willy, those pictures are amazing! your photos are much better than mine. you must have a very good camera. when you get it running i will link to your project in post #1. very good job.

Originally Posted by willy163

Program check
After powering up, I pressed the switch tree times.
Then I've read EEPROM and I've found 03 at location 01. Great: program is working!

i dont know why esu had trouble with that part. maybe his switch was flaky or he had extra wires on that pin picking up noise. nobody else had trouble there. i may follow his suggestion and add a "blink verify" routine if i can free up some code space.

Originally Posted by willy163

Communication speed measurement
To check serial communication speed, I've used a soundcard scope (THIS) connected to AVR tx pin. I know that sampling rate of sound card (44.1kHz) isn't enought to display the 115kBaud of tx speed but... just to have an idea!
And... DAMN! There is the radio tx burst!!! (1.6ms burst each 20ms)

a soundcard scope may be able tell you whether theres an rs232 signal or not but a pc with hyperterminal is a good way to actually see the data. remeber that the t6 continuosly transmits stick position which is a good test signal. also dont forget you need to put a baud rate number in ee 0x3e-0x3f.

Originally Posted by willy163

I think the connector pinout in the 1st post is wrong.
Dave, please check it. AVR TX should be connected to RADIO RX (in this post is lower right).

i think my diagram is ok for a couple reasons. first iirc that photo from the t6 megathread was shown to be wrong in later posts. also several people built working units using the diagram so its probably ok. of course if you get it working using the other pin that will prove me wrong.

it should only take a minute to check the pinout using hyperterminal because the t6 continuously transmits out that jack. i think its very important you verify or show a problem so others reading the thread wont go astray. good idea to remove the diagram if its not right. please keep us tuned in.

Hi dave.
It's not so easy to trim AVR TX speed trying different values connected to Hyperterminal. I've tried to fill blank FLASH space with easy readable values to check immediately when speed is synchronized.
Finally I've found two working limits: 2ED1 and 3CC3. So I've programmed 35CA.

Now communication with PC works pretty good.

Question: is it correct that 1st model data is stored in FLASH between 0x1C0 and 0x1FF? Data is identical to .CFG file used expect for last three bytes that are missing in flash.
If this is correct, I think I'm close to our goal.

Communication between T6MMX and T6 doesn't work. Maybe TX speed has to be more fine trimmed. I'll borrow a real oscilloscope to check baud rate.

I have to stop until next week.

Bye,
WIlly.

PS.: I don't think my camera is better than your. Because I don't use a camera!
Google translation of my italian post where I describe my little secret to take supermacros without a camera.

Dave1993 - I've been contemplating buying a 9x but you've nearly talked me out of it.

The big problem I see with the CT-6B is the lack of expo, and shortage of mixes. I've been flying 3D and F3P type planes without expo, so it's not a total deal breaker, but I'd like to have the option of dialing a little in.

Originally Posted by willy163

Sorry but... In my opinion it's better to trim models mechanically. This way you'll have TX trims close to neutral position for each model.

Willy.

A lot of the 3D and F3P planes have non-adjustable linkages, and the foam warps enough between flights to require retrimming every session.

Originally Posted by pchewn

The ideal perfect trim methodology was on the JRX347 and JRX388.

Mechanical trim levers -- Fast and easy to trim.

"Autotrim" function after landing. Press the Autotrim function and the trim values are stored in the radio (PER MODEL) and then you can return the mechanical levers back to the center points.

Perfect in every way. You could center the trim levers and know your plane was in trim. You could dial in some temporary down-elevator trim for flying faster and then just feel the trim lever click back to neutral for normal speed.

Sounds ideal!

I've tried flying multiple models on the CT-6B, but they are mostly of the non-adjustable linkage variety. Writing down the position of the trim levers for each model is a hassle. I ended up only using one model on the CT-6B.

Today I went to work early so I've verified all with a real scope (but don't say that to anybody )

Single bit duration
One bit transmitted by T6A lasts 8.65us (115 600 Bps)
One bit transmitted by T6MMX lasts 8.62us (116 000 Bps) that is 0.35% of difference.
They should be able to communicate.

Connector pinout
I really don't understand. I've checked with the scope signals at T6A trainer port connector. They agree to this:

At pin 1 I see PMM signal. At pin 2 +5V. At pin 3 +5V (RX pin pulled up to VCC). At pin 4 continuous TX burst of channels.

Furthermore, every document I've found says same thing (check THIS, THIS is yours, THIS, THIS and THIS). T6A rx pin is close to +5V.

I see in your picture you've connected T6MMX TX to opposite pin. I can't understant how it can work.
BUT... yours works. Mine not. Maybe I'm wrong. Next week I'll try to connect my T6MMX TX pin according to your diagram.

