RC Joystick NG^2 [Archive]

alessio70

Sep 12, 2006, 03:55 AM

for those who are interested in our little project of an RC to USB interface...

- from firmware version 3.2.0 the interface become a full speed USB device to get around constraints on speed and report size imposed for low speed devices by USB specifications; this means that at last we can have more proportional axes per joystick without loss in resolution: the current version implement two joysticks with 8 proportional axes each (up to about 5000 points resolution each) and 8 buttons for a total of 16 proportional axes and 16 buttons (32 controls). Anyway the traditional low speed interface firmware binary with 4 axes per joystick is still available inside the software package and is compilable by just setting a define; polling interval for interrupt transfer is now set to 5 ms (and could also be set to 1ms) to minimize input latency. The new version does NOT need any hardware mod... only firmware changes.

a summary of the other features:

- smd (4x2cm) or through hole (6.5x2.5cm) PCB and very few components

- uses the new chips from microchip 18f2455/2550 that have a flash program and data eeprom and then are programmable more than once (useful for programming firmware updates available on our site)

- assignments of channels and joystick controls are configurable by editing a txt file configuration and programming it directly via usb with a simple utility; assignments configuration remain stored inside PIC non volatile data eeprom so there is no need to reprogram it after disconnecting

- it should be compatible with virtually all radios equipped with a PPM output, and it's also able to decodify some type of PCM signal (for now Futaba PCM1024 and Sanwa/Airtronics PCM1/2)

- it should be compatible with virtually any RC or flight simulator that can make use of the system joystick

- the firmware work also under linux (and the assignments utility should be quite easily ported on linux linking against libusb instead of LibUSB-Win32) and as an HID peripheral should work under every OS with HID USB support

- an extensive manual is provided on our site, with hardware, software, assignments, and also modulations description

- the source is extensively commented for the ones who need to personalize it

- the project is totally free and suggestions and reports about it are very appreciated :)

you can find it at:
http://rcjoyng2.qstep.net/

enjoy it :)
alessio e andrea

alessio70

Sep 22, 2006, 10:29 AM

some news on the project :)

the firmware now also implement ADC mode: if someone has a radio not equipped with PPM/PCM or has old unused or malfunctioning radio, now has the possibility to directly connect its potentiometers/switches to PIC analog inputs AN0..AN11 (JP4 and JP5 on the new through hole PCB); firmware uses the PIC internal ADC to sample these inputs (voltage in the range 0-5V) with a resolution of 10 bits (1024 levels); up to 9 channels can be read in this way with 28 pins PIC18F2455/2550 (up to 12 channels could be read using 40/44 pins PIC18F4455/4550). See manual on our site for details.

alessio e andrea

alessio70

Sep 27, 2006, 03:39 AM

news on the project: set_assignments and get_assignments ported on linux with very little modifications to the source: the same source work now on windows (linked against LibUSB-Win32) and on linux (linked against libusb). Linking them against libusb they should be smoothly ported on all OS supported by this: FreeBSD, NetBSD, OpenBSD and Darwin/MacOSX. Also corrected a bug with usb error messages...

alessio e andrea

alessio70

Sep 27, 2006, 03:55 AM

P.S. for linux there are at least two programs that deserve a try in my opinion... FlightGear (a flight simulator) and CRRCsim (an RC simulator... thanks again Flieslikeabeagle for make me know this little pearl and the beauty of flying gliders :) )... both available for windows too.... about CRRCsim try also the latest beta for windows you can download from:
http://crrcsim.sourceforge.net/wiki/index.php?n=CRRCsim.TaskList

alessio

shim

Sep 29, 2006, 05:51 AM

Not trying to be cheeky, but does the cable work with RealFlight G2/G3 and Reflex XTR?

alessio70

Sep 29, 2006, 12:41 PM

Not trying to be cheeky, but does the cable work with RealFlight G2/G3 and Reflex XTR?

it works with virtually any simulator that can use system joysticks....

alessio

Norman Adlam

Sep 30, 2006, 03:51 AM

Not trying to be cheeky, but does the cable work with RealFlight G2/G3 and Reflex XTR?

Reflex XTR, later (non-Interlink) Realflight G2 and G3 all have their own special dongle, and don't allow joystick input - so the interface won't run with them.

Cheers,

Skylined

Jan 08, 2007, 08:01 PM

Great job Alessio! ;)

I won't use the ICSP header.
Do I still have to leave the 10K pull up resistor (R4)?

Can you please tell me the current draw of this circuit?
I plan to power the radio from Vdd and Vss and I just want to make sure that the total current draw is under 500mA, which is what USB supports.

