[Solved] Voice Mod -AudioMega by Oliver - pls. help to debug

[LTMNO]

What memories. I had 4 of these.

[andrewju]

4? I thought OSHPark produces boards in multiplies of three!

[LTMNO]

I actually build them from scratch, I was the ginny pig, during the development cycle.... I then sold them to someone in the UK for practically nothing...I couldn't get any if then to work. Then I built my first Hans board and all was good. That is what is in my radio now. Then I built another for fun. .

[LTMNO]

Great looking board. Have to say. Oliver designed it beautifully.

[jhsa]

I also built the Hans board. two of them for both my radios.

Ok, concerning the regulator, I think I would only remove the 5V regulator, get the Battery voltage for the audio board from the cap as I have it now, and just make a connection from the 5V on the voice board directly to the 5V line on the main board. Like that the radio should also work when powered from the trainer port.
This needs experimenting of course.

João

[LTMNO]

The few and the proud. Sound like a beer commercial.

[andrewju]

I actually build them from scratch, I was the ginny pig, during the development cycle.... I then sold them to someone in the UK for practically nothing...I couldn't get any if then to work.

Did you ever find out what the issue was?
I built three boards. The first board I built had issues with Atmega, and only the third chip worked well on it: one Atmega programmed Ok, but didn't go past bootloader, and second one had problems with BL control. Third one works good though, and I use that board in my radio now.
The other two boards that I built worked Ok right from the beginning (I have tested these, but I currently don't have enough radios to put these boards to).

Ok, concerning the regulator, I think I would only remove the 5V regulator, get the Battery voltage for the audio board from the cap as I have it now, and just make a connection from the 5V on the voice board directly to the 5V line on the main board. Like that the radio should also work when powered from the trainer port.

When I think about it now, I do not see a reason why this option will not work.
So go for it, I'm quite sure it will work fine!

[jhsa]

Question, is there a problem to parallel the regulators? I mean not removing the radio's regulator?

João

Sent from my GT-I9195 using Tapatalk

[andrewju]

What's the point of doing this? It will probably work, but one regulator will output slightly higher voltage than the other. Even if it's just a few mV, it will still be enough to make the current flow the opposite direction. And I'm not sure how the other regulator will react on that. I guess they're not designed to work this way.

If you're just worried about VR's capability to power the entire board, I think the stock regulator is still fine even with Atmega 2561. Because the current consumed by the mainboard itself is still less than 100mA. With the voice board installed, the BL and all voice features are powered from the VR on the voice board, so they make no additional load onto the stock VR.

[jhsa]

Not if you get the Taranis gimbals with opamps for example. The current might increase over 100mA.

I was just curious about paralleling the regs. Wouldn't do it of course. Thank you for your answers.

João

Sent from my GT-I9195 using Tapatalk

[andrewju]

Not if you get the Taranis gimbals with opamps for example. The current might increase over 100mA.

I do not think opamps will consume that much power (I guess it should be less than 10mA), although it may be interesting to check.
Anyway, I think all recent Taranis sticks come without the amps. So it shouldn't be an issue at all (if you don't already have the sticks with amps that you plan to put to your 9x).

I was just curious about paralleling the regs. Wouldn't do it of course. Thank you for your answers.

Maybe someone more knowledgeable will step in and correct my hypothesis regarding paralleling regs. What I wrote above is just an assumption.

Please post back when you do the power mod. I'm also interested to know how it works for you!


Thank you for your support!

[jhsa]

One advantage I see (trying to find an excuse to do it ) is that I think the reg on my voice board is a low drop. At the moment I'm using a 6 cell pack on my radio and I can't fully discharge it as I have to set the alarm at about 7.2V. If that is true I could have more flight time. I will have to check the datashet. I'm using an eneloop pack that I made.

João

Sent from my GT-I9195 using Tapatalk

[andrewju]

Yeah, that is a good reason. My radio worked from 6V when I measured the current yesterday.
Also, having 1 reg instead of 2 will slightly increase the efficiency.

And ... it's fun! So there are simply no reasons not to do it!

[jhsa]

You're absolutely right. That is why both my 9x's have most of the mods done. I think I only didn't do the haptic and dsm. Reason? don't like electronics shaking and plane crashers.
Next mod will be the m2561. That means redoing all the mods again. Fun....

João

Sent from my GT-I9195 using Tapatalk

[andrewju]

Yeah, I know that feeling! I now find myself looking to buy another radio just to do the upgrades again... Especially since I have a couple of spare voice boards laying around...

[ReSt]

regarding paralleld VRs.
I have an ESC that went dead.
The BEC circuit consisted of three parallelled 78D05AL VRs

Reinhard

[andrewju]

The BEC circuit consisted of three parallelled 78D05AL VRs

Hmm, so maybe I'm wrong and the VRs are designed to work in such configurations. Then leaving stock VR on the board will increase redundancy (at a cost of an increased power consumption, but this is not that important).

