Ok when you are discussing performance issues please do not use SysInfo... if you can show me a video stuttering or something thats a real example. Sysinfo is rubbish.8 Bit Dreams wrote: ↑Tue Sep 10, 2019 8:21 am These are just numbers shown in sysinfo, same that 0MB on TF330 are shown yet ram is present..
TF360 Status
Moderators: terriblefire, Terriblefire Moderator
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: TF360 Status
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: TF360 Status
As Jens says, sysinfo was made to sell accelerator boards. But its actual testing abilities are really bad, especially the transfer rate speeds.
I would recommend trying something like the good old Disk Speed utility (available on aminet). Or maybe the drive speed section of the SysSpeed utility.
I would recommend trying something like the good old Disk Speed utility (available on aminet). Or maybe the drive speed section of the SysSpeed utility.
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: TF360 Status
SysSpeed gives me a good result.
TBH i want to write my own. I would do it if i could find an example of how to read raw disk sectors. AmigaOS is utterly horrific.
EDIT: Ok here is my half arsed attempt at a diskspeed util.
compiles with vbcc
vc -c99 +aos68k -fpu=68881 -I/opt/vbcc//NDK_3.9/Include/include_h/ -c -o testdisk.o testdisk.c
For some reason the IO requests just lock up between reads. I guess i'm missing some step.
TBH i want to write my own. I would do it if i could find an example of how to read raw disk sectors. AmigaOS is utterly horrific.
EDIT: Ok here is my half arsed attempt at a diskspeed util.
compiles with vbcc
vc -c99 +aos68k -fpu=68881 -I/opt/vbcc//NDK_3.9/Include/include_h/ -c -o testdisk.o testdisk.c
For some reason the IO requests just lock up between reads. I guess i'm missing some step.
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: TF360 Status
Quick update: got the new power supply I linked above, it works, doesn't heat at all, however my 060 is still unhappy at 75 mhz, but on the plus side now I also have a 133 crystal so will solder that soon to see if there is any improvement.
EDIT: works great at 66 Mhz on rev 6... will try rev 1 later.
EDIT: works great at 66 Mhz on rev 6... will try rev 1 later.
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: TF360 Status
Yeah i think we might be pushing the decoupling at 75Mhz. Do you have multilayer caps on under the CPU?
Did you try the IDE speedup firmware yet?
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: TF360 Status
I have the caps but they are 1uf... the values are petty random to be honest. Didn’t get a chance to try the ide firmware yet, I had only 20 mins before having to go get the kids. Will try it tomorrow.
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: TF360 Status
No worries. I'm expecting chucky's in the post soon and then i'll send one back to supaduper.... Hopefully get to sorting out the Rev1 in the next week or two... then the fun begins.
Generally interested in the things that make building it a pain too.
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: TF360 Status
I have no general complaints about the current state of the 360, it's fairly easy to build.terriblefire wrote: ↑Tue Sep 10, 2019 11:56 pm Generally interested in the things that make building it a pain too.
You have probably already thought of most of the the following, but this would be on my wishlist of things to change:
- If possible, move the ESPHEADER and the jumpers further away from the IDE connector. Right now, when the cable is plugged it overlaps the space, so you technically can't use both at the same time.
- If possible, reduce at least one of the CPLDs to a cheaper version.
- Add .7mm to the edge connector which will make it a LOT easier to hand-bevel as we tried with the 330 (I couldn't get anywhere with JLCPCB, they keep not doing it).
- Finally, change the regulator footprint for one that matches the standard DC-DC converters with 3 leads spaced at 2.54mm and the VIN GND VO pinout
-
- Moderator Team
- Posts: 5389
- Joined: Mon Aug 28, 2017 10:56 pm
- Location: Glasgow, UK
Re: TF360 Status
Yeah this is on the list. I think the ESP will turn through 90 degrees.alenppc wrote: ↑Wed Sep 11, 2019 1:34 pm I have no general complaints about the current state of the 360, it's fairly easy to build.
You have probably already thought of most of the the following, but this would be on my wishlist of things to change:
- If possible, move the ESPHEADER and the jumpers further away from the IDE connector. Right now, when the cable is plugged it overlaps the space, so you technically can't use both at the same time.
IMG_8651.png
Unsure if this is possible but i'll try and I agree. they're expensive things. For the first version i just gave myself all the room possible.
- If possible, reduce at least one of the CPLDs to a cheaper version.
This was done i think for the 360... But JLCPCB made such an arse of the TF360 prototype im not sure i'll ever use them again.
- Add .7mm to the edge connector which will make it a LOT easier to hand-bevel as we tried with the 330 (I couldn't get anywhere with JLCPCB, they keep not doing it).
Yeah i had no idea what id use on this board so i left a big set of pads.- Finally, change the regulator footprint for one that matches the standard DC-DC converters with 3 leads spaced at 2.54mm and the VIN GND VO pinout
———
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
"It is not necessarily a supply voltage at no load, but the amount of current it can provide when touched that
indicates how much hurting you shall receive."
Re: TF360 Status
Usually when you add it they leave a strip of plastic without the gold contacts which wasn't on the one you sent me, but regardless this is really the biggest issue with JLCPCB. On the other hand the boards don't present issues like they do with PCBWay.terriblefire wrote: ↑Wed Sep 11, 2019 2:09 pmThis was done i think for the 360... But JLCPCB made such an arse of the TF360 prototype im not sure i'll ever use them again.
- Add .7mm to the edge connector which will make it a LOT easier to hand-bevel as we tried with the 330 (I couldn't get anywhere with JLCPCB, they keep not doing it).
I have also used Elecrow and they are also very good, so I could recommend trying them, maybe they will finally bevel them.