Shop our Airplanes Products Drone Products Sales
Thread Tools
Dec 27, 2016, 11:40 AM
Registered User
Thread OP
Discussion

DIY based on Eachine VR D2 Goggles with Diversity and DVR (by Batlapin)


Initially that text was located at the "Eachine VR D2 Diversity FPV Goggles with DVR" thread but was asked to make new thread keeping all that info.

Hello everybody!

Going to share information on VR D2 and to present mine "Dream VR D2".
Are there any questions after reading, feel free to ask.

And sorry for the long reading.

Internally VR D2 combined (with minimum efforts) of three INDEPENDENT modules:

1) Diversity board: 2x RX5808 + some 16-pin Chinese MCU;
2) DVR module (that is the Genuine Eachine ProDVR, version depends on you luck);
3) Display module based on Mstar MST706;

The buttons on the VR D2 front are three independent groups each connected to own module: 1st row "FR"/"CH" -> Diversity, 2nd row from left to right: 3 buttons -> DVR (shared with on-board DVR buttons hidden inside) and 3 buttons -> Display. Again: all button groups are absolutely independent from each other and every module operation is independent from other modules. That means that it is absolutely not important do you switch DVR record ON while monitoring AV2 or AV1. More that that, DVR is independently powered on when you connect the battery and it stays on all the time. As do the Diversity and the Display. The only you can do is to switch Display part to sleep mode - just because Display controller board is made this way. No influence on/from DVR/Diversity.

1) Diversity board

Two connectors: A/V/Power out to Display and pins to connect on-board DVR buttons/power to DVR. The DVR Video Out is connected to Diversity but only to pass-by to the A/V/Power to Display.
RX5808 receivers are SPI-controlled by some Chinese MCU, and MCU is switching Video out signal based on RSSI levels from both RX5808. Additionally MCU displays tuned band and frequency on 2-digit LED and manages FR/CH buttons (yes, they are connected in wrong order - that can be easily corrected by cut/rewiring).
The board routes Audio out from both RX5808 and even switches them with Video Out signals but further the Audio Out is passing by to the nowhere.
The board is powered by external 2xLiIon battery, there is the StepDown IC that outputs 5V to power 2x RX5808 + MCU + 2-digit LED with shift register + output is passing by to DVR.
The external power itself is routed to the Display connector.

2xRX5808 + Video/Audio switches seems to be quite nice base for Atmega328P based receiver with OLED instead of that low-functional 2-digit LED. For the moment that is the future project.

2) DVR

Not too much to say - it is just ordinary Eachine ProDVR with buttons routed to the VR D2 front. Even the LED is at its place hidden behind the screen. Powered by 5V provided from the Diversity StepDown converter. You can control DVR any time battery connected, independently from what AV input selected on Display. The only thing that confuses many users - DVR provides VideoOut in ~8 sec after been powered on. That is why so many people tries to switch Display to AV2, no signal, back to AV1, AV2 again - oh, here is the video! And decides that there are special press sequences to activate the DVR - long AV1/AV2 press or something else. No way. DVR does not know what you are doing with Display buttons! It just starts up on battery connected and always on after ~8 sec delay.

3) Display

That is the cheap (~15-18 USD Aliexpress) Chinese 5" TN TFT CarTV-monitor module. The best feature is the snowscreen (no video drop-out on weak signal). It is powered by external battery directly passing by Diversity board. The Display Buttons located on the Diversity board are connected to the Display board in usual resistor-dependent way, every button shorts the connected with MST706 ADC input common single wire to the ground with different valued resistors and on that the scaler (MST706) detects the exact button pressed. There is even the AV2 priority circuitry presents on the Display board (to switch AV2-view when car rear cam is activated - as I said, that is the CarTV-monitor module) but it seems to be disabled in MST706 firmware as there is always active video signal from DVR.

All together is the easy-to-design solution with a lot of cons (and some pros if we say on the combined proce etc):

1) There are NO inputs/outputs. No way to view receiver output on external monitor, nother possible to use external DVR or use the in-build DVR/Monitor with external receiver.
2) All three modules are power-hungry and always on. Even when you do not need DVR, it discharges the battery (quite fast).
3) Propietary charger. Personally I hate propietary charging solutions. That is what you can forget at home going to vacation.
4) As that is the connect-and-sell design, nothing done to reduce power consumption. Is there anybody who is luck to use that goggles for 4 hours from one charge?
5) Everything is already said on optical scheme.
6) That WEAK Power Connector!