Malc C

Jan 09, 2007, 05:03 AM

Do I still have to leave the 10K pull up resistor (R4)?

Yes you will still need to tie MCLR high via the 10K resistor

I plan to power the radio from Vdd and Vss and I just want to make sure that the total current draw is under 500mA

The interface will draw next to nothing on its own, however I don't think its possible to power a radio from the USB's limited 5v supply, you will still need to have the transmitters 9.6v battery charged and installed.

perttime

Jan 09, 2007, 05:51 AM

Aha, it is an "interface to connect a Radio Controller Trasmitter to your computer via USB."

(had some trouble figuring that out just by reading this thread... engineers ... :) )

Malc C

Jan 09, 2007, 07:16 AM

Err.. yeah ! its an interface to allow you to connect any commercial RC transmitter to a PC via a USB port so you can control a virtual model in a simulator... what did you think it was ??

Skylined

Jan 20, 2007, 06:58 PM

Thanks Malc C

As I've already told Malc C, and just to inform the rest, you can power the radio from USB.

Some considerations:
5V is going to be able to power up the radio, but seems that it isn't enough for it to transmit through the antenna.
USB's amperage rating is 500mA, check how much your radio draws.

I've seen the two layer PCB design, but I wanted it to be a single layer PCB, so after a little bit of time, and after getting rid of the ADC pins I got this.

http://members.lycos.co.uk/herdkilling/RC/aleusb.jpg

I think that it can be a little bit smaller, but I'm waiting for my 18F2550 to check if everything fits.
If anybody wants the Eagle file of this PCB, just PM me.

Alessio feel free to use this if you want.

Malc C

Jan 21, 2007, 05:23 AM

Yes it is possible to shring it down a bit. This was the layout we came up with for our interface detailed in the sticky post in the main section. The only changes you need to make it compatible with Alessio's interface is to remove the Jumper to pins 10 - 13 and re-route the LED as Alessio uses a different pin.

The sticky post contains all the eagle files should you want to play about with the layout further. Personally I have my reservations about powering a tranny from the interface using the USB port, especially if you include the transmission module or xtal. I find that if you remove these its a long time before you need to re-charge the TX batteries - However I'll agree to disagree on this occasion :)

alessio70

Mar 10, 2007, 09:28 AM

sorry guys haven't read anymore this thread... :p

anyway there are some news on our page:

- PCB completely redesigned: both the through hole and smd PCBs are now single layer... removed two resistors (R4 and R2 in the old design)... the USB plug can be soldered directly to PCB like in a dongle... both have also ADC connectors... dimensions are slightly reduced

- qpicprog: this is a software (bootloader and programmer) that can be used for programming the pic directly through USB so also RC Joystick NG^2 is now programmable directly via USB

projects homepage:
http://projects.qstep.net/

ciao :)
alessio

jmralves

Mar 10, 2007, 05:00 PM

Thanks Alessio

Jorge

alessio70

May 03, 2007, 08:29 AM

added support for Walkera PCM :) you can find it in the latest software version (3.3.4); tested only on WK-0701 but could work also on others greater models... if someone tries it with the others please let us know how it works...

http://alessioandrea.altervista.org/rcjoyng2.html

ciao :)
alessio e andrea

TheSteve

Feb 16, 2008, 12:29 AM

Just wanted to say I built your interface and it worked first try with my Futaba 9CAP. I also tried it with my Jr x9303 which needed the pullup resistor. Great work and thanks so much for sharing!

alessio70

Feb 17, 2008, 06:38 AM

Just wanted to say I built your interface and it worked first try with my Futaba 9CAP. I also tried it with my Jr x9303 which needed the pullup resistor. Great work and thanks so much for sharing!

:-)

fred73

Feb 17, 2008, 05:48 PM

HI

Ive tried built your usb interface but no what what after several attempt, i can get any signal from my radio. The interface would be recognised by windows but cannot calibrate or do whatever.

I've tried changing most components but still, so finally i built Malc usb interface, it works fine only with my futaba 4 channel but with my DX7 its works a millesec and then goes off.

