Problems with terminal

Ken Peterson

I'm having a bugger of a time trying to communicate with the propeller over the P30/P31 serial port on USB. Is there a trick to it?· I'm using HyperTerminal in Windows.

Also, just a suggestion for the next version of Propeller Editor/Developer.· Why not add a menu item to open a terminal window on the same USB port?· Wouldn't that be convenient?

Ken

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Paul Baker

The next IDE version will have a terminal, and in anticipation of your next question·I don't have a timeframe for when it will be ready. In the meantime try using Ariba's PropTerminal, most people feel it's much better than HyperTerminal.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

parsko

Ken,

Not sure exactly what you are doing, but the Parallax-DAQ has been working pretty good for me. It's an excel spreadsheet with macros that Parallax set-up. It really is a lifesaver for me who knows too little about VB. If you're just showing values, it may not be so good though... But, if you have a stream of data, it's awesome!

-Luke

Ken Peterson

I'm just trying to get something that will work with the Monitor Demo. So far unsuccessful with any program. I get garbage going down, but nothing coming up. I double checked the baud rates. I think something's mis-configured but for the life of me I can't figure out what it is. I'm sure it's not a hardware problem because I'm having no problems putting code onto the chip.

I have a Rogue Robotics MP3 player that I'm controlling through the serial port on a different set of pins, and that works! By the way, does anyone know of any similar products out there?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Kit Morton

The only thing that I know of that is similar is the MOD-MP3 from Olimex. www.sparkfun.com/commerce/product_info.php?products_id=7832

Or you could just use the base chip for both those board on a custom board.
www.vlsi.fi/vs1033/vs1033.shtml

But I don't know of any other mp3 decoder chips or boards.

-Kit

Paul Baker

Without seeing your code it's difficult to determine what you are doing wrong, a common mistake is not using (or setting up properly) an external crystal. The internal RC oscillator is not accurate enough for serial communications.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

mcstar

Micronas makes one, from their PDF...
http://www.datasheetcatalog.com/datasheets_pdf/M/A/S/3/MAS35X9F.shtml

The MAS 35x9F is a single-chip, low-power MPEG
layer 2/3 and MPEG2-AAC audio stereo decoder. It
also contains the G.729 Annex A speech compression
and decompression technology for use in memorybased
or broadcast applications. Additional functionality
is achievable via download software (e.g., CELP
voice decoder, Micronas SC4 (ADPCM) encoder/
decoder).

I found this chip in my Cybiko Xtreme MP3 player. It's laying in my parts bin just begging to make it work with the propeller! Since the company folded and the batteries are dead, parts are all the Cybiko is good for anymore.

Ken Peterson

Paul,

You can see the code. It's the Monitor Demo that comes with the prop. Only thing I modified is to try different baud rates. All I wanted to do is verify that the com port works before using it to debug my software, but I didn't get past that.

Ken

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

rjo_

Ken,

First of all... I'm a mac guy. We don't have terminals on macs... no serial ports... no need for terminals. (Ok... we got them... but nobody uses them and you have to search for them to find them). So, it was a real treat to find something on the Windows side that isn't readily available on a Mac... usually it is the other way around.

I had a similar problem... my terminal program was working. I didn't use it for a week or two... forgot what I was doing and didn't know what I forgot.

First, put everything in your code back to the way it was... use multiple Ctrl-z to do your multiple undo-s.

Then see if it works... remember to load to your code into your eeprom... so you can run it automatically on reset.

Then start the terminal.

If it doesn't work... you have a setting problem.

There is a nice ... fully illustrated, step-by-step description for setting up the terminal in one of the first Propeller Education Kit labs... You probably have a setting wrong...

