SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
Old Aug 08, 2008, 02:05 PM
Registered User
Daemon's Avatar
Lakewood, Colorado
Joined Aug 2002
28,663 Posts
Quote:
Originally Posted by JettPilot
Not True,

You are thinking 455 KHZ intermediate frequency for the RC RX, which is 1000 times lower than 455 MHZ. It should not be a problem.

JettPilot
Sorry, you're right. Brain fade.

ian
Daemon is online now Find More Posts by Daemon
RCG Plus Member
Sign up now
to remove ads between posts
Old Aug 08, 2008, 02:12 PM
Registered User
thomasscherrer's Avatar
Danmark, Hovedstaden, Ballerup
Joined May 2007
3,213 Posts
>Hi thomas whatīs the servo latency response for this system?

same as any other PPM system +10mS
it samples the PPM frame, while sending the bytes in double speed, when checksum is ok, all servo positions will be updated with new values.
unlike the PCM 1024 or 2048 bits systems that need many bits pr servo = slow,
we have gone the other way and simply use 8 bits pr servo,
we have carefully tested this resolution using any servoes we could get our hands on,
and none have higher mecanical resolution anyway.
we have not added any hysteresis or pulse filterings, and no alias filters needed,
since TX and RX pulse detector and generators run in a phase lock !
this is needed for crips and fast update.

another fact from the radio world:
less bytes to be transfered at lower speed= less bandwidth used = longer range
so the trick is to use as low baud rate as possible, and get still get the needed information.
we use a 32bit lead in, and 16 bit checksum, we have also calculated a few bytes of extra payload, since the system have an extra serial port we can use later for other funny things we might need to control, like extra channels using channel expander boards. but right now 8ch seemed like a nice number to begin with.
thomasscherrer is offline Find More Posts by thomasscherrer
Reply With Quote
Old Aug 08, 2008, 02:45 PM
Registered User
Daemon's Avatar
Lakewood, Colorado
Joined Aug 2002
28,663 Posts
8 bits (256 steps) is pretty low resolution, especially considering that
most flying only uses a few % of the total throw of servo.
I've definitely got digital servos with much higher mechanical resolution
than that.

ian
Daemon is online now Find More Posts by Daemon
RCG Plus Member
Old Aug 08, 2008, 02:53 PM
Registered User
Elcheapo's Avatar
Joined Oct 2007
1,310 Posts
As allways, your work its really imppressive. Planned to do something similar, but this is beautifull, i will buy this system asap.
Elcheapo is offline Find More Posts by Elcheapo
Reply With Quote
Old Aug 08, 2008, 03:00 PM
fast
fmkit's Avatar
ISRAEL
Joined Aug 2007
2,641 Posts
I'm using similar system for 1 year now,
because ALL modern video receivers have intermidiate frequency 480MHz you must keep your control transmitter atleast 20ft away from video RX.

added: also 450mhz 2nd harmonic kills 900mhz video RX completly
fmkit is offline Find More Posts by fmkit
Last edited by fmkit; Aug 08, 2008 at 03:28 PM.
Reply With Quote
Old Aug 08, 2008, 03:02 PM
Free Flying
anci3nt's Avatar
Joined Nov 2006
525 Posts
VERY NICE WORK THOMAS! Been wondering when someone was going to do this. Two questions:

1. Do you plan to sell these packages with the various RC transmitter training plugs? That would be more convenient for users so they don't have to hunt down the plug and pinouts, and/or solder directly to their transmitter.

2. For US folks, the ARRL band plan has 433 MHz as auxiliary/repeater links, and 444 MHz as ATV repeater inputs. I don't know if there is a 70 cm band specifically for RC control (there are some around 50 MHz). I think it would be helpful to have some frequency control or at least a frequency option(s) in the case 433/444 are being used. As unlikely as interference may be (since most of us aren't flying in congested cities scapes right??), I'd like to be a good radio samaritan especially if using a PA. I'm also not sure what the callsign requirement would be if operating high power in a non-RC control specified band.
anci3nt is offline Find More Posts by anci3nt
Reply With Quote
Old Aug 08, 2008, 03:33 PM
Registered User
thomasscherrer's Avatar
Danmark, Hovedstaden, Ballerup
Joined May 2007
3,213 Posts
>8 bits (256 steps) is pretty low resolution, especially considering that
>most flying only uses a few % of the total throw of servo.

people that say that, have most likely not tried to measure this in real life,
I have: if you take the normal 1 to 2 mS pulse and divide this area down to 8 bits,
well we have added a normal 10% margin on both sides, so you can trim out of range, that is also normally possible on many transmitters, then we have tried to make software that step one position up, wait 1 sec then step 1 position down wait a sec and so on,
not possible to see feel or measure any kind of movement on many digital servoes, why ! they have added a hysterisis to avoid the jitter from analog radios :-)
ok back to analog servoes, some of the best actually generate a tiny click, not enought to really move the shaft so you can see it, we have mounted a 3 inch long servo arm to really prove if any useable movement, but the mecanical cock-to-cock distance gives a tiny movement that is always over 1 bit anyway, and no mater how you connect the servo arm to a rudder surface, you will add another bit, so yes, 8 bit is perfect !

>you must keep your control transmitter atleast 20ft away from video RX.

yes that is right, with the 7w booster I will recommend to use a long coax cable
and mount the booster with big antenna on a hight pole or stick to achive even better range.

>Do you plan to sell these packages with the various RC transmitter training plugs?

no, I plan to cut the cable at about 10 incl long, and leave that up to all you,
it is simply, you own the transmitter, you know what kind of plug, you have the manual, you can look up the connections, GND, power out, PPM out, that is all you need.

you see if I have to get 20 different plugs and look up all the connections,
and also test it, imagine I have to get access to all kinds of transmitters too,
web sites exist with trainer plugs pinouts , I think it is called tx2tx ? or something like that.

>I'd like to be a good radio samaritan especially if using a PA

I understand, but the funny thing is, you cant hear or see this system using normal ham equipment, the frequency hopping and bandwidth used makes this system invisible to ham users, even if they sit on the bench right next to you, they still cant find out what frequency you use, even if you told them where to look :-)

but to get the best range and best undisturbed control, I suggest you find a free and best if also legal frequency band in your local area/contry, this system can use any frequency in the 410 to 450MHz with a few kh resolution, I just need at least two "channels"
The more spaced they are, the better imune you will be to jammers or noise bursts from outher space or whereever.

>added: also 450mhz 2nd harmonic kills 900mhz video RX completly
true, so use 410MHz then if legal/free
I have harmonic kill filter, however it is always impossible to opperate a TX and RX in the same area on exactly the double or trible frequency,
take any dual band ham radio that covers 145MHz and 435MHZ
try to use both those exact frequencies at the same time, impossible.
even 6th harmonic will kill a 23cm channel, but then go a few channels away,
now the RX works fine again, and problem solved.
thomasscherrer is offline Find More Posts by thomasscherrer
Last edited by thomasscherrer; Aug 08, 2008 at 03:39 PM.
Reply With Quote
Old Aug 08, 2008, 03:58 PM
Registered User
Daemon's Avatar
Lakewood, Colorado
Joined Aug 2002
28,663 Posts
Quote:
Originally Posted by thomasscherrer
>8 bits (256 steps) is pretty low resolution, especially considering that
>most flying only uses a few % of the total throw of servo.

people that say that, have most likely not tried to measure this in real life,
I have. In testing to find a good digital servo for smooth panning
of the onboard camera through 180 degrees, I put a 1 foot long
extension on my Tx stick, and a similarly long extension to the servo
arm and then ran it one "step" at a time from end to end, basically
counting every detectable movement of the servo with the finest
movement of the control stick, always proceeding in one direction.
Most analog servos, particularly those pushed to 180 degrees, show
only 30-60 steps per half throw of the servo. With my digital servo,
I lost count somewhere around 200, per half throw. That's over 400
for a full end to end throws, and well over 8 bits of resolution.

You see these steps on the pan servo very clearly in the video,
which is why it matters to me, and why I've tested it.

Quote:
then we have tried to make software that step one position up, wait 1 sec then step 1 position down wait a sec and so on,
not possible to see feel or measure any kind of movement on many digital servoes, why ! they have added a hysterisis to avoid the jitter from analog radios :-)
In other words you've created a test which just confuses the servo,
instead of actually just seeing how fine of a movement the
servo is capable of. I don't jiggle a control back and forth while flying.