Any Advice please

Thanks

fred

TheSteve

Feb 17, 2008, 06:19 PM

Did you add the pullup resistor recommended for Jr radios in the docs, I am sure the DX7 needs it.

fred73

Feb 18, 2008, 12:34 AM

H

Yeah evern with the 220K pull up resistor it doesnt work ,about alessio project its not working at all though recognised by windows, tried all firmware also but in vain :confused:

Thanks

alessio70

Feb 18, 2008, 04:26 AM

H

Yeah evern with the 220K pull up resistor it doesnt work ,about alessio project its not working at all though recognised by windows, tried all firmware also but in vain :confused:

Thanks

when you connect the radio and turn it on the led also turn on?

if the led does not turn on it could be a problem with input stage... don't know dx7 in detali but think that it shoud have and you are using a ppm output signal... I don't know if it has or not an open collector output and then requires or not a pull-up resistor.... have you made the schematic/pcb on our page (http://rcjoyng2.qstep.net) or something else?

another possibility is that modulation in the firmware is not correctly set to ppm.... the possible reason if you have just programmed the interface is that the data eeprom is not correctly programmed.... look at the last byte of the 256bytes data eeprom: it should be 01 and not FF... if it is not 01 let me know...

fred73

Feb 18, 2008, 06:58 AM

Hi

The leds is pretty dim and i've got signal at all from the futaba and the DX7

I've tried change all components and the pic too, i must be doing something wrong i guess. Also try previous firmware 3.3.3 but still nothing.

Is the Adapter able to receive signal from the Dx7 spectrum as with Malc usb its works fine with my futaba but only a millisec with the DX and goes dead, tried the 220K pull up resistor, no result

Thanks for all your input, i'll try pcb another one and see how its goes

Malc C

Feb 18, 2008, 02:40 PM

Fred,

Done some googling and came up with a few snippets from other forums

The "quickness" of the DX7 is because of an all digital link between the processor and RF sections. All other transmitters produce either a PPM or PCM 'stream' that goes to the RF section. This stream introduces some latency that the 7 bypasses with it's all digital link. However, the DX7 digital link does NOT apply to the trainer port and therefore not to Reflex. The DX7 trainer port is pure PPM, just like any other transmitter.

This suggests that although the sticks are digitally encoded and sent to the RF stage, they are also encoded and sent as a PPM stream to the trainer port. However IF that statement is incorrect and the output from the trainer port is in PCM format then that would possibly go some way to explain the issue.

The manual of the DX7 on page 36 says the following :
Trainer
The DX7 offers a programmable Trainer function that allows the transmitter to operate in three different
Trainer modes. Either the left or right rocker can be programmed as the trainer switch.
1) Normal:
The transmitter can be used as a master or slave but the slave transmitter must have the same programming (i.e.
reverse, travel adjust, dual rates, mixes, sub trims, etc.) as the master.
2) P-Link:
In Pilot Link the master transmitter maintains control of all secondary functions (i.e. dual rate, expo, gear,
flaps, etc.) and only the primary stick controls (aileron, elevator, rudder and throttle) are transferred to the slave
transmitter when the trainer switch is pressed.
3) Slave /P-Link:
In the Slave mode, the DX7 is used as a slave radio in conjunction with a Spektrum® radio that is used as the
master that is in P-LINK mode; there is no need to match the slave’s programming to the master transmitter’s
programming in this mode.

In "Trainer mode normal" the slave TX must have exactly the same settings and mixings as the master DX7 (less or more throw is not a problem if the pupil can handle it).

In "P-link mode" any simple non-programmable 4 channel slave TX with the same channel assignment (JR) can be connected to the master DX7. The master DX7 will take care of all throws and all mixings and settings. Very easy ! This is really the ideal program for easy buddy boxing, without hasstle. And the buddy lead is a simple mono to mono jack 1/8".

The slave/P-link is to use the DX7 as slave.

It might be that you have to configure the trainer port correctly so it sends data rather than receive -

And some further links
http://users.belgacom.net/TX2TX/tx2tx/english/tx2txgb1.htm

And lastly on a forum about realflight where someone was experiencing similar issues

I've used the DX7 with RealFlight and actually had that same problem but the deal was I had turned the radio on via the power switch - you have to just plug the interlink into the DX (power off) and it will power on automatically and then RealFlight should see it.

alessio70

Feb 19, 2008, 03:42 AM

Hi fred, as malc c says the radio should be able to output a PPM signal (I've seen also in a thread that it works with reflex xtr so should output a ppm signal and then should work with rcjoystickng2 too... I've also seen in another thread that the pull-up resistor does not seem to be needed...). as malc c says on page 36 of the dx7 manual in pdf you should find the instructions on how to set the three modes for the trainer port.... the manual also talk about the assignment of the trainer switch to the left or right rocker....

but if the rcjoyng2 interface you have built is not able to see the futaba signal I think there is something wrong with the input stage (transistor or resistor or connector or polarity...): the led should be off until you connect the radio (perhaps you don't need to turn the radio on... don't know about dx7, but other radios automatically switch on without activating the rf stage when you simply plug in the connector.... even if for doing that they normally need a bridge between two pins of the connector to plug in the trainer port) and then turn stable on when you connect the radio....

