[the_bo]

Is there a flirc type case that is all black?

I'm a sucker for all my tech being black, blends in better under the TV

[speedwell68]

Is there a flirc type case that is all black?

I'm a sucker for all my tech being black, blends in better under the TV

 Being made of aluminium it would be real easy to make it all black.  Flat it off with some really fine grade wet n dry paper to get a good key for the paint.  Then clean it with some meths.  Then use a direct to metal paint like Plastikote.

[rob77]

I've just bought the 3b+ today and so far so good running libreelec. The only issue I have is with the wireless, it seems really poor. Max I can get out of it is 23Mbps although it is connected to my ubiquiti access point on 5Ghz at 150Mbps.

I had a tplink 2.4ghz usb dongle attached to my pi 2b and that would achieve 30Mb+

[Milhouse]

I've just bought the 3b+ today and so far so good running libreelec. The only issue I have is with the wireless, it seems really poor. Max I can get out of it is 23Mbps although it is connected to my ubiquiti access point on 5Ghz at 150Mbps.

I had a tplink 2.4ghz usb dongle attached to my pi 2b and that would achieve 30Mb+

Hmm odd. The 3+ WiFi performance as reported on the blog was done by me with a 3+ running LibreELEC, connecting to a Netgear DGND4000 access point which supports 11n on 2.4GHz (up to 300Mbps, according to Netgear) and 11a/n on 5GHz (up to 450Mbps, again according to Netgear) - product sheet (pdf).

The AP is in a hallway closet while the uncased 3+ is in an adjoining room, so there's a distance of about 20-25 feet between them, and the signal has to pass through two walls (partitions). Testing TCP with iperf3 I get a solid 100Mbps on 5GHz and about 50Mbps on 2.4GHz.

It's difficult to know why your 3+ and Ubiquiti combination is performing below expectations - maybe the performance of your Ubiquiti compared with my Netgear AP has something to do with it (although my Netgear is several years old, so I'd expect at least as good performance from a newer device, particularly a Ubiquiti AP), or it could be your WiFi environment (noisy neighbours?)

[noggin]

I've just bought the 3b+ today and so far so good running libreelec. The only issue I have is with the wireless, it seems really poor. Max I can get out of it is 23Mbps although it is connected to my ubiquiti access point on 5Ghz at 150Mbps.

I had a tplink 2.4ghz usb dongle attached to my pi 2b and that would achieve 30Mb+

Hmm odd. The 3+ WiFi performance as reported on the blog was done by me with a 3+ running LibreELEC, connecting to a Netgear DGND4000 access point which supports 11n on 2.4GHz (up to 300Mbps, according to Netgear) and 11a/n on 5GHz (up to 450Mbps, again according to Netgear) - product sheet (pdf).

The AP is in a hallway closet while the uncased 3+ is in an adjoining room, so there's a distance of about 20-25 feet between them, and the signal has to pass through two walls (partitions). Testing TCP with iperf3 I get a solid 100Mbps on 5GHz and about 50Mbps on 2.4GHz.

It's difficult to know why your 3+ and Ubiquiti combination is performing below expectations - maybe the performance of your Ubiquiti compared with my Netgear AP has something to do with it (although my Netgear is several years old, so I'd expect at least as good performance from a newer device, particularly a Ubiquiti AP), or it could be your WiFi environment (noisy neighbours?) 

 I wonder if the OP has his Pi 3B+ in a metal case like a Flirc? (Might that make a difference?)

[rob77]

Thanks guys

I don't know why im seeing such low speeds. The pi is in the official red and white plastic case and attached to the back of the tv in the bedroom. The access point is only in the front room below. It's showing as me having excellent signal strength in both the libreelec settings and the ubiquiti app.

My Galaxy s8 phone for example can easily achieve 200Mb download in the same area of the house as the pi.

[Milhouse]

 I wonder if the OP has his Pi 3B+ in a metal case like a Flirc? (Might that make a difference?)