Remember to "echo" to the screen. That's a setting somewhere, which I forgot to set... and now that I have it set right, can't remember where it is[noparse]:)[/noparse]


Rich

Paul Baker

Ken, can you get it to work with the default baud rate?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Ken Peterson

Not yet. I might play with it a little over the weekend, but for now I'm too busy with work. I just tried the Monitor Demo out of the box and tried using it with HyperTerminal, PropTerminal, and some Java terminal program that I downloaded. No luck with any of them.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

rjo_

Ken,

You also have to be using the same COM port that the Spin Tool reports as it is downloading your program to eeprom of your Propeller (hit the F11 key to download to the eeprom).

Work? I didn't know anyone worked anymore[noparse]:)[/noparse]

By the way... the Propeller is addictive and I am happy to report that Suboxone is totally ineffective against it.

Rich

Ken Peterson

Thanks, Rich.

Yeah, I checked the com port. I chose "Identify Hardware" F7 to see what port the Spin Tool is using. So it's not that, but it still might be something obvious!

Ken

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Ken Peterson

Another problem I'm having...when I use Full Duplex Serial, I can get my mp3 player to work, but when I use Simple Serial, I get garbage. All I did is substitute one for the other, remove the mode parameter, and change all rxcheck calls to rx. I assume both are using 8N1, and I had them both set for 9600 baud. Any ideas here?

Example: I send the "version " command which is V<cr> on the tx line to the mp3 player, then I copy the response from the rx line character by character to the display using tv_text.out(c). With Full Duplex Serial, I get the proper response "110.10 SN:UMM1-0000-0001" but if I use simple serial, I get garbage on the screen instead.

some code:

CON
_clkmode = xtal1 + pll16x
_xinfreq = 5_000_000

tx_pin = 0
rx_pin = 1
prompt = $3E
CR = $0D

OBJ
serial_u : "simple_serial"
tv_text : "tv_text"

PUB start | c

serial_u.start(rx_pin, tx_pin, 9600)
tv_text.start(12)

serial_u.str(@version)
repeat
c := serial_u.rx
if c <> -1
tv_text.out(c)
until c == prompt

DAT

version byte "V", CR, 0

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Mike Green

FullDuplexSerial is buffered and the receive (and transmit) routine runs in its own cog so it won't miss any characters on receive. SimpleSerial is not buffered although it can be made into a buffered receive routine. In processing the characters for the display, it's probably dropping a bit or two and getting out of sync with the serial stream so you're receiving garbage because of the dropped bits.

Ken Peterson

hey...that's a good point. Thanks, Mike. So I should do nothing but read the bytes and buffer them until I get back to my prompt before I try to do anything else with them. I guess I sort of assumed that at 9600 baud I wouldn't fall behind that much.

It's a simple command / response interface with no unsolicited messages from the player, so I hope I can make this work somehow without having to use a cog.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
----


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Mightor

Ken,

If you're looking to use your serial debugging from multiple cogs/objects, I would suggest SerialMirror. I am using it now on my project and it's great. Basically it is a modified FullDuplexSerial object, so most of the functions will be the same. The comments contain a little example of how you could use it.
This is the documentation section from it:

SerialMirror.spin : V07.05.10
              
Derived from: Full-Duplex Serial Driver v1.1  
              (C) 2006 Parallax, Inc.