if the interface is recognized in windows it should work ok... only the input stage could be wrong.... or the 256 bytes eeprom data not correctly programmed (in this case the interface could be not programmed for ppm and theno not able to receive anything!!).... this sometimes happens because of errors in the programmer software settings.... I repeat... check the last byte of the 256 bytes data eeprom with your programmer software and verify that it is 01 and not FF or something else....

alessio

Malc C

Feb 19, 2008, 12:33 PM

Alessio, I think fred is saying that the code Mark and I have listed, and your latest offerings both work fine when used with his Futaba TX, but when he plugs in his DX tx it stops working after 3 seconds or so. If it works with his Futaba then this would suggest that there is no problems with the programming of the PIC or the build of the PCB. My guess (and I hope you could help confirm this) is that for some reason the PPM signal from the DX transmitter is:

1) - Sent at a differnt timing frame, so after the first few seconds it looses the sync pulse and confuses the interface

2) - The amplitude of the PPM signals are too small to fire the tranny or below the levels that the PIC can register, or being pulled below the level to register.

Fred - can you post screen captures of the PPM traces from the winscope for both TX's without making any adjustments so we can compare the outputs

alessio70

Feb 20, 2008, 04:44 AM

Fred does my interface work with futaba?

fred73

Feb 20, 2008, 06:06 AM

[IMG]Hi

First would like to thank you guys for all your help and concern, really helps.

So to summarise, i first tried your rc usb adapter Rgn^2 alessio but didnt get it working, as i said will be recognised by windows xp but no signal from any radio, futaba and dx7. checked all components, new pcb, etc transistor BC 107B, Bc 109 and finally BC 547 but still nothing. Im using this programmer by the way

http://www.nbglin.com/jdm.htm

So then i built malc usb adapter and it worked but only with my futaba, and with the DX7, it works a few millisec only but the goes dead. tried 220k pull up resistor, no go,

Also as advice change trainer mode, normal, Plink/slave. slave still nothing
DX7 stays off but when plug in mono jack it will switch on, signal works few millisec on game controller and nothing

please find the scope screen shot hope it makes sense what i did capture

thanks for all guys, very nice project by the way ;)

Malc C

Feb 20, 2008, 12:40 PM

Fred,

Assuming you didn't change any setting of the software, its obvious that the signal from the Futaba is about twice the voltage (or amplitude) of the DX, which was my assumption. Also, it looks as if the DX is refreshing the signal at around 30ms intervals, compared with the Futaba's 22(ish) ms, however you would really need a proper scope to confirm this.

I'm no expert, but I think that its the firmware that will need re-writing to match the refresh rate, but the hardware will also need changing to provide better gain when using the DX, but then you may have a problem with too much gain when using the futaba.

In a nutshell, it would appear that the interface is just not compatable with the DX due to tollorances in their design (especially as it won't work with either of our designs)

alessio70

Feb 20, 2008, 01:30 PM

hi fred, the signal seems to be an 8 channel PPM, the timings seem to be correctly recognizeable by rcjoyng2 without any mod to the firmware, the only thing I can't extimate is the signal amplitude.... rcjoy needs a signal of at least 0.7 volt than if it was half than futaba (that I think is 5V) should be enough... but from the pics you have made the futaba signal seems to extend over the screen (over -100%) so can't value the proportions between the two....

for the possible programming issue if you are using winpic800 look at the option:

Settings -> Software -> .hex -> File .HEX -> Data 18Fxxx Address * 2

it must be unchecked otherwise the 256 bytes data eeprom would not be programmed correctly...

alessio

fred73

Feb 20, 2008, 03:59 PM

Hi guys,

Thanks for all input, i think i will buy one of these usb adapter and see how it go
http://www.r2hobbies.com/proddetail.php?prod=rcps81919

Its only cost 7$ so hope i can get something out of it :rolleyes: as no mention of number of channel etc.

will keep you updated once i got it

Thanks guys

Malc C

Feb 20, 2008, 05:37 PM

That's the thing now, its far cheaper (but less rewarding) to purchase than to build. Fred, you have to remember that when we started these projects (around two - three years ago) most interfaces were serial or parallel, and those USB interfaces that started appearing were typically £50 +, so it was cheaper to build ones own. Now they seem to give USB interfaces away with cornflake serial :)

TheSteve

Feb 20, 2008, 07:22 PM

From what I have read the one listed by fred73 has only 7 bit resolution - 128 steps, ouch!
I have used several, Malc - yours has a great feel to it, which is why I am building a few more.

