Hacking a DX5e?

[gobigkahuna]

I plan on using a Spektrum DX5e with an Endurance RC PCTx interface to computer control my RC robot. The whole thing will go into a small Pelican case. I'm a bit of a hack at electronics so this will be easier for me than trying to build my own from scratch. I want to dissect the DX5e and remove whatever components that won't be needed since the PCTx sends its data directly through the trainer port. Any idea what components can be removed? I'm going to try disconnecting bits one by one, but thought maybe one of the gurus here might be able to give me some guidance before I get started. Tops on my list of bits to remove are the joysticks since they take up so much room. Any idea if the DX5e will initialize properly without joysticks (pots) connected?

[F-111 John]

Since all you appear to be interested in is the DX5e RF transmitter, Instead of hacking a DX5e have you thought about using a Spektrum TX module (http://www.spektrumrc.com/DSM/Products/airModuleSystems.aspx) instead? Or a Corona module (http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=7380&Product_Name=Corona_2.4Ghz_JR_Module_&_Rx_(V1_DSSS))? Or an Assan module (http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=8287&Product_Name=X8_2.4GHz_Combo_pack_for_JR_W/_Module_&_RX)?

Or better yet, the cheap 4 channel transmitter from Hobby King (http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=8338&Product_Name=Hobby_King_2.4Ghz_4Ch_Tx_&_Rx_(Mode_2)) has a standalone 2.4 GHz TX module inside, and others have removed it to use as a hack module for their own radios.

Here is a thread on using the Hobby King as a hack module:
http://www.rcgroups.com/forums/showthread.php?p=11732251#post11732251

[gobigkahuna]

Thanks John. I followed that thread on hacking the HK radio for a while and decided that, at least for now, hacking that was beyond my current electronics know how. On the other hand, the DX5e + PCTx works and there's even software already written that will do pretty much what I want... and I already have a DX5e. I may re-investigate hacking one of those modules in the future once I've gotten my feet wet.

[David T]

Any idea if the DX5e will initialize properly without joysticks (pots) connected?
The middle wire on the pot is likely to be the input to the MCU for that channel. If this is not connected to anything it will be 'floating'. This will typically yield random results on that channel but is not likely to affect anything else. If this is a problem for you then there are simple ways to fix it.
David.

[gobigkahuna]

Thanks David. I'll be getting the PCTx in a couple days and can start testing then, but I've already determined that I'll need to keep the trainer switch (since I'll need that to bind and to connect to the PCTx). But I think I can remove the 5th channel switch as well as the servo reverse and rate switches since those will be bypassed by the PCTx. I plan on also stripping down a small wall wart 5 vdc supply to replace the batteries (the rest of the robot is powered by 115 vac by way of a portable generator). I'll post again with what I find after I've had a chance to test this.

[richg99]

...re the Wall Wart.....REMEMBER ... JR and Spektrum charging circuits are often the REVERSE of the "normal" power inputs for their competitors. i.e. Don't charge ( or do anything remotely connected to the charging circuit) unless you check polarity first. Rich

[gobigkahuna]

...re the Wall Wart.....REMEMBER ... JR and Spektrum charging circuits are often the REVERSE of the "normal" power inputs for their competitors. i.e. Don't charge ( or do anything remotely connected to the charging circuit) unless you check polarity first. Rich
Got it, thanks Rich. Fortunately, when you open the case the wires from the battery circuit are clearly -red- and -black-, unless they've done something really weird I should be OK.

[richg99]

Yes, I figured that if you go directly inside and attach the wires there all will be well. However, someone might try to use the charger port to by pass the battery. If so, that's where the issue MIGHT come up. rich

[David T]

You probably just need the trainer switch and leds to see radio status. The trainer switch is used for both binding and trainer mode so that probably needs to be changed to a latching toggle switch. You will have to power up with trainer/bind off and then flick it on once the radio has established it's link.

With floating inputs this means your controls will be bound to random settings which will manifest themselves in failsafe conditions. They will go to random settings between the time you switch on and switch over to PC control. So I suspect you will need to make the inputs 'not floating'. The gear channel can probably be tied to ground. Although the 'pot' channels probably can be as well, I suspect you would do better to replace the pots with resistor arrays (know what I mean by that?).

