khalsans

My Feedback: (7)

Analysis of Hobbyzone/Parkzone electronics

I thought that perhaps someone on this forum might be interested in this. I recently tore down a Firebird Commander receiver, then a Aerobird Challenger and Xtreme. Then I looked at the Parkzone J-3 Cub and P-51D. (I'm an electrical engineer in my spare time when I'm not flying

All receivers have similar front ends: an RF amp stage consisting of a FET and a couple of tuned RF transformers. The 'Q' of this circuit is not so tight that if the crystal frequency is changed, the receiver is not noticeably less sensitive. I've swap crystals and had the unit work fine.

From the RF amp stage, the signal is passed to a low-power narrow-band FM IF system chip. The label is scraped off, but it is probably an LM3361 or MC3371. They are all quite similar. A local oscillator set to 455KHz below the channel frequency feed the mixer that converts the 27 MHZ signal to the 455KHz IF signal. The signal can be seen by probing pins 1 & 2 of the chip with a scope.

From there, the IF signal passes through a ceramic filter, the little black box near the IC. The demodulated signal appears at pin #9 of the IC, then is amplified by an op amp in the chip at pins 10 & 11. The signal is now a digital one, and it is passed to IC2, the microprocessor.

If you probe this signal on a 3-channel receiver (Challenger, Aerobird), you will see 4 negative-going pules, 250 microseconds wide: one start pulse and three data pulses. These repeat every 20 milliseconds. THe first data pulse is the throttle, 2nd is joystick vertical and 3rd is joystick horizontal. The distance between the pulses changes as the throttle or joystick is moved, from 1 to 2 milliseconds.

The "fire" button on the top of the transmitter does not go to a 4th channel data pulse, but rather seems to have been added as an afterthought - it just changes the rep rate of the data pulse train.

The 2-channel receivers (Firebirds) are different. There is just a square wave signal. Throttle varies the freqency (starts at 66Hz) and joystick varies the pulse width. So you can't use a Challenger remote on a Firebird.

Parkzone receivers are pretty much the same circuitry, but converted to surface mount. Even the RF cans are smaller.

The PIC processor, IC2, decodes the digital signals and sends pulse-width modulated signals to the motor and servos. The motor controller is typically a FDP6035L low-threshold power FET. Some models have heat sinks (Xtreme has a big one, Challenger has a small one), some have none (Commander, Parkzones). I've never seen them use heat grease on the transistors! They use a light-weight 1N4001 diode across the motor to clamp the reverse EMF from the motor. So one of these items, transistor or diode, will eventually fail. Often, the transistor fails shorted, causing the motor to run continuously.

The servos are driven from an 'H' bridge of 4 small-signal FETs. The servo is just a motor and a pot (no other electronics), so there are 5 lead: 2 big ones for the motor and 3 ribbon-connector wires for the pot. The 3-channel receivers use the A/D converter on the PIC to read the pot and determine servo position. The 2-channel ones use a slope converter (ref Microchip AN513, "A/D conversion using the PIC16C54." They have a big low-temco cap on the back of the pcb for the timing cap.

The PIC processors on the 2-channel receivers use the on-chip 2 MHz oscillator, observable on pin #6. The 3-channel receivers use a ceramic oscillator element to fix the frequency. This can fail mechanically in a crash, causing the processor to drop dead. It's fairly cheap and avaiable from Digikey, so it's worth replacing.

The accessory port is dirt simple- when the plane is "hit" by a competitor pushing his fire button and activating the sonic module, a pulse is sent to the squelch input of the demodulator chip, killing the received signal for a period, and the processor shuts down the controls.

Parkzone J-3 Cub and P-51D receivers differ slightly. The P-51D uses a bigger motor control FET, and the board that looked at had 3 jumper blocks instead of the 2 on the Cub board, but this could be just an older rev, not sure. I think the jumpers are for enabling V-tail mixing. More on this later, as I dig into Parkzone deeper.