Malc C

Feb 21, 2008, 02:08 AM

Cheers Steve, glad to know you like it

alessio70

Feb 21, 2008, 03:15 AM

hi fred in case you take the other interface let us know how it works...

anyway I'm really curious to see if rcjoystickng2 can work with this interesting radio so if you could help me... you can also write me on msn messenger (you can find my msn in my rcgroups profile) or in email (you find here: http://alessioandrea.altervista.org/email.html )... I could try also to make a debugging test firmware...

ciao
alessio

TheSteve

Feb 21, 2008, 03:42 AM

alessio70

Feb 21, 2008, 06:43 AM

The DX7 is pretty standard(and hugely popular) There should be no issues with it working, its trainer output is the same as all Jr radios. It requires no setup config, if you plug a cable in it will turn on and output PPM. If anything perhaps a slightly lower value resistor as a pullup.

if it has a open collector output like other jr radios probably you are right, and fred could try smaller resistor values.... 100k for example.... even if I have read somewhere that the resistor was not needed.... I'd really like to know if someone knows for sure if it is open collector or not :)

anyway it's quite strange that rcjoyng2 does not work with futaba.... there are dozens of people who reported it works on futaba.....

fred73

Feb 22, 2008, 06:20 AM

Hi guys

Should get the new rc usb adapter by next week so finger cross to see how it goes.

BTW this is the one im using right now, built it a year ago and it works both on my futaba and Dx7 straight away, only problem is, its only game port. Thus the reason i wanted to go usb :D

http://www.rcdesign.ru/eng/electronics/rc2joy

The guy has now done a usb version, maybe u will need to translate with google
http://www.rcdesign.ru/articles/electronics/rcusb

will keep you updated

ciao
fred

JDill

Mar 09, 2008, 07:52 PM

alessio70

Mar 10, 2008, 07:55 AM

I ran into a problem when building up both M&M's interface and the NG2 by using an "equivilant" transistor. My old Futaba transmitter worked but not my newer 6EXA. See post #142 (http://www.rcgroups.com/forums/showpost.php?p=9271563&postcount=142) and those nearby for my exploits.
Jeff

thanks for the info and measurement... very useful :) (unfortunately I don't have a scope :( )

rcjoystickng2 schematic requires indeed a transistor with a high hfe (but not darlington if you want to keep the Vbe threshold low...) because of the high resistance (R1=22k) that I have put in input as a safer protection for the radio.... high hfe permits to saturate even with a small current at the base.... then a bc107b should be better than bc107a or a 2N3904 and even better should be a bc108c or 109c (and should be also very common).... and the hfe can also change a lot from a transistor to another even in the same family...

anyway in case of problem another solution would be trying a lower value for the input R1=22k resistor (for example replacing with a 10k or even 4.7k one).... another solution would be using an external pull-up for the transistor collector (instead of the internal portb pull-up as in rcjoystickng2 design) and try to use values for example of 22k or more.... but perhaps the most simple solution is to change the transistor or input resistor....

alessio

Andy64

Apr 16, 2008, 01:35 PM

I also posted this in the "other" USB joystick thread. I think I know why people are having trouble with the DX7 (and possibly other transmitters as well).

I just took a look at the output of my DX7. Its PPM and about 22mS cycle time.

However, its only about 1.4 Vpp and apparently AC coupled. With the DX7 output connected to the direct coupled scope (10 Meg impedance) the signal goes between +.4 and -1.0 V. The levels are determined by the duty cycle of the signal. That probably explains why it could work for a few seconds (while the coupling cap gets charged) then quit.

I believe the input circuit will need to be revised. A simple diode (Schottky?) clamp might be a solution.
The PIC and XTAL for my NG2 should arrive today so I should be able to take a closer look.

Can anyone comment on whether or not the 22mS time should be OK?

Malc C

Apr 16, 2008, 05:09 PM

Andy,

Did you read Jeffs post in the link above

Problem solved! :D It turned out to be the apparently not-so-equivilant transistor I had used.

When building up the NG2 I just moved over the 2N3904 I had used in my M&M build, and had similar results. The old Futaba FG worked but the newer 6EXA didn't. The 2N3904 was listed as being a cross for the BC550C.

So then I went with a BC107 as listed on Alessio's schematic (which I had) and it worked for both builds! I then also tried a 2N2222 and it worked just fine too.

Back to the scope. As I mentioned previously the amplitude of the signals at the base of the transistor were the same. But, at the collector/pin 21 of the pic there is a difference. With the BC107 it is 4.9v with both transmitters on both builds. With the 2N3904 it is 4.2v (FG) and 2.8v (6EXA) on the M&M build. On the NG2 it is 3.8v (FG) and 1.7v (6EXA).

