13BRV3

It sure is quiet here

Seriously, I can wait. I won't enjoy it, but I CAN wait

Rusty

Old Man Mike

You won't have to wait long. Last parts arrived today. I will do the final conversion in a few hours and will post the results in the early morning hours. I have the parts list, schematic and pictures ready for posting. I need to write up details on the assembly process but those interested will be able to see the design and order parts if they wish.

Mike

13BRV3

Outstanding news Mike. For some silly reason, I've been out in the garage this afternoon (cooking my brain) looking at the board. I was just about to go out and get my scope from the van to take a look at the gate signal on the MOSFETs, but it will be a lot easier to just wait for your next post

Maybe I'll give my neighbors a thrill, and (finally) cut my grass instead

Regardless of the outcome, thanks for all your work on this.
Rusty

Old Man Mike

All four motors have been converted.

Here is a video showing the Brushless DF flying with a 21 oz payload:

[link=http://www.af9y.com/Brushless21oz.wmv]Brushless 21 oz flight[/link]

This all took a little longer than I expected so I did not finish up the documentation. Here is a start showing the Parts List and the most difficult part of the conversion - building a little circuit board for the optical isolators:

[link=http://www.af9y.com/BrushlessConversionRevA.mht]Documentation RevA[/link]

(hope the above Powerpoint format will work for everyone)

You should be able to click on the supplier name in the parts list and it will take you right to the part. Unfortunately I see the motors are already out of stock so you might have to look around for other suppliers. Also, I added five additional 0.22 caps that are shown in the pwb layout but have not yet been added to the schematic.

I'm dead tired and will start working an updated package with much more detail tomorrow.

Mike

13BRV3

Looks great so far Mike [sm=thumbup.gif] I guess you've earned some sleep

I'll hold my questions until you get the documentation closer to being finished, since that will probably answer most of them.

Thanks,
Rusty

Dirk Schooner

WOW!!! Amazing, top work, you deserve some sort of award for this project.

Gives the old DF a new lease of life.

One thing I can imagine though is power drain, I imagine much bigger battery (looks like a TP 1320 on there) is needed for the motors and this sort of weight for any reasonable time, I will wait until you have some more flight tests and some sleep.

Thanks so much for this.

Old Man Mike

Thanks for the kind words.

Dirk: This conversion does not require a larger battery but it might be worth considering for even longer flying time. The motors were picked so that current draw is close to the original motors. Since brushless motors are more efficient, the same amount of current can provide greater lifting or longer flying time. At max throttle, the avg current is about 33% higher (along with higher thrust) so the fuse has to be changed from a 10 amp to a 15 amp. With my normal camera payload of 7oz, I expect to get longer flying times since the brushless motors should be loafing along at about 1/3 total capacity.

I flew it outside (without a payload) this evening in 5 to 10 mph winds just to get an idea how long the standard 1320 mah battery would last. Even with the wind, I could tell it was at least as smooth and stable as the old brushed motors. It flew for aprox 15 mins before the reduced throttle mode kicked in. The HD brushed motors typically lasted around 10 mins. At the end of the flight, the brushless motors were just luke warm.


I am adding the additional documentation as I complete it. Here is an update which now includes the block diagram showing the interface connections:

[link=http://www.af9y.com/BrushlessConversionRevD.mht]Updated Documentation RevD[/link]

Is this powerpoint formating working OK? I've started adding notes below each slide which I hope will help answer questions. I use a wide screen display so I'm not sure how well this style works with smaller displays.

Mike

13BRV3

Hi Mike,

First, the powerpoint type page is working fine for me, even with my monitor in portrait mode.

The block diagram really clears things up, and it appears you've done a clever job of making this all work. I also think it's a very reasonable approach to extra power, basically using the power that's already available more efficiently.

Do I see a Berg receiver there? Any idea where to pick off the throttle signal for a non-Berg board?

