I guess you've seen that the "dust havn't settled yet". 😉
Heat issues, gpu issues, audio OSes fighting to catch up....
Nothing that wasn't expected. 😉
I'll stick to my plan not to jump on the train before September/October.
Enjoy.
Heat issues, gpu issues, audio OSes fighting to catch up....
Nothing that wasn't expected. 😉
I'll stick to my plan not to jump on the train before September/October.
Enjoy.
I don't know, I've had mine for a week now running MPD with Archphile's GUI (on Raspbian) with no issues. It's running just fine as an audio transport, feeding my USB / XMOS dac.
But I have not had the chance yet to take it out of the workshop and up to the living room to make a serious SQ comparison with the RPi 3.
But I have not had the chance yet to take it out of the workshop and up to the living room to make a serious SQ comparison with the RPi 3.
Of course it'll work on common tasks. Especially if you're able to configure a system yourself.
pCP has e.g. just launched a 6.0 Beta with Pi4 support and that after releasing 5.0 just a couple of weeks ago. They simply didn't see it coming.
pCP has e.g. just launched a 6.0 Beta with Pi4 support and that after releasing 5.0 just a couple of weeks ago. They simply didn't see it coming.
A lot of USB C cables also won't work well with the raspberry Pi 4 right now: Pi4 not working with some chargers (or why you need two cc resistors) – The blog of Tyler Ward (aka scorpia)
I'll jump in when they're on the next hardware revision. Thanks to my phone I only have the e-marked style cables.
I'll jump in when they're on the next hardware revision. Thanks to my phone I only have the e-marked style cables.
It's obviously not a cable or power supply issue.
The USB-C issue is a real PI4 HW design issue.
The USB-C issue is a real PI4 HW design issue.
Yes, it is, but I don't think that such an issue should discourage any one of us.
It's not like we'll be using any plain old SMPS and usb cable to power it.. 😉
It's not like we'll be using any plain old SMPS and usb cable to power it.. 😉
Pi4 works with a Pi3 official supply and a usb-micro to usb-C adapter provided the total draw from the usb outlets is less than 500mA.
If more then use a hub....or two...one for usb3 and one for usb2...
Cheaper still...buy the official Pi4 supply 😉 or build your own...
If more then use a hub....or two...one for usb3 and one for usb2...
Cheaper still...buy the official Pi4 supply 😉 or build your own...
The real design "flaw" of the Pi 4 seems to be the lack of heat management and an SOC that runs way too hot.
This is a pretty stupid move by the Raspberry Pi developers. The one thing about the Pi that separates its form factor from a mini PC is the ability to install "HATs" via the GPIO pins. The HAT sits just above the board where the SOC is located. This means that a heat sink will occupy the space where HATs are isntalled, but without one the SOC heats up so badly that it can reach 80C EVEN WITH THE CPU AT IDLE if the Pi is in any kind of case! At least that was what I took away from this temperature test series:
Cooling options for the hot Raspberry Pi 4 - Martin's Explorations
See also:
Raspberry Pi 4 Hot new release – Too hot to use enclosed - Martin's Explorations
So... now we have a Pi that is faster on paper (still not fast by any means) but if you try to make use of that speed the thermal management will just slow down the CPU. What good is that? If you add a fan and/or large heatsink to improve thermal dissipation, HATs are blocked. And you have the issue of noise from a tiny fan, not a selling point for audio applications. This is a real problem.
One workaround would be a board that attaches to the backside and that reproduces the GPIO and other headers, so that "HATs" could then be stuck on the bottom side, leaving room for a fat heatsink on the top. Maybe then HATs will become knows as A S S es...
