Thread Tools
Jan 07, 2016, 11:48 PM
Registered User
Originally Posted by flybabo
Panasonic produced the highest capacity 18650 (NCR18650B: 3400mAH, 3.7V) and equivalent cells are powering Tesla Model S.
I'm also using 3.6V 2400mAH (CGR18650) cells.
Both are good and delivered claimed mAHs.
Note that most no name 18650 cells sold in eBay have fake mAHs.
I'm feeding a balance line to 6pin DIN connector found on old Kraft/Proline and charging two cells with an external balanced charger.
Good to know the brand. I checked mine and found smaller Panasonic CGR17500's in my very old Futaba (18650 didn't fit in the 8U).
Long abandoned MP8K's still have three ENERGETEC STG GR18650GR's. I used to charge them with a lab power supply, but as the price of Li-Po chargers came down to a parts cost, and not having a stand alone charger, I moved to Li-Po. I will definitely go back to Li-Ion's if I get a charger.

One question on the on-boad dc-dc is the peak voltage at the flywheel. it may call for a caution against electrical shock. I looked at some $1 dc-dc on eBay. It uses 50V cap the rectifier. I don't think it is a real problem since the current drain is 100mA or so for the charger, and the pulse width should not be so large at high voltage.
But, it would be nice to know whether I will get a shock when I play with the board
Sign up now
to remove ads between posts
Jan 08, 2016, 12:53 AM
Registered User
flygear's Avatar
Originally Posted by flybabo
I bought those and use it everywhere
All you need is a buck converter chip, a 100uH inductor and several passive components.
Every FrSky RF modules have one of such.
Or more expensive end product like this:
Certainly, they are far expensive that I expected.
could be much lower by putting those circuit on the board but still They are far too pricey to use them I think.

How about, have some options or pin headers for extra power sources? then you can attached any of those on your own.
Jan 08, 2016, 02:39 AM
Registered User
Originally Posted by flygear
Certainly, they are far expensive that I expected.
could be much lower by putting those circuit on the board but still They are far too pricey to use them I think.

How about, have some options or pin headers for extra power sources? then you can attached any of those on your own.
Far too pricey? It costs $0.86 per unit including shipping!
9XR-PRO has an on-board switching regulator.
Last edited by flybabo; Jan 08, 2016 at 02:48 AM.
Jan 08, 2016, 11:49 AM
Registered User
Is the 9XR-PRO regulator integrated into the same board of the logic unit? Developers normally avoid sharing the ground plane between power and signal units due to a risk of common mode noise...a separate board is much easier to control thru a single point power feeding.

Yes, the price of DC-DC is outrageous! How they can make it sooooo cheap?
I think we should just buy it and make it as a daughter board instead of risking the logic unit.
As you know, the noise characteristics are quite different between the step-down and step-up converters.
The choice depends on the voltage from the DC mains. If you want to charge the TX thru USB, you need the latter, :roll eyes:

Well, that's my two cents.
Last edited by signalone; Jan 08, 2016 at 12:09 PM.
Jan 08, 2016, 12:55 PM
Registered User
flybabo and folks,
Sorry, I just remember a simple fact; we do NOT operate TX while charging
So, please disregard the noise issues and you can keep my 2 cents

However, the failure rate goes up 2 to the power of the heat rise, if I remember correctly. I prefer to have a spare power unit rather than to repair the main board. I already burned an input cap on one of 9XR's
Jan 08, 2016, 01:04 PM
Registered User
Since I don't own any Chinese made plastic transmitters, I can't take a peek
But this photo (post #3) shows HK put a switching regulator module in upright position.
It appears that 7805 pin compatible type of module.
Typical buck converter type switching regulator switches under 100KHz and
I guess it shouldn't cause any problem if you have a good ground.
As a matter of fact, every FrSky 2.4GHz RF module has on board DC-DC down converter.
I put one of those on my FPV setup and it doesn't seem to affect the quality of video.
Jan 08, 2016, 05:19 PM
Registered User
Originally Posted by flybabo
As a matter of fact, every FrSky 2.4GHz RF module has on board DC-DC down converter.
I put one of those on my FPV setup and it doesn't seem to affect the quality of video.
Sounds good!

From the picture, the regulator in the XR9 does look like a step-down DC/DC. A large black SMD should be the plastic molded inductor (normally not used in a step-up). Again, if the PSU is for charging, it doesn't matter what/how install the PSU.

Generally speaking, harmonics of the switching frequency themselves are rarely an issue, except for operating on LF. In many cases, issues come up when the receiver front-end (or the mixer) is intermodulated by the PSU noise (or 60Hz in old days) through ground loops. That is for 2-way radios. If TX is only transmitting, or not sharing the PSU with the telemetry RX, the PSU can be quite noisy.

I am a little bit confused now...are we talking about DC/DC for a Li-Ion charger? Or, the voltage regulator for the logic unit?
Jan 08, 2016, 06:06 PM
Registered User
flygear's Avatar
Originally Posted by flybabo
Far too pricey? It costs $0.86 per unit including shipping!
9XR-PRO has an on-board switching regulator.
Ops! The price was for 10 units.
Surely its low cost.
I don't see any reason not to considering this one.
Jan 08, 2016, 07:32 PM
Registered User
I'm talking about on-board voltage regulator powering the logic board.
That cheapo has a small pot adjusting the output voltage that you don't need it.