It has been extended in the following ways:
1) Make allowance for different buffer sizes. I've set the default buffer to
   64 bytes as this is large enough to contain a complete incoming/outgoing
   message for my protocol.
2) Check for a stop bit!! If a stop bit is not present then the received byte
   is dropped. No flags of any sort are set. It's up to a higher level of the
   protocol to check the message for integrity.
3) Variables have been moved to DAT space.
   The significance of this is not obvious at first, but what it means is that
   every instance of this object shares the same variable space. Why do this?
   If you have multiple nested objects, then serial comms can be added to all
   of them. ONLY the top object must call the start method. The nested objects
   then have full comms access. A simple example is:
   ┌────────────────────────────────────────────────────────────────────┐
   │  APP_SerialMirror.spin                                             │
   ├────────────────────────────────────────────────────────────────────┤     
   │CON                                                                 │
   │  _clkmode      = xtal1 + pll16x                                    │
   │  _xinfreq      = 5_000_000                                         │
   │                                                                    │
   │  BaudRate      = 38400         'Baud                               │
   │  RxPin         = 31            'For RS232                          │
   │  TxPin         = 30            'For RS232                          │
   │  CommsMode     = %0000         'See SerialMirror.spin for meaning  │
   │                                                                    │
   │OBJ                                                                 │
   │  Debug : "SerialMirror"                                            │
   │  Test : "TestObject"                                               │
   │                                                                    │
   │PUB Start                                                           │
   │  Debug.start(RxPin, TxPin, CommsMode, BaudRate)                    │
   │  Test.SayHello                                                     │
   └────────────────────────────────────────────────────────────────────┘ 
   ┌────────────────────────────────────────────────────────────────────┐
   │  TestObject.spin                                                   │
   ├────────────────────────────────────────────────────────────────────┤
   │OBJ                                                                 │
   │  Debug : "SerialMirror"                                            │
   │                                                                    │
   │PUB SayHello                                                        │
   │  Debug.str(string("Hello",13,10))                                  │
   └────────────────────────────────────────────────────────────────────┘
4) This object is NOT suitable for driving multiple serial ports.
   If you want multiple serial ports each with single object access, then
   use FullDuplexSerial.
   If you want serial comms through the same port from multiple objects,
   then use THIS object.
   If you want multiple serial ports each with multi-object usage, then copy
   and rename this files for the number of unique serial ports.
5) Added GetNumber which will receive a number (positive or negative) from
   the serial port in either decimal, binary or hexadecimal notation.
6) Added extra comments to the assembler section - for the newbies. I've been
   doing this propeller stuff for a month now, so I guess that makes me an
   old hand.
7) Added CrLf which does just that - purely a conveniance function for me.

If you have questions, don't hesitate to ask. SerialMirror can be downloaded from here

Gr,
Mightor

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
| To know recursion, you must first know recursion.

Post Edited (Mightor) : 8/18/2007 6:07:29 AM GMT

Ken Peterson

Thanks for the post, Mightor. I think I might be able to make use of SerialMirror. I have a project that I'm working on that would have several threads potentially sending over the same serial port, and I was wondering how I was going to do that. The fact that the dat section allows me to define shared variable space is also a very useful tip!

Ken

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Ken Peterson

OK...I've had a chance to work on this serial thing some more. I'm using Simple Serial and I'm reading a stream of characters from a MP3 device at 9600 baud.· Here's my code fragment.· "serial_u" is my name for the simple serial object.


PUB

· ...
· i := 0
· serial_u.str(@version)
· repeat
· until (inbuf[noparse][[/noparse]i++] := serial_u.rx) == ">"
· inbuf[noparse][[/noparse]i-1] := 0
· tv_text.str(@inbuf)
· ...

DAT

······· version········ byte··· "V", CR, 0
I intend to keep reading into a buffer until I get a ">" character.· I get the first 2 or 3 characters OK, but the rest is garbage.·· Is there a quicker way to read a string of characters from·Simple_Serial in SPIN without dropping bits (I assume that's what's happening)?

Ken

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Ken Peterson

Has anyone tried to double up FullDuplexSerial to run two serial ports with one cog? I guess if I could do that it would help with my problem too.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


The more I know, the more I know I don't know.· Is this what they call Wisdom?

Peter Jakacki

Ken,

Just a thought, what does your board use for a serial connection? If you are trying to use the simple transistor circuit then it won't work with standard serial coms programs as the DTR is taken positive by coms programs. There is a very simple fix for this.

For terminal software I always use Teraterm as it supports baud rates up to 921600, com ports up to, and also TCP/IP connections. It also includes a debugging mode and has very high throughput (i.e. it keeps up with the traffic).

*Peter*

Wayne Bartnick

Ken Peterson said...
Has anyone tried to double up FullDuplexSerial to run two serial ports with one cog? I guess if I could do that it would help with my problem too.

