Re: XRAM notes

From: Clockmeister <clockmeister_at_internode.on.net>
Date: Tue, 11 Mar 2014 22:05:47 +0800
Message-ID: <531F183B.7010108@internode.on.net>
On 11/03/2014 7:55 PM, silverdr@wfmh.org.pl wrote:
> ------------------------------------------------------
> From: Clockmeister
>>   
>> On 11/03/2014 4:30 AM, silverdr@wfmh.org.pl wrote:
>>> When trying to deal with some 4168 XRAM chips, I couldn’t get them to work properly. After
>> studying the datasheet I found in the notes:
>>> http://dl.dropboxusercontent.com/u/58002657/4168_notes.jpg
>>>
>>> Point [2.] seems important to me but I seem to have problems interpreting it unambiguously..
>> or for some other reasons these RAMs still don’t want to work..
>>> “An initial pause of 2ms is required after power up, followed by any 8 _CE cycles and 64
>> _RFSH cycles before proper device operation is achieved. Read, write, and external
>> refresh cycles may be used as _CE dummy cycles for initialisation. The 64 refresh dummy
>> cycles can be performed before or after the 8 _CE dummy cycles. Both dummy cycles must
>> be within AC parameters. See figure 1, below."
>>> Figure follows as on the picture linked above. How do you understand the above? How should
>> it be interpreted? Please note that e. g. here:
>>> http://www.smspower.org/uploads/Development/SegaMasterSystemIIServiceManual-1715922A-1.png
>>>
>>> the _RFSH is tied up to VCC..
>>   
>> I'm sure I've replaced these with standard SRAM (and/or vice-versa).
>> When you say they don't work, what are you trying to get them to work in
>> and in what way don't they work?
> Some time around the middle of last year I had an idea of doing exactly what you wrote. I wanted to put XRAM in place of SRAM onto my DD2 and DD3 boards because I have a good bunch of those 4168 XRAM chips that lie unused (taken off a working industrial equipment during scrapping). But they don’t work in the sense that the system hangs when I put the XRAM chip. I tried doing the same as on the RAMBOard, i.e. connecting _RFSH to PHI2 and adding one more cap for delaying the _OE/_CE lines. I even connected CS to PHI2 as in the RAMBOard schematic (just to be sure) but it still didn’t help. After scratching my head for a long time I eventually found the note in the data sheet that made me suspicious whether I shouldn’t be doing something extra to make them work..
>
> --
>

Ok, I follow now.

Have you seen this datasheet for the compatible NEC chip?

It talks about the different refresh modes (self refresh/Pulse/external 
refresh). Not sure if it helps though.
http://pdf1.alldatasheet.com/datasheet-pdf/view/129893/NEC/UPD4168.html





       Message was sent through the cbm-hackers mailing list
Received on 2014-03-11 15:00:04

Archive generated by hypermail 2.2.0.