Devo 12s 3in1 / 4in1 RF module neat installation.

@goebish,
I tested again again the 4in1 module with last multiprotocol code.
Now it is binding ok Hubsan and Cx-10 blue
There were some modifications in the main code and a bug fix on CX10.
Remembering back I used before V911 module which has only PA and working ok with Hubsan(Hubsan original has Rf module with PA only).
So it remain to fix in 4in1 module full range Hubsan telemetry(LNA),which can be done as a HW hack on existing module.

Cut a trace from 4042 chip pin 12 to Vcc and solder one wire from the same pin12 to A7105 GIO1.

@midelic, did you actually try this hack? Because looking at tightly packed 4in1 module I'm a bit afraid to solder anything there. Also you should remove pull-up for TXEN to enable LNA for nRF.

just got an email that my back ordered 3 in 1's have shipped, so I just checked bg and they still have some available in stock (oops they have gone back to "ships in 10-15days" also they have gone up in price from when I placed my order), I presume these will be the latest revisions with the required fixes

victzh wrote: @midelic, did you actually try this hack? Because looking at tightly packed 4in1 module I'm a bit afraid to solder anything there. Also you should remove pull-up for TXEN to enable LNA for nRF.

In my eyes there is more simple modification to the 4in1 module.
Here are the results of my findings:
Multiplexer CD4052 pinout:
Pin 11 <- CYRF6936.XOUT => selected for CYRF
Pin 12 <- 3v3 => selected for A7105
Pin 14 <- CC2500.GDO0 => selected for CC2500
Pin 15 <- NRF24L01.VDD_PA => selected for NRF

Pin 13 -> RFX2401C.TXEN through a 1K resitor and a 4K7 to VCC

The modification idea I have is to cut the trace going out of 4052.pin13 which is easy to do as there is a lot of space. Connect one side of a wire to the RFX2401C.TXEN 1K/4K7 resistor which is easy to do as well and then the other side to either 4052.pin11=CYRF or 4052.pin14=CC2500. Optionnaly add a second wire from the RFX2401C.RXEN 1K/4K7 resistor to the CYRF6936.PACTL or CC2500.GDO2.
Then change the code with a #define 4in1_module in the few rf functions used to switch RX/TX/Standby states to instead use the CC2500 or CYRF6936 GPIOs which is really easy to do.

- Pascal

Pascal,

yes I've seen it being discussed somewhere. It's easy to do for your multiprotocol module, not that easy for Deviation - it has configurable CSs for different chips. I think that the damage to the early adopters (me included, I bought these modules from TaoBao early on) is not so high we should compensate for it in the main code base. To do so in UniversalTX seems to be quite logical - no such configurability there.

One question..
I buyed a 3in1 module from the first batch on banggood...

The first batch of modules have problems??

HappyHarry wrote: just got an email that my back ordered 3 in 1's have shipped, so I just checked bg and they still have some available in stock (oops they have gone back to "ships in 10-15days" also they have gone up in price from when I placed my order), I presume these will be the latest revisions with the required fixes

Mine was shipped as well.
I am still not quite sure if this second batch 3-in-1 from BG has issue. I thought its NRF24L01 had range issue in certain protocol due to its low receiving sensitivity. Then someone pointed out that was because the higher data rate.... Does anyone have its schematic?

the issue with the 3in1 was that the lna wasn't wired up properly for one of the chips, this has since been fixed so the 3in1 has no more issues, the 4in1 also has some issues and these are being worked on at the minute. also the issue you are referring too was mentioned due to data rate, not tx or rx power/sensitivity and it will be the same whether you use a 3in1 or 3 seperate modules as it's the code not the hardware that sets it

ursus69 the first batch had an issue where the pa/lna for the cc2500 wasn't connected properly so this resulted in poor range for protocols usiong this chip, but dc59 informed the manufacturer and he has fixed this error, so this current batch of 3in1 are the good ones, but you can modify the original batch to work, if you search this thread you will find the info

there was mention of a power supply issue when using these modules in the 7e, in that sometimes one of the rf chips don't initialise on startup, i was wondering if there is something we can do to fix this? put in a better ldo vreg or similar? anyone have any thoughts or suggestions? thing is i'm about to knock up an 7e-256 with a 3in1 so i thought while i've got it open for surgery i'd tackle this at the same time

