EmbeddedRelated.com
Forums

[OT?] a SIMPLISTIC RS232 data logger using some flash device --- does it exist commercially?

Started by Richard Owlett June 21, 2005
Everett M. Greene wrote:

> Kelly Hall <khall@acm.org> writes: > >>Richard Owlett wrote: >> >>>BTW my intended real of application is recording data from a GPS receiver. >> >>I googled for "NMEA data logger" and got 11700 hits. >>I googled for "GPS data logger" and got 163000 hits. > > > Now all you have to do is find the time to examine all > 175K of them to see if one of them is optimal...
Actually I've previously posted to a GPS group Response i got was product XYZ satisfies ALL needs Actually my post had been triggered by product XYZ *not* meeting needs But general thanks to this group. The various public responses have shown how I failed to specify my goals. I've received private responses that open up vistas of attacking my problem that I had not considered. I'll make a more careful repost
On Tue, 21 Jun 2005 15:03:09 -0500, Richard Owlett wrote:

>Caution: I explicitly stated "simplistic" on purpose. > I did not say simple, elegant, robust or many other adjectives. > >The input sequences will be a rs232 level 4800 baud data stream of <=100 >characters terminated by a <CR><LF> pair no more frequently than once >per second. > >I wish this unedited data to be stored on a 'compact flash' device which >may be read by a USB device having relatively standard interface to a >WinXP system. > >I *explicitly* do not require this flash device to have a file system. >A file system might actually be an impediment. >I envision each <CR><LF> terminated "sentence" being stored in a >sequential block. [ Reading software responsible for ignoring any thing >after <CR><LF>. ]
I believe all XP supported USB card readers expect the card to have a file system on them, because they identify themselves to XP as a mass storage device type. But there might be an alternative: you already have an RS232 interface, with data coming in at about 1kbps . Why don't you use the same interface to play it back? hook it up to a PC, at 115 kbps it will playback 100 times faster than recording. Depending on how long you plan to record, this might be acceptable. There might be serial to USB devices that can go even faster. You could even write the firmware so that during playback it (playback) can be controlled from the PC, e.g. going directly to a certain point. Some of the PIC micros have built-in UARTS and run on 3.3 V. Mat Nieuwenhoven
On 2005-06-23, Mat Nieuwenhoven <mnieuw@dontincludethis.zap.a2000.nl> wrote:

>>I *explicitly* do not require this flash device to have a file system. >>A file system might actually be an impediment.
Note that explicitly not requiring the device have a filesystem isn't the same as explicitly requiring the device not have a filesystem. The device can either use or not use a filesystem and meet the stated requirement.
>>I envision each <CR><LF> terminated "sentence" being stored in a >>sequential block. [ Reading software responsible for ignoring any thing >>after <CR><LF>. ] > > I believe all XP supported USB card readers expect the card to have a file > system on them, because they identify themselves to XP as a mass storage > device type.
I don't see the reasoning behind that statement. There is no connection between a device being a mass storage device and a device requiring a filesystem. I've used dozens of various card-readers and none of them expected/required the card to have a filesystem even though they were all mass storage devices. It's possible that XP is too stupid to deal with cards that don't have a filesystem, but that's Microsoft's damage: it has nothing to do with the USB mass storage device expecting to find a filesystem. -- Grant Edwards grante Yow! Give them at RADAR-GUIDED SKEE-BALL visi.com LANES and VELVEETA BURRITOS!!
On Thu, 23 Jun 2005 13:41:52 -0000, Grant Edwards wrote:

> -- > Grant Edwards grante Yow! Give them > at RADAR-GUIDED SKEE-BALL > visi.com LANES and VELVEETA > BURRITOS!!
Just curious. Are the quotes include in your .sig from Zippy the Pinhead? They sound oddly familiar... Bob
On 2005-06-23, Bob Stephens <roberts@dcxchol.com> wrote:

> Just curious. Are the quotes include in your .sig from Zippy > the Pinhead? They sound oddly familiar...
Yup. I found a big file of Zippy quotes once upon a time (as part of the Emacs "yow" command, I think). I wrote a shell script that picks one at random for each posting. It's surprising how often they seem related to the thread. Zippy's like that: you're not always sure what he's saying but it somehow seems relevent anyway. Once upon a time, many years ago, I also had random Zippy quotes attached automatically to e-mails (even internal work-related ones). Once e-mail got to be more mainstream, I decided that wasn't a good idea. I particularly remember one recipient's reaction to a Zippy quote about the pope and a bubble-bath... -- Grant Edwards grante Yow! I'm thinking about at DIGITAL READ-OUT systems visi.com and computer-generated IMAGE FORMATIONS...
On Thu, 23 Jun 2005 15:11:14 -0000, Grant Edwards wrote:

> On 2005-06-23, Bob Stephens <roberts@dcxchol.com> wrote: > >> Just curious. Are the quotes include in your .sig from Zippy >> the Pinhead? They sound oddly familiar... > > Yup. I found a big file of Zippy quotes once upon a time (as > part of the Emacs "yow" command, I think). I wrote a shell > script that picks one at random for each posting. It's > surprising how often they seem related to the thread. Zippy's > like that: you're not always sure what he's saying but it > somehow seems relevent anyway. > > Once upon a time, many years ago, I also had random Zippy > quotes attached automatically to e-mails (even internal > work-related ones). Once e-mail got to be more mainstream, I > decided that wasn't a good idea. I particularly remember one > recipient's reaction to a Zippy quote about the pope and a > bubble-bath...
Fun idea! My vote goes to Zippy for moderator;) Bob
Grant Edwards wrote:
> Mat Nieuwenhoven <mnieuw@dontincludethis.zap.a2000.nl> wrote: >
... snip ...
> >>> I envision each <CR><LF> terminated "sentence" being stored in a >>> sequential block. [ Reading software responsible for ignoring any >>> thing after <CR><LF>. ] >> >> I believe all XP supported USB card readers expect the card to >> have a file system on them, because they identify themselves to XP >> as a mass storage device type. > > I don't see the reasoning behind that statement. There is no > connection between a device being a mass storage device and a > device requiring a filesystem. I've used dozens of various > card-readers and none of them expected/required the card to > have a filesystem even though they were all mass storage devices.
You might do well to emulate the CP/M system, as used for script (submit) files. The storage was originally based on the 128 byte sector size of an 8 inch SSSD floppy (and subsequent other formats were mapped into this). The submit file consisted of one line per sector, <cr> terminated. Although the file was a part of the file system, the only addressing necessity was sequential sectors. -- "If you want to post a followup via groups.google.com, don't use the broken "Reply" link at the bottom of the article. Click on "show options" at the top of the article, then click on the "Reply" at the bottom of the article headers." - Keith Thompson
My thanks to the group for providing useful input in response to a 
poorly defined question.

I had broken my problem statement into three parts.
1. what I wanted to accomplish
2. a view of possible solution
3. thoughts on what hardware might satisfy needs/desires/aesthetics

Unfortunately my original post pureed all into amorphous glob.
My apologies.

I'll break things down in above groupings.

1. What am I looking for
It might be described as "data logger" or a "very large FIFO".

The input will be an RS232 data stream terminated by <CR><LF> of no more 
than 100 bytes at 4800 BAUD occurring once per second.

It shall buffer/store at least 10 mega bytes

The output shall be via USB
  1. I wish to take advantage of USB speed
  2. My computer has available USB port(s)
  3. My computer does *not* have available serial port(s)


2. Possible solutions
Flash devices are large and do not require power to maintain data.
There are environments such as FORTH which can presume any mass storage 
device/system to be a sequence of nnn byte blocks/sectors/whatever

3. available hardware
Let's just forget how badly I mangled this one ;{


Is this a better problem definition?
Grant Edwards wrote:
> On 2005-06-21, Paul Carpenter <paul$@pcserviceselectronics.co.uk> wrote: > > > You want a USB CF card to be WinXP compliant *without* a file > > system? That is an oxymoron like Military Intelligence... > > Leave it to MS to make the simplest, most obvious use of a > device impossible.
Actually, there's a simple solution. Format the CF card on a pc and create a single binary file which fills the entire capacity of the card. Your micro simply has to skip over the the file system information and then is free to write whatever it wants to to the bytes constituting the actual contents of the file. You host application simply opens the file representing the whole card (less filesystem headers) and accesses it according to whatever organizational model you chose. Not quite as simple as opening the device file representing the card itself under linux, but once you figure out how to skip past the filesystem tables not much different in actual practice. To verify that everything is where you think it is, write a PC program that creates the filler file and fills it with incrementing numbers, then use a dump program under either linux or on your micro to examine the card and make sure the data is where you think it is (and where the filsystem tables aren't).
Grant Edwards wrote:
> On 2005-06-21, Paul Carpenter <paul$@pcserviceselectronics.co.uk> wrote: > > >>You want a USB CF card to be WinXP compliant *without* a file >>system? That is an oxymoron like Military Intelligence... > > > Leave it to MS to make the simplest, most obvious use of a > device impossible.
Does Win XP provide any way to read/write the raw sectors of the device? If not, that's the real Microsoft fsk.