Is anyone running a USB Tera drive directly using a Propeller? These drives are dropping in price and to have a thousand Megabytes which can be used as nonvolatile memory and storage is a great asset.
I would not use USB for this. ATA is a lot simpler, beside most USB HDDs are nothing more than an ATA hooked to a usb capable microcontroler, all stuffed in a nice little case.
And heater:
I hate these new standards, that round to the nearest power of 10, that is wrong. The standard has always been powers of 2 (1KB=2^10, 1MB=2^20, 1GB=2^30, 1TB=2^40, 1EB=2^50, etc...). And what is this TebiByte stuff that you speak of? Does this refer to the new standard of rounding to a power of ten?
Phil:
I refer to pATA (I do not think the Prop can do sATA). done correctly it takes 12 pins on the Propeller plus four 74HC244, or four 74HCT244 buffers. A little later I will try to dig out the schematic. Plus two latches.
Phil:
I refer to pATA (I do not think the Prop can do sATA). done correctly it takes 12 pins on the Propeller plus four 74HC244, or four 74HCT244 buffers. A little later I will try to dig out the schematic. Plus two latches.
Hey ... that's cool. I would REALLY like to see how you hook up those ATA drives to prop. I have a number of old 40GB HDDs I would like to make use of them in the near future and .... codes
Sorry Humaniod, saw his thread & couldn't resist ...
Err, no, a tera byte is 1024 giga bytes. A giga byte is 1024 mega bytes. A mega byte is 1024 kilo bytes.
See here for correct standard units http://en.wikipedia.org/wiki/Terabyte (I hate that stupid IEC kibibyte, mebibyte, tibibyte stuff)... how long will it take to read all of a 1.5 TB drive at the 1.43MB/S of USB 2?
Have you seen any drive not containing some driver that leaves the full amount of 1024 on the drive? I have not. Therefore I was simply approximating for the space consumed by the driver and what might be left as space on the hard drive. A TB drive no longer holds a full TB.
The reason I mention a USB drive is because they're readily available in this neck of the woods. However, if ATA would work better and has a Prop interface, that's good too. Where can I find sources that show how to connect ATA drives to a Propeller chip?
I also doubt any application that I have in mind, other than a possible one time formatting, would continuously read and write the drive sectors from start to finish and fill the drive. Rather, the app would more likely save many small files, over a short period of random time, not in any lengthy continuous fashion.
Hey ... that's cool. I would REALLY like to see how you hook up those ATA drives to prop. I have a number of old 40GB HDDs I would like to make use of them in the near future and .... codes
Sorry Humaniod, saw his thread & couldn't resist ...
MacTuxLin, no problem, join right in.. and I have the same question..
I cannot fill a 2GB microSD card for prop code so why on earth would I want 1TB. Besides the fact the microSD is simpler and uses less pins and is already done well.
However, for backups and TV recording, I bought a pair of 320GB USB (Western Digital Elements) drives for A$48 each retail. Not much can beat that. IIRC a 1TB was only A$79 for USB2.0 & A$89 for USB3.0. Currently A$1~US$1. Makes my 1976 10MB washing machine sized disk drives for $16,000 each sound a touch expensive - I had 3 of them on my mini - ouch!!!
Based on this, I expect 1EB (ExaByte) disks to be free in Cereal packs in less than 30 years
Now the question about how long would it take to read a tera byte was not meant to imply that one would want to do that on the Prop. It's just hinting that the thing is so big and slow that it is kind of pointless.
As it stands the Prop can perform as a USB with a 12Mb/s data rate. Or 1.43MB/s. Probably a lot less once all the protocol overheads are added on. So I reckon on it taking 9 days to get from one end to the other!
Not to mention that to drive the USB sucks up 4 COGs so there is not much Prop left to work with all that data.
ATA is interesting but then that eats a lot of pins which makes the Prop a less than ideal platform for it.
It is taking quite a bit of time to Draw my paper schematics into a BMP file. So I will give the functional logic here.
I am not listing the level conversions (3.3V to 5V and 5V to 3.3V), though ATA is 5V.
ATA D0-D8 to Latch0 in and latch1 out.
ATA D8-D15 to Prop P8 - P15.
ATA /IORD to Prop P20.
ATA /IOWR to Prop P21
ATA /A0-A2 to Prop P16-P18
Latch0 LE = A3 AND NOT IORD.
Latch0 /OE = NOT (NOT A3 AND NOT IORD).
Latch1 LE = NOT A3 AND NOT IOWR.
Latch1 /OE = NOT A3 AND IORD.
ATA CS1 = ATA CS2 = NOT ((NOT IORD OR NOT IOWR) AND A3).