Quick googling brought up this thread. As last message states, "in practice , it does seem to work". But if one wants to do it right, it's not as easy as just a parallel connection of VRs' outputs:

The problem comes in when one wants to supply, say, 8.03V, and the other wants to supply 8.05V. The 8.03V one will do most of the work and then overheat and shut off. So simply hooking them up in parallel is not really a good idea.

But there may be a way around this. I would try the following (just as an experiment)

1) Put a small series resistor on the output pin of each regulator. One ohm or so should be OK. Just makse sure the resistors can handle the power.
2) After the 1-ohm resistor, tie the outputs together.
3) Give each voltage regulator its own feedback circuitry, but put an adjustable resistor (potentiometer) in each pack.
4) Crank up the amps with a load of some sort, and adjust so that each regulator drives the same current.

It still might not work, but this stands a chance of working.

I guess it gets even more complicated if we use two totally different VRs designed for different current. Also, what will happen if the stock VR drops out, while the one on the audio module will continue working? Will we get a voltage spike when that happens? I think if one decides to go for this solution, it will need a lot of testing before flying a real model with it...

[andrewju]

Just wanted to share my findings regarding battery voltage connection point (in fact, it is not directly related to a voice mod). There is a simple mod that fixes an issue of the stock 9x where the beeper does not work when a simulator cable is plugged in. This can be done by cutting a track and making a connection nearby:



With this mod done, the beeper will work in both the 'regular' mode and the 'trainer' mode. RF module power behavior will not change - power to RF module will be supplied when the radio is powered on with the switch, and will not be supplied when just a simulator cable is plugged in.

In case if someone decides to modify the voltage regulators as discussed earlier in this thread and take +5V from an audio board to power everything (except an external RF module) - the green line on the picture is a good place to take a battery supply voltage from. The +5V line should be connected to the 'mainboard' as João suggested above.

[jhsa]

Won't the beeper's regulator be always powered even if the radio is off?
The voice module would also be on all the time. I think that the middle pins from the switch are direct battery supply?
Joao

Sent from my GT-I9195 using Tapatalk

[Kilrah]

Yes this is direct battery voltage. You'll drain your battery in a few days even when powered off.

[jhsa]

You want to power your voice module from the capacitor just under the 5V regulator. There you have batt power, but switched by the trainer plug and/or the main switch.

Joao

Sent from my GT-I9195 using Tapatalk

[andrewju]

No! The battery voltage is the lower right pin of the switch on the picture above.
Those middle pins are connected to the capacitor under 5V regulator João mentioned, so in fact there's no difference whether one takes battery power from the capacitor on the mainboard of from these middle pins of the power switch.

It took me two days to understand how it works.

But now I'm quite certain about it. So with the modifications as on my picture everything will power ON and OFF as it should. Fully tested on my system!!!

[jhsa]

If I remember well the lower pin conmects the charger port
to the battery so you don't charge the battery when the radio is on.

Joao

Sent from my GT-I9195 using Tapatalk

[andrewju]

The picture above is 'upside down' (sorry, this was the only one I found).
So charging pin is lower right when you open your radio, but on the picture above it's upper left.

[Kilrah]

Yes. With board the right way up Middle = battery, when switch is off it's connected to bottom = charge port, when on it's connected to top = radio supply rail.

[jhsa]

the power coming from the trainer port's switch doesn't pass by the main switch circuit.

[andrewju]

This post has been updated:

the power coming from the trainer port's switch doesn't pass by the main switch circuit.

It does, but it's not coming by a dedicated cable/pin. The power from the trainer port comes directly to VR on the mainboard, which is always connected to middle pins on the power switch.

I spent a lot of time on this, and I'm quite sure I'm right.

Here, I found another photo (it's not upside down this time) and marked pins on the power switch (this is the default, unmodified board):



Note: the picture is updated.

Upper left pin is the battery power, it is always connected.
Lower left pin is wired to the Charging connector.
Middle pins are connected to the mainboard, and may get the power when a trainer jack is plugged in, irrelevant to the power switch state.

The only mystery still remaining is how a battery is supposed to be charging. I really don't understand how it works, as there is no continuity between the positive pole of the charging port (after the diode, so it had no influence) and a positive pole of the battery.

So when power switch is OFF, the mainboard may get its power from the trainer jack. Whenever training cable jack is in, the motherboard will be powered up (and there will be power at the middle pins of the power switch). When the power switch is ON, the mainboard will get its power from the battery, and the RF module is powered via the power switch then as well.

[Kilrah]

Ah, better photo. On the other one we couldn't see that the 2 top pins weren't connected together.

Then yes it makes sense, if you've really beeped the connections and are sure of your markings it should be OK.

[jhsa]

Yep, and I just saw something? don't plug the charger if you are powering the tx through the trainer port. that means if you're flying sim.. Or am I seeing it wrong?

João

[andrewju]

Looking at what I posted, I now wonder how the power is supposed to get from the charging port to the battery at all... Hmm...

As I use LiPo, I never ever used charging port in the radio. So I'm not even sure it works.
I'll make some more tests with the power connector tonight.