RE: [UDB-dev] Re: FDTI USB Serial breakout board problem

[Marcus Fahlén]

John,

Do you use a wide, or narrow beam device in your tests?

Sincerely.

Marcus Fahlén



I'm out of office at the moment.

The information contained in this message may be CONFIDENTIAL and is
intended for the addressee only. Any unauthorised use, dissemination of the
information or copying of this message is prohibited. If you are not the
addressee, please notify the sender immediately by return e-mail and delete this message. Thank you.

Sent from my Sony Ericsson XPERIA™.

---

From: John McClelland <mcclell...@gmail.com>
Sent: 7. februar 2011 00:49
To: uavhel...@googlegroups.com
Subject: Fw: [UDB-dev] Re: FDTI USB Serial breakout board problem

 

----- Original Message -----

From:John McClelland

To: Adam Barrow

Cc: John McClelland

Sent: Sunday, February 06, 2011 4:49 PM

Subject: Re: [UDB-dev] Re: FDTI USB Serial breakout board problem

Just wanted to close this out...

 

I added an inverter to the sonar output and hooked up to the RX of the serial port, Vcc, GND and TX from sonar are the 3 wire connection.

 

Hooked FTDI with 3 wire, Vcc, GND and RX from FTDI with RX going to TX of port. 

 

This configuration has no floating pins on the serial port.

 

Still using old MP serialIO code (not updated to Bill's patch.

 

All works fine now.  From this I conclude it was a floating pin issue on the seriral port.  No pullup needed on TX of FTDI.

 

Just FYI, I tried the sonar detector on various reflective surfaces (textured carpet, towels, wicker, irregular plastic) and all returned good data.  I don't think we will have an issue with different surfaces as we use this for hover altitude control on the Quad (and eventually heli).  Open loop tests of the sonar and altitude control code looks good.  Flight tests hopefully in a day or two.

 

Thanks for all the assistance.

 

John

----- Original Message -----

From: Adam Barrow

To: uavdevb...@googlegroups.com

Cc: <uavdevb...@googlegroups.com> ; John McClelland

Sent: Saturday, February 05, 2011 3:53 PM

Subject: Re: [UDB-dev] Re: FDTI USB Serial breakout board problem

John,

 I realize that you already have the patch Bill is referring to, but wanted to point out that it may be the same issue. When we were developing is patch, I tried numerous times to reproduce the error on my UDBs with a three wire connection but wasn't able to. We decide that it was a combination of the open pin and enough EMI to generate noise on the pin. I would imagine that Jerry, like me, is just not triggering the condition rather than being immune to it.

Good luck integrating the sonar with the UDB, I'm hopefully going to be able to start working on my own integration here next week.

Regards,

Adam Barrow

On Feb 4, 2011, at 6:38 PM, "John McClelland" <mcclell...@gmail.com> wrote:

Thanks Bill

 

When you say "the code", you are talking about MatrixPilot FW?   Jerry and I are using the same Quad FW and he is able to make the 3 wire connection to FTDI work, but has a newer  version of the FTDI.   Were you suggesting there is new FW for the FTDI?

 

Best,

John

----- Original Message -----

From: William Premerlani

To: uavdevb...@googlegroups.com

Sent: Friday, February 04, 2011 5:02 PM

Subject: Re: [UDB-dev] Re: FDTI USB Serial breakout board problem

Hi John,

There was a bug in the older versions of the code that required a 4 wire connection between the UDB and an external serial port such as OpenLog or FTDI. If you used a 3 wire connection, noise on the unconnected pin would cause problems in the communications. The problem was eventually fixed so that you could use a 3 wire connection, I am not sure what version that was.

Bill

On Fri, Feb 4, 2011 at 6:51 PM, John McClelland <mcclell...@gmail.com> wrote:

Thanks Paul

The problem was when I removed the Tx line on the FTDI....does that also need to be pulled up?  The Rx line on the FTDI is hooked to the serial Tx.

John
----- Original Message ----- From: "Paul G." <relbo...@hotmail.com>
To: "uavdevboard-dev" <uavdevb...@googlegroups.com>
Sent: Friday, February 04, 2011 4:09 PM
Subject: [UDB-dev] Re: FDTI USB Serial breakout board problem

FTDI's need the input line pulled up or it may think its getting
characters all the time.

On Feb 3, 5:27 pm, "John McClelland" <mcclelland.j...@gmail.com>
wrote:

Maybe someone on can help with a problem I am having. Jerry and I are developing a sonar/baro altitude control routine for our Quads. Interfacing the sonar is through serial, analogue, or PWM. Our configuration requires serial, at least for development. We will go another route long term. This means we need to use the Rx on the serial port for the sonar at the same time we use the Tx to grab data for debugging into the PC through the FDTI serial/USB board. Jerry was able to make this work with a later version of the breakout board. Mine is behaving badly.

I have this FTDI USB/serial board:

http://www.sparkfun.com/products/718

This is the one the wiki tells how to hook up:

http://code.google.com/p/gentlenav/wiki/SerialConnection

I have the jumper open (removed solder blob)

I am supplying power to the UDB thru a servo cable (5V). Vcc on the serial port reads 4.97V.

I hook up GND from the serial port to GND on the FTDI, Tx from the serial to Rx on the FTDI, and serial Vcc to VccIO on the FTDI.

Buad on serial and COM port set to 9600

Only if I connect the Rx from the serial to Tx on FTDI do I get good data out. If I remove the serial Rx connection I get jibberish. But of course I need the Rx for the sonar.