Tramsmitter
___________

The transmitter circuit is a miracle of engineering - cost engineering, that is. There's only one chip, It's part number has been scratched off. I tried to figure out what it could be. I looked long for a single-chip 3-channel FM transmitter chip. It does not exist. Then I realized that all you need are 4 data pulse modulating the RF oscillator, so you could use a bunch of 555 timers, like an LM558 (quad 555 chip). But they went cheaper than this. The fancy chip with the number scraped off is a lowly CMOS hex inverter, configured as a bunch of cascaded one-shots, and diode-OR'd to to make a pulse train that modulates a singe FET crystal oscillator. The R/C time constants of the one-shots are changed by throttle and joystick control pots. This signal feeds an RF power transistor that has a tunable tank circuit (RF can) on the antenna output.

There’s also a voltage comparator connected to a beeper to indicate low battery condition. Not much else on the board.

critterhunter

[sm=surprised.gif].........................(coughing)......Um, uh...my plane has stuff in it that makes it go.

Seriously, this is the most detailed overview of the parkzone/hobbyzone electronics I've ever seen. I'm sure it'll be useful to others with the ability to take advantage of it. Great job and thanks for posting.

Wastegate

Yeah, I thought that was how it worked

flyingace451

*ahem*........................................well ...........................I didn't understand some of it because I'm in high school, but I was actually able to understand most of it apart from the names and the 'Q' connectors and what not.

I'm with Critter on this one. That was the most detailed review EVAR!!! Very nice job. Except for the part that some of us aren't very fluent in the electrical language.

treedog

ummm...... way more info than i needed to know at this time of the morning ......,. get that dog outta my plate!!!!!!!!!!

khalsans

My Feedback: (7)

..<snip>.........................I didn't understand some of it because I'm in high school, but I was actually able to understand most of it apart from the names and the 'Q'

The Q factor of the circuit bascally relates to how sharp you can tune it. If it has a high Q (ideal situation) then you have to tune it dead-nuts on to get a good signal out of it. But manufacturers find it cheaper to make the tuning range more broad so they can swap crystals without retuning. Of course, the performance is not as good - it does not reject adjacent channels as well.
And if you detune a transmitter circuit enough, you could , in theory, get "splatter", where you interefere with another channel. That's why it is illegal for users to change transmitter crystals. But truth be told, a friend who worked at the Hitec factory once told me that they rarely, if ever, retune a transmitter after changing its crystal , but that's a little secret I wouldn't dare divulge to the public.

gigelus2k3

He he,

Nice seeing other people interested in prying open their toys just to see how they're made...

Regarding the particular way Hobbyzone/Parkzone electronics are made (obviously to reduce the cost as much as possible), your post leads to some practical observations:
1) it confirms what others said about how to reverse the servos (I just read another thread regarding a RTF Stryker with reversed ailerons)
2) the electronics are unusable outside the original plane. Other RTF plane makers use standard components in theirs; if one plans to go beyond the first plane, buying a RTF from Parkzone is too costly.
3) it is not obvious at all why they had to make a custom servo system instead of going for the standard one. Two more wires and wider connectors on both sides easily offset the cost of the opamp-based integrator a standard servo uses.

Kudos to the engineers who had to design the RC system to work OK given the sh*tty cost constraints!

Serban

khalsans

My Feedback: (7)

<> Kudos to the engineers who had to design the RC system to work OK given the sh*tty cost constraints!

I don't think you'd want to use the radio systems except in park flyers anyway, because they are single conversion receivers and not paricularly designed with selectivity in mind. So you'd want to keep the plane close where it has plenty of signal, or interference would eat your lunch. Or your plane, I should say.

The era of cheap ( < $1) PIC processors with A/D converters has made it possible to strip the servos of their electronics and the receiver of its timed-pulse circuitry to drive them. This is probably the future of servos, unless they go the opposite direction and put a processor in the servo. If that were the case, you would just daisy-chain them on a 'network' like computer USB devices. This is what is happening in automotive sensors.

But the cheap processor has also enabled the brushless technology to develop.

I just finished tearing down a Chinese knock-off of the Challenger. It has no processor, but has 4 times as many chips on the circuit board! Plus, it's flakey as hell.

khalsans

My Feedback: (7)

<> Kudos to the engineers who had to design the RC system to work OK given the sh*tty cost constraints!

