Elev-8 Altitude Hold = Not Possible?

tranger

I submitted a ticket to Hoverfly:

An ultrasonic sensor was recently added to my quadcopter in order to use Altitude Hold flight mode. However, I have not been able to get it functional - the controller LED does not change color from green during flight and the behavior does not change.
The quadcopter is an Elev-8 kit purchased from Parallax. It was built in August of 2013 and has been functioning normally otherwise.
The sensor, a Maxbotix MB1000 LV-MaxSonar-EZ0, was tested on the bench prior to mounting on the quad using a logic analyzer and an oscilloscope.
The sensor is powered from the Hoverfly open connection.
The sensor output was verified while mounted and powered on the quad using a logic analyzer. The connections are +5(red), GND(black) and PW(white).
The radio transmitter is a HiTech Aurora 9 channel.
The receiver is a HiTech 7 channel unit.
The primary gain is set to 25%.
The altitude hold gain is set to -25%.
With the Hoverfly open connected to the Setup Utility (v1.5.0.60), the Altitude Hold Enabled checkbox follows the transmitter command (“gear” channel).
The Altitude slider on the General tab of the Setup Utility does not change or respond to changes in object position in front of the ultrasonic sensor.
The Altitude label is grayed out in the Sensor Test area on the Diagnostics screen of the Setup Utility.
Hoverfly Flash Firmware is Version 2.0.
I suspect the firmware is somehow an issue. Please advise what other troubleshooting steps can be taken.

And got this reply:


The HoverflyOPEN firmware v. 2.0 only includes flight control functions. It does not include any functionality for third-party accessories.
While we encourage users to experiment with adding functions to the HoverflyOPEN, it is fundamentally a do-it-yourself product and adding accessories will usually require writing your own firmware for the board. For assistance and advice in creating your own functions, please visit the Projects forum at Parallax.com.

My reply:


Thank you for the quick reply. However, there is some kind of mistake here. The user manual clearly has a section, 5.7.1 on page 54, devoted to Altitude Hold. The gain settings, sensor model and connections and other associated information related to using this mode is described and specified in detail.
User manual: 80000_HoverflyOPENUsersGuideV1_0.pdf found both on Parallax's site and on hoverflytech.com (static copy).
I've never seen nor can I locate a different version of the manual.
What is the full story behind this discrepancy?

I get the feeling I'm screwed on this one.

Ok, now I cannot fly at all. The quad wants to flip forward hard on take-off. I think I could probably get it off the ground with a LOT of up elevator, but it wouldn't last long. I tried trim, but it is way beyond trim settings. So I reflashed - same result. I also noted that every time I open up the setup utility it is showing Altitude Hold checked! I can flip the switch to disable, but it always returns after a power cycle on the controller board. I tried initializing and then going in and out of Altitude Hold, but no dice. This is a royal PIA....

-Russ

Ken Gracey

Hey Russ,

There's a strong chance you've got the ESCs or receiver port mis-wired. I've been here a bunch, as have other forum members. Can you verify the connections? Are you certain that you didn't mis-wire the OPEN connections when you removed the Maxbotics sensor?

Also, you need to be certain that you take off with altitude hold NOT enabled. Can you confirm?

And I concur - there's a discrepancy here that we need to sort out. I've called the CEO of Hoverfly to look at this thread. It's my understanding that the ultrasonic sensor you identified is indeed compatible with our hardware.

Hope to return with some real solutions for you shortly.

Ken Gracey

tranger

The wiring is good. Since the last time I flew only the ESC connectors had been removed. I've checked and double-checked. I really wish it were wiring - it would be simple to fix.

Yes, Altitude Hold is disabled while attempting to take off. As you probably know the board won't initialize with Altitude Hold enabled. In any case, I've tried switching it on and then off with no success either.

After the initial installation of the sensor I was able fly without any issues, although the Altitude Hold did not function. Last night I reflashed the controller firmware hoping to "fix" the Altitude Hold mode.. It was late so I didn't try to fly at the time. This evening when trying to fly I had the flip forward issue.

FYI - I sent these pics with the original support ticket.

I've downloaded Jason Dorie's code and am considering loading it just to get back up and running. There are some issues I need to work thru before I can do that however.

Thanks,

-Russ

Duane Degn

I also followed the HoverFly instructions and added a Maxbotix ultrasound (the model the manual specified). I took off with altitude hold off and when I switched it on about 10 feet in the air the ELEV-8 flipped and crashed. I thought I might have had a loose wire that moved in front of the sensor but I was just looking for something I could have done wrong.

I don't recall reading anyone having anything but trouble when trying to get the altitude hold to work.

tranger

"... when I switched it on about 10 feet in the air the ELEV-8 flipped ..."

Maybe this could be used during aerobatic maneuvers to intentionally flip.

Duane - were you using the OPEN board or the SPORT board?

-Russ

tranger

Duane Degn wrote: »

I don't recall reading anyone having anything but trouble when trying to get the altitude hold to work.

