[Arthur P.]

Mr.RC-Cam, on Nov 9 2007, 02:11 PM, said:

Arthur P., I have seen some of the rcgroups posts regarding your efforts. The build details definitely look fascinating.

If the HC221 interface does not work out then there are other solutions to try. The next attempt would probably be best solved with a dedicated microcontroller at the helm.

I may well try a dedicated controler. I have a few Holger ESCs which refuse to work properly as ESC but which have perfectly fine ATMega8's sitting around. I may well remove the FETs and parts related to driving them. That leaves me with a nice small ATMega8 PCB with ISP, serial and I2C ports. And with the info here http://www.rcgroups....ad.php?t=714299 on the digital signals coming off of the two Spektrum AR7000 satelite receivers it should be possible to just feed them the serial datastreams and put out a real PPM signal which would not be at risk of reordering of the channels as each digital datavalue for a channel includes the channel number.

[MorenoEnNYC1]

Hi,

The layout below represents Arthur's final design. I'd like to move forward in having it etched and completed. The 1st layout includes a 3.3v regulator @ 400ma..not sure if this will be a good solution. The 2nd is without the regulator.

BTW, Terry over at RC Groups produced both layouts.

Any feedback is greatly appreciated.

Thanks

Attached thumbnail(s)

• 

[MorenoEnNYC1]

without 3.3v regulator

Attached thumbnail(s)

• 

[Mr.RC-Cam]

The one with the 3.3VReg would probably be best. I don't see the usual loop stability caps on the Vreg, so be sure to check its data sheet very carefully.

[MorenoEnNYC1]

Mr.RC-Cam, on Nov 12 2007, 07:38 PM, said:

The one with the 3.3VReg would probably be best. I don't see the usual loop stability caps on the Vreg, so be sure to check its data sheet very carefully.

Will do, thanks Mr, RC Cam!

[chargenut]

Hi Mr. RC CAM,

Good point on the Caps on the VReg... I'd like to say I checked this before bolting the component onto my design... but I didn't, having said that... it appears from the Datasheet that Caps aren't required on that VReg (but can be added for good practice/better stability).

Looks like I got lucky in not adding them!!!

What I wasn't sure of, is whether a 400ma VReg would be sufficient to run the circuit...

Anyway, if anyone spots any other issues let me know,

Cheers,

Terry.

[Mr.RC-Cam]

Quote

What I wasn't sure of, is whether a 400ma VReg would be sufficient to run the circuit...

The circuit will draw only a few mA's.

[brashley]

I am is the same boat as Arthur with DX7 (but just starting on MK). Has anyone made any more progress. Is this realy a stable solutions?

Thanks,
Richard

[brashley]

I think I need a little bit of help here. I made the circuit and eventually (after 2 mistakes) got it working. It was not reliable though. Some of the PPM channels would drop out (on the scope they would flicker on and off) if I just let it sit. I started checking things and found that the PWN signal was only 2.2 V max. If I unplugged it from the converter it was back to 3V. The HC221 is running at 3.1V and the PWM signals after the diodes were now down to sub 2V and I was thinking that that might start creating problems for the HC221 to reliably trigger. I suspected (I am not a EE) that the 2.2K resistor at R3 was too much of a drain on the AR7000. ( As an interesting point, the channel seven pulse from the HC221 would show up at ~3V after the diod). I decided to change the resistor at R3 to a 10K. Voom, the PWM out of the AR7000 was now at ~3V and appeared very stable. Even the MK started loving the PPM. (with the 2.2k at R3 the channels would bounce around and sometimes the MK would reports signal loss, that’s when I hooked up the scope and could see the channels randomly dropping out on the PPM side)

The only problem is that now I lost a channel. (only 7 pulses in the PPM signal) The channel seven signal out of the second vibrator was not generating a PPM pulse (from what I can tell). How do I get this back to work and stable. Should I increase the R2 to generate a longer pulse?