This is the same iperf3 test from a FLIRC encased 3+ (running LibreELEC 9.0 build #0318) to DGND4000 on 5GHz channel 44 (signal strength is shown as 71 in LE):

text:
rpi22:~ # iperf -c 192.168.0.9 -t 60 -i10

Connecting to host 192.168.0.9, port 5201

[  4] local 192.168.0.52 port 41498 connected to 192.168.0.9 port 5201

[ ID] Interval           Transfer     Bandwidth       Retr  Cwnd

[  4]   0.00-10.00  sec   105 MBytes  88.5 Mbits/sec    0    380 KBytes

[  4]  10.00-20.00  sec   109 MBytes  91.5 Mbits/sec    0    595 KBytes

[  4]  20.00-30.00  sec   109 MBytes  91.6 Mbits/sec    0    897 KBytes

[  4]  30.00-40.00  sec   110 MBytes  92.0 Mbits/sec    0    897 KBytes

[  4]  40.00-50.00  sec   109 MBytes  91.1 Mbits/sec    0    897 KBytes

[  4]  50.00-60.00  sec   109 MBytes  91.2 Mbits/sec    0    897 KBytes

- - - - - - - - - - - - - - - - - - - - - - - - -

[ ID] Interval           Transfer     Bandwidth       Retr

[  4]   0.00-60.00  sec   651 MBytes  91.0 Mbits/sec    0             sender

[  4]   0.00-60.00  sec   649 MBytes  90.7 Mbits/sec                  receiver



iperf Done.

So a slight 5-10Mbps reduction in performance compared with my original testing, most likely due to surrounding the Pi3+ in a metal can.

[Milhouse]

Just for comparison, if only to demonstrate the bandwidth available from this AP over 5GHz, here's an Intel NUC6i5SYH (Skylake NUC) sitting about 8 inches away from the Pi3+, with AC 8260 built-in WiFi, running LE9 #0318 on 5GHz to the same Netgear AP:

text:
NUC:~ # iperf -c 192.168.0.9 -t 60 -i10

Connecting to host 192.168.0.9, port 5201

[  4] local 192.168.0.20 port 51634 connected to 192.168.0.9 port 5201

[ ID] Interval           Transfer     Bandwidth       Retr  Cwnd

[  4]   0.00-10.00  sec   215 MBytes   181 Mbits/sec    0    892 KBytes

[  4]  10.00-20.00  sec   232 MBytes   195 Mbits/sec    0   1.86 MBytes

[  4]  20.00-30.00  sec   166 MBytes   139 Mbits/sec    0   2.83 MBytes

[  4]  30.00-40.00  sec   231 MBytes   194 Mbits/sec    0   2.83 MBytes

[  4]  40.00-50.00  sec   226 MBytes   190 Mbits/sec    0   2.83 MBytes

[  4]  50.00-60.00  sec   179 MBytes   150 Mbits/sec    0   2.83 MBytes

- - - - - - - - - - - - - - - - - - - - - - - - -

[ ID] Interval           Transfer     Bandwidth       Retr

[  4]   0.00-60.00  sec  1.22 GBytes   175 Mbits/sec    0             sender

[  4]   0.00-60.00  sec  1.22 GBytes   174 Mbits/sec                  receiver



iperf Done.

So my Pi3+ is not limited by the AP, and it's perhaps unlikely that a Pi3+ will achieve better than 100Mbps on a 5GHz/11n WiFi connection. Not sure if 11ac connections will offer any improvement, as my AP is not 11ac compatible.

Still no idea what is wrong with the Ubiquiti results. Maybe the Ubiquiti is using 11ac in which case try disabling 11ac (if that is possible) and use only 5GHz 11n - I've never tested an 11ac connection.

[rob77]

I've found out what the issue is. I have hdmi over Ethernet running to the tv upstairs from the freeview box downstairs. When it is turned on and playing a channel my download speed drops on the pi. Turning it off and it maxes out my internet connection of around 55Mb. So it's some interference from the hdmi receiver.

Edit: I have just checked all the wiring at the back of the tv and the pi is running at full wifi speed even with the box downstairs on now. Weird! Can only assume that it was one of the cables not seated correctly and was causing interference. I'm much happier now it's all working correctly woohoo!

[Vimes]

For those who have the new 3+ Pi and one of the Flirc cases does it fit ok with the older thermal pad being a little thicker than desirable.?
I'm not sure if that will have any implications on its cooling capability.
Generally is there a good improvement in cooling using those cases over a std plastic type.? My Pi3 summertime gets toasty when playing back 720p hevc encoded material, as expected. I have noted Milhouse quote 15 to 20c, which would be amazing if I could expect that. 
I'm not sure if retailers like Amazon who sell those cases will differentiate them when Flirc start sending them out with the two pads, as noted.