Thank you,
Willy.

omg! i just checked with a scope, something ive not done since starting the project, and you are right. also my prototype in the first post works but the "improved" version i did for pics does not. i also spoke with one of the local fellows who built one and he said the same thing about my diagram. wish he had mentioned that. it turns out he copied my protoype which was correct and the other two locals who built one copied from him. this is one case where not wanting to read schematics worked in their favor.

i wonder what happened with the ones i sent out who did not have a chance to copy. i havent heard from them but better email them anyway. this may be esus problem too. thanks for finding that. i apologize for any trouble my error caused and will make corrections to the diagram and photos. thanks again.

to be clear again pin 3 of the t13 goes to the lower right pin on the radio, not the lower left pin.

edit: post #1 has been corrected. i see now where the error crept in. it turns out designers of the usb/serial ebay dongle in post #1 labeled it wrong. that may be why it was so cheap. the pic shows the pin labelled "rx" connected to the t6mmx which makes no sense. but that is actually tx and therefor my photo was correct wiring for that dongle. i was aware of a faulty label in one of the diagrams but instead i thought it was the t6 jack photo that was wrong and switched its tx/rx instead of the ones on the dongle. that explains why i thought there was an error in that link too. things just got worse from there. again, i apologize for my mistake.

Originally Posted by Jack the R

Dave1993 - I've been contemplating buying a 9x but you've nearly talked me out of it.

The big problem I see with the CT-6B is the lack of expo, and shortage of mixes. I've been flying 3D and F3P type planes without expo, so it's not a total deal breaker, but I'd like to have the option of dialing a little in.

A lot of the 3D and F3P planes have non-adjustable linkages, and the foam warps enough between flights to require retrimming every session.

i bought a 9x and do not regret doing so because it is a great deal for the money. once trims are set its my favorite for use with copters. i do recommend it for rc "power users" and specially helisbecause its primarily designed as a helicopter radio. my gripe is many recommend it to model plane noobs and casual park flyers like me where imo its overkill and can actually discourage participation in the hobby. i saw this happen locally.

you are oviously not in this category and may benefit greatly by having a 9x. my comments on the subject relate to myself and 99% of the other rc hobbyists who are small parkflyers where its advanced features may get in the way. and there is that issue with the trim tabs which could just be a personal thing. in these cases the t6 is hard to beat for the money.

one thing i am critical of is the 9x lipo conversion craze. imo too many non-technical individuals are led on a wild goose chase which has high risk and little benefit. i do believe in freedom of religion but think sometimes the "missionaries" go too far and should be more upfront about dangers involved.

Originally Posted by Jack the R

I've tried flying multiple models on the CT-6B, but they are mostly of the non-adjustable linkage variety. Writing down the position of the trim levers for each model is a hassle. I ended up only using one model on the CT-6B.

like i said before if i had one of those jr models this project would not have existed. it was fun though and id recommend you try it. it would certainly solve your trim issue until 9x radios become available again at a reasonable price. never hurts to have backup either.

i can not get away with one t6 model myself because i switch between conventional and v-tail often in the same flying session. and some of my hxt500 servos are reversed (those chinese!) so i need either model memory or the old 5 switches.

I tested why the expander does not store data in my transmitter HK-T6A. Of course, I also found an error in this scheme.
I compared the data from flash memory T6config. It seems that the data were stored correctly.

Picture 1 - Original data
Picture 2 - Data captured by the terminal sending to the transmitter.

It is obvious that the error in position 52-byte. Then shift by +1 byte of additional data.
The data was not taken to the transmitter.
I set a different value 0x1E, 0x1F EEPROM, but the error was the same.
Unable to admit. The data must be transferred all right!

Edit: 01.03.2012
I apologize to everyone.
This error is actually a mistake HyperTerminal. Or inaccurate transmission speeds.
For this purpose, I started using Linux a great program
jpnevulator and kermit settings. After setting the correct rights to the port / dev / ttySx can be read port that send procesor.

The command is
# jpnevulator --ascii --timing-print --tty /dev/ttyS0:AVRDAVE --read

thanks esu, im looking into it.

i tested my own prototype unit again just now and its working fine. ive been communicating with 2 of the people that were sent preprogrammed chips and it seems one guy found the wiring error days ago but the other fellow needed to be informed by me. both of them are working now. if anybody has trouble please let me know asap, either here or offline.

next step is for me to check the posted hex file. its the same size as the one in my prototype but needs to be downloaded and tested again. ill do that today.

(document control is not pretty)

Originally Posted by dave1993

and there is that issue with the trim tabs which could just be a personal thing.

I haven't used a radio with digital trim tabs yet, so I wouldn't know. It's hard enough for me to work the analog trims while flying - I don't have much "straight and level" space to work with. If I've got to work a screen, there's just no way. Analog trims FTW.

