Page 8 of 106

Re: er9x development

Posted: Fri May 11, 2012 10:45 am
by ReSt
MikeB wrote: The 'custom' altitude alarm is currently settable from 0 to 1020 (3346 ft) metres in steps of 4 metres (13ft). Is this actually too high for an alarm? Would 0-255 m (0-835 ft), in steps of 1 m (3ft) be better?

Mike.
255 m as maximum altitude is to low, for me.
I normally get up higher with my electric gliders.

I don't see a problem with 4m steps.

Reinhard

Re: er9x development

Posted: Fri May 11, 2012 11:12 am
by MikeB
The 255 m would only be the max height at which the alarm sounds, the reading, and max height reached will still be as they are.
Currently, there is only 1 height alarm setting (400ft, 122m).
Do you need an alarm at a height higher than 255m?

Mike.

Re: er9x development

Posted: Fri May 11, 2012 11:21 am
by ReSt
At these altitudes (of around 350m and more) it's hard to guess the real altitude of a model. And it's an altitude, where an uplift may be rather strong. So I would like to have that alarm limit to tell me, I have to go down immediately to stay in a safe (visible) range.
(My electric gliders are really 'vintage' models. 25 and more years old that have to be brought down carefully.)

Reinhard

Re: er9x development

Posted: Fri May 11, 2012 11:58 am
by MikeB
Fine, how about splitting the difference, 510m (1673 ft) max alarm height, in steps of 2m (6.5ft)

Mike.

Re: er9x development

Posted: Fri May 11, 2012 2:38 pm
by gohsthb
MikeB wrote:I've got a new version of er9x, still a work in progress, but lots of changes to telemetry. These are the changes:
Everything is working as I expect. The display update seems slow if you are looking at it. But I really shouldn't be staring at that while flying anyway. I think the filtering will be quite beneficial.
-Gohst

Re: er9x development

Posted: Fri May 11, 2012 3:27 pm
by ReSt
No problem at all for me to have larger steps. even 10m per step would be fine (for me)

Reinhard

Re: er9x development

Posted: Sat May 12, 2012 2:47 am
by ShowMaster
As long as the first 400 ft is smaller incrimints for us FAA hight limit please.
SM

Re: er9x development

Posted: Sat May 12, 2012 3:44 am
by ShowMaster
Er9x stock board only SP board r760 switch error?
I flashed a friend board and got a switch error that that throttle and ail switches were in the wrong position but they were? We found that setting them both towards the sticks made the errors go away. This is opposite of normal. The diagnostic screen says the switches are wired correctlly and not reversed. We had no Internet but both had downloaded the eepe file for the stock 9x board earlier and saved it.
I only use frsky er9x so I didn't try this non frsky version on my 9x.
Any ideas of anything else to try? Maybe an older version FW if there's a link.
SM

Re: er9x development

Posted: Sat May 12, 2012 4:32 am
by gohsthb
There are options now in the general radio settings to select the specific positions of the switches at start up. This is probably what has caused this.
-gohst

Sent from my LG-P999 using Tapatalk 2

Re: er9x development

Posted: Sat May 12, 2012 6:16 am
by ShowMaster
gohsthb wrote:There are options now in the general radio settings to select the specific positions of the switches at start up. This is probably what has caused this.
-gohst

Sent from my LG-P999 using Tapatalk 2
Thanks. I hope it's that simple. I searched each menu for such a switch setting in my er9x frsky r760 tx and couldn't find anything to set or change. A screen number with this setting would be great. I don't have the stock er9x tx to try so maybe it's only in that version FW?
Keep the ideas comming. An older version FW will be next. I think I can find the source. My er9x frsky works normally without the switch warnings.
SM

Re: er9x development

Posted: Sat May 12, 2012 8:05 am
by Camboui
MikeB wrote:Fine, how about splitting the difference, 510m (1673 ft) max alarm height, in steps of 2m (6.5ft)

Mike.
I use DHT-U as is, but I do not like the too big 10m step.