Quote:
ok back to analog servoes, some of the best actually generate a tiny click, not enought to really move the shaft so you can see it, we have mounted a 3 inch long servo arm to really prove if any useable movement, but the mecanical cock-to-cock distance gives a tiny movement that is always over 1 bit anyway, and no mater how you connect the servo arm to a rudder surface, you will add another bit, so yes, 8 bit is perfect !
Rudder is the one surface with the greatest angle of throw, and least
need for high resolution. Ailerons and elevator on a high performance
aircraft may only swing through a few degrees, during normal high speed
flight, and you need high resolution through that tiny range of movement.
You can't fix it with "geared down" linkage, because you may still need full
throws for flaperons/spoilerons etc.

Anyway, I'm sure it's flyable, but I'd love to see what it does on
the 180 degree pan servos and hear how it feels on something
faster than an EZ*.

Quote:
but then go a few channels away,
now the RX works fine again, and problem solved.
Has it been tested with a 900Mhz video system yet?

ian
Daemon is online now Find More Posts by Daemon
RCG Plus Member
Old Aug 08, 2008, 04:32 PM
Registered User
Impulze's Avatar
Sweden
Joined Jul 2004
712 Posts
How to connect to different radios.

http://users.belgacom.net/TX2TX/tx2t...2txgb3.htm#top

//Matte
Impulze is offline Find More Posts by Impulze
Reply With Quote
Old Aug 08, 2008, 04:51 PM
Registered User
thomasscherrer's Avatar
Danmark, Hovedstaden, Ballerup
Joined May 2007
3,213 Posts
thanks for the tx2tx link I mentioned

yes 180 degree servoes are special, and you need 9 bits to give you 512 positions,

you are also right about the back and forward 1 pulse if often either filtered away or hidden in a hysterresis routine,
your test showed 400 positons on a 180 degree servo, normal servoes go about 90 degree for the same 1-2mS pulse, so that is 200 positions, easy to solve with 8 bits, and that is why I use 8 bits,
as I said a few overhead bytes are possible, you have a transmitter with total freedom to assign the pan servo to any channel you like I assume :-)
then I could make two of the channels in 12 bit resolution,
the units can be field upgraded, so if/when I get time to solve this in sw, it can be released and tested on the other side of the globe in minutes.

it have not been flight tested with 900 MHz video yet,
I dont own 900 MHz video gear since it is not legal in europe due to our GSM band use 900 and 1800
I have not heard from any customers about 900 yet, but I am sure they sneak-listen here, so please come forward if you have/can test it with 900 MHz video TX located like 10-20 inch away from the RX.
I have all sorts of signal generators and can easily generate any frequency and any modulation and even any RF power I like :-) so I can can see how many dB needed on the RF connector to jamm/block it.
thomasscherrer is offline Find More Posts by thomasscherrer
Reply With Quote
Old Aug 08, 2008, 06:36 PM
Registered User
Carlyle Harper's Avatar
Joined Feb 2007
1,278 Posts
Is there a RSSI output option for signal strength? Not exactly sure how this information is generated, but could it be as simple as a varying voltage output connected to a voltage sensor on an OSD that you could monitor during flight?
Carlyle Harper is offline Find More Posts by Carlyle Harper
Reply With Quote
Old Aug 08, 2008, 07:13 PM
Registered User
Carlyle Harper's Avatar
Joined Feb 2007
1,278 Posts
I have a 9ZAP WCII and currently use a tm-7 FASST module in it with absolutely no problems for normal flying.

However it has been reported here:

http://www.rcgroups.com/forums/showt...3&highlight=9z

that some 9z's have a problem with the ppm stream. I have no idea if mine is affected and I'm not exactly sure what a ppm pulse is or consists of. I just know it has been known to cause a problem with non futaba modules, but I'm sure you know exactly what the pulses are. So do you know if your system will be able to work with a 9z?

Update: The 9ZAP will work with Thomas's system. No glitching observed.
Carlyle Harper is offline Find More Posts by Carlyle Harper
Last edited by Carlyle Harper; Sep 24, 2008 at 06:48 AM.
Reply With Quote
Old Aug 08, 2008, 08:55 PM
FPV Flyer
wavess's Avatar
Madrid, Spain
Joined Mar 2006
763 Posts
Quote:
Originally Posted by thomasscherrer
>Hi thomas whatīs the servo latency response for this system?

same as any other PPM system +10mS
it samples the PPM frame, while sending the bytes in double speed, when checksum is ok, all servo positions will be updated with new values.
unlike the PCM 1024 or 2048 bits systems that need many bits pr servo = slow,
we have gone the other way and simply use 8 bits pr servo,
we have carefully tested this resolution using any servoes we could get our hands on,
and none have higher mecanical resolution anyway.
we have not added any hysteresis or pulse filterings, and no alias filters needed,
since TX and RX pulse detector and generators run in a phase lock !
this is needed for crips and fast update.

another fact from the radio world:
less bytes to be transfered at lower speed= less bandwidth used = longer range
so the trick is to use as low baud rate as possible, and get still get the needed information.
we use a 32bit lead in, and 16 bit checksum, we have also calculated a few bytes of extra payload, since the system have an extra serial port we can use later for other funny things we might need to control, like extra channels using channel expander boards. but right now 8ch seemed like a nice number to begin with.
Thatīs great, this is the first long range system with that refresh speed. But isnīt it the standard ppm refresh 20ms about 40hz?
Great work very tempting to buy one
wavess is offline Find More Posts by wavess
Reply With Quote
Old Aug 08, 2008, 09:00 PM
Registered User
Thailand
Joined Apr 2006
317 Posts
http://users.belgacom.net/TX2TX/tx2t...2txgb3.htm#top


Yes, I also ever sent that link to Thomas , but when we open Futaba T9CHP and check it signal, what it show in the link and what we get from test is make me more worry. I ever plan to sent my TX to Tomas for his test.

But finally give up and I just learn again that Thomas do not like to spent his time to busy with many type of Remote TX. When I recieve this product from Thomas I feel that his work is too nice to let me take any risk on connector mistake, that's why I finally order Multiplex to use with it.

Personally , if Thomas can make his system with Futaba, JR it should be good enought to sell his product to many interesting customer.
ThaiskyDigital is offline Find More Posts by ThaiskyDigital
Reply With Quote
Old Aug 09, 2008, 02:57 AM
Registered User
thomasscherrer's Avatar
Danmark, Hovedstaden, Ballerup
Joined May 2007
3,213 Posts
>Is there a RSSI output option for signal strength?

yes, actually the receiver board have component mounting options,
that will feed RSSI out the on the 8th channel,
however later channel hopping was invented and then the RSSI jumps up and down like crasy, also when switching antennas to find the strongest signal the rssi will jump again, so all channel hopping or diversity recevers have internally RSSI, but it looks like useless noise to users, sorry. A really smart engineer user, could sneak out the signal, and peak detect it and filter it, and that will make it usefull again.

>So do you know if your system will be able to work with a 9z?

no I dont yet, I sure could try to borrow one and try, I am a member of a good large club, but that name number dont even ring a bell here :-)
also I would need the special connector and that is not always easy to get ? help ?

>But isnīt it the standard ppm refresh 20ms about 40hz?

yes I said, the normal PPM repeat + 10mS
the receiver is in sync, like in phase with the ppm generator, ok
so they both run 40Hz, but the receiver is simply 10mS behind constantly !
we have tried to use the fastest servoes we could get, and assigned a switch to it,
then moved the switch as fast as possible, even the smallest possible pulse will be seen on the servo, we had a normal PPM 35MHz link on the same TX on the same table, connected to the same source ppm frame, not possible to see or feel this 10mS delay, but ok you can see it on a scope :-)
also if you use helicopters the swash is critical, some systems with higher resolution sends the bytes over several pakages, and this adds a phase difference delay to fast moving swashes, I dont have that problem, since ALL 8 channels are send during EACH ppm frame and in sync with it.

>but when we open Futaba T9CHP and check it signal,
>what it show in the link and what we get from test is make me more worry.

yes I have an idea, send me the trainer connector you have my address,
or direct my to a place I can purchase this special connector,
then I try to borrow a T9CHP from a club member, then I measure the PPM pulses and test if compatible, if not I find a workaround.
thomasscherrer is offline Find More Posts by thomasscherrer
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Long range flight using new DMD 868Mhz RC equipment tron FPV Talk 500 Feb 25, 2014 02:22 AM
Idea Long Range RC pmarek UAV - Unmanned Aerial Vehicles 22 Nov 07, 2011 06:17 PM
Long Range RC Link? cloudy Radios 2 Aug 09, 2008 12:11 PM
long range radio system suggestions?? oracle_9 Thermal 15 Aug 09, 2008 10:44 AM
Discussion Long Range 2.4 Ghz Wireless RC Camera edplane Aerial Photography 8 Apr 18, 2008 08:24 PM