If the size of the aerial is a problem for you you can replace it. All you need is 30mm of unshielded wire. If it has a U.FL connector then the easiest would be to get another with a cable attached and cut it as required. Old wifi routers often have them inside.
David.

[David T]

Another thought. The Dx6i saves settings on power down although I don't think the DX7 does. So you just need to check whether the main on/off switch cuts power to everything or is just used to put the Tx into sleep mode like the Dx6i. Depending how it handles this may affect the implications of removing the battery as the primary power supply.

[gobigkahuna]

Thanks David. Good point about the floating input issue, I hadn't thought of that. I guess the simplest fix would be to pull the pots and super glue them in their "neutral" positions, or measure the resistance at that spot and replace them with fixed resistors. There's a fair amount of empty space in a DX5e radio case, the electronics aren't all that big. If worse came to worse, I could just put everything in my new case as-is.

[Andy2No]

Has anyone opened up a DX5e to see how the pots are connected? Are they connected by wires or just soldered directly to the PCB?

I'm interested in building a custom mixer to do elevon mixing. Yes, it already has a mixing switch but it's of limited use:

http://www.rcgroups.com/forums/showthread.php?t=943411

I want to use it with a Vapor brick and the mixer function makes the servos that are built into it move in opposite directions with elevator input (and the same direction with aileron input) so it's next to useless for micro planes with elevons. The reversing switches reverse the inputs to the mixer, not the outputs as they should, so there's no arrangement of the switches that makes the servos move the right way.

I found a circuit for the sort of mixer I have in mind here (designed by John Nooyen):

http://www.uoguelph.ca/~antoon/gadgets/mixer.htm

It uses op amps and should work with any Tx that senses a voltage from the pot. Because it's done with analog electronics, there'd be no noticeable latency (microcontroller based mixers would add a small delay). It could just as easily connect to elevator and rudder to make a v-tail unit, of course.

It looks like the LM3900 quad op amp chip used is in an SOT (surface mount) package, so it would be hard for the likes of me to solder, but the LM3900N is the same thing in a "normal" sized DIP version (if anyone knows different, please correct me).

Some cheaper Tx's that come with RTF toys only have two wires connected to each pot so it might not work in that case.

Does anyone have pictures of the insides of a DX5e? If it's completely hopeless, I won't bother trying to pull mine apart... Are there any of those annoying little clips that make it hard to get it apart?

[gobigkahuna]

Has anyone opened up a DX5e to see how the pots are connected? Are they connected by wires or just soldered directly to the PCB?
They are connected by wires to strip connectors on the main circuit board.

[Andy2No]

After drawing some truth table style diagrams of what happens with the elevon mixer and what I'd like to be able to make happen, it occurred to me that all I really need is to be able to swap over the elevator and aileron controls.

That inspired me to take the back off - it's six philips screws and no clips. The wires to the back panel are a reasonable length. If it's in your lap with the antenna facing away from you, just tilt the back towards you.

The pots are connected by wires. There are three wires per pot and the outer two wires are common to all four pots. Woohoo! All I need is a nice little DPDT switch that routes the wiper wires either directly or swapped over.

There are blanking plates and holes for extra switches - the two nobbly bits at top right are held in with a blob of hot glue on the inside of the case, and there are a total of four switch mounting holes behind the cover at the top, including the one that's in use by the dual rates switch.

I was thinking of a slider switch but the holes are round ones, for mounting toggle switches, so I may go with one of those instead... I just have to make the handle short so I don't use it by accident.

I think I'll move the dual rates switch to the throttle side too. Mine's mode 2 (left throttle) and I don't like having to let go of the elevator/aileron stick to switch rates.

The op amp mixer method above ought to work too, and it's a nicer solution because it's an adjustable mixer. This switch method will do fine for me though.

[Richard Cox]

The problem with that is that the trims will not move. So the elevator trim buttons
will affect the ailerons and vice versa.
I sell a mixer for the Vapor and Kyosho transmitters that allow servo reversing,
servo swapping, elevon mixing, and 4 model memory. See the Micro section for
details.
It has the trim problem as well.

[Andy2No]

Good point, Richard.

To do it properly, I'd need to switch some more wires over to swap the trims at the same time as swapping the control assignment.