I don't think you'd want to use the radio systems except in park flyers anyway, because they are single conversion receivers and not paricularly designed with selectivity in mind. So you'd want to keep the plane close where it has plenty of signal, or interference would eat your lunch. Or your plane, I should say.

The era of cheap ( < $1) PIC processors with A/D converters has made it possible to strip the servos of their electronics and the receiver of its timed-pulse circuitry to drive them. This is probably the future of servos, unless they go the opposite direction and put a processor in the servo. If that were the case, you would just daisy-chain them on a 'network' like computer USB devices. This is what is happening in automotive sensors.

But the cheap processor has also enabled the brushless technology to develop.

I just finished tearing down a Chinese knock-off of the Challenger. It has no processor, but has 4 times as many chips on the circuit board! Plus, it's flakey as hell.

critterhunter

I cut my teeth on the Challenger, stock Stryker, & Extreme and fly with several friends who have as well. While I know "cheap" is the key word in parkzone/hobbyzone electronics, we've been pretty happy with all these planes. We've crashed them, abused them, made them do things they just were not supposed to do, yet the components stood up to the abuse. In fact, I'd say we've only lost a few Stryker servos, one Stryker RX, and about three or four Extreme boards (after some really bad wrecks). The Challenger boards have never failed that I can recall, and what is interesting is that the Extreme board is identical to the eye yet has suffered more failures. I'd chalk this up to a weak FET and the added momentum of impacts due to the Extreme's heavier weight. On the other hand, we've flown all these planes as far (and further...lost a few) than the eye can see without a loss of TX/RX control. Very impressed with the range, fairly impressed with the punishment these planes can take, and for the price they are great planes to start out with. After you've grown and gone on to standard electronics you can rip the guts out of one of these and install it into other projects you don't want to risk "good" standard electronics on. Foam toy stoy gliders, flying wings, homemade foam builds, etc. It makes for a great test bed. I for one am not sorry I started out on these planes. Even if I started with standard electronics I would have took very little with me into the realm of brushless motors and lipo batteries I use now on most of my planes.

khalsans

My Feedback: (7)

<>the Extreme board is identical to the eye yet has suffered more failures.

I think it's a heat management issue, to a large degree - the motor is a bigger, and so is the current.
On the Xtreme that I rebuilt, I replaced the FET with a bigger one, IRL3803, the diode with a higher-voltage one, 1N4005, and a heat sink that was bigger and coated with heat grease.

gigelus2k3

I was just expressing my sympathy for the engineers. In my case, I'm using ARFs and standard equipment, having had from the beginning enough insight into those things to avoid the RTF path.

Regarding the servos, my point was that electronics are even now cheaper than the wires and connectors; anything that reduces the mechanical/connectivity complexity at the cost of a slightly more complex electronics will be embraced. In my opinion, for at least the next few years, nothing will beat the simplicity of the 3-wire servo link; you really need the 3 wires and the header-type connector is dirt-cheap (heck, who buys Y-splitters and extenders when they are so easy to make?).

Electronics, especially things that are used in other large-size consumer markets as well (the radio chip being the exception), are dirt cheap (damn! I still need to earn a living from those things, it's not good they're so cheap)

B.R.

Serban

jmir

My Feedback: (1)

critterhunter,

I agree with you. If it wasn't for Parkzone, I would probably not be in the hobby now. The cost for that first plane that included everything was just right for me to determine whether or not I wanted to go into this fascinating hobby. That was a J-3 Cub which I still fly almost every week (the second version -- don't ask what happened to the original one!). I also purchased a Stryker and a Typhoon. All three planes fly totally different and I still enjoy each one of them.

I did come accross a used Typhoon receiver which I installed on the Stryker along with a ESC I had. This way I only had to carry one TX for these two planes.