I looked at the data sheets for these transistors but I don't know enough to tell what might cause the difference.

At least it made me get out and play around with the scope again. It's been fun building these up, and I might even use them with a simulator now! ;)

Jeff

Changing the transistor seems to make the difference

alessio70

Apr 17, 2008, 04:43 AM

I also posted this in the "other" USB joystick thread. I think

However, its only about 1.4 Vpp and apparently AC coupled. With the DX7 output connected to the direct coupled scope (10 Meg impedance) the signal goes between +.4 and -1.0 V. The levels are determined by the duty cycle of the signal. That probably explains why it could work for a few seconds (while the coupling cap gets charged) then quit.

...

Can anyone comment on whether or not the 22mS time should be OK?

The 22ms is perfect...

if your measurements are correct you could be right about AC coupled output signal.... even if it seems quite strange because someone seems to get his dx7 working... anyway if you are right a workaround could be to try to invert the polarity (-0.4 to 1V) and try to lower the value of r1=22k if needed (you could try for example with a trimmer...) and if possible use a transistor with a high hfe.... (bc109c for example... test it with a multimeter if you can)

let me know....

alessio

Andy64

Apr 17, 2008, 04:53 AM

JDill is comparing two different vintage Futaba transmitters. I also have an old 4 channel Futaba and it puts out 5 volts peak to peak with the the bottom of the pulses at essentially 0V. It should work fine. I don't know what JDill's newer Futaba puts out but, if its just lower amplitude, a higher beta transistor could make a difference. Or, if it puts out 5V p-p and is AC coupled, the output signal would go from approximately +1.4 to -3.6 and that could work but might require a higher beta transistor than the original (my guess is that's what's happening).

In the case of the DX7, look at the picture from several posts ago. That's very similar to what I see if the zero graticule line in the waveform is also zero volts. The signal needs to go at least .7V above zero for the circuit to work. Ain't gonna happen with the present design. I'm sure there's a simple, reliable fix.

As soon as I get a chance, I'll take a look. Unfortunately, I found out today that I need to update my PICSTART Pro programmer's hardware to work with these chips. Its on order but probably won't be here 'til next week.

JDill

Apr 17, 2008, 10:42 PM

My older transmitter does indeed put out a +5v peaked signal. The newer 6EXA is just a lower amplitude, starting at 0v and peaking at 3.9v.

Andy64

Apr 18, 2008, 12:01 AM

Here are a couple of mods that work well with my DX7 and old Futaba. I don't have my board working yet (waiting for update to my programmer) but I breadboarded these simple circuits for the input stage and they both work.

The circuit with the two added resistors simply biases the input stage so that the signal from the DX7 swings from approximately 0V to 1.4V. When a transmitter with a direct coupled output is attached, 0.1 ma is applied to the transmitter- should be no problem. Note also that this circuit would probably work with a transmitter that has an open collector output (the 39K resistor provides the pullup).

The one with the diode is just a simple clamp. I used a common 1n4148 general purpose diode. However, a diode with a lower forward voltage would be better - such as a Schottky diode. I also tried a 1N5818 (a fast recovery schottky rectifier diode) and it worked almost perfectly.

alessio70

Apr 18, 2008, 02:43 AM

Hi Andy thanks for the tests.... I prefer the first solution (the one with the bias) even if it adds two resistors.... I was tempted to implement it since the beginning but had choosen not to implement it to keep the number of components low... thinking that no radio would have had an AC coupled output... but really it seems that it is not the case....

alessio

Andy64

Apr 18, 2008, 12:24 PM

My concerns about the biased version is that I'm not familiar with all the possible versions of output signals from the radios. That circuit works well with my DX7 but I don't know what the signals from e.g. 12 channel radios or PCM radios look like. Are there radios that put out positive rather than negative pulses? The values selected for the bias resistors depend on the amplitude and duty cycle of the signal. The higher the amplitude and the lower the duty cycle, the more bias that's needed. I just took a look at the signal from a Futaba 4 CH AC coupled signal (I put a .22 cap at the output of the TX). The signal is about 5Vpp. It works OK. In the process, I also changed the bias resistors to 100k and 390k which is a better choice. Bottom line: the bias circuit would probably work OK with all radios - I don't know for certain.

If I were designing it from scratch rather than making slight changes to an existing design I would probably put a cap at the input (AC couple the signal) then use a Schottky diode for the clamp. That way I wouldn't need to worry about the signal as long as its at least 1V pulses. A pullup resistor could also be used at the input.

I want to thank you guys for all the work you've put into this project.

alessio70

Apr 19, 2008, 03:18 AM

My concerns about the biased version is that I'm not familiar with all the possible versions of output signals from the radios. That circuit works well with my DX7 but I don't know what the signals from e.g. 12 channel radios or PCM radios look like. Are there radios that put out positive rather than negative pulses?
yes, and with directly coupled signal the firmware should be able to read both types without problem.... but with AC coupled input it all depends on the bias and capacitor value... if the capacitor is quite great the signal will look like the directly coupled one inside the 22ms cycle so should not be any problem: just put the bias voltage near 0.7V and that's all.... but if the capacitor is very low you should put into consideration that the firmware for a PPM signal is set to receive interrupts on the signal falling edge (RB0/INT rising edge because of the transistor inversion).... so should be better a bias slightly over the 0.7V threshold (as you made in your mod) for both positive and negative PPM signal (don't know if I have been clear enough here :-) but it is difficult for me to explain without drawing the signal....)

anyway I think your design should work for all inputs... I prefer the 390k/100k values too...

...
If I were designing it from scratch rather than making slight changes to an existing design I would probably put a cap at the input (AC couple the signal) then use a Schottky diode for the clamp. That way I wouldn't need to worry about the signal as long as its at least 1V pulses. A pullup resistor could also be used at the input.
project.
AC coupling in input would be the best choice for isolating the radio but unfortunately would cause some problems in recognizing PCM signals.....

alessio

chintancm

Jun 05, 2008, 12:06 PM

if any one has tested these above method then please let me know is it working or not..for me both the method is not working...i've not tried with 390k/100k..

alessio70

Jun 05, 2008, 12:12 PM

if any one has tested these above method then please let me know is it working or not..for me both the method is not working...i've not tried with 390k/100k..

have you a dx7 radio?

chintancm

Jun 06, 2008, 07:03 AM

ya i have dx7 radio and i did above test on dx7...but no success...ur original ckt is working fine with futaba and hitech radios..

so i need the modified and tested input stage for DX7..

Andy64

Jun 06, 2008, 07:57 AM

Both circuits work well with my DX7 and Futaba. If the circuits work with your Futaba and not the DX7, it is likely that the transistor you are using has a lower beta than a 2N2222. The DX7 signal amplitude is lower than for the Futaba. You might try changing the 22K resistor to e.g. 10K or change the transistor.

alessio70

Jun 06, 2008, 08:34 AM

ya i have dx7 radio and i did above test on dx7...but no success...ur original ckt is working fine with futaba and hitech radios..

so i need the modified and tested input stage for DX7..

if the same circuit works with futaba and hitec then the problem should be the input stage so you could try to do a test: use the 390k/100k solution replacing the 100k with a 220k trimmer and try different values near 100k (near the center of the trimmer) to adjust the bias point.... (or if you prefer 39k/10k replacing the 10k with a 22k trimmer)

alessio

chintancm

Jun 10, 2008, 06:32 AM

Sonex128

Jun 10, 2008, 03:35 PM

Has anybody that built this got parts from Digikey or Mouser? I was trying to make a list of what to buy. Anyone have a list of what they bought? Is there a better place to buy the parts in the USA

Brian
Claremore, OK

Skylined

Jun 16, 2008, 11:32 AM

I've tried the circuit last weekend with a couple of friends and we couldn't get it to work.
We didn't have much time, so we're going to check the circuit this week.

We're using the art2003 (http://www.foxdelta.com/products/art2003.htm) programmer together with WinPic800 (v 3.63b).
Do you have any idea of which settings should we use under WinPic800?
It seems that the PIC got everything right, but we want to double check everything.

We've connected the radio (Futaba 9C) to the circuit and then to an osciloscope, and we got some data, it seemed that the circuit and the radio were working.
Can you please tell me which values should we get?

alessio70

Jun 18, 2008, 02:44 AM

We've connected the radio (Futaba 9C) to the circuit and then to an osciloscope, and we got some data, it seemed that the circuit and the radio were working.
Can you please tell me which values should we get?

when you connect the radio and turn it on the led also turn on? the interface is recognized by windows?

alessio

Malc C

Jun 18, 2008, 08:24 AM

We're using the art2003 (http://www.foxdelta.com/products/art2003.htm) programmer together with WinPic800 (v 3.63b).
Do you have any idea of which settings should we use under WinPic800?
It seems that the PIC got everything right, but we want to double check everything.