It's a problem I can live with though. Instead of not realistically being able to fly elevon models with my Vapor bricks, I'll now be able to do it, but with the irritation of remembering the trims are swapped over. If it bothers me too much, I'll do something about it. The trim switches are each on a PCB but there's wiring from there to the main PCB. From the ones I can easily get to to check, it seems each has three wires, with two of them in common to each switch. Either way, it just means more DPDT switches, ideally banked together with just one lever.

I'm interested in mixers like yours though. My first thought was a PIC based add on board between the controls and the radio, so I could do anything I wanted, e.g. add exponential. I realised that would add a bit of latency, though hopefully not enough to matter. That reminded me that people used to do analog mixers with op amps, and those are very fast (almost no latency). I like the switch idea for now though. It's simple and anyone can do it without any real outlay.

I've got as far as moving the dual rates switch to where I want it, and I've found a suitable toggle switch that fits the hole where that was before. The wire for the DR switch was just long enough, once I untwisled it a bit (it's two wires twisted together)... More later.

[Andy2No]

Well, I did it and it works.

I now have a switch that can swap the aileron and elevator controls over, which lets me set up elevon mixing on my DX5e as I need it for the Vapor brick. Yes, the trims are on the wrong control when I do that but I could fix it if I put the effort in. The channel reverser switches are then labelled wrong too, but that's not a problem. It's easy to see when you've changed the right ones.

The trims thing will be a bit confusing, and not much use for in-flight trimming but I can live with that for now. A second changeover switch would solve that problem.

Here's some pictures. My dual rates switch is now on the left side - on the right side in the wiring photos. The control swap over switch is top right, where the dual rates switch used to be. I used wire from a twisted servo lead extension. Presumably, the use of twisted pairs helps with EMI.

[Richard Cox]

Nice job! It will take a 4PDT switch to fix the trims, since there are essentially
2 switches for each. May not be worth the effort.
This mod does not fix the original fault that I started a thread on, that is the
reversing is applied after the mixing.
Shame on Horizon.

[Andy2No]

Thanks. I'll just leave it as it is now then, and put up with the trims being in the wrong place when I use the new switch.

Agreed, shame on Horizon.

Although it doesn't allow reversing of the outputs, I think it does give all the possbile combinations. In one position you get {A+E,A-E} and in the other you get {E+A,E-A}. The input reversing switches give you all the other combinations (total of 8 possible combinations). It seems to me that's effectively the same thing as being able to reverse the outputs becauase I can set each servo to move in either direction with either stick input.

[Richard Cox]

I only use my DX5E for some indoor planes. I may apply your solution on mine.
Thanks for the effort.

[Andy2No]

Yes, that's what I bought mine for really. You're welcome.

[Andy2No]

Has anyone figured out if the channel 5 switch can be changed to an analog control?

[Richard Cox]

Probably not possible, unless it is connected to an analog to digital
converter input, AND the guy who wrote the code made an extra effort
to use that feature to program the channel 5 pulse width. If he did,
then the switch one way looks like a fully CW pot, and the other way
looks like CCW, with the channel 5 pulse following the input.

As a programmer of such things, I wouldn't have, unless the code was
borrowed from another very similar transmitter. It obviously was not,
since no other TX I know of handles the mixing the way it does.

I wouldn't hurt to try it , however. Disconnect the center terminal
of the switch, connect it to the wiper of a 5K pot, add connections
from the two outer switch terminals to the outer pins of the pot.

If there is only one outer pin, then jumper the pot pins to the
outer pins of another control pot.

If it doesn't work, make sure you reconnect the switch. Digital inputs
don't like to be connected to odd voltages.

[Andy2No]

Yes, that makes sense. Thanks, Richard.

The only DSM2 receivers I have are two Vapor bricks and I'd have to solder a socket onto the pads for CH5 to try it... Maybe I will, one day. I was just curious, really.

[bozonghare]

Has anyone opened up a DX5e to see how the pots are connected? Are they connected by wires or just soldered directly to the PCB?

I'm interested in building a custom mixer to do elevon mixing. Yes, it already has a mixing switch but it's of limited use:




elevon/v-tail mixer unit

http://static.rcgroups.com/forums/attachments/7/9/7/0/3/a2482854-249-mixer2.gif

http://static.rcgroups.com/forums/attachments/7/9/7/0/3/a2482858-221-mixer1.gif



Hello Dear friends!

These design not work aright!!!
I make these but not work ?!

Have you any design of V-TAIL MIXER ???