EDIT: I didn't read the original question!!! The answer is 'NO' -- I used 3 cogs for 3 RS232 channels...

Yes, I've successfully used 3 instances of a FullDuplexSerial. Well, technically, one instance of the FullDuplexSerial object (for debugging on pins 30/31 via my Prop Plug) and 2 instances of the Extended_FDSerial (on pins 12/13 and 14/15 via a MAX232).

I am running all 3 RS232 ports @ 9600N81.

I will say that I have made the very same mistake that Paul cautions against below [noparse][[/noparse]er, above]:

Paul Baker (Parallax) said...
...a common mistake is not using (or setting up properly) an external crystal.

But I see from your posted snippet that you are setting the clock properly.

And if all else fails, reboot your PC.

-Wayne

Post Edited (Wayne Bartnick) : 8/22/2007 10:51:27 PM GMT

deSilva

Ken Peterson said...
Has anyone tried to double up FullDuplexSerial to run two serial ports with one cog?

This would be an "anti Propeller task" As I said the other day: Everything is simpel when you have a free COG; most things are a nightmare when you have not.

Do you understand the ping-pong multitasking in the machine code part ...*shudder*

Well it IS possible to extend it to another pair of pins... max bit rate will be cut by halves of course.

Ken Peterson

geez....I know what you mean about the ping-pong part, deSilva. That's why I didn't take on the task right off the bat. Hoping someone else has already solved the problem [noparse]:)[/noparse] If I have to, I'll take a crack at it myself!

Cheers!

BTW: Is anyone using the propeller in mass production yet? I'm looking at proposing to use it for an automotive product, but everyone I work with is wide eyed and shaking their heads!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


The more I know, the more I know I don't know.· Is this what they call Wisdom?

deSilva

Ken Peterson said...
...but everyone I work with is wide eyed and shaking their heads!

A very good description of the situation :-(

Peter Jakacki

By mass production you are talking 10,000s to millions I guess so the answer for me is no at present but this I can say about the Propeller, in this context I have found it to be a hardy and reliable chip. All those errata that we are used to with other chips doesn't really seem to exist here. Usually all those fancy peripheral functions on other chips will have little quirks (that can turn out catastrophic) but the prop's symmetrical multi-processor approach is far more universal and easy to "fix" in software. I know some have zapped the pll but that seems to be on improperly designed layouts for inductive loads (you wouldn't do that!). Most chips would and do just die completely.

Just interested, what would you define as the minimum requirements of a CPU for automotive product?

*Peter*

deSilva

To my knowledge there are three:
- robustness (especially temperature, not fomally but factually better than MIL!!)
- low price
- existing know-how in engeneering staff

The last one is a vicious circle...

Paul Baker

I personally know of several commercial applications the Propeller is in, many companies like to keep a low profile about what technology they use in thier products, so we don't go around saying who they are.

deSilva, what do you mean by formal? Running the ESC demo from cold start at -73 C and 190 C isn't enough formal testing to prove the chip operates well beyond MIL spec temps? There is no other real world application that better tests the capabilities of the chip than the demo: hub access, spin execution, assembly execution, counters, PLLs, video generators are all being excercised in the chip. If what you are looking for is to see in in print, just wait a little while until I rev the datasheet from preliminary to non-preliminary, but I can tell you now it will be rated for military temp.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Ken Peterson

deSilva,

Add this to your list:

- history

Nobody in Automotive trusts anything unless it's been done before. Somebody has to be first, but no one seems to want to be!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


The more I know, the more I know I don't know.· Is this what they call Wisdom?

deSilva

@Ken, yes, that's what I expressed badly.
They will or will not require the parts fulfil MIL standards. What they want is that the parts do better, known from their own long time experience!

You see a car is a much worse environment than an aircraft or any "military" item can be. Aircrafts are very regularly maintained. The electronic parts are held in compartments luxuriously climatized by oversized fans. Most or even all are double or tripple redundant. Pilots will never really stress their aircrafts for fear of loosing their licence.