I might not be concerned anytime soon by the 9xFrSky mods, but I may suggest, may I ?
So steps of 2m up to 254m, and steps of 4m from 254m to 762m.
Or
Steps of 2m up to 254m, steps of 4m from 254m to 510m, and steps of 8m from 510m to 1014m.
Precision is less needed at higher altitudes. And I like non linear scales :mrgreen:

Re: er9x development

Posted: Sat May 12, 2012 10:35 am
by MikeB
I'll see what is possible. The problem is, with the way custom switches work, we only have a an 8-bit value (0-255) with which to compare the height.
We definately need an alarm that may be set at 400 ft (122m). There are several places where this is necessary. This needs to be fairly accurate.

Mike.

Re: er9x development

Posted: Sat May 12, 2012 10:46 am
by Kilrah
On the other hand, it's not like you'd lose a lot of valuable airspace if it sounded at 119m, or be judged guilty of airspace intrusion because you were at 124m, which nobody could prove... and the accuracy of the altimeter is often much less than that in flight anyway due to ram air, filtering delay,... I have yet so see any barometric model altimeter that's more precise than +/- 5m or so at short term during maneuvers.
I doubt steps <4m are of much use.

Re: er9x development

Posted: Sat May 12, 2012 3:46 pm
by MikeB
I'm trying 0-200m in steps of 2m, 202-800m in steps of 4m at the moment.

Mike.

Re: er9x development

Posted: Sun May 13, 2012 10:25 am
by Camboui
I saw you have an 8 bit storage limit.
I suggest 4 ranges that uses each 64 values (64x4=256, that's within 8 bit limit)
0- 63m in steps of 1m (we are limited to 50m in some places in EU)
64-190m in steps of 2m (ok for 122m)
192-444m in steps of 4m
448-952m in steps of 8m (enough IMO for extreme gliders)

Here is the code, if I'm not wrong, that gives altitude in the above ranges from an 8 bit value

Code: Select all

unsigned altitude(unsigned char x)
{
unsigned d=x/64;
return ((x % 64)-(1<<(6-d))+64)*(1<<d);
}
Hope this helps :)

Re: er9x development

Posted: Tue May 15, 2012 12:43 am
by ZeroPitch
MikeB wrote:Rob: It does provide that functionality. Partly I wanted to avoid the need for the resistor mod when wanting to run a simulator, but PPMSIM does indeed allow a PPM stream to the trainer port that does NOT go to the module pins whene the Tx module is plugged in. So you choose, 8 PPM channels to the module (PPM), 8 PPM channels to the trainer PPMSIM) or 16 split 8 to each (PPM16).
Brilliant :-)

Question, if PPMSIM is enabled on a model, does the module still transmit RF (and suck power)?

Re: er9x development

Posted: Tue May 15, 2012 1:21 am
by MikeB
If you switch the power on, then the module is powered, but with PPMSIM it won't see any PPM signal. If you power on using the trainer socket, then the module is NOT powered.

Mike.

Re: er9x development

Posted: Tue May 15, 2012 1:27 am
by MikeB
r762 now available, quite a lot of changes for telemetry, summary of changes:

New method of processing and storing Hub data
New method of averaging analog (A1,A2,RSSI,TSSI) data
Fix a bug or two with Current measurement offset and zeroing
If no telemetry data, force TxRSSI to zero
Change alarm checking scheduling
Changes towards better operation in Companion9x
New telemetry inputs to custom switches
Changes towards a custom telemetry display
Unused safety switches may now be used to trigger alarms
Some changes to save some flash space (I've used up some for these changes)
Change the timer code to allow timers as custom switch inputs
Fix position of model name on main screen
The model structure in the eeprom has changed slightly, but should be compatible
The length of a LONG key press is a bit longer
Fixed a bug in the mixers trim choice
Height scaling for telemetry in custom switches changed

I've tested as much as possible, Gohsthb has also done some testing (the r761 above) , there may be some things I haven't caught though. I suggest making sure you have a backup of the EEPROM, and keep a copy of your current flash hex file just in case you need to put it back.

Mike.

Re: er9x development

Posted: Tue May 15, 2012 2:39 am
by jhsa
Thank you !!!

Re: er9x development

Posted: Tue May 15, 2012 6:18 am
by Kilrah
MikeB wrote:If you switch the power on, then the module is powered, but with PPMSIM it won't see any PPM signal. If you power on using the trainer socket, then the module is NOT powered.
I think that anytime the module is powered it should receive a signal. That would also allow double transmission of the same signal via both the module and the trainer port... If you don't want it to transmit you can just power the radio off anyway.

Re: er9x development

Posted: Tue May 15, 2012 9:20 am
by MikeB
But some people have an extra switch on the power to the module, so they can switch between modules. This would also use up some extra code space.
If you want the same signal on both, use PPM16, and CH09=CH01, CH10=CH02 etc.

Mike.

Re: er9x development

Posted: Tue May 15, 2012 9:21 am
by jhsa
thatÅ› what PPM 16 does.. transmits via both module and trainer port.

Re: er9x development

Posted: Tue May 15, 2012 9:22 am
by jhsa
oops, Mike was faster.. :mrgreen:

Re: er9x development

Posted: Mon May 21, 2012 7:51 am
by man-bis
Hi guys!
To flash on the r762 Frsky-Noht, calibrated sticks and potentiometers in the relevant section of the menu. When you turn on the transmitter, I noticed that the channels "not zero", but in 20-80%. Corrected section limits in eepe this is called "offset". So, every time you turn on or off the transmitter, the offset rises at an arbitrary position along the channel ELE AIL RUD except for THR. Why is it so?
All my friends from the Russian RC forum is the same bug.
It appears only on the current model, due to the indicator "offset" in the limits. In this calibration does not help. I have 2 remote telemetry converted by Frsky, the same bug on different consoles with firmware r762 ER9x-Frsky-N. This happens only on the current model, the rest are stored and loaded fine, but if again off and on the remote, then the offset value for this model goes far back in the minus or plus. Can someone help?

Re: er9x development

Posted: Mon May 21, 2012 8:30 am
by MikeB
I'm running the NOT version and I don't see any problems. Could you post your EEPROM file here so I can take a look at it? A lot of changes went into r762.

Mike.

Re: er9x development

Posted: Mon May 21, 2012 7:21 pm
by man-bis
This file contains two models eprom. When you turn on the transmitter offset parameter is set randomly through the elevator, ailerons and Ruder. If you fix the menu limits, then the inclusion of a new transmitter is repeated. Firmware Version V1.5651-erez date14/05/2012 FRSKY NHT
Thank you Mike.

Re: er9x development

Posted: Mon May 21, 2012 8:05 pm
by gohsthb
The instatrim switch is set to !TRN. It needs to be changed to something else. Here is what happens: when you switch on your transmitter the instatrim is enabled. This resets the trim of the model. There is no bug in the firmware. You can test this out by holding the TRN switch on and turning on the transmitter. The limits will stay where they are.
-Gohst

Re: er9x development

Posted: Mon May 21, 2012 8:27 pm
by man-bis
Oh, yes! Thank you very much. So it depends on the instatrim. Very good, it's not a bug in the firmware. I am very pleased. Thanks for the correction.

Re: er9x development

Posted: Thu May 24, 2012 4:10 pm
by Camboui
In stat screens some values are overlapping text.
EDIT: stat2 screen more specifically. Design feature or mistake ? ;)

Re: er9x development

Posted: Mon Jun 18, 2012 6:14 pm
by MikeB
Does anyone find the throttle TRACE display, in the STAT screen, useful, or even look at it?
It appears to show the throttle setting (averaged over 10 seconds) for the last 20 minutes.
I ask because if not useful/needed removing it will save some flash and quite a bit of RAM.

Mike.