Searching \ for 'Serial RAM??' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/io/serials.htm?key=serial
Search entire site for: 'Serial RAM??'.

Truncated match.
PICList Thread
'Serial RAM??'
1997\05\30@015909 by David Gould

flavicon
face
I am considering a data logging project that would sample a number of sensors
and record the data for later playback into a PC host for analysis. Minimally
it would need to record one or two hundred bytes per second for several
minutes, ideally quite a bit more. So something like 32K to 256K of memory
would need to be written and then later read back. There is almost no
requirement to randomly access this memory, just need to write it as data
comes in and read it for upload to the PC.

I know that I could build a data and address bus out of I/O pins, but that
seems like overkill and would probably use too many pins anyhow.

I have seen serial EEPROMS and they would almost be perfect except for the
slow programing and power consumption. Is there a serial SRAM available?
or a simple way to use commodity SRAM chips?

Any other thoughts?

-dg

David Gould           spam_OUTdgTakeThisOuTspamillustra.com            510.869.6383 or 510.305.9468
Informix Software (formerly Illustra)  1111 Broadway #2000  Oakland, CA 94607
- I realize now that irony has no place in business communications.

1997\05\30@021220 by Mark Hitchings

flavicon
picon face
David,
TRY PCF8570 256 x 8 Static RAM made by Phillips which interfaces to the I2C
bus. There are larger than 256 bytes but I don't know who makes them.
Regards,
Mark

At 22:47 29/05/97 -0624, you wrote:
{Quote hidden}

----------------------------------------------------------------------
Mark Hitchings
R&D Engineer
Intelligent Control Systems Laboratory
School of Microelectronic Engineering
Griffith University, Nathan, Qld. 4111
ph: +61 7 3875 5067
fax: +61 7 3875 5384
email: mhitchinspamKILLspamme.gu.edu.au
----------------------------------------------------------------------

1997\05\30@031640 by TONY NIXON 54964

flavicon
picon face
If you can't find a serial RAM chip to do the job and you have
sufficient board space you might like to try binary counters such as
a CMOS 4040 for the address decoder. These will work fine if you do
not need random access. Just 2 control pins, clock and reset.
Use 2 chips for 256K. I've used this method to get 12 bit data stored
into Battery backed RAM at 10K samples/sec.

Tony.


Just when I thought I knew it all,
I learned that I didn't.

1997\05\30@034333 by Andrew Warren

face
flavicon
face
David Gould <.....PICLISTKILLspamspam.....MITVMA.MIT.EDU> wrote:

> I am considering a data logging project that would sample a number
> of sensors and record the data for later playback into a PC host for
> analysis. Minimally it would need to record one or two hundred bytes
> per second for several minutes, ideally quite a bit more.
> ....
> I have seen serial EEPROMS and they would almost be perfect except
> for the slow programing and power consumption. Is there a serial
> SRAM available? or a simple way to use commodity SRAM chips?

David:

You could rig up an SRAM with a counter chip on its address lines,
then just clock the counter with the PIC and present data 8 bits at
a time to the SRAM's data bus... But I'd probably recommend using
the serial EEPROMs, instead.

Microchip makes large (64K) EEPROMs, and they can be powered from a
pin on the PIC (so you can power them down when you're not actually
accessing them).

As you've noticed, the write time IS relatively slow (10 milliseconds
max), but you can get around that with just a little bit of
cleverness:

   1.  The large EEPROMs allow "block writes" of 8 or 16 bytes at a
       time.  Writing a block takes the same amount of time as
       writing a single byte, so using this feature drops the write
       time to 1.25 ms/byte, max.

   2.  If you use multiple EEPROMs, you can interleave writes to
       them.  That is, you can write a block of 8 bytes to one
       chip, then (while that chip is occupied in its 10-ms write
       cycle), you can write the next block of 8 bytes to another
       chip.  If you use two EEPROM chips, this cuts the write time
       in half, to 0.625 ms/byte.

At 0.625 milliseconds per byte, you can store 1600 samples per
second.

-Andy

=== Andrew Warren - EraseMEfastfwdspam_OUTspamTakeThisOuTix.netcom.com
=== Fast Forward Engineering, Vista, California
=== http://www.geocities.com/SiliconValley/2499

1997\05\30@050704 by Keith Dowsett

flavicon
face
At 22:47 29/05/97 -0624, you wrote:
>I am considering a data logging project that would sample a number of sensors
>and record the data for later playback into a PC host for analysis. Minimally
>it would need to record one or two hundred bytes per second for several
>minutes, ideally quite a bit more. So something like 32K to 256K of memory
>would need to be written and then later read back. There is almost no
>requirement to randomly access this memory, just need to write it as data
>comes in and read it for upload to the PC.
>
>I know that I could build a data and address bus out of I/O pins, but that
>seems like overkill and would probably use too many pins anyhow.
>
>I have seen serial EEPROMS and they would almost be perfect except for the
>slow programing and power consumption. Is there a serial SRAM available?
>or a simple way to use commodity SRAM chips?
>
>Any other thoughts?