I've had no problems with the TX/RX combination on these planes.

flyingace451

If it wasn't for that Firebird Scout my dad got me a year a go, I wouldn't be on this forum! (Don't buy a Scout though, they are crap) After 3 Scouts, even though I only payed for 1 (HZ replaced one....gotta love that company) I moved up to a Challenger. She still flies, except teaching my dad how to fly has been the most difficult thing of my life. He's beaten that thing up way too much. Now I have a Stryker and I'm hoping to go brushless with it soon. Then, I'm gonna give 3D a try.

khalsans

My Feedback: (7)

<>Regarding the servos, my point was that electronics are even now cheaper than the wires and connectors;

I agree, the 3-terminal standard is firmly entrenched and will be available for a long time. Besides, the biggest problem with reliability in electronics is usually connectors, so the fewer, the better.
But we have the technology now to do a lot more than we are doing. I have a lot of ideas re: 'smart' planes - it's fun to speculate.
But my experience with aircraft electronics is that simpler is better - fewer things to go wrong!
When the ignition system went out in my powered paraglider (I was crossing Interstate I-25 at the time), I was wishing I'd had mechanical points instead of an electronic module!
My only gripe about Hobbyzone electronics is the heat management in the motor driver circuitry for the bigger motors (Xtreme).
I'd gladly pay the extra buck or two for a bigger heat sink and better component layout.
Otherwise, the cost engineering is amazing, and the reliability much better than the knock-offs I've seen from Chinese factories.

p0stal

Wow. This post has way to much reading in it for me.

critterhunter

Yes, I've also replaced the FET on a few Extreme boards with a better one after they blew. This fixed that problem, but after a few hard crashes the boards still puked. One had a hair line crack in a trace going to one of the servos. Another I couldn't figure the problem on after extensive use with a multimeter. Still, the main failure of these boards on the Extreme is a weak fet/diode due to (as you said) increased amp draw of the 540 motor.

Hagar69

Hey, I have a broken F-27 Stryker RX, it allways runs the motor at full power. Sounds like my FET is blown. Anyone have a part number for this? Is it worth fixing?
I allready bought a new RX (LiPo compatible) so maybe I should just forget about the old one. But sometimes it is very satisfying to be able to fix something

Hagar

khalsans

My Feedback: (7)

Yes, it's worth fixing, if the board isn't cracked like Critterhunter's.
If not, sell it to me. I have so many folk's planes here that need crystals, etc.
Makes me thing I should charge for my service

A motor that runs continuously is usually a shorted FET.
A motor that does not run at all (but the servos work) is usually an open FET.
The diodes can short, but they usually take out a trace, or the FET, or both, since the diode is in parallel with the motor. The FET is in series with the motor.

I use the IRL3803 as a replacement FET that I get from Digikey or Newark.
It costs about $3, I think. Digikey sells small quantities; Newark would rather not be bothered.

It has to be a low-theshold FET that turns on at the 3.6 volts coming out of the processor. The IRLZ44 might work pretty well.
I'm not sure if Radio Shack carries any FETS with low threshold gates. If you use a FET that does not turn on fully at this voltage, it will look like a resistor instead of a switch, and it will get hot (and your motor will run slower).

The diode used is usually a 1N4001, but I put in a bigger one. I have a bag of surplus 100 volt 3 amp diodes (UT4010) that I use, but others will work. The 1N4001 is a 50 volt diode, so I like to go higher in voltage if it fails from voltage spikes from the motor, like a 1N4005. Does not cost much more. Radio Shack has some of the 1N400x series diodes, I believe. The bigger the last digit, the higher the voltage.

Parkzone does not usually put heak sinks on their FETs in their receivers. It ruins their slim figure. But I would, if it will fit. Every 10 degrees (Celcius) increase in operating temperature halfs the life of the semiconductor (rough rule of thumb).

thermos

My Feedback: (3)

Hello All,

I have an ABC that is buggered. I had switched my batteries over to Deans plugs and was charged up and ready to go. I got to the field and realized that I had not switched over my aerobird to a Deans plug. Desperate to fly, I cut the connector appart and inserted the bullet type ends into the battery plug hoping to get a decent connection. I forgot to turn on the trans first and the motor started up and the plane jerked. I crossed the wires as a result and buggered the board up pretty good. I am not sure why this happened. Wouldn't this have only shorted the battery? I am not 100% sure but I am pretty sure that I did not switch the pos and neg. I went home and put the proper plug on the bird. Here is what is happening now:

One of the servos is turning when I connect the battery. This servo does not respond to the trans and turns the same direction when the battery is connected. I almost broke a control line before I disconnected them. I completely removed the board to test things out. I noticed that the solder on the black wire to the motor was almost melted free at the board. When I was testing the controls the motor seemed to work OK but then the diode started to smoke. I disconnected the battery and desoldered the diode. I checked it and I got current in both directions. Not good. I also removed the FET.

