er9x development

[Flaps 30]

No, flaps.. it isn't.. Don't forget that the resistive center of the pot is not necessarily the center of the mechanical part..
That is why the sticks still need to calibrate the center..

So you are telling me that the potentiometer electrical centre might not be correct due to the pot being mechanically rotated so that it doesn't give what should be mid voltage when the springs centralise the stick/knob or whatever?

[pmackenzie]

In an ideal world the centre value of the pot would be the average of the value at the two end points.
But in the real world it might not be.

Having the stick calibration routine get three points to use in interpolating the stick position from the analog input will be more accurate than just using the two end points.
Particularly when neutral/centre stick is so important.

[jhsa]

So you are telling me that the potentiometer electrical centre might not be correct due to the pot being mechanically rotated so that it doesn't give what should be mid voltage when the springs centralise the stick/knob or whatever?

That's correct.. that's exactly what I'm sayin..
And if Im not mistaken that is the reason why we also have an option to calibrate the center of the sticks..

João

[MikeB]

Such a joystick is not a flight control. If the 'centre' is off a bit, then you can always change the calibration values in eepe. You should not need to calibrate very often,if more than once!

Mike.

[Kilrah]

Not to mention the number of people using that holds in about 2 fingers

[jhsa]

True, those can't play with 2 sticks at the same time with only one hand..

Fot that you must use your thumbs..

[Oliver]

Idea:
You specify an input value, e.g. stick, pot, channel output, telemetry input, maybe even the output of another scaling function!
You specify an offset, multiplier and divisor, giving an integer output (these values may be negative).
Mike.

Is it only me thinks about it like about.... MIX? All of that looks like standard mix, when you have telemetry input as source and some variable (GVar?) as output. If we would have mixes like that... So it would be like VoiceSwitch in SafetySwitches menu. You can select, how many mixes use channel as output and how many - use 'variable' as output. Sorry, my english is quite poor, but GoogleTranslate, I think, will not understand, what I mean. You should.

[jhsa]

Crazy Idea:

Any chance of having Real Time Clock on the stock board also? I would guess we could buy the boards already? of course I would prefer to build it

As I said, just a crazy idea.. but I smashed my new alarm clock this morning. I think I would think twice before doing the same to the radio..

No. I'm joking, but it would be nice..

João

[kaos]

Crazy Idea:

[jhsa]

http://www.ebay.com/itm/New-Real-Time-C ... 3f19dba290



João

[ReSt]

Or this one
http://www.ebay.de/itm/400337658645?ssP ... 1436.l2648

with 56 bytes battery buffered Ram

Reinhard

[jhsa]

Yeah, I wonder how accurate this stuff is..

Mike, possible? In case we don't have free pins on the m64/128, could the sound module handle it?

João

I have the mouse cursor on the "buy now" button

[MikeB]

It may well be possible, but I don't have the time to implement anything for it.
You may have seen this post viewtopic.php?f=23&t=994&start=810#p41749. The general idea is an addon board for the programming, telemetry and voice, but as long as we put a powerful enough processor it may well be able to do all the main processing as well, and just use the M64 as an I/O processor (sticks, switches, display). I'm thinking about its possibilities, so it may get a RTC if it doesn't add significantly to the cost.

Mike.

[jhsa]

So, we will have all to buy the daughter board if we want some more coll features, right??

well, I'm not undoing all the mods I did now on my 9x's. I'm going to forget about the RTC.... for now..

João

P.S. - Hmmm, I might buy a couple just in case. maybe someone in the future decides to implement it, as it happened with the voice for example..

[jhsa]

Forget it, just found out the RTC is useless for what I had in mind..

[MikeB]

I'm thinking about the Voice and safety switches. Over time, voice output has 'happened', not quite fully designed. Until it has really been used it wasn't clear what was needed.
So, we've ended up with several ways of doing things, including some audio alarms for those that do not have voice, or want a different alarm.

We also have a possible 16 safety switches, I think this is more than are really needed, and the safety function can be handled by a 'REPLACE' mix if extra ones are needed.

Keeping the EEPROM structure stable I have the following proposal:

1. Add 4 new 'Function' switches. These may be used to provide the safety switch function, and other things still to be defined.
2. The existing 16 safety/voice switches will be changed into 10 voice functions.
These will have 4 parameters
a) a switch, if switch is ---, then function is disabled
b) a mode, 4 possible, only two defined so far SWITCH and VARIABLE
For SWITCH mode,
c) File to play when switch turns ON, 0 for no file
d) File to play when switch turns OFF, 0 for no file
For VARIABLE mode
c) File to play as prefix, e.g. "Flight Battery", "Altitude", 0 for no file as prefix
d) Variable

If you want regular repeating messages you define a custom switch as a timer.

