uart speed

[jyb]

hi all
i need help
used picaxe is 40x2 3.3v
i need two uart port running a 115200 bds and i planned to use hserout/hserin commands for one link and serin/serout for the other one link
one link for a gyro/accelerometer other link for xbeepro
the problem i get is :
hserout/hserin commands and serout/serin commands doesnot deliver same speed and are incompatible
gyro/accelero and xbee need 115200bauds speed each but
the giro/accelero's 115200bd speed is unmodifiable
hserin/hserout commands can speed up to 115200 bd max will be for it

xbeepro's speed is modifiable but it don'offer slower speed compatible
with serin/serout commands which can speed up to 76800bd max

some idea what i have to do ?

thanks by advance

 

[hippy]

There are a number of possibilities but a lot depends upon exactly how the links need to be used.

If you are only receiving from the gyro / accelerometer and only sending to the XBee with the same baud rate you can use the HSERIN for one HSEROUT for the other.

You can also operate the PICAXE at a speed which allows 115200 baud for the gyro / accelerometer and also allows SERIN / SEROUT to provide a baud rate compatible with XBee; the XBee isn't locked to 115200 baud.

 

[ciseco]

Don't know if it will be of interest, it might be worth considering, an XRF (will sit in same xbee socket) supports all common serial speeds between 1200 to 115K. The latency is lower than the XBee , so you could soft serial into this at a lower speed yet it "appear" to be running faster than the XBee. The overhead of constructing/de and transmitting an 802.15.4 packet means even if you send just 1 byte, what actually goes over wireless is many many 10's of bytes.

For really low latency it has a useful feature, you can specify a packet length to commit before a time-out happens (the xbee usually waits for 3 x bytes time before transmitting).

What do I mean? An Xbee will sit and wait until 3 blank bytes are seen before committing the packet. If your packet is 6 bytes then in simplistic maths the xbee will be sitting around doing nothing for 33% of the time.

Here's an example on an XRF using aProtocol (a design for wireless networks) the packet length is always 12 bytes so we set commit after 12. This then doesn't waste any time and transmits at the point the last bit of byte 12 was received. You can set the number of bytes (3 to 255) at which it forces a transmit without waiting for an "end" period. Any data sent after this point is buffered and either gets transmitted in another x byte chunk or is committed after a time-out. What you see the other side is usually contiguous data as the over the air rate is much higher than the serial rate.

Just a thought.

Miles

 

[jyb]

it is what i try

There are a number of possibilities but a lot depends upon exactly how the links need to be used.

If you are only receiving from the gyro / accelerometer and only sending to the XBee with the same baud rate you can use the HSERIN for one HSEROUT for the other.

You can also operate the PICAXE at a speed which allows 115200 baud for the gyro / accelerometer and also allows SERIN / SEROUT to provide a baud rate compatible with XBee; the XBee isn't locked to 115200 baud.

yes hippy it was what i try but the higher speed common to the picaxe and xbee is 38400 which is less what i need because too much latency,upper this speed the possible other speed are different

on other side the picaxe is receiving data from giro/accelero and receiving data from xbee (itself receiving from main command panel)

 

[jyb]

XRF ? what that's ?

Don't know if it will be of interest, it might be worth considering, an XRF (will sit in same xbee socket) supports all common serial speeds between 1200 to 115K. The latency is lower than the XBee , so you could soft serial into this at a lower speed yet it "appear" to be running faster than the XBee. The overhead of constructing/de and transmitting an 802.15.4 packet means even if you send just 1 byte, what actually goes over wireless is many many 10's of bytes.

For really low latency it has a useful feature, you can specify a packet length to commit before a time-out happens (the xbee usually waits for 3 x bytes time before transmitting).

What do I mean? An Xbee will sit and wait until 3 blank bytes are seen before committing the packet. If your packet is 6 bytes then in simplistic maths the xbee will be sitting around doing nothing for 33% of the time.

Here's an example on an XRF using aProtocol (a design for wireless networks) the packet length is always 12 bytes so we set commit after 12. This then doesn't waste any time and transmits at the point the last bit of byte 12 was received. You can set the number of bytes (3 to 255) at which it forces a transmit without waiting for an "end" period. Any data sent after this point is buffered and either gets transmitted in another x byte chunk or is committed after a time-out. What you see the other side is usually contiguous data as the over the air rate is much higher than the serial rate.

Just a thought.

Miles

perhaps a good track to follow ,could you tell me more about xrf, reference and dealer shop ? possible compatibility with 2.4 Ghz xbee ?


thanks a lot

 

[hippy]

For maximum throughput one option may be to use two PICAXE ...

From Gyro -> (HSERIN) PICAXE (HSEROUT) -> To XBee

From Xbee -> (HSERIN) PICAXE (HSEROUT) -> To Gyro

and add an extra link between the two PICAXE. Or have each PICAXE handle Gyro or XBee and create a single or dual high-speed parallel link between the two.

With a single PICAXE you will be limited to 115200 baud via high-speed serial and 76800 baud maximum via SERIN / SEROUT.

Without a full specification of data transfer, maximum latency required, quantity and frequency of data, and interaction between the parts it's not possible to really give other than generic advice.

 

[ciseco]

Hi,

The data sheet for the XRF is nearly ready, we are taking back orders at the moment. It runs at 868.3Mhz so wont talk to an xbee. Unusually you can set 315/433/915Mhz in software (and it'll work) As the circuit is tuned for 868Mhz the range at these settings is much shorter than the 1/3km+ we have tested the 868.3 to, we will release 315/433/915 versions in the new year. When set to 433 (in very limited tests) the 868 still outperforms an XBee standard for range.

Basic spec is, around half the power of xbee, same physical size, 3 x distance of S1 XBee, half the cost. We are just putting the finishing touches to the firmware updater. You can re flash the firmware with plain old two wire serial, no flow control needed like the XBee does, you will be able to re-flash using a picaxe cable and a resistor or two.

I suspect (but haven't tried yet) it'll work on an AXE210 very well and allow for re-flashing of the firmware without any mods. Wireless PICAXE programming is pencilled in for Q1 next year, we are just working on arduino over the air programming first. The new features will be available as free downloads as they become available.

Full details are just days away. As to dealers, nothing in France yet. If you put XRF into ebay you'll find more info.

Regards

Miles

 

[jyb]

Hi,

The data sheet for the XRF is nearly ready, we are taking back orders at the moment. It runs at 868.3Mhz so wont talk to an xbee. Unusually you can set 315/433/915Mhz in software (and it'll work) As the circuit is tuned for 868Mhz the range at these settings is much shorter than the 1/3km+ we have tested the 868.3 to, we will release 315/433/915 versions in the new year. When set to 433 (in very limited tests) the 868 still outperforms an XBee standard for range.

Basic spec is, around half the power of xbee, same physical size, 3 x distance of S1 XBee, half the cost. We are just putting the finishing touches to the firmware updater. You can re flash the firmware with plain old two wire serial, no flow control needed like the XBee does, you will be able to re-flash using a picaxe cable and a resistor or two.

I suspect (but haven't tried yet) it'll work on an AXE210 very well and allow for re-flashing of the firmware without any mods. Wireless PICAXE programming is pencilled in for Q1 next year, we are just working on arduino over the air programming first. The new features will be available as free downloads as they become available.

Full details are just days away. As to dealers, nothing in France yet. If you put XRF into ebay you'll find more info.

Regards

Miles

hi ciseco
i found this on a french dealer web site
http://www.matlog.com/wireless/modeles-digi-maxstream/modules-de-communication-xbee-pro-868mhz.html
is it what you tell about ?

 

[ciseco]

Hi,

No that's a different type of XBee by Maxstream, it uses an Atmel chipset, it does run at 868Mhz, this is the only similarity. The XRF is by Ciseco and uses a Texas Instruments chipset.

Here in the UK an 868 XBeePRO is from Farnell £60.29 + VAT, an XRF is £9.90 + VAT

They are quite different products. The TI datasheet for the IC on the XRF is here http://focus.ti.com/docs/prod/folders/print/cc1110f32.html you'll be able to compare technical details.

It's the software running on the XRF which makes the CC1110 work in a way that's similar to an XBee, it is NOT an XBee. There are many features you will not find in an XBee.

Regards

Miles

 

[jyb]

did i misunderstood ?

Hi,

No that's a different type of XBee by Maxstream, it uses an Atmel chipset, it does run at 868Mhz, this is the only similarity. The XRF is by Ciseco and uses a Texas Instruments chipset.

Here in the UK an 868 XBeePRO is from Farnell £60.29 + VAT, an XRF is £9.90 + VAT

They are quite different products. The TI datasheet for the IC on the XRF is here http://focus.ti.com/docs/prod/folders/print/cc1110f32.html you'll be able to compare technical details.

It's the software running on the XRF which makes the CC1110 work in a way that's similar to an XBee, it is NOT an XBee. There are many features you will not find in an XBee.

Regards

Miles

please Miles, would you give the farnell reference for XRF i will order a pair to try on my system
thanks by advance
jyb

 

[ciseco]

Hi,

Sorry, my answer does look like farnell do both, I didn't mean that to happen.

The XRF is so new it's only available on back order at the moment, this is direct from us on ebay. It is shipping in 2-3 weeks from now. After then you will be able to buy from ProtoPIC, Crownhill and HotSolder in the UK as well as us direct. We will soon be looking for more resellers in more countries France being one for sure.

From what I hear Farnell expect to make more more profit than we could offer them without putting the price up a lot. So the smaller hobbyist resellers are the ones we will go to first.

I hope this news isn't too disappointing?

Regards

Miles

 

[jyb]

there i not a matters

Hi,

Sorry, my answer does look like farnell do both, I didn't mean that to happen.

The XRF is so new it's only available on back order at the moment, this is direct from us on ebay. It is shipping in 2-3 weeks from now. After then you will be able to buy from ProtoPIC, Crownhill and HotSolder in the UK as well as us direct. We will soon be looking for more resellers in more countries France being one for sure.

From what I hear Farnell expect to make more more profit than we could offer them without putting the price up a lot. So the smaller hobbyist resellers are the ones we will go to first.

I hope this news isn't too disappointing?

Regards

Miles

i must tell that i am french and there is most english words that i do not understand in their entire meanings
this is the cause i misunderstood sometimes what i am reading

about XRF
i will wait ,overall if this product is foot print compatible,gives same features and performances
£9.9 is really the price?

regards
jyb

 

[ciseco]

Hi,

It will be 2-3 weeks for the first ones to arrive.

Le prix, oui, cest moins cher

Regards

Miles