Does the Rx line on the FTDI need to be grounded or pulled up?

Any idea what I am doing wrong?

John

[John McClelland]

It is fairly narrow beam...I haven't fully characterized it, but a couple feet from the sensow pointing down, I start interfering with the reported distance when I get less than a foot from the center line.

 

This is it

 

http://www.sparkfun.com/products/8502

 

 

 

----- Original Message -----

[Marcus Fahlén]

Ok, that's the widest beam available from "MAXBOTICS", probably a good choice when the heli is flying forward or banking not to miscalculate the altitude due to the pitch or roll angle.

I have looked at these for a while with difficulty to make up my mind of which one to choose. I'll guess this one is the most suitable if you don't find out otherwise ...

Cheers!


/Marc

I'm out of office at the moment.

The information contained in this message may be CONFIDENTIAL and is
intended for the addressee only. Any unauthorised use, dissemination of the
information or copying of this message is prohibited. If you are not the
addressee, please notify the sender immediately by return e-mail and delete this message. Thank you.

Sent from my Sony Ericsson XPERIA™.

---

From: John McClelland <mcclell...@gmail.com>

Sent: 7. februar 2011 02:35
To: uavhel...@googlegroups.com

[John McClelland]

We should know shortly how it performs.  The idea is to use this for a hover to landing transition for helis and quads.  Hopefully minimal roll and pitch, although we could correct for it with the DCM if we are confident that the sonar data is reliable.  I'm afraid that we may not be able to rely on sonar return at angles far from normal to the ground due to beam width and reflective geometry.  We will find the overlap reagion where the sonar and baro both work, and tansition back and forth within this band gracefully.  The sonar maxes our at 6.5meters.  We know there are turbulent effects with the baro at low altitudes.

[John McClelland]

I am changing the thread name for this discussion since we have laid the FTDI to rest.

[Marcus Fahlén]

John,

I am awaiting the baro chip from Spark (not the complete breakout assenbly), which I will test with the UDB. I'm also going to order a Maxbotics sonar from them, and the PPM converter together with a magnetometer from DIY Drones.
Hopefully I can afford this in the next few weeks to be able to follow your development.

Did you find out why my re-structured code didn't compile on your computer? I have used it all day,. But must say that it doesn't behave very well. This might be my UDB that's causing it, but it seems to me that there is "lag" in the swash mixing, making the swash servos travel "asymmetrically". Have you noticed strange swash behavior when you tried the "merged" code (MP-H/MP3.X)?



Sincerely.

Marcus Fahlén


-------------------------------------------------------------------------

I'm out of office at the moment.

The information contained in this message may be CONFIDENTIAL and is
intended for the addressee only. Any unauthorised use, dissemination of the
information or copying of this message is prohibited. If you are not the
addressee, please notify the sender immediately by return e-mail and delete this message. Thank you.

Sent from my Sony Ericsson XPERIA™.

---

From: John McClelland <mcclell...@gmail.com>

Sent: 7. februar 2011 03:33
To: uavhel...@googlegroups.com
Cc: John McClelland <mcclell...@gmail.com>

[John McClelland]

We will probably go with the PPM input as well in the long run.  I need to buy that board and start looking at it.

 

I haven't done anymore on the MP-H migrated code since I observed the oscillations on takeoff.  I will look to see if I observe the lag you describe.  My guess was that with oversampling and no FILTERSHIFT I need to run the gains down considerably.  As soon as we get some good weather I will try it.

 

It will be important to pour through the migrated code to see if it all makes sense.  I know there are some errors, but not sure if they impact performance yet.

 

While I think Steven did a great job in getting us started, I would not recommend use of this code until it is well tested.

 

These threads will be very useful to others in the future, so let's try to respond to the appropriate threads if possible.  This one should go under the existing migration thread.

[Daniel Tan]

Hi John and Ben,

Can you email to me your full address and contact again so I can send you one of the latest 12chnnel PPM board with higher clock rate which in effect gven a better resolution.

 

Like to let you test it out to see if this work well with UDB. I already using this in the MK based system and the current software tool from MK do shows a much smoother or less drift from this new board.

 

Also like to know if it is possible to use the Compass designed for MK as well since it is also a ready product.

 

Regards,

Dan

[Marcus Fahlén]

John,

I did the restructuring to make it easier to do a diff using SVN client (Tortoise). This should be a fairly easy affair. Then we have your MP-H3 work to integrate...

It will take a while to have everything tied together. Here it's been -17C, very windy and lots of snow. It will take quite a while before I can do any testflying, but I will try to do some tethered tests indoors. Just have to build a "smart" tether system that allows for some movement without causing resonation feedback... I have at least repaired the helicopter now after the last test flight using the UDB which ended in a sideways crash after loosing tail authority. (not directly caused by the UDB). It was a consequence from a series of hard landings after UDB malfunctioning. This time I will go forward a lot more careful, not "flying" until proved stability in the control loop.



Cheers!

/ Marc

-------------------------------------------------------------------------

I'm out of office at the moment.

The information contained in this message may be CONFIDENTIAL and is
intended for the addressee only. Any unauthorised use, dissemination of the
information or copying of this message is prohibited. If you are not the
addressee, please notify the sender immediately by return e-mail and delete this message. Thank you.

Sent from my Sony Ericsson XPERIA™.

---

From: John McClelland <mcclell...@gmail.com>

Sent: 7. februar 2011 04:07