To quote their website
Programmer takes its power from Parallel Port pins 2-9. VPP for this programmer is 5V and hence "LVP" must be selected in programming software.

That's the obvious one, and as its through the LPT port I'm not sure if you would get any feedback if the PIC failed to program.

Skylined

Jun 22, 2008, 03:50 AM

Malc C

Jun 22, 2008, 04:18 AM

Malc, yeap, I've used LVP, but there are other settings and I'm not sure if they're right.

Looking at the PDF the settings are shown as a screen shot. I take it you've set winPIC800 up as shown ? If you have are you able to varify that the HEX code has been programmed correctly. My guess is that the programmer isn't actually programming the chip.

Recommendation: - buy yourself one of the PicKIT2 programmers. It seems to do nearly all the chips Microchip produce, and is fairly cheap.

http://ezkits.illumicon.nl/pickit2/pickit2skblue.jpg

The alternative, if you really want to get into this hobby, have a look at the EasyPIC5 development board.

http://www.circuit-ed.com/easypic5/images/easypic5_550_3.jpg

It makes project development a breeze, no need to keep removing the PIC from the programmer and plugging it into the proto board, just re-load the hex, write to chip and test ! - Well worth the investment. Its now made my three other programmers redundant !

alessio70

Jun 22, 2008, 03:11 PM

alessio, the PC tries to recognize what that thing is, but it could never recognize it.
I don't remember about the LED, but when does it light up?

Malc, yeap, I've used LVP, but there are other settings and I'm not sure if they're right.

My friends are trying to upgrade their Microlabs programmer to accept this PIC.

if the pc does not recognize the interface it is probably a programming issue.... have you build the through hole pcb on our site? the led should light up when the interface recognize the radio signal.... but here the problem seems to be elsewhere....

Skylined

Jun 22, 2008, 04:05 PM

Malc, I don't really want to spend money on that kind of stuff right now, unless I get into this PIC world, but anyways, thanks for the recommendation.

Alessio, I've built the hole trough version.
Tell me in which of these situations should the LED light up.
1. Radio connected, USB disconnected.
2. Radio disconnected, USB connected but not recognized by PC
3. Radio disconnected, USB connected and recognized by PC

My main point is to troubleshoot this without bothering you nor wasting your time. :)

Malc C

Jun 22, 2008, 04:50 PM

No problem, just thought you might like a few suggestions if you were planning on getting into this game ;)

Oh and I'm sure you're not wasting Alessio's time. This is what this forum is for, helping people out when things don't quite go as planned.

One thing that I came across with any verion of USB interface (see the sticky above) is that if the crystal is the wrong value, has a dry joint, or the two capacitors are wrong value or suffering a dry joint this can stop the PIC running and cause the sorts of errors you are having. - Try checking the board for dry joints

I asked if you were able to read back and varify the code after programming, can you confirm this ?

The final thing to try is reversing the D- and D+ connections on the board. I know it sounds silly, but I found that can also work in getting the device recognised by the PC.

alessio70

Jun 24, 2008, 03:33 AM

Alessio, I've built the hole trough version.
Tell me in which of these situations should the LED light up.
1. Radio connected, USB disconnected.
2. Radio disconnected, USB connected but not recognized by PC
3. Radio disconnected, USB connected and recognized by PC

none of the above :) it light up only when the Interface is configured from the host and there is a signal in input.... in other words when all seems to work ok... (anyway I've put the led mainly for testing the input signal because you can see on the pc if it is recognized and configured correctly by the host.... for example in windows you can see it in control panel.....)

I don't know your programmer but you should be able in winpic800 to see if the code has been programmed correctly don't you? (with verify function.... or simply read it)... anyway double check also the usb cable and connector.... and also the pinout... I have had also a problem in the past with a pic that worked correctly but had a bad usb...

alessio

Skylined

Jun 24, 2008, 09:46 AM

The crystal is 4MHz, and it's working, we've checked it with an osciloscope.

We've always verified the PIC and it was always OK.

I'm going to put to test all the recommendations that you gave me. :)

NoX_sp

Jun 27, 2008, 03:19 AM

Hi, all my problem with DX7 has been solved. in my case i started with 39k/10k and 2N2222 method but initially it was not working, then i changed 10k with 22k preset resistor and started testing with different values...and i got the success near by 17k instead of 10k...later i removed preset and put 22k resistor.....

so, if any one have the same problem like me then try this 39k/22k 2N2222 method..

Andy and alessio, Thanks for ur efforts

Hi,

I think this is a great project, and I'm going to build it, I have the DX7 so, as I see the schematic needs some modifications, just for clarifying, the circuit for the DX7 is this, right?

Sergio