This is a pretty stupid move by the Raspberry Pi developers. The one thing about the Pi that separates its form factor from a mini PC is the ability to install "HATs" via the GPIO pins. The HAT sits just above the board where the SOC is located. This means that a heat sink will occupy the space where HATs are isntalled, but without one the SOC heats up so badly that it can reach 80C EVEN WITH THE CPU AT IDLE if the Pi is in any kind of case! At least that was what I took away from this temperature test series:
Cooling options for the hot Raspberry Pi 4 - Martin's Explorations
See also:
Raspberry Pi 4 Hot new release – Too hot to use enclosed - Martin's Explorations
So... now we have a Pi that is faster on paper (still not fast by any means) but if you try to make use of that speed the thermal management will just slow down the CPU. What good is that? If you add a fan and/or large heatsink to improve thermal dissipation, HATs are blocked. And you have the issue of noise from a tiny fan, not a selling point for audio applications. This is a real problem.
One workaround would be a board that attaches to the backside and that reproduces the GPIO and other headers, so that "HATs" could then be stuck on the bottom side, leaving room for a fat heatsink on the top. Maybe then HATs will become knows as A S S es...
Last edited:
From what I understand,
The Pi4 heat issues are only if it is being taxed
Other than that, it can perform the same, heat wise, as the Pi3
The eliminated bottlenecks alone are a selling point to myself.
The Pi4 heat issues are only if it is being taxed
Other than that, it can perform the same, heat wise, as the Pi3
The eliminated bottlenecks alone are a selling point to myself.
The tests I quoted were with the Pi IDLE in a case with no heatsink or fan - it reached 77C! That is with it doing NOTHING at all.
There are certainly many improvements to I/O and speed, but the platform is just not as "usable" in general because of the heat issue.
It only reduced temps by about 3C overall according to what I read.
The TomsHardware article is with the board in free air. If you use a heatsink, that is still a viable option, but again the HATs are then blocked...
And they conclude the article with:
Pimoroni makes a fan/heatsink and passive heatsink
review and measurements here:
Raspberry Pi 4 Cooling Review: Pimoroni Heatsink and Fan Shim Tested - Tom's Hardware
On ebay, I found this heatsink, which is larger so should be better than the pimoroni passive one
New CNC Aluminum Radiator Protective Case Enclosure Heat sink for Raspberry Pi 4 | eBay
I wonder how loud the fan is on the active heatsink, that could be the deal breaker here.
I'm sure companies will come out with more heatsinking options, looks like you will need something with a pi4.
Randy
review and measurements here:
Raspberry Pi 4 Cooling Review: Pimoroni Heatsink and Fan Shim Tested - Tom's Hardware
On ebay, I found this heatsink, which is larger so should be better than the pimoroni passive one
New CNC Aluminum Radiator Protective Case Enclosure Heat sink for Raspberry Pi 4 | eBay
I wonder how loud the fan is on the active heatsink, that could be the deal breaker here.
I'm sure companies will come out with more heatsinking options, looks like you will need something with a pi4.
Randy
There are also other conclusions. 🙄
Such as:
3° can make a hell of a difference.
If this device now works properly under ""normal"" load conditions, I probably could live with that.
How it'll look with a HAT attached - we'll see.
IMO similar situations we've seen on earlier RPIs.
Last edited:
The 3+ had plenty of thermal issues as well, mine would throttle itself back or simply lock up until I got a fan. I designed my hat to hang off on the side rather than over the pi.
My RPi4B's have been running piCorePlayer for over a week. They are "open" and temperature is not an issue, being a constant mid to high 50's. The player has a DAC hat. LMS has a USB drive. This is 20 degrees higher than previous RPi's.
Using my basic configuration, CPU usage is negatable, squeezelite maxes out at 0.4% and LMS maxes out at 0.5%.
I wouldn't stick one in an enclosure without adequate ventilation!
Using my basic configuration, CPU usage is negatable, squeezelite maxes out at 0.4% and LMS maxes out at 0.5%.
I wouldn't stick one in an enclosure without adequate ventilation!
Ad hats - why not using some pin header extender ( e.g. Stacking Header for Pi A+/B+/Pi 2/Pi 3 - 2x20 Extra Tall Header ID: 1979 - $2.95 : Adafruit Industries, Unique & fun DIY electronics and kits ) + rotating the RPI by 90deg, so that the warm air from the heatsink fins can rise?
Last edited: