Will definately report back, i have a scope but a bit of a n00b when it comes to actually using it, :-), although i managed to find the PPM trace on a old 35mhz Rx and solder the wires to get it working as a wireless sim box.

Looking to try build up one of these: http://electronics-diy.com/avr_programmer.php

http://www.rcgroups.com/forums/showthread.php?t=1177548

read that ok starts bad but ends well.
you only need 25dsub and cable rest not needed at all.
and after that build an usbtiny. cheap and you can use it in avrstudio.

Only issue is that locally the "ATAVRISP2" is around $83, vs the $35 from Digikey (granted buying it from there would be cheaper but shipping would probably make it as expensive).

Looking for a fairly cost effective way to make this up as im probably only looking to make 1 or 2 (one for the sim and maybe a second for the flightbox as a buddy box), :-)

Originally Posted by helidave1

Hi,
You do not need an external oscillator with this chip. The onboard one does the job just fine. If you use the fuse calculator (easy to find on the net) just pick the standard default settings without the clkdiv setting.
I used Ponyprog to load the hex file into the chip and set the fuses and it worked well. Simple parallel port programming method. The only reason it took me so long to get up and running with this was work got in the way!
since sorting out my signal order and fuse settings I now have this chip working with almost all the RX units in my box. Only problem ones are those with non sequential signals (Graupner IFS) and my 35mhz JR Spcm although I think I just need to modify the code for this one.
Cheers
David

@Helidave1 and SadSack ( and anyone that can help)

Well i think i was already kind of rusty when it comes to "pure AVR" programming and burning and AT90S2313 is so old and simples there even isn´t a lot of options and fuses.
After a while i finally found a fuse to enable the external XTAL and now at least i get a coherent waveform. ( yes i´m kind of ashamed of myself too.. :S)

Here´s a video:

PPM SUM Converter - uC Approach - Waveform Response (0 min 50 sec)

It doesn´t look good does it?
How can i get nice clean square pulses out of it?

Also, it is inverted.. is there an easy way to invert them back again? Ithink it might be easier then trying to find the right place to alter the MultiWii code ( specially because the way Alex deals with the ints is still a little over my head.

Anyone... ??

bad rabbit!!!
stay in your basket

Originally Posted by SadSack

these may not work and i'm almost 100% sure they will not work

But hey...
No2 could be the one LOL

Failing that go a page or 2 back. I had same problem and how i got around it.

Should one work, tell me and i'll delete or get rid of them thanks

Nope... none of the 3 hex files worked for me..

The third one gives the "less ugly" output, but it still is an ugly output.

@all the people that got this working, how does your output waveform look like? Did you manage to get nice clean square pulses like what comes out ou the Rx ,as i think it should be? Or you managed to make it work even with such an ugly and messes up output ?

SadSack, i´ll check to see if a have any 4.7V Zeners around and maybe i´ll try that hardware inverter of a page ago.

But i still would like a way to get a clean output if that´s possible.

not pretty but works


so signal didn't invert then? wasn't trying to make pretty signal.

Originally Posted by SadSack

not pretty but works


so signal didn't invert then? wasn't trying to make pretty signal.

LOL i know SadSack! You are being of much help to me already!

What i meant about the ugly signal is that maybe if it wasn´t clear for me, it could not be clear to the microcontroller.. so to speak..

Anyways it looks better now, and your third hex is the one that has come closer to what i expect.

But i still cannot make MultiWii detect it , aparently. I´m using the ground station program to check it and it just does not detect the sum signal.

well if using your scope signal didn't invert. Would be fair to say multiwii wouldn't see it at all because nothing changed

To be fair i'm the last person to help out with my random code changes lol

More a code butcher

It did invert.

With your hex , and after some adjustments on my setup, the signal is very coherent now. As you can see in the picture.

I´m trying to detect it with MultiWii ground station but it doesn´t

Originally Posted by richardkalaf

It did invert.

With your hex , and after some adjustments on my setup, the signal is very coherent now. As you can see in the picture.

I´m trying to detect it with MultiWii ground station but it doesn´t

sorry just re read post. but don't remember which edit was on v3.
better get out soldering iron later see if i can change mine
thanks for input.

Hi,
I'm not sure if this helps or not but it seems to me that the problem may well be due to the fact that the signal shown on your scope is simply not "clean" enough for the edge trigger to be seen by the software.
The signal I get on my unit is a good square wave and the software I am using (shrediquette) sees it wether the signal is inverted or not, even though the software engineer says in needs to see a negative pulse.
Why do you need to use the external xtal? I thought this chip had an inbuilt oscillator circuit. As I mentioned in a previous post, mine works perfectly without xtal and with fuses set for 8mhz (no clkdiv8).
If you need a photo of my setup (just the chip on a veroboard with link wires, no other components) or if I can help in any way let me know.
David

Originally Posted by helidave1

Hi,
I'm not sure if this helps or not but it seems to me that the problem may well be due to the fact that the signal shown on your scope is simply not "clean" enough for the edge trigger to be seen by the software.
The signal I get on my unit is a good square wave and the software I am using (shrediquette) sees it wether the signal is inverted or not, even though the software engineer says in needs to see a negative pulse.
Why do you need to use the external xtal? I thought this chip had an inbuilt oscillator circuit. As I mentioned in a previous post, mine works perfectly without xtal and with fuses set for 8mhz (no clkdiv8).
If you need a photo of my setup (just the chip on a veroboard with link wires, no other components) or if I can help in any way let me know.
David

That's what i meant! I beleive that my signal is not clear enough for the software to recognize it.

That's why i asked if anyone knows of any analog filters to make the signal nice and clear.

What program are you using to burn the AVR ? I'm using PROGISP and i havent found anywhere to set the fuses, it's only "half-working" now because i found a new version of PROGISP that has a "sort of fuse setting" that says "Enabled XTAL".

I'll try a little further to make it work without the xtal and see if that outputs a better signal.

Hi,
If the signal on the inputs is clean then the output signal should also be clean as the circuit is so basic I don't see where distortion can be generated (assuming you are testing the unit on its own and not connected to any other electronic gear).
I used PonyProg (free software) and a very, very simple parallel port program unit based on a few resistors and a 25 way D type connector. Works like a charm on a wide range of stuff and costs almost nothing to make. I think Sadsack showed a similar unit in an earlier post.
In addition the PonyProg software allows you to read the fuses/security bits and program them. I initially thought I had mis-read the order of the RX channels I wanted to use but in the end I noticed I had set the ckldiv8 bit and this was causing my problems. If the signal you are seeing on the output pin is a correct assimilation of the sum signal (all the channels are there and they all change when you waggle the relevant stick) then I think your problem is most likely the fuse settings.
David