ATA IORDY to +5V through 22K ohm.
ATA DMARQ to GND through 22K ohm.
ATA DASP to +5V through diode and 1Kohm.
ATA RST to Prop P22.
I hope this helps until I can finish copying the schematic. They are always faster to draw on paper.
I was not specific in low verses high active states in the above. IORD, IOWR, CS1, CS2 are active low.
I think I corrected the errors in the above, sorry about that.
Check some of the recent threads regarding free/cheap schematic drawing software. It will save you gobs of time, and the results will look much better than what you can achieve with a paint program.
Thank you Phil. I think I will just install DIA on the Linux side and use that. The options mentioned in those threads are not free. One seems to be a flow chart only app, and DesignSpark is not free, It is crypleware (you must register the product to save).
Diptrace works very well under Wine.
It is free for non professional users up to a rathre high number of pins, and gadget gangster is offering for free many project files. Parallax itself is going to release diptrace files of ther open sourced projects.
I use Diptrace both in my ubuntu netbook and in my win7 pc, and using dropbox I work on the same files and library on both. Up to now with no issues.
Would every one be happy with a list of connections? I usually do not attempt to draw a schematic into a file, and It is taking way longer than expected. I could probably create a text file of the connections fairly quickly, if that would be good enough. There is a reason I do my schematics on Paper (probably should get a scanner). I wasted almost all of yesterday on this (I made a couple of errors that I had to correct, while copying it in), and still have a bit to do.
Yesterday eve I tried a couple of the schematic tools. Some of them I could not create a custom Part, and the one I could (Fritzing), I created the needed 74HCT00 and when I tried to connect to pin 2 on the 7400 it connected to pin 14 instead. I could not find most of the basic gates in the library for any of them.
Wikipedia has a description of the ATA interface. This would need to be multiplexed down to fewer pins for the Prop. David, maybe all you need to do is describe how this would work. Your pin list from a previous post suggest that you would multiplex two ATA pins onto one Prop pin. A simple block diagram is probably all you need. You could do this in the Prop tool using the special graphics characters.
Dave Hein:
Thank you. And even with out multiplexing you only need 23-pins, 16 Data, 4 control, and 3 address. My design simply multiplexes ATA-Data0-7 with ATA-Data8-15, by using 2 latches. How do you get the Parallax custom font on BST-Linux? I tried Windows for a couple days, and quickly remembered why I do not use Windows. And for a better view of the ATA Registers see http://wiki.osdev.org/Category:ATA
I run BST on Windows without any problems. I don't know how to get the Parallax font onto Linux. Maybe one of the Linux gurus can provide assistance.
You could always use the lowest common denominator -- ASCII graphics! Check this site out for some interesting ASCII animation -- http://www.asciimation.co.nz/ .
EDIT: Use the ">>" or ">>>" buttons to watch in fast forward in you don't want to spend a lot of time watching it.
or in your desktop environment (gnome or kde, for others I do not know if they provide this feature) right click on the ttf font, it should give you the possibilty to show it with the font viewer (do not know the exact english name, my linux is set to german) and install it with that.
Problem is that once you have installed the Parallax font into Linux you find out it looks horrible both in BST and PZST. I can't stand to work with it.
Problem is that once you have installed the Parallax font into Linux you find out it looks horrible both in BST and PZST. I can't stand to work with it.
Is font smoothing (subpixel hinting) enabled to improve the appearance of the vector fonts?
Comments
See here for correct standard units http://en.wikipedia.org/wiki/Terabyte (I hate that stupid IEC kibibyte, mebibyte, tibibyte stuff).
Now, how long will it take to read all of a 1.5 TB drive at the 1.43MB/S of USB 2?
And heater:
I hate these new standards, that round to the nearest power of 10, that is wrong. The standard has always been powers of 2 (1KB=2^10, 1MB=2^20, 1GB=2^30, 1TB=2^40, 1EB=2^50, etc...). And what is this TebiByte stuff that you speak of? Does this refer to the new standard of rounding to a power of ten?
-Phil
I refer to pATA (I do not think the Prop can do sATA). done correctly it takes 12 pins on the Propeller plus four 74HC244, or four 74HCT244 buffers. A little later I will try to dig out the schematic. Plus two latches.
Hey ... that's cool. I would REALLY like to see how you hook up those ATA drives to prop. I have a number of old 40GB HDDs I would like to make use of them in the near future
Sorry Humaniod, saw his thread & couldn't resist ...
Have you seen any drive not containing some driver that leaves the full amount of 1024 on the drive? I have not. Therefore I was simply approximating for the space consumed by the driver and what might be left as space on the hard drive. A TB drive no longer holds a full TB.