Keep up the good work [sm=thumbs_up.gif]

Cheers,
Rusty

Sky High

Even though I don't understand schematics, it is great achievement. Could you post some close up pictures of the new components on the airframe? I just want to see what the wiring and connections look like even though I couldn't replicate any of it with my limited knowledge.

Old Man Mike

Hi Sky,

Take a look at the RevC package. I guess I'll just keep editing my earlier post with the updated revision rather than making new posts.

Mike

Nicoleto

Hi Mike, congratulations you’ve made an excellent job. I still have a question. Does this conversion works with the old integrated DF Boards? (The ones that have the RX integrated to the board).

Thanks for your help

Regards
Nick

Old Man Mike

Thanks Nick,

I see no reason why it would not work with the older DF boards using the integrated receiver. The only challenge would be to find the line carrying the PPM throttle signal so that you could tap it and feed it to the optical isolator board. Someone with a older board and an oscilloscope should be able to probe around and find the signal. You can do this without arming the DF since the signal is there anytime the power switch is on and the transmitter is transmitting. The signal will be a simple 1ms pulse with a low duty cycle that varies in width as the throttle is increase. By the way, here is GREAT Oscilloscope which plugs into your PC via the USB Port:

[link=http://www.parallax.com/detail.asp?product_id=28014]USB Oscilloscope[/link]

Mike

13BRV3

Hi Mike,

Very cool, and cheap scope. If my employer didn't already supply me with a nice scope, I'd sure think about buying one of those.

As for the throttle signal location on the older DF boards, I'll look for that this weekend. I already had the scope on it yesterday looking at the gate signals to the hexfets. Unfortunately, the conversion required to use that the gate signals would be harder than your idea.

If anyone here has a really good understanding of these brushless motors, are there other good options available? The Feigao seems perfect, and the price is right, but I wonder about availability.

Cheers,
Rusty (trying to resist checking the DF now)

Old Man Mike

Hi Rusty,

Yeah, I was quite surprised to see that they are currently out of stock at BP Hobbies. They certainly had the best price out of the many suppliers I checked.

Selection of the right brushless motor was actually one of the most difficult aspects of the design. There are few motors that have their peak efficency in the 2 Amp range with a 11 volt supply. In addition the KV rating needs to be in the right range to match the gear ratio in the DF. This is one of the sites I used during the search:

[link=http://www.peakeff.com/MotorDetails.aspx?MotorID=2330&v=10.5]Motor Data[/link]

Unfortunately most of the data on brushless motors is for props much smaller than used with the DF. I had to extrapolate data and eventually ended up testing several candidates. I also recommend that you limit your search to inrunners since mounting the outrunners in the current DF frame would be a significant mechanical modification.

Mike

13BRV3

Hi Mike,

I "probed" the DF board (sold just before they changed to the Berg RX) for an hour or so, and I can't find any signals that appear to be separate channels. I did find the composite that shows 7 channels, and changes as you'd expect it to when you move the TX sticks. This signal is on one of the pins that goes from the vertical board to the horizontal board. I found some TI sensor signals, and the signal that actuates the red LED on top, but no separate channels.

Note that I didn't attempt to scope the pins on the square chip (label sanded off) on the horz board. Since the vert board has the RX, I'm assuming the separate signals will have to come from the vert board, if they exist at all.

My guess (really all it is) is that they use the composite RX signal as the input to the main controller on the old style boards. On the Berg boards, they must either use the individual signals at the controller, or re-combine them on another chip. This "might" be bad news for owners of the older board, but none of this is guaranteed to be accurate, so don't panic yet.

On a side note, if the Berg delivers only separate channels to the vert board, then I don't see why we can't use other receivers (2.4 GHz) as well. Mike, does the Berg RX output anything a standard RX doesn't? I'm guessing (there's that word again <g>) they used it because it's small and light, not because it has any special operational properties.

Cheers,
Rusty (way too many projects, but I can't resist)

PS- Thanks for the motor info. I'll get back to that (hopefully).