I've been thinking along similar lines. So here goes,

EEPROM has several important features:

   Relatively low power (I can turn it off when I'm not using it)
   Not affected by power glitches (except when writing)
   SLOW (1ms write time)
   Only one supplier of 128k x 8 chips (that I've found)
   Expensive (10 x cost of PIC)

SRAM has some important features too:

   Eats power like it's going out of fashion
   Wiped by power glitches
   Damn FAST
   Lots of suppliers
   Fairly cheap (2 x cost of PIC)

DRAM entered briefly into my considerations:

   Pretty low power
   Wiped by some power glitches
   Pretty fast
   Needs refresh electronics (real killer)

For the sort of data rates you are talking about EEPROM should be quite
feasible. If you need to write more than 500 bytes/second you might have to
work a bit harder.

Rather than have a proper address bus how about using a counter? You are
only going to want to read/write sequentially anyway.

What I need now is a couple of months to design the board and write some
(really quite simple) code for it. Sadly there are lots of other projects
mounting up on my desk at the moment, so let me know how you get on.

Keith.



==========================================================
Keith Dowsett         "Variables won't; constants aren't."

E-mail: kdowsettspamspam_OUTrpms.ac.uk
  WWW: http://kd.rpms.ac.uk/index.htm

1997\05\30@065847 by Andy Kunz

flavicon
face
At 10:47 PM 5/29/97 -0624, you wrote:
>it would need to record one or two hundred bytes per second for several
>minutes, ideally quite a bit more. So something like 32K to 256K of memory
>would need to be written and then later read back. There is almost no
>requirement to randomly access this memory, just need to write it as data
>comes in and read it for upload to the PC.

RAMTRON http://www.ramtron.com makes what they call FERAMs.  These are pin
and software compatible with I2C EEPROMs, but the 10mS delay we are all
used to using is _!NOT!_ required.  Also, you can write the entire chip in
sequence after a single addressing command!

This sounds like it would be ideal for your application.  I am awaiting a
few samples.

Pricing is pretty good, too - very competitive to EEPROM for what you
advantages you get, imho.

Andy

==================================================================
Andy Kunz - Montana Design - 409 S 6th St - Phillipsburg, NJ 08865
         Hardware & Software for Industry & R/C Hobbies
       "Go fast, turn right, and keep the wet side down!"
==================================================================

1997\05\30@071511 by Mike

flavicon
face
At 06:51 AM 5/30/97 -0400, you wrote:

>RAMTRON http://www.ramtron.com makes what they call FERAMs.  These are pin
>and software compatible with I2C EEPROMs, but the 10mS delay we are all
>used to using is _!NOT!_ required.  Also, you can write the entire chip in
>sequence after a single addressing command!

Odd - that link gets me to:-   http://www.sni.net/sni/index.html

and not the FRAM company's home page.

Netscape does not report a domain server error or any exception, can
you confirm that address actually gets you to RAMTRON ?

Rgds

Mike
Perth, Western Australia

PS: And yes I did a cut and paste as well - it was entered correctly.


Some say there is no magic but, all things begin with thought then it becomes
academic, then some poor slob works out a practical way to implement all that
theory, this is called Engineering - for most people another form of magic.
                                                                      Massen

1997\05\30@072310 by Mike Smith

flavicon
face
> From: David Gould <@spam@dgKILLspamspamILLUSTRA.COM>
> To: KILLspamPICLISTKILLspamspamMITVMA.MIT.EDU
> Subject: Serial RAM??
> Date: Friday, 30 May 1997 14:41
>
> I am considering a data logging project that would sample a number of
sensors
> and record the data for later playback into a PC host for analysis.
Minimally
> it would need to record one or two hundred bytes per second for several
> minutes, ideally quite a bit more. So something like 32K to 256K of
memory
> would need to be written and then later read back. There is almost no
> requirement to randomly access this memory, just need to write it as data
> comes in and read it for upload to the PC.
>
> I know that I could build a data and address bus out of I/O pins, but
that
> seems like overkill and would probably use too many pins anyhow.
>
> I have seen serial EEPROMS and they would almost be perfect except for
the
> slow programing and power consumption. Is there a serial SRAM available?
> or a simple way to use commodity SRAM chips?
>
> Any other thoughts?
>

Yeah.  I'd think about doing on-the-fly compression of your data - ie RLL,
or maybe an ADPCM-like system.  Given that most transducers change slowly,
this'd save space over saving absolute value every time.