The reason I mention a USB drive is because they're readily available in this neck of the woods. However, if ATA would work better and has a Prop interface, that's good too. Where can I find sources that show how to connect ATA drives to a Propeller chip?
I also doubt any application that I have in mind, other than a possible one time formatting, would continuously read and write the drive sectors from start to finish and fill the drive. Rather, the app would more likely save many small files, over a short period of random time, not in any lengthy continuous fashion.
However, for backups and TV recording, I bought a pair of 320GB USB (Western Digital Elements) drives for A$48 each retail. Not much can beat that. IIRC a 1TB was only A$79 for USB2.0 & A$89 for USB3.0. Currently A$1~US$1. Makes my 1976 10MB washing machine sized disk drives for $16,000 each sound a touch expensive - I had 3 of them on my mini - ouch!!!
Based on this, I expect 1EB (ExaByte) disks to be free in Cereal packs in less than 30 years
Oh yes, hard drive manufacturers are famous for twisting around with the units they use to make their products look more capacious.
The TearDrive seems to be a case in point. If you mean this one: http://www.supertalent.com/products/ssd_category_detail.php?type=TeraDrive. It's called a "tera" drive but the biggest one is only 480GB.
Now the question about how long would it take to read a tera byte was not meant to imply that one would want to do that on the Prop. It's just hinting that the thing is so big and slow that it is kind of pointless.
As it stands the Prop can perform as a USB with a 12Mb/s data rate. Or 1.43MB/s. Probably a lot less once all the protocol overheads are added on. So I reckon on it taking 9 days to get from one end to the other!
Not to mention that to drive the USB sucks up 4 COGs so there is not much Prop left to work with all that data.
ATA is interesting but then that eats a lot of pins which makes the Prop a less than ideal platform for it.
Give me a couple hours. I have the schematics in front of me, and am working on drawing them into a BMP. Right now they are on paper.
I am not listing the level conversions (3.3V to 5V and 5V to 3.3V), though ATA is 5V.
ATA D0-D8 to Latch0 in and latch1 out.
ATA D8-D15 to Prop P8 - P15.
ATA /IORD to Prop P20.
ATA /IOWR to Prop P21
ATA /A0-A2 to Prop P16-P18
Latch0 LE = A3 AND NOT IORD.
Latch0 /OE = NOT (NOT A3 AND NOT IORD).
Latch1 LE = NOT A3 AND NOT IOWR.
Latch1 /OE = NOT A3 AND IORD.
ATA CS1 = ATA CS2 = NOT ((NOT IORD OR NOT IOWR) AND A3).
ATA IORDY to +5V through 22K ohm.
ATA DMARQ to GND through 22K ohm.
ATA DASP to +5V through diode and 1Kohm.
ATA RST to Prop P22.
I hope this helps until I can finish copying the schematic. They are always faster to draw on paper.
I was not specific in low verses high active states in the above. IORD, IOWR, CS1, CS2 are active low.
I think I corrected the errors in the above, sorry about that.
-Phil
Thank you. Do you have links to any of these threads?
http://forums.parallax.com/showthread.php?127243
http://forums.parallax.com/showthread.php?126643
http://forums.parallax.com/showthread.php?107645
http://forums.parallax.com/showthread.php?110797
http://forums.parallax.com/showthread.php?109285
It is free for non professional users up to a rathre high number of pins, and gadget gangster is offering for free many project files. Parallax itself is going to release diptrace files of ther open sourced projects.
I use Diptrace both in my ubuntu netbook and in my win7 pc, and using dropbox I work on the same files and library on both. Up to now with no issues.
Massimo
-Phil
Thank you. And even with out multiplexing you only need 23-pins, 16 Data, 4 control, and 3 address. My design simply multiplexes ATA-Data0-7 with ATA-Data8-15, by using 2 latches. How do you get the Parallax custom font on BST-Linux? I tried Windows for a couple days, and quickly remembered why I do not use Windows. And for a better view of the ATA Registers see http://wiki.osdev.org/Category:ATA
You could always use the lowest common denominator -- ASCII graphics! Check this site out for some interesting ASCII animation -- http://www.asciimation.co.nz/ .
EDIT: Use the ">>" or ">>>" buttons to watch in fast forward in you don't want to spend a lot of time watching it.
sudo mv ~/Desktop/parallax.ttf /usr/share/fonts/truetype/
sudo fc-cache -f -v
or in your desktop environment (gnome or kde, for others I do not know if they provide this feature) right click on the ttf font, it should give you the possibilty to show it with the font viewer (do not know the exact english name, my linux is set to german) and install it with that.