So you could have:
SWITCH ELE 107 0 - equivalent to the existing ON ELE 107
SWITCH AIL 107 108 - equivalent to the existing BOTH AIL 107, except the second message does not need to be the next in sequence.
VARIABLE GEA 120 A1 - equivalent to using two existing switches one to play file 120 as the prefix and a second to play the variable A1

I might manage 11 of these Voice functions.

If voice is NOT enabled in the speaker mode, I'll might change things to allow selection of an audio alarm instead of a voice, or I might add an AUDIO mode to SWITCH and VARIABLE.

Mike.

[pmackenzie]

Stupid question - what does the safety switch do than a replace mix doesn't?

I have never used them , and have no idea why I would want to.

Pat MacKenzie

[jhsa]

saves a mix?? There are only 32 allowed, right?

[gohsthb]

Safety switches are calculated after the mixer, limits, and channel reversing. So say you setup a safety switch on the throttle channel, that holds the channel at -100. You also set your usual replace mix on the throttle channel, but using a different switch. Now with the replace mix activated it holds your throttle channel at -100. You then go to the limits menu and accidentally reverse your throttle channel. The esc will now see the signal for full throttle. With the safety switch it would stay at -100, since it is calculated after the limits and reversing.
-Gohst

[pmackenzie]

If that is all it does, then I am not sure it is really needed. Or at most there should be only a couple of them.

[rabono]

What i really do need is some variable for the voice naming since i got a six positions switch instead of P3,like in this post viewtopic.php?f=35&t=1486
my friend Jhsa got the voices working for each 6 positions of the p3 ,we got ID1 and ID2 override P3 and do the announcement for the override but i need a variable for the ID0 since it should play the one of the variable 6 positions of P3
I am wondering if this can be implemented?

[HC1969]

These are good ideas, But I need to limit the number of audio switches?

[ReSt]

And safety switches have one disadvantage.
It may cause a problem to use the channel output of a safety switched channel as input to another mix.

e.g.
you have throttle stick mixed to channel 03 and a safety switch to channel 03 with -100
If the safety switch is active, the channel signal is -100 according to the safety switch setting, but the channel value depends on the stick position and may be +100
So, if you use channel 03 as input to another mix, that mix is working with the +100 value whilst the bar graph display shows -100 for channel03

Reinhard

[MikeB]

I should issue a new revision of er9x later today:

Fixes/Features:
Correct voice of negative values
Correct Voice of exactly 100
Raise voice volume for inactivity warning
Add differential option for mix curve
Use timer 2 for high speed audio, timer 1 just for 10mS
Add channel outputs as inputs to GVARS

Mike.

[jhsa]

Use timer 2 for high speed audio, timer 1 just for 10mS

I'm not getting this one??

drinkin too much beer lately I suppose

[pmackenzie]

How does the differential option for the mix curve work?
Is it a way of getting differential without using a curve, so different rates "up" and "down"?

I am going to have to dig into the GVARS, have not even looked at them yet and it seems like it could be quite powerful.


Pat MacKenzie

[pmackenzie]

Use timer 2 for high speed audio, timer 1 just for 10mS

I'm not getting this one??

drinkin too much beer lately I suppose

This probably refers to the mcu's timer peripherals and how the firmware uses them, nothing that a user would ever have to know about.

[jhsa]

AAHHH !! Ok, only for the Gods, not for the mortals..

I thought I was drinking too much.. No worries then

João

[MikeB]

Yes, the timer 1 for audio thing is the processor hardware timer allocation. I mention it just in case something odd happens with the audio. It should all just be better.

Pat: Yes, different rates up and down. You always have full rate in one direction, but set a percentage reduction in the other direction, no curve needed. This percentage may be a fixed value, or a GVAR to allow it to be variable. The direction reduced depends on the sign of the differential value.

I have allowed a channel output to be the value of a GVAR. So a channel goes from -100% to +100%, the corresponding GVAR goes from -100 to +100. One use of this is to limit the range a GVAR could take. You could use, say, P1 as the value of the GVAR, but this would go from -100 to +100, but you might want to limit the GVAR. If you use the GVAR to control your dual rate, you could set P1 such that you disabled the control (DR=0). Instead, you could use a curve on a channel to limit the range of the GVAR to, say, 50 to 100. Then you can't turn the D/R down below 50% so you always have control!

Mike.

[jhsa]

that is a good idea Mike.