I am desperate for some help.

How can I test the FET?

Would the diode effect the servo? This doesn't seem to make sense to me. Any other ideas what may be causing the servo problem. The servo is obviously not completely dead if it turns when I connect the battery.

PLEASE HELP!

Calm days to come...

khalsans

My Feedback: (7)

<>One of the servos is turning when I connect the battery. This servo does not respond to the trans and turns the same direction when the battery is connected. I almost broke a control line before I disconnected them.

I have a board like this. The little surface mount FETs that drive the servo are probably ON all the time, causing the servo to run. I'm currently looking for a replacement for these, but they are very hard to swap out without surface mount soldering equipment. I pulled a couple from another dead receiver, spent an hour or two soldering them on my board and managed to switch the N-type with the P-type, and they went POOF when I plugged in the battery.
I pulled the bad parts off again, but damaged the pc board traces in the process, so I think I need a new receiver. And anger management lessons.
These are the FET replacement cadidates I an going to try (available from Digikey):
N channel: FDV303N specs: 25V, 680mA
P channel: FDV304P specs: 25V, 460mA

gigelus2k3

khalsans,

What soldering iron are you using? Even if the PCB is crappy pertinax, you should be able to swap a few times the SOT23 parts. The nice thing about this package is that it's easy to de-solder: simply put your iron tip on the drain (the pin which is alone on its side), wait a little (just a little) for the solder to melt and, applying a constant pressure into this pin, pull-up the iron tip. The part will flip up, with the other two pins acting as hinges. After that, flip up-down the part a few times and the pins will break. You can then remove the pins off the landing pads and place the new one.

You can do perfect SMD soldering with the most ordinary 30W iron. The secret is to use solder flux. Avoid the rosin thing, which is not so good. There are water-based fluxes that make the solder joint to look just about perfect.

Regarding the NMOS vs. PMOS transistors, if you can't read their marking, you can use a meter to determine their type. You can check the body diode with the diode function of your meter. The diode is between the drain (pin #3) and source (pin #2). For a PMOS you should see 0.7V when placing the + lead on the drain and overrange when placing it on the source. For the NMOS the reverse stands true. Pin numbering for SOT23 is almost universally as follows: looking at the package from the top, pin #3 is the lone pin, then clockwise, pins #2 and then #1.

Good luck next time!

critterhunter

Khalsans, think you asked if anybody has any boards they'd like to sell. I won't sell them to you but I'll give you them. They are beyond my repair ability (something other than the FET has gone bad and my basic diagnostic knowledge couldn't find the problem). I've got two or three Stryker boards, Extreme boards, and Challenger boards you can have. A friend also has a a few more Extreme boards to that I'm sure he'll donate. PM me with an address (if you live in the US) and I'll ship'em out when I get the chance at no cost.
This will be minus the motor as I've got other projects for them.

thermos

My Feedback: (3)

The servo does not run continuously. It turns about half a turn when I plug the battery in and then stops. The trans has no effect on that servo. I see the fet that you are talking about. That looks pretty hard to solder. I will give it a go.

Thanks,
Ian

thermos

My Feedback: (3)

Can you change channels by simply changing crystals? I just bought a used aerobird with 2 fuses wings etc on channel 3. I have 2 complete planes on channel 3 but only one trans on channel 3. I have a good channel 2 trans and toasted channel 2 board (which I could send to you for the cost of shipping after I remove the crystal) that has a good crystal. Could I simply swap out the chan 2 crystal to one of the channel 3 fuses? Would there be any other tuning needed?

Thanks,
Ian