Old Man Mike

It could be that the pulses are separated on the horizontal board or the processor just did it in software. For the latter case, all is not lost. It should not be that hard to design a circuit to strip off the third pulse.


Actually I initially used a separate Berg RX receiver for the single motor conversion. You could certainly do that if you can't find a separate throttle signal on the old board and don't want to design a circuit to strip off the pulse. You only need the small 4 channel receiver which will add an additional $40 to the cost of the conversion (you will still have use the optical isolators).

The Berg receiver used in the Draganflyer appears to be a 4channel unit that has been modified to extract a fifth channel for the TI enable signal. I have been thinking about changing out the DF receiver to the Berg 7 channel unit to give me an extra control channel to do an altitude hold enable. But first I need to determine how that TI enable is extracted from the current receiver and emulate it with the 7 channel receiver.

Mike

13BRV3

Hi Mike,

Sounds like there are a few other options indeed. I'm going to let my tiny little brain keep thinking about this until morning, and see if it comes up with anything useful

Rusty

aerohelipilot

WOW this is so cool MIKE thanks for the great info I got all the parts ordered except for the motors now its the wait game again unless someone can find a replacement I see them on other websites but for 40 a piece.

Old Man Mike

Torture Testing:

This evening I decided to do a little torture testing by flying that 21 oz weight until the 1320 mAh battery gave out. Here's what I found out:

1) Total fly time was 3 mins. With the 21 oz payload, the total weight up in the air during the test was 41 oz (over 2.5 lbs!)

2) I was flying at about 90% throttle until near the end when I moved up to 100%.

3) The DF made a gradual decent at the end of 3 mins.

4) After the flight I felt everything on the DF for heat. The battery and four motors were the only warm items. Both felt about the same temperature which was not too hot to grasp continuously with fingers. The ESCs and the hexfets were not even warm.


Based on the above test, I think a more reasonable flying payload should be kept under 1 pound. This is still more than twice the payload of the original HD brushed motors and should give about the same flying time. It may also be worthwhile to consider moving up to a 2200 mAh battery for even longer flying times.

I think things can be optimized a bit more and I will list those improvements in future documenation updates.

Mike

msarid

This is great info, MIKE thank you for your amazing project.
I have 4 castlecreations 18 thunderbird ESC i programed them to Nicad cutoff to lower the cutoff to zero and cancel the break. this is the only parammetes that I can change via the USB interface ,so my question is if I can use this ESC as i calibtated or should i buy the ESC according your recomandation.

Have a great weekend
Moti

Old Man Mike

Hi Moti,

Thanks! It is hard to say if the castlecreations ESC will work as well. I recommend you buy one of the recommended ESC units to compare. As will be explained in assembly notes, I recommend that anyone doing the conversion first do it with just the one rear motor. You should be able to fly the DF with just one converted motor as I demonstrated at the beginning of this thread. (You will have to set the trim to compensate for a more powerfull back motor). Once that is working, you can then try the castlecreations ESC to see if it flys as well. Also, you can test all the positions of the optical isolator board by plugging the motor in each of the four outputs.

Mike

msarid

Hi Mike
THanks for your quick answer

have a nice day
Moti

msarid

Hi
I orderd all the parts but I can't find the motor's, do you know were i can find it or there is alternative motor.
Thanks
Moti

VRNowhere

So, for a dumb question of the day, how do you know if you have the "older" style board or one of the Berg boards?

Thanks.

13BRV3

The Berg boards are quite new, so unless you got yours in the last few months, it's likely you have the "older" style. I couldn't find a great pic of the new Berg board, other than one that was posted about the glue coming loose. The old boards had a receiver built into the vertical board, but the new boards use a tiny Berg (brand name) receiver that's glued to the vertical board. I'll attach the best pics I have.

Rusty

PS- I notice that my file names aren't visible, but it should be easy to tell that the one on the left is the non-Berg, and the one on the right is the newer Berg type.