Another idea would be to dedicate a low-end PIC to generate address/data
for a parallel  memory device, and communicate with it via serial lines
from your main PIC.  Lower device count than using logic chips, and its
smarter.  Also leaves main PIC with plenty of free lines. (which you need
for a logger) Consider making it detachable from the logger, and with an
inbuilt lithium battery, and you could use SRAM instead of EEPROM (cheaper
& available larger), and you could make up a 'reader' unit for a PC.  Unit
could be disposable when battery went flat, or replace battery(or recharge
- personally, I'd use lithium - they last longer than NiCd in that sort of
app anyway...).  With datalogging, you should monitor the battery, so you
don't lose data.

What language would you use for logger?  Most systems I've seen have really
brain-damaged schemes for programming them.  I'd be interested in hearing
your ideas.  I'd use something like an '84 with EE^2 for program memory,
write a C (or maybe basic) like language on a PC based text editor, then
have the PC compile that into a hex file and d/l it to program memory.
Saves having an interpretive based system that has to fetch tokens from
somewhere, and allows easy use of library functions.

My 2c worth of what a decent datalogger should do - I might do it one
day...

MikeS
<RemoveMEmikesmith_ozTakeThisOuTspamrelaymail.net>

1997\05\30@074935 by Mike Smith

flavicon
face
> From: Mike <spamBeGoneerazmusspamBeGonespamWANTREE.COM.AU>
> To: TakeThisOuTPICLISTEraseMEspamspam_OUTMITVMA.MIT.EDU
> Subject: Re: Serial RAM??
> Date: Friday, 30 May 1997 20:45
>
> At 06:51 AM 5/30/97 -0400, you wrote:
>
> >RAMTRON http://www.ramtron.com makes what they call FERAMs.  These are
pin
> >and software compatible with I2C EEPROMs, but the 10mS delay we are all
> >used to using is _!NOT!_ required.  Also, you can write the entire chip
in
> >sequence after a single addressing command!
>
> Odd - that link gets me to:-   http://www.sni.net/sni/index.html
>
> and not the FRAM company's home page.
>
> Netscape does not report a domain server error or any exception, can
> you confirm that address actually gets you to RAMTRON ?
>

I'd love to blame NetScape, being an IE user, but...

That one caught me out a while ago - what you do is goto that page, click
on findit, and you see the following sentence...

(Note: If you were expecting to see another company's web site, but got
SuperNet's web site instead, click here for a possible explanation.)

doing that gets you this - (excerpt only)

1) The URL you entered was not complete. If the web for which you are
searching is hosted on one of our shared-web machines, then omitting any
required part of their URL (for example, going to http://www.domain.com/
when the correct URL was really something like
http://www.domain.com/domain/ or http://www.domain.com/~user) could send
you to our home page, since in this case their web site may in fact be
hosted on the same machine as our web page. (They would need to upgrade
their web site to one of our Virtual Domain hosting services for the
previous simpler URL to work as expected.)

Now we're getting somewhere! Try -

http://www.ramtron.com/ramtron/

It worked for me - I just tried it again.


> Some say there is no magic but, all things begin with thought then it
becomes
> academic, then some poor slob works out a practical way to implement all
that
> theory, this is called Engineering - for most people another form of
magic.
>
Massen

Any sufficiently advanced technology is indistinguishable from magic
                                               Niven or Asimov - I think it was
the former.


MikeS
<RemoveMEmikesmith_ozspamTakeThisOuTrelaymail.net>

They're a friendly enough company - sent me a databook  - all the way down
here to Oz.

1997\05\30@080536 by Stuart Tyrrell

flavicon
picon face
Mike wrote:

> >RAMTRON http://www.ramtron.com makes what they call FERAMs.

> Odd - that link gets me to:-   http://www.sni.net/sni/index.html

Try http://www.ramtron.com/ramtron



Stuart.

1997\05\30@092428 by John Payson

flavicon
face
> At 10:47 PM 5/29/97 -0624, you wrote:
> >it would need to record one or two hundred bytes per second for several
> >minutes, ideally quite a bit more. So something like 32K to 256K of memory
> >would need to be written and then later read back. There is almost no
> >requirement to randomly access this memory, just need to write it as data
> >comes in and read it for upload to the PC.
>
> RAMTRON http://www.ramtron.com makes what they call FERAMs.  These are pin
> and software compatible with I2C EEPROMs, but the 10mS delay we are all
> used to using is _!NOT!_ required.  Also, you can write the entire chip in
> sequence after a single addressing command!

