I tried the overlock using Extrem mode.
Recalbox OS (not only ES) crash after several second each time (and I need to make an "hard restart" by unplug my Pi).
Rpi3b+, Superpi case, SFC30pro
@vincent-84 I use extrem too and everything is fine. I guess it depend on some boards that can support better oc than others
That's why there is a Warning when you select Turbo/Extrem O/C.
There are many Pi manufacturers and they have all different quality checks. Some chineese manufacturers have no QC at all.
When playing with O/C options, I highly recommend to put a good cooling system (at least a large heatsink).
Sarting immediately with the Extrem mode is not a good idea. Start with the "High" mode, check if everything is fine, then try the next level and so on...
@bkg2k you're wright : that was an heat issue (solved by opening my superpi case and use and external fan for test).
I had to add some direct fan on the pi (and change my case).
Rpi3b+, Superpi case, SFC30pro
@vincent-84 you can see the temperature with the webmanager, over 80 degres your pi will become instable. I have my B+ in a snespi case with heatsinks and fan, lt's oc at 1500 and it's 60 degres.
I have the latest version of RecalBox Beta and Raspberry Pi 3B+
I do not know how to make remount partitions as RW:
@substring said in Testers wanted: major system upgrade (Pi3b+, KODI 17 and much more ...):
- know the very basics of Linux:
- remount partitions as RW
So most errors are likely because of this.
The following errors I noticed:
@Gaetan thanks for the tip, I didn't know that (good) tool.
Last crash (with case close) doesn't see any anormal temperature, have to try again.
Maybe I was just lucky during my test with open case....
What is the main objective of the "step by step OC" ? Seeking the maximum value of the OC for each pi or try to stabilized the OCing mode ?
P.S.: my objective (and test) : running goldeneye64 (it seems to work but only at 1500).
A bug has been identified with some pads (8bitdo for example), making them work in Emulation Station but not in Retroarch emulators. We're working on fixing it, but here is a "quick fix" as a workaround:
Then your pad should work 
I confirm, when we add a game in favorites, the screen remains blocked on a black screen (below the build number) during a reboot or during the update of the lists of games or white screen during the recalbox stop. In the latter before unplugged the RPI3B +, during reboot, loss of connectivity bluetooth my gamepad 8bitdo SF30Pro and inpossible to re-pair.
But strangely, the fact of removing favorite games, the gamepad works properly
@oyyodams have your try without attribute any buttons for L3R3 ?
(That's my configuration and I still have deconnexion entering retroarch)
@oyyodams My sfc30 pro are pairing and I can map them correctly the first time, no problem.
If i map L2 R2, I can't play any retroarch games, if I map everything but L2 R2, I can play retroarch games.
With my PS3 pad, i had to remap 3 times the pad to have my configuration right. I noticed that i had to map the pad very fast to make it. It's like, if i take to much time to configure it, I will have an axis instead a button in the configuration.
Hope you anderstand this part
Once it's done right, i can use it in game too.
So i confirm, L2 R2 are messing up something with retroarch games
Boards : Now: Nuc i7 7567u // Past: Rpi3 B, Rpi3 B+, Odroid xu4
Controllers : Ps3 dualshock, 8bitdo SFC30 pro, Snes wired controller
@gaetan Same on 8bit Nes 30 pro L1 R1 is sometime see as axes as L2R2 for SFC 30 pro.
Bluetooth and Wifi problem seems gone for me.
Hello,
I confirm. Despite several tries to plug, unplug, remap, forget bt config, remap etc..., connection stil remains lost between ES & Retroarch and this version is not playable with SFC30pro in wireless mode when combined with dongle bt (mine at least) supposed to reduce inputlag.
Let's see on next update.
Thanks
@vdeolive I have absolutely no problem with my SFC30 Pro on the internal Bluetooth of the 3+, as long as you make sure that the R2 L2 are correctly map as buttons and not axe.
I have the 1.29 firmware on the gamepad
Hello Choum
And with a bluetooth dongle, it works for you?
I am using it to play with less input lag for game such as Super Mario World of Street fighter II Turbo where timing is essential. You don"t have any input lag on internal BT?
Thanking you
Regards
I don't use a dongle.
I have made some try on SF2 Turbo snes.
with the l2ping command (to see ping time from the gamepad)
Internal Bluetooth with wifi on : 7.75ms
Internal Bluetooth without wifi : 7.99ms
tested during around 140s (2-3 round each)
It seems that the input lag caused by wifi like on previous 3b model are gone on 3b+.
@choum your conclusion about lags gone on B+ could deserve to be verified. On the B it was horrible
Hi, following emulators did not work in the latest beta version:
BLUEMSX - Returns to the menu
FMSX - Stop on a black screen
HATARI - Returns to the menu after error message
EDIT: RPI3 version
@choum I could try the bluetooth input lag performance on B and B+, could you tell me how you did to mesure it ?
Boards : Now: Nuc i7 7567u // Past: Rpi3 B, Rpi3 B+, Odroid xu4
Controllers : Ps3 dualshock, 8bitdo SFC30 pro, Snes wired controller
@gaetan on putty :
hcitool con -> To list gamepad mac adress
You should see somthing like "ACL AA:BB:CC:DD:EE:FF handle 11 state 1 lm MASTER"
Then
l2ping AA:BB:CC:DD:EE:FF
This will run a infinite ping on the gamepad each second.
A bug has been identified with some pads (8bitdo for example), making them work in Emulation Station but not in Retroarch emulators. We're working on fixing it, but here is a "quick fix" as a workaround:
Then your pad should work