Li-Ion cells must be charged/balanced by an external charger.
Jan 08, 2016, 08:18 PM
Registered User
Now I got it.
For logic board, an on-board step down DC/DC is fine.
Jan 09, 2016, 02:38 PM
Registered User
Similar problem caused by 0.1u cap on poorly designed original Turnigy 9x board (not 9XR).
It appears that the aileron trim switch ground is connected to the analog ground not to the digital ground.
Since the analog ground is isolated from the digital ground via 10uH choke to suppress the processor switching noise,
clicking aileron trim will produce pretty big ground bouncing.

I also noticed that the ground wire (black) of ISP connector of 9XR board is incorrectly soldered to the analog ground which is isolated by 10uH inductor labeled as L2.
If you program the 9XR with higher baudrate, it may cause a problem.
Jan 09, 2016, 02:44 PM
Registered User
Since I will eliminate 4 analog trims and use two 5-way NAV switches for digital trims. Also I will eliminate 6 menu switches, and use the same pair of NAV switches for Menu selection with a PROG/RUN switch selecting the ground.

So, I now have ;
Analog input: 4 for joysticks, 4 from old trims, 3 AUX, total 11 inputs
Digital input: 8 for new trims, 3 for D/R, 2 for MIX, 2 of 3 pos. witches (total 4), 4 push buttons, total 21 inputs.

I have following questions;
(1) With your new board, how many inputs can be used?
(2) Presume that your board is capable of controlling 16 channels over the air. Is there any plan/option to expand the input count to utilize all the 16ch?

Since your board is cheap and small, I can use two in the TX since I do not have to worry about the internod between the two RF modules at 2.4GHz, which is quite difficult to achieve at 72MHz. Only? problem is that I need two LCDs for the two boards.

So, next questions are;
(3) what is the output spec on the LCD driver? can I tie signal lines together and select one board by turning the other off during programming? What is the driver's loading characteristics? Will diodes on the signal lines work in terms of the threshold margin?

I only need to see the timer when I fly, so I do not need display from the "sub" transmitter.

(4) The board is no longer plug-in compatible with XR9 w.r.t. the size and connectors, and seems targeting more general users. Is it possible to change the LCD interface to the serial SDI so that we can get simpler interface and more availability/variety on LCDs?

The industry is tasting the water of the dual transmitter, or at least expanding the channel count. We used to invest enormous amount of money to fly a 4-ch model plane. Now, everything is 1/10 of cost of what used to be in 1970's ( or whenever you started flying...). Wanted channel count is limited only by our imagination, as someone said

Any way, your board inspires me a lot.
Jan 09, 2016, 04:56 PM
Registered User
I'm not the "flygear" but I can answer some of your questions
To support your feature set, it requires firmware modification.

Since current hardware model of all the 9x firmwares (er9x/ersky9x/...) only supports 7 analog pots, 8 trim switch inputs, 8 switch inputs assigned to 6 2-pos switches and 1 3-pos switch, adding the new hardware resource requires major effort and creates many backward compatibility problems.

[Slight off topic...]
As for the er9x front, I re-architected switch handling and tightly integrated the voice card (aka MegaSound) as a coprocessor on er9x firmware from ground up.
I added 2 additional analog pots for left/right sliders, 4 additional 3-pos switches and 2 push buttons, telemetry data logging, RTC support...
Its's up and running more than a month and is close to release as an extended version of er9x-v2.

If you don't have a voice card, adding a coprocessor to a 9XR board is really cheap and easy.
Here's a picture of my test setup based on Arduino ProMini board ($2.50) with a MicroSD card adapter($1), an RTC module($1) and Taranis' slider pots($7).
You can add an audio amp to enable "talking" feature later as shown in the second picture.

[Back to original topic]
As I don't have any vested interests in ersky9x firmware since all my works are based on 9XR/er9x, I have never paid any attention on it.
As I got two AR9Xs and maybe more in the future, I have to take a look

If you put those extra hardware resources, you can quickly fill up 16 channels using various mixings.

As for the serial LCD, I wrote a serial LCD and an OLED drivers for er9x about two years ago for my first Proline project.
OLED looked really cool indoor but it's useless under the direct Sunlight.
Serial LCD requires just 5 wires including PWR and GND.
But, it's about 8 times slower refreshing the screen even though you won't notice it.
You can select it from eepe with a custom hex on my blog page.
I'm planning to include it in er9x-v2 official release.

I'll back port LCD related code to ersky9x if it's necessary.

BTW, did you check 5-way guitar tone selector whether it is a real 5 way switch?
Jan 09, 2016, 07:19 PM
Registered User
The 5-way guitar switch is actually DP3T, and it can't be used as a direct replacement of SP5T rotary.
I thought about a diode matrix to change the coding, but I chose the shortest way; I cut the pattern.
If I had steady hands, I could have made the cut so that it had make-before-break contacts. But at that time, I was not thinking about the contact gaps in a rotary switch. Sorry for the blurry picture.The contact numbering is not in order since I chose to use the existing pattern to the terminals. Actually the picture only shows the cuts on the upper half, but the bottom half also needs to be cut.

Since I still use EVO12, it was my first target on the modification.
After many tries, I successfully modified one of the slide POT housings. Everything was in place after time consuming realignment ...THEN I found the TX back lid didn't close The height of the switch was 5mm too high...

So, I am back to MP8K and started thinking how to make cuts for a make-before-break switch.
Also started thinking how to calibrate the 5-way if FW assumes 6-way. May be I add another switch for NORMAL- OTHER 5 MODEs.
Jan 09, 2016, 07:54 PM
Registered User
This is the final cut. Because of the roughness of the cut, the reliability of the wiper surely goes down. I wonder if some one could make a PCB for this switch and implementing make-before-break. If there were register pads for a voltage divider on that PCB, I guess people would be happy.

Quick Reply

Thread Tools