Please help.

Thanks,
Richard

This post has been edited by brashley: 06 January 2008 - 07:36 PM

[Mr.RC-Cam]

Using a 10K at R3 is reasonable.

(1) Since you have a scope, can you monitor all the pins on U1B and report what you see?

(2) With the circuit connected, use a Y-cable and add a standard servo to the Rx's CH-8 port. Does the servo work too?

(3) Grab the magnifying lens. What is the HC221's exact part number and brand.

[brashley]

Thanks for the quick response. After working on it all day yesterday I just type up the message and walked away.

for (3) I used CD74HC221M96 from TI: http://focus.ti.com/...k/cd74hc221.pdf

I will rehook it back up and test the rest now.

Thanks, Richard

[brashley]

for (2): servo works with Y in 7th channel (last channel in sequence and the one that is hooked up to U1B - A)

This post has been edited by brashley: 07 January 2008 - 04:56 PM

[Mr.RC-Cam]

Quote

(3) I used CD74HC221

When powered with 3V, that is a compatible part. When using 5V, the CD74HCT221 seems to be the better pick in this application.

Quote

(2): servo works with Y in 7th channel

Ok, that is good info.

Seems like it is time to get out the scope and snoop around all the pins related to U1B.

[brashley]

Ok, working on it. I am going to take some pictures of the scope as well. Easer to show than explain. (and yes, I am using a 3.1V regulator for the HC221)

This post has been edited by brashley: 07 January 2008 - 05:45 PM

[brashley]

(part 1) Here are some pictures (the scale is one volt per division) for the PPM-out picture the 0V is at the bottom, for the rest it is about one division up from the bottom. I only have one good probe so the bottom trace is accurate. For the pictures with 2 traces, the top trace is from my bad probe for reference and is set on the PPM-Out (U1A-Q-pin13) so that you can at least see the signal timing in relation to the output. Except for the first picture (ppm-out) the time scale is 1ms/devision.

PPM-Out:

U1A-B(pin2):

U1B-A(pin9):

This post has been edited by brashley: 07 January 2008 - 07:31 PM

[brashley]

(Part 2) Here are the rest of the pictures. Other pins: U1B-C(pin6)=Ground, U1B-R(pin11) & U1B-B(pin10) = 3.1v

Thanks for the help, Richard

U1B-Q(pin5):

U1B-R/C(pin7):

This post has been edited by brashley: 07 January 2008 - 06:26 PM

[Mr.RC-Cam]

U1B-5 has the correct pulse with good amplitude. That would suggest that there is something wrong with D9 or the wiring related to it.

Try this with your 2-channel scope. One probe on U1B-5, the other on D9 cathode (that is, the D9/R3 junction). Sync on U1B-5. You should see an instance of that pulse at both places. The amplitude will be one diode drop less at D9, but other than that, it should show up the same on both places. Of course D9 will have other rising edge pulses from the PPM train, but just ignore them.

Post a photo of what you see.

[brashley]

I will hook up both probes and take a picture. Actually though, you can see that (just not in the same picture) above. The D9/R3 junction is essentially U1A-B(pin2) and I have the U1B-Q(pin5) picture as well. The PPM-Out at the top syncs them up. You can see the Pin5 pulse at the D9/R3 junction (at the end ~.2v above the rest).

[brashley]

Here it is- the top trace is D9/R3 junction and the bottom is pin5. The dip down in the top trace is my bad probe, it does not show up with the good one. Both are set to 1V/devision. Top trace 0 is at the middle, bottom trace 0 is at the bottom.

Attached thumbnail(s)

• 

[brashley]

Here is a picture of my board (not as pretty as it once was). Some how I messed up the layout of the voltage regulator (got the pins in the wrong order) and it caused me to have to cut some traces, flip the part around and jumper it back in with a white wire and some solder bridges. Live and learn.

Richard

Front: Back: