The TF536 arrives

68030 + SDRAM + IDE

Moderators: terriblefire, Terriblefire Moderator

terriblefire
Moderator Team
Moderator Team
Posts: 1360
Joined: Mon Aug 28, 2017 10:56 pm
Location: Glasgow, UK
Contact:

Re: The TF536 arrives

Post by terriblefire » Sat Aug 17, 2019 6:13 pm

richx wrote:
Sat Aug 17, 2019 6:06 pm
Don't think you are not wrong, most motherboards use dual leaf sockets and using the top solder side pins (mine are ~0.58mm) would not be a good idea for dual leaf sockets. For comparison, regular flat IC pins are around 0.28mm thickness. Machined/turned pin standard is around 0.44-0.46mm I think, same as the thinner bottom side pins on mine.

But we never plug directly into the motherboard because there are too may obstacles to clear.. we always (or at least I do) put turned pin socket strip in the board to gain some height... this is where the issue was.. the skinny pin rattle around in the these sockets.
Could any stability issues be due to the power delivery to the board, with less surface area of the round pins in contact with the dual leaf socket (plus corrosion)? What is the voltage drop from motherboard to board? Would testing with a floppy or AMP type cable soldered to the board be a good idea to ensure power delivery?
No it was simply HALT/VMA/DTACK were floating low simultaneously locking the damn thing up. I could see it on the LA.
As a side note, I did see these regular flat IC pins on ebay.com/ic-china: items 254251573942 and 254251585095. Not sure how easy these would be to solder, but much cheaper than some other regular flat pin alternatives I've seen..
———
I get cranky when asked to repeat myself.

richx
Posts: 5
Joined: Sun Aug 11, 2019 5:03 pm

Re: The TF536 arrives

Post by richx » Sat Aug 17, 2019 6:45 pm

terriblefire wrote:
Sat Aug 17, 2019 6:13 pm
richx wrote:
Sat Aug 17, 2019 6:06 pm
Don't think you are not wrong, most motherboards use dual leaf sockets and using the top solder side pins (mine are ~0.58mm) would not be a good idea for dual leaf sockets. For comparison, regular flat IC pins are around 0.28mm thickness. Machined/turned pin standard is around 0.44-0.46mm I think, same as the thinner bottom side pins on mine.
My original quote to Alen should NOT read "Don't think you are not wrong", looks like you replied before my edit was approved :)
terriblefire wrote:
Sat Aug 17, 2019 6:13 pm
But we never plug directly into the motherboard because there are too may obstacles to clear.. we always (or at least I do) put turned pin socket strip in the board to gain some height... this is where the issue was.. the skinny pin rattle around in the these sockets.

No it was simply HALT/VMA/DTACK were floating low simultaneously locking the damn thing up. I could see it on the LA.
Interesting..

At first glance my A2000 mb using a 12mm height round pin strip it looks like it should fit, but I don't have a board to really know :)

I've just run into wire AWG issues, ATX cables (due to cheap copper wire and crimps I suppose) etc dropping voltages lately, so was curious.

terriblefire
Moderator Team
Moderator Team
Posts: 1360
Joined: Mon Aug 28, 2017 10:56 pm
Location: Glasgow, UK
Contact:

Re: The TF536 arrives

Post by terriblefire » Sat Aug 17, 2019 10:42 pm

It doesnt clear the roms on my A2000 without a bit of height...

But in any case the A2000 is not the target machine and its not officially supported (although i'll usually do a quick test). Is the A500 which needs a relocator and height.
———
I get cranky when asked to repeat myself.

terriblefire
Moderator Team
Moderator Team
Posts: 1360
Joined: Mon Aug 28, 2017 10:56 pm
Location: Glasgow, UK
Contact:

Re: The TF536 arrives

Post by terriblefire » Wed Sep 04, 2019 10:49 pm

@alenppc has his TF536 back finally. Hopefully some reports soon.
———
I get cranky when asked to repeat myself.

User avatar
kludge
Posts: 334
Joined: Thu Nov 08, 2018 2:05 pm
Location: Sweden
Contact:

Re: The TF536 arrives

Post by kludge » Wed Sep 04, 2019 11:19 pm

terriblefire wrote:
Wed Sep 04, 2019 10:49 pm
@alenppc has his TF536 back finally. Hopefully some reports soon.
Holy fudge that took ages!
A kludge is a workaround or quick-and-dirty solution that is clumsy, inelegant, inefficient, difficult to extend and hard to maintain.

