Page 12 of 23
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 26 Apr 2020 10:54
by LADmachining
No change with a different CPU, or on a completely different A500 board (rev 6a, also running KS3.1).
Going to reflow the CPLD again to see what effect that has.
What value resistor networks did other use for the missing ones on the underside? I used 2K2.
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 26 Apr 2020 20:55
by 8 Bit Dreams
Spent some time with TF536 rev.2 today,
Got very good results, card works just nice even with cheap SAMSUNG K4S561632-UC75 ram installed (according to datasheet this ram works @110Mhz, seems we are within the specs with 100Mhz) tried to run different demos, Frontier and Nemac, no problems noticed, now waiting on FedEx to try out buffered ide
20200426_214444.jpg
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 27 Apr 2020 00:24
by dalek
8 Bit Dreams wrote: 26 Apr 2020 20:55
Spent some time with TF536 rev.2 today,
Got very good results...
....
Nice! What are your CPLD temps like? warm? not warm?
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 27 Apr 2020 13:41
by alenppc
8 Bit Dreams wrote: 26 Apr 2020 20:55
Spent some time with TF536 rev.2 today,
Got very good results, card works just nice even with cheap SAMSUNG K4S561632-UC75 ram
Where did you buy this RAM? UTSource? It doesn't seem available on either Digikey or Mouser.
Btw, I confirm that the Xilinx chips I got from UTSource are ok.
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 28 Apr 2020 09:29
by 8 Bit Dreams
Ok guys,
got buffers on TF536 and removed RN6 - RN10.. What can i say..It isn't working, card starts with flashy bars, but then power LED starts to pulse with reboot followed..
Will swap ram chips later to see if i get further..
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 28 Apr 2020 09:32
by PaulJ_2.0
8 Bit Dreams wrote: 28 Apr 2020 09:29
Ok guys,
got buffers on TF536 and removed RN6 - RN10... What can i say... It isn't working, card starts with flashy bars, but then power LED starts to pulse with reboot followed..
Yeah it doesn't work for me either, I would like it to, so I might look into it a little deeper to try and see what might be the issue.
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 28 Apr 2020 09:45
by 8 Bit Dreams
oh,
ok, good to know, so i can save this step with swapping ram.. :geek:
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 28 Apr 2020 09:58
by terriblefire
PaulJ_2.0 wrote: 28 Apr 2020 09:32
8 Bit Dreams wrote: 28 Apr 2020 09:29
Ok guys,
got buffers on TF536 and removed RN6 - RN10... What can i say... It isn't working, card starts with flashy bars, but then power LED starts to pulse with reboot followed..
Yeah it doesn't work for me either, I would like it to, so I might look into it a little deeper to try and see what might be the issue.
Try just buffering the data bus pins.. .if that doesnt work just buffer the control lines. etc.. (control lines are the buffer chip inside the 68000 socket).
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 28 Apr 2020 10:07
by 8 Bit Dreams
Ok, removed chip for control lines and placed back RNs - got boot now :thanksyellow:
plain ide speed is a little bit slow (880kb/s), muFastrom does it much more comfortable to use: 4,3MB/s wich is more than enough for most of existing tasks. im very pleased, Thank You Stephen!
Next step is to add 30cm cable and see how stable will it be, but im already loving it!
SIDENOTE!
have noticed one detail while removed RNs: some pins wasn't tinned at all, seems they are using low quality soldering paste, have resoldered all RNs now..
Re: TF536 Rev 2 Design Complete (all gone)
Posted: 28 Apr 2020 11:15
by terriblefire
Here is the control buffer logic schematic..
I think its possible you need to lift IOR/DIR pin on the control buffer and tie it high or low to make it work.. must have overlooked this.
buffer.JPG
EDIT: Pin 1 on this chip needs lifted and tied high to make the buffer work. Will fix it on the next batch (2b) if someone confirms it working. Should be possible to lift the pin and tie it to pin 2 which is also high.