From: Jan S. <jan...@vo...> - 2007-12-23 20:03:06
|
Hi, the important properties are: - pin compatibility, size, data width (>=32 bits) - flowthrough mode support - cycle time < 10ns - data setup time <= 2ns - data hold time <= 2ns I did not spend too much time with it but your second option MCM69F737TQ7.5 seems to be usable. Jan Myscha napsal(a): > Hello, > > thank you for welcoming me and thanks for the answer. > > What SRAM-properties are important? That should mean: which criterias are > important when choosing a SRAM? > I found a number of possible replacements, but I'm not sure, whether they > will do or not. > > Regards, > Michael > > > ----- Original Message ----- > From: "Jan Smrz" <jan...@vo...> > To: <min...@li...> > Sent: Sunday, December 23, 2007 1:01 PM > Subject: Re: [Minila-forum] Some questions about MiniLA > > > >> Hello, >> >> be welcome. Please see answers below: >> >> Myscha napsal(a): >> >>> Hello, >>> >>> I want to build a MiniLA and I'm searching for the parts therefore. >>> >>> Does anyone have parts he doesn't need, especially the PCB (with USB), >>> the CPLD, the SRAM and the oscillator? >>> >>> >>> And I have some other questions, too: >>> >>> 1. Can the XC95288XL be replaced with a XC95144XL? I know this >>> questions has already been asked, but it was not answered in English. >>> >> Unfortunately not, the XC95144 is too small. I tried to make some >> limited version of miniLA for this device but the limitations were so >> high that no -144 version of firmware was released. >> >>> 2. Is it neccessary to have a pipelined burst SRAM or will one without >>> pipeline also do? >>> >> The pipeline mode is not used. SRAM runs in flowthrough mode. Some >> people were having problems obtaining the memory and they successfully >> tried an alternative - IS64LF12832A-7. >> >>> 3. What is limiting the speed? The CPLD runs with 200MHz and I >>> have also seen AS7C33128 with 200MHz. And the sample rate is 100MHz. >>> >>> >> Speed is mainly limited by the CPLD. 200MHz is an ideal situation with >> very simple (or no) logic implemented. We need 16-bit trigger, 2x >> 17-bits counters etc. and it slows it down. It was very difficult to >> make it run at 100MHz. >> >>> 4. Are there alternatives for the oscillator? I have no old PC >>> mainboard where I could take it from. >>> >> I don't know, but it has shown up that most 5V oscillators worked at >> 3.3V too. Try it with standard 5V oscillator, which is not difficult to >> obtain. >> >> >> Jan >> >> |