Maybe Post #8 in this thread..... http://forums.parallax.com/showthread.php/147432-Altitude-hold

Publison

It's interesting that on the HoverFly site, under definitions, the AH (Altitude Hold) is described as using a barometric sensor on the HoverFlyPro. No mention of the OPEN board.

https://hoverflytech.zendesk.com/hc/en-us/articles/201993003-AH-Altitude-Hold-function

Duane Degn

tranger wrote: »

Maybe Post #8 in this thread..... http://forums.parallax.com/showthread.php/147432-Altitude-hold

Thanks for the link. I had forgotten about Nick's success with altitude hold.

I was using a Sport board.

I know when I experimented with ultrasound on a small quadcopter the noise from the motors interfered with the sensor. I think the small motors must have had a strong ultrasound component.

Publison

Duane Degn wrote: »

Thanks for the link. I had forgotten about Nick's success with altitude hold.

I was using a Sport board.

I know when I experimented with ultrasound on a small quadcopter the noise from the motors interfered with the sensor. I think the small motors must have had a strong ultrasound component.

I'm guessin' they don't sell the Sport Board anymore?

http://www.shop.hoverflytech.com/collections/controllers

tranger

From HoverFly Support:

"The HoverflyOPEN is very similar to the old HoverflySPORT board, which did use the MaxBotix sensor. It looks as if that chapter of the HoverflySPORT manual was erroneously copied into the HoverflyOPEN. I've requested an update to remove it."


The OPEN does "read" the sensor. How do I know this? The log files downloaded from the board have Altitude entries when the sensor is connected and don't when not.

-Russ

tranger

Publison wrote: »

It's interesting that on the HoverFly site, under definitions, the AH (Altitude Hold) is described as using a barometric sensor on the HoverFlyPro. No mention of the OPEN board.

I figured that if the Maxbotix sensor worked then it wouldn't be too hard to hack in a baro sensor. Another micro could read the baro sensor and supply a PWM to the OPEN - it wouldn't know the difference. Looks like that idea is in the tank though.

-Russ

dmagnus

I can confirm that - only with the PRO board, which was sent to Hoverfly to be checked. They said they calibrated it and when I got it back, it did the same thing - flipped violently when engaging altitude hold. It has been sitting ever since and is now for sale... I have no complaint with Parallax, of course, but I'm not real happy with my $900 PRO/GPS purchase...

tranger

I'm back flying again - no changes made... However, it still acting very weird.

It still wants to flip forward on the initial take-off. If I modulate the throttle it will actually balance on the front two landing gear. To get a successful take-off requires almost 100% up-elevator to compensate for the flip. After a few times of counteracting the flip and letting is set back down it will then want to back flip. At that point I can get it off the ground and bring it under control. Workable, but not smooth.

Once in the air it is very twitchy. Somewhat noticeable in the attitude stability, but most noticeable in an audible oscillation in motor speed while hovering. It seems like what you might get with a high gain. So I changed the gain from 20% to 15% and saw no discernible difference.

I really happy to fly again. I've been flying this thing four or five times a week, a battery at at time. It is a lot of fun and I miss it when I can't do it. However, there is still something amiss and I need to get that resolved somehow.

The problem is that I don't know how to troubleshoot or diagnose issues related to the controller. Since I've seen a step change since trying to run the Ultrasonic sensor and reflashing the controller, I'm pretty sure that is where the problem is coming from.

In the past I've had instances where the rudder axis bias has been off after initialization - it would start rotating immediately during take-off. This was addressed by simply setting back down and restarting (power cycle on the controller). After this would happen there was no abnormal behavior in the subsequent flight. Between front flips, back flips and rotation I have to wonder if the gyro is a little flakey or if it just isn't calibrated correctly. I also wonder why it is now nervous at 15% gain and was happy at 20% before. There are flight log data files available to dump from the controller, but I have no idea what the data is or how to analyze it.

Take-off clip here: http://youtu.be/sqjuxV_L9Ys Doesn't really show the twitchyness very well, but the take-off flippage is obvious.

-Russ

tranger

I resolved my flipping on take-off issue and it is back to smooth-as-butter again. Actually it just got to the point it was too hard to fly. So I went back to square one and checked everything. The trick was re-synching the ESC's.

Here are are the steps for my radio (HiTech Aurora)

1. Take off props (safety)
2. Set the EPA on the radio to 100% (I've been running 80%)
3. Turn off radio alarm for throttle-not-at-zero
4. Power up radio
5. Throttle full up
6. Enable transmission from radio
7. Power on HoverFly (receiver and ESC's all connected - flight-ready configuration)
8. After two beeps from ESC...
9. ...Move throttle to zero
10. Three beeps
11. ESC's are synched
12. Power down HoverFly
13. Set throttle EPA on radio from 100% to 80% (or as desired)
14. Turn on throttle-not-at-zero alarm
15. Put on the props

All set to fly. To start-up and arm

Radio on, transmitting, throttle @ 0, gain is positive (gear switch)
Power on Quad
After initializaion beeps...
Throttle @ 0, full left rudder
Wait for blinking green
Throttle @0, full right rudder
Wait for solid green

Armed and ready....


-Russ