In some applications I've done using FRAM devices, there sometimes seems
to be a memory-loss problem when the system is powered down and back up.
I don't know whether this is due to the system's CPU going bonkers and
writing something it shouldn't as it loses VDD, but I'd suggest anyone
using the FRAM devices take care to ensure that their system shuts down
nicely when it loses power.

Also, another important thing to note about FRAMs: they perform a store
operation as soon as they get the last data bit of a byte, even if the
write is not terminated properly.  This is in contrast to most EEPROMs
which will abort a page-write cycle without writing anything if, e.g.,
they receive a stop in the middle of a byte.

1997\05\30@102148 by mike

flavicon
picon face
In message  <9705300547.AA17247EraseMEspam.....hawk.illustra.com> EraseMEPICLISTspamMITVMA.MIT.EDU
writes:
> I have seen serial EEPROMS and they would almost be perfect except for the
> slow programing and power consumption. Is there a serial SRAM available?
> or a simple way to use commodity SRAM chips?

You could try the serial FRAMs available form Ramtron. They are
non-volatile, write as fast as SRAM (for all practical purposes)
have a high write cycle endurance etc.

Regards,


Mike Watson

1997\05\30@122316 by Geoff Smith

flavicon
face
The nicest ram I have seen is the Dallas semiconductor RAMport ram,
unfortunatly the largest size is 2k X 8. It requires around 700nS to
read/write - it uses parallel 8 lines to communicate to the memory, but the
clever thing is that it only uses 2 io control lines. The 8 lines of
data/addr bus are fed thru the chip - so you still can use these lines as
io. When you select the ram the output lines are latched, and the ram
accessed.

Geoff.

http://www.dalsemi.com/Prod_info/Memory/ramport.html


{Quote hidden}

Geoff Smith
---------------------------------------------------------------
Home (UK)                      |  Interval Research Corp.
                              |  Palo Alto, California
RemoveMEgsmithaEraseMEspamEraseMEcix.compulink.co.uk    |  RemoveMEgsmithspam_OUTspamKILLspaminterval.com
                              |    (415) 842-6159

1997\05\30@140419 by Mike

flavicon
face
At 09:27 PM 5/30/97 +0930, you wrote:

>> >RAMTRON http://www.ramtron.com makes what they call FERAMs.

>http://www.ramtron.com/ramtron/

>It worked for me - I just tried it again.

May I suggest that when you reference a URL it be in full so we don't
waste both our both valuable time.

Rgds

Mike

Doin what yah always did will get yah what yah always got.


Some say there is no magic but, all things begin with thought then it becomes
academic, then some poor slob works out a practical way to implement all that
theory, this is called Engineering - for most people another form of magic.
                                                                      Massen

1997\05\30@214307 by Mike Smith

flavicon
face
> From: Mike <RemoveMEerazmusTakeThisOuTspamspamWANTREE.COM.AU>
> To: EraseMEPICLISTspamspamspamBeGoneMITVMA.MIT.EDU
> Subject: Re: Serial RAM??
> Date: Saturday, 31 May 1997 03:34
>
> At 09:27 PM 5/30/97 +0930, you wrote:
>
> >> >RAMTRON http://www.ramtron.com makes what they call FERAMs.
>
> >http://www.ramtron.com/ramtron/
>
> >It worked for me - I just tried it again.
>
> May I suggest that when you reference a URL it be in full so we don't
> waste both our both valuable time.
>

Hey, Mike, I didn't post the original message - just the one with the
corrected URL.  I just had a sense of deja vu about the problem, so thought
I'd help :)


MikeS
<RemoveMEmikesmith_ozKILLspamspamrelaymail.net>


'Serial RAM??'
1997\06\06@134809 by Sami Khawam
flavicon
picon face
As I remember the NM29A080 and NM29A040 FEEPOM
from National are fast and have 8Mb and 4Mb,
and use a 3-wire interface. But they aren't
more produced....but you can still find them.

--
Sami Khawam
sKhawamSTOPspamspamspam_OUTbigfoot.com
spamBeGonea9501901STOPspamspamEraseMEunet.univie.ac.at
http://unet.univie.ac.at/~a9501901

More... (looser matching)
- Last day of these posts
- In 1997 , 1998 only
- Today
- New search...