the seller did mention about the over heat issue from the regulator
i have been using my 7e with 1s supply for years with zero problem. with a mere 4.2V, the heat generated should be minimal. too bad that he only get one prorotype on hand and cannot spare one to me to test

they mention about the issue of 'missing modules' which the text says is caused at startup by too much power being pulled through the vreg, and iirc on the 7e the 3.3v ldo reg is only rated for 250ma peak, I'm pretty sure a couple of years back there was a thread on here about someone replacing it but the search hasn't turned anything up. so if the 3in1 and 4in1's pull too much juice the outcome could be nasty, premature failure of the ldo is one possibility and dropping in to thermal cutout/protection mode while flying is another. perhaps a second vreg just for the multi boards might be an option instead of replacing the stock one?

use 1S supply in the 3in1 with zero issue so far

ok it turns out the 3.3v ldo is only a 100ma version, and looking at the datasheet (i've attached it here) at the rated 100ma the vdroop is only 0.2 so as long as you stop at ~3.6V you should be fine with your 1S lipo regarding voltage but as for temperature who knows, the thing is that we are pulling more than 100ma through this chip when using additional modules (see here for some info >> www.deviationtx.com/forum/6-general-disc...w?limitstart=0#24216 ) plus the 3in1 and 4in1 have additional chips using more power. the ldo's power dissipation is 220C/watt hence the overheating.

myself (and i presume others) like the convenience of using nimh cells (especially the fact i can plug it into a wallwart and let it trickle charge overnight) so the ldo is going to be pushed harder when using 6V and may even drop into protection mode. i think replacing it with a more powerful version would be a good idea and i've attached datasheets for some possible options. what are peoples thoughts on these options? any better suggestions?

Wouldn't a simple AMS1117 do the trick here, or am I missing something? You'd need 3 wires due to the different package type, but it'd also dissipate heat better.

HappyHarry wrote: i think replacing it with a more powerful version would be a good idea and i've attached datasheets for some possible options. what are peoples thoughts on these options? any better suggestions?

I actually started a topic about upgrading the regulator here , unfortunately the topic didn't quite get much attention.

Anyhow, The current regulator is the LP2981. It has a PPO of 100mA. I was thinking of doing one of 2 possible solutions:

1. Replace the SOT-23 regulator with a MIC5219 (PPO 500mA)
2. Hook up am AMS1117-3v3 (PPO 1A) to pin 6 of the current RF module to provide another isolated 3v3 regulated supply for extra modules.

Solution 2 should be easier for everyone to accomplish, but I am slightly concerned about possible ground loops (if any) and am hoping someone with more experience can give some feedback on this.

What's speaking against replacing the existing regulator with the AMS1117? That'd give you 800-1000mA, so plenty of headroom for any modules, with no worries about groundloops.

i opened up my 7e (running FrSky protocol @150mW with the 3in1 module) .
Touching the L04B regulator and it barely has any temperature at all. May be because the Vin is low (1S).

Furthermore, I touch every components around in the pcb and nothing gets hot either..

Please test the current.
My 7e uses nearly 200mA (frsky at 100mW) for the whole tx at 5V level.
(CC2500 /blue, from BG) No other additional modules attached at the time.

Arakon wrote: What's speaking against replacing the existing regulator with the AMS1117? That'd give you 800-1000mA, so plenty of headroom for any modules, with no worries about groundloops.

TBH, I don't really know if there is anything wrong with removing the original regulator and replacing it with the AMS1117 module through 3 wires. I don't have enough experience.

I do remember reading about PCB designs for ground return paths and "guiding" it through better design to try to keep analog ground and digital ground separate for the main processor. This can limit "cross talk" between our analog inputs (pots in the gimbals) and the other digital IOs. I'd assume that the Walkera PCB designer considered these when designing the board, so removing the existing regulator to the processor and changing the ground return paths could have some effects (they could of course be negligible).

As such adding another regulator that shares the same VBAT and GROUND pins that the RF module's regulator uses may maintain the ground path integrity around the processor, as those points are already designed for a separate regulator to power RF type devices.

I might be reading into it too much, but like I said, I don't have the experience/knowledge to come too a conclusion. Hence why I asked the questions in the other post.