My lack of focus:
[ 3 * Amiga 500 ][ Amiga 500+ ][ Amiga 600HD ][ Amiga 2000 ][ Amiga 4000/030 w/ 64 MB RAM hack ][ CD32 (NTSC) ][ VIC-20 ][ 4 * C64 Breadbin ][ 2 * C64C ][ 2 * C128 ][ Mac Classic ][ Mac Classic II ][ Mac Colour Classic ]

alenppc
Moderator Team
Moderator Team
Posts: 314
Joined: Thu Nov 08, 2018 12:59 pm

Re: The TF536 arrives

Post by alenppc » Thu Sep 05, 2019 2:01 am

Tested it tonight for about 30 mins. Big improvement!
It’s still not completely stable for me but much much better than before. I run frontier for a while, doom in a window on workbench and novacoder’s wolf3D ECS port. I get crashes from time to time but I only tried one cpu and for a relatively short time. It might work better with a different one.

Time permitting I am going to test a bit more in-depth tomorrow and hopefully build a second one.

Dano
Posts: 1
Joined: Tue Aug 27, 2019 1:48 am

Re: The TF536 arrives

Post by Dano » Thu Sep 05, 2019 10:51 am

Hi all,
New to the forum, this card sounds like a treat for my A500 and A500+.
Thanks TF for all your hard work in developing these accelerators. :thumbup:
Hopefully it will get off the ground and get through the beta testing.
alenppc wrote:
Thu Sep 05, 2019 2:01 am
Time permitting I am going to test a bit more in-depth tomorrow and hopefully build a second one.

Good luck with your testing :coffee:

Damn Koalas are grunting like mad outside the window tonight mating season :poke:

User avatar
arkadiusz.makarenko
Posts: 92
Joined: Wed Jun 19, 2019 7:36 am
Location: Edinburgh

Re: The TF536 arrives

Post by arkadiusz.makarenko » Thu Sep 05, 2019 11:41 am

alenppc wrote:
Thu Sep 05, 2019 2:01 am
Tested it tonight for about 30 mins. Big improvement!
It’s still not completely stable for me but much much better than before. I run frontier for a while, doom in a window on workbench and novacoder’s wolf3D ECS port. I get crashes from time to time but I only tried one cpu and for a relatively short time. It might work better with a different one.

Time permitting I am going to test a bit more in-depth tomorrow and hopefully build a second one.
That is good news :)
Do not trust people. They are capable of greatness.
~ Stanislaw Lem

go0se
Posts: 94
Joined: Sun Nov 25, 2018 7:55 pm

Re: The TF536 arrives

Post by go0se » Thu Sep 05, 2019 10:41 pm

Hi Alen,

Can I ask what version(s) of Kickstart you are running on your 500/+?

When building some TF534s previously I had issues very like the ones you describe when running 37.175 on an A500 and A500+. The cards themselves all checked out ok with Diagrom but when the kick ROM was swapped out they all booted ok but crashed or developed graphic corruption after a short while - anything from 2-10 mins usually on AmigaTestKit or booting a game.

When I swapped out the Kickstart for 37.350 the problems disappeared.

I appreciate the 534 and 536 are very different but if you have the facility to swap out the ROM it might be worth a try if you're currently on an earlier version of Kickstart to see if it makes any difference to the issues?

For the record the 37.350 ROM I was running was on an AM27C400 120DC EPROM, The 37.175 was a stock MX ROM.

If the other Beta testers could chip in with what versions of Kickstart they are running that might be interesting?

Apologies if this has already been covered.

Cheers.

terriblefire
Moderator Team
Moderator Team
Posts: 1360
Joined: Mon Aug 28, 2017 10:56 pm
Location: Glasgow, UK
Contact:

Re: The TF536 arrives

Post by terriblefire » Thu Sep 05, 2019 11:09 pm

go0se wrote:
Thu Sep 05, 2019 10:41 pm

Apologies if this has already been covered.

Cheers.
I pretty much only test with KS 3.1.

I havent tested much on the A500 with this board. Its been mostly the CDTV. It might be ATA timings.. Its very odd. I ran Alen's TF536 for hours and no crashes. :/
———
I get cranky when asked to repeat myself.

Post Reply

Return to “TF536”