[speedwell68]

For those who have the new 3+ Pi and one of the Flirc cases does it fit ok with the older thermal pad being a little thicker than desirable.?
I'm not sure if that will have any implications on its cooling capability.
Generally is there a good improvement in cooling using those cases over a std plastic type.? My Pi3 summertime gets toasty when playing back 720p hevc encoded material, as expected. I have noted Milhouse quote 15 to 20c, which would be amazing if I could expect that. 
I'm not sure if retailers like Amazon who sell those cases will differentiate them when Flirc start sending them out with the two pads, as noted.

 Not got mine yet.

[Milhouse]

For those who have the new 3+ Pi and one of the Flirc cases does it fit ok with the older thermal pad being a little thicker than desirable.?

I had no issues installing a Pi3+ into a FLIRC gen1 case with the thicker thermal pad. Mind you, I had been using it with a Pi3 so the pad was a little squashed already. I just went easy on the screws when putting the case back together and didn't over-tighten them.

This original FLIRC case that I have also isn't optimal for the Pi3/Pi3+ as the heatsink lug isn't entirely centred over the SOC - the original case is designed for the Pi2 and the Pi2 SOC is in a slightly different position to the Pi3, but the latest "gen 2" FLIRC case now fully covers the Pi3 SOC (and presumably Pi3+) so case temps might be even better than what I observed.

[Vimes]

For those who have the new 3+ Pi and one of the Flirc cases does it fit ok with the older thermal pad being a little thicker than desirable.?

I had no issues installing a Pi3+ into a FLIRC gen1 case with the thicker thermal pad. Mind you, I had been using it with a Pi3 so the pad was a little squashed already. I just went easy on the screws when putting the case back together and didn't over-tighten them.

This original FLIRC case that I have also isn't optimal for the Pi3/Pi3+ as the heatsink lug isn't entirely centred over the SOC - the original case is designed for the Pi2 and the Pi2 SOC is in a slightly different position to the Pi3, but the latest "gen 2" FLIRC case now fully covers the Pi3 SOC (and presumably Pi3+) so case temps might be even better than what I observed.

That reads to be quite impressive.
As it seems the only difference is going to be an inclusion of a thinner thermal pad I can imagine that the older type can be flattened somewhat, or at least care taken on tightening the screws on the case as you note.

Maybe thermal pads of the correct thickness could be bought from ebay, if required.

[speedwell68]

For those who have the new 3+ Pi and one of the Flirc cases does it fit ok with the older thermal pad being a little thicker than desirable.?

I had no issues installing a Pi3+ into a FLIRC gen1 case with the thicker thermal pad. Mind you, I had been using it with a Pi3 so the pad was a little squashed already. I just went easy on the screws when putting the case back together and didn't over-tighten them.

This original FLIRC case that I have also isn't optimal for the Pi3/Pi3+ as the heatsink lug isn't entirely centred over the SOC - the original case is designed for the Pi2 and the Pi2 SOC is in a slightly different position to the Pi3, but the latest "gen 2" FLIRC case now fully covers the Pi3 SOC (and presumably Pi3+) so case temps might be even better than what I observed. 

I recently had to buy a tube of thermal paste to reseat the heat sink on an Intel Atom.  Would that be an alternative to the thermal pad?

[Vimes]

For those who have the new 3+ Pi and one of the Flirc cases does it fit ok with the older thermal pad being a little thicker than desirable.?

I had no issues installing a Pi3+ into a FLIRC gen1 case with the thicker thermal pad. Mind you, I had been using it with a Pi3 so the pad was a little squashed already. I just went easy on the screws when putting the case back together and didn't over-tighten them.

This original FLIRC case that I have also isn't optimal for the Pi3/Pi3+ as the heatsink lug isn't entirely centred over the SOC - the original case is designed for the Pi2 and the Pi2 SOC is in a slightly different position to the Pi3, but the latest "gen 2" FLIRC case now fully covers the Pi3 SOC (and presumably Pi3+) so case temps might be even better than what I observed.  

Would I be right in assuming that the case doesn't allow you to see the LED's on the Pi3 or the newer + model...?

I do rely on being able to see those and watch the sequence of it powering down so that I then can use my remote to fully power it off.
I do need to fully power it off as the powering on and off wakes / puts to sleep my media server. 

If that is the situation I wonder how effective the cooling would be on a std plastic type case, even with the top off. considering that it has a higher clock speed than the toasty Pi 3....?

Lastly with using your nightlies for the Pi, Kodi L, would I be good to go in terms of just swapping over the SD card to the newer 3+....?

Thanks