takeaway from his tests was that a well-ventilated case with a heat sink should be sufficient. Although he was able to get it significantly cooler with a fan, it was not significantly faster than with the heat sink alone, indicating that most of the throttling gains come from keeping it under about 82°C.  If you look at his numbers, his video encoding test took 41 minutes @90°C with no heat sink in the official case, 11 minutes @81°C with a tiny heat sink and the top off of the case, and 10 minutes @56°C with a huge heat sink, fan and custom case.

[The Latinist]

Now that I have a 4K TV, I kind of want to get a RPi 4 to try it out on.

[BAWRFRS]

Now that I have a 4K TV, I kind of want to get a RPi 4 to try it out on.

You probably know this, but just in case (or for other readers) -

I suspect there's a difference between being able to play a 4k video file in a given format from local storage (e.g., via USB 3 flash or SSD), and being able to watch a video from Youtube or some other streaming service in 4k. Not just from potential speed difference between internet connection and flash memory read, but also things like hardware acceleration, encoding, etc etc. A similar reasoning, I think, with why the Pi 3 model(s) can play 1080p video, but only via an OS like LibreElec, not via a browser window in Raspbian.

That said, I wish you success, and look forward to hearing how it turns out.

[BilLumberg]

Update on the raspberry pi NAS. The pi and two external drives (1TB, and 5TB) only pull 10 watts at idle, and 20 when streaming from both drives simultaneously. Big drop from the 300 watts at idle it replaced. I was overthinking the openmediavault login. All that is needed is the IP address and then everything is done through the browser interface. I initially thought it needed to be logged in directly to start it. So far everything has worked flawlessly. Which has been my experience with most linux based systems. They just work. The best part is I was able to do this with hardware I already had. Of course most thanks go to the people who put in the work developing the openmediavault software.

Now I think I will see how well a pi-hole works for network ad blocking. After all, I have a extra micro sd card just sitting there doing nothing.

[The Latinist]

I’ve seen really good reviews of the Flirc case. The case is aluminum and the whole thing acts as a heat sink. It gives access to all of the connectors with the exception of the GPIO (though it does provide room for a ribbon cable).  Tests I saw indicated it can keep the maxed-out CPU temp at 60°F with no fan.

The only draw-back seems to be reduced Wi-Fi and Bluetooth range (for obvious reasons).