Originally Posted by dave1993

one thing i am critical of is the 9x lipo conversion craze. imo too many non-technical individuals are led on a wild goose chase which has high risk and little benefit.

I dunno about the 9X, but I converted a twenty year old Challenger 720 to Lipo. It was easy, worked well, and MUCH cheaper than replacing the stock battery. 9X may be a totally different story.

Originally Posted by dave1993

it would certainly solve your trim issue

Maybe I need to look at the CT-6B again, but I didn't think there was a way to copy the trim setting from the analog levers to the digital file. I suppose if you know how many subtrim clicks it takes to equal one lever click, you could manually edit the file settings.

Originally Posted by dave1993

thanks esu, im looking into it.

i tested my own prototype unit again just now and its working fine. ive been communicating with 2 of the people that were sent preprogrammed chips and it seems one guy found the wiring error days ago but the other fellow needed to be informed by me. both of them are working now. if anybody has trouble please let me know asap, either here or offline.

next step is for me to check the posted hex file. its the same size as the one in my prototype but needs to be downloaded and tested again. ill do that today.

(document control is not pretty)

Hi dave.
I've found a little difference between data sent by T6config and data sent by T6MMX.
T6MMX sends one byte less before CRC and appends CR+LF.

I attach the two log files, so you can compare them.

Willy.

crlf was tacked on as a debug luxury for use with hyperterminal. i see them in esu dump too. extra bytes are ignored by t6 so that wouldnt be a problem. im a little concerned though because it did take up a couple flash locations and i thought i removed those.

i checked on your previous comment that 3 bytes are missing from flash compared with the files and that is expected. the last data byte is fixed but not stored in flash because theres no room for 65 bytes per model due to limited flash. it is transmitted though. and the files do have 2 extra bytes that are always 0 but never transmitted. im not sure why they are there but t6config dont send them out either so thats ok.

in summary there should be 69 bytes sent: 2 command, 65 data, and 2 checksum. esu mention of the extra byte being duplicated is most serious. im having some trouble interpreting his dumps because for some reason its 17/line instead of the easier to read 16 and the addresses seem wrong. your log format is also slowing me down a bit because i cant read binary easily with the computer im on atm. but the input is appreciated.

i must get to the bottom of this because i got 3 more emails today asking for chips and those are highest priority. hopefully the posted hex is the same file as in my prototype which works and the chips i mailed already which were also ok. things are a little slow because this aint my normal pc but im sure everything will be fixed by tomorrows mail. thanks for the assistance so far.

Originally Posted by dave1993

i must get to the bottom of this because i got 3 more emails today asking for chips and those are highest priority.

Busy busy - your next task is to add expo and more mixes . . . but mostly expo.

Originally Posted by Jack the R

Busy busy - your next task is to add expo and more mixes . . . but mostly expo.

It isn't a limit of memory expander: it's a limit of the radio.
This means T6 firmware rewtite!

Willy.

Originally Posted by dave1993

[...]
in summary there should be 69 bytes sent: 2 command, 65 data, and 2 checksum. esu mention of the extra byte being duplicated is most serious. im having some trouble interpreting his dumps because for some reason its 17/line instead of the easier to read 16 and the addresses seem wrong. your log format is also slowing me down a bit because i cant read binary easily with the computer im on atm. but the input is appreciated.
[...]

Here's a pic with readable logs. Cursor position shows missing byte.

I was thinking... I own a FS-CT6A (rebranded HK-T6A V1). Maybe different firmware version of my radio (CT6A vs CT6B) dislikes missing byte. This could explain why my T6MMX doesn't work and yours do. Which is version of yout CT6? A or B?

Willy.

Originally Posted by willy163

Here's a pic with readable logs. Cursor position shows missing byte.

thanks for that. it is similar to what i see when downloading and running the posted hex myself. strangely it conflicts with eus earlier symptoms which showed a problem earlier in the file.

i played around with hardware and files all night and have a handle on the problem now. yesterday 2 people asked for assembled units and 3 wanted chips. atm im building and programming because they have highest priority but as soon as i send them out i will get back here and fix things.

Originally Posted by willy163

I was thinking... I own a FS-CT6A (rebranded HK-T6A V1). Maybe different firmware version of my radio (CT6A vs CT6B) dislikes missing byte. This could explain why my T6MMX doesn't work and yours do. Which is version of yout CT6? A or B?

i dont think that can be it. like you say fsct6a=hkt6av1 and fsct6b=hkct6av2. but that relates to the rf board protocol and i dont think there is any difference in the radio pcb or file format. i am onto the real problem now and will take care of it asap. thanks for the samples guys, it was helpful.

i have hkt4v2 and fsct6b which have same v2 rf protocol and same file format. i also have a hk6df which is v1 rf and does not bind with v2 rx.