Now it seems that I have resolved issues 1, 2, 3 and in minor the 4th one. And 5th one, but that was the easist. The 6th disappeared as process part.

Starting from power curcuits to rebuild device to be powered from 1 LiIon cell (that can be easily charged from USB adapter) I have added 3 power switches and 4 miniJack connectors. Now I can switch on/off each of Diversity/DVR/Display in any combination and have ability to: view (and hear) Diversity output using external devices, input external sources to be recorded by the internal DVR and monitored on the Display, output internal DVR Video/Audio Out signals, and AV2 can now be used to monitor second external source. Every internal module can be replaced by external that way. Example usage: 1) view/record 1st external source been able to switch to 2nd external source = external receiver+external DVR.

Much efforts was made to minimise power consumption. Not the huge gap but now goggles consumes ~1W less and that means the temperature drops down. Just nothing can be done with DVR (it is nice standalone solution that can be replaced by another ProDVR in case it is broken so I have decided to change nothing on that module), consuming upto 1.2W (5V) or ~1.4W powered by StepUp or StepDown. But checking the internal RX5808 power curcuit I have found LDO (quite good Richtek RT9193) that generated (5-3.3)*0.16*2=0.5W of heat for both modules. I do understand why the analog curcuits preffer the LDO regulators to the PWMs (output ripples) but that seems to be wrong solution for the portable device (and the PWM output ripples could be reduced to aceptable level by LC filters). And the second power-hunting point is the Display board. It is designed to be powered from the Car 12V source when nobody count the milliwatts but the VR D2 is not the Car. At first the Display power curcuit steps 2xLiIon voltage down (by XL1509 IC with very low typical efficiency=76% due to the datasheet and with measured efficiency ~80-82%) to 5V, after that the scaler chip and TFT-logic are powered by... LDO 3.3V (down from 5V) output and the LED backlight is ... powered up again from 5V to ~17.5V/34mA. Every step produces heat, lower efficiency => more heat.

As i have told, mine solution was to power the goggles by single LiIon (3-4.2V). Each module received its own StepUp (boost) converter to (a) allow individual module usage and (b) redice power per converter (that increases the efficiency for that converters type).
Now the power scheme is:
1) Diversity: LiIon -> StepUp 4.9V (eff. 87%) -> ( MCU+LED+StepDown 3.1V (eff. 88%) -> 2x RX5808 ).
2) DVR: LiIon -> StepUp 5.15V (eff. 87%) (and most of the DVR curcuits are powered by internal PWM StepDown chips so really not so much idea to review something there). How can one decide are there LDO or PWM inside? Easily. You power the curcuit arying the voltage (in allowed range) and check the current/power. If the power is close to constant - the PWM is used. If the current is constant - the LDO.
3) Display. The LiIon directly powers LED Backlight StepUp circuitry (that part is stable powering even by 2.7V and the efficiency is ~88% in all voltage range) and there are two PWMs: LiIon->5V and 5V->3.1V.

Why double voltage conversions, LiIon->5V and 5V-> 3.1V, used lowering the efficiency? Because I have failed searching for someting better providing 3(.3)V from single LiIon cell. There are the universal StepUp/Down modules available but their efficiency is even lower compared to double-solution. And 3.1V outputs instead of usual 3.3V as that allows reducing power consumption even further because some MST706/RTC6715 internal voltage regulators are LDOs and lowering the voltage 3.3->3.1V means +6% efficiency for that circuits. And less heat, YES.

Additionally I have moved the power wires output spot behind one of the antennas from the previous location where wires obstructed microSD access. And located the DVR board angled to the supporting it Diversity board. The last was done as I was going to incline down a little the 5" TFT as that way TN display is brighter and there is space inside goggles allowing that, but later have rejected that idea as the project already took too much time.

Oh, on optical solution. That was the easiest part, I have used some +4.00 glasses breaking the frame and installed it in spaces formed just behind the goggles face-side surface. Fixed with cut of some soft material (Eachine use it to pack the Cameras/Transmitters).

Now to the fun part, pictures (some of them).
There are 2x StepDowns glued to the Diversity board shown by one photo - later had to move all converters to the TFT backside as there were screen-visible noise, the place initially choosen for the converters is too close to the Diversity video path.
PS Yellow LEDs are power-effective each consuming only 2-3mA.

By the way, had to CAD-design VR D2 (at least the main parts) as that was not obvious will power switches fill or not due to curved surface and har-to-measure internal space. During that process understood that possibly designing from scratch could be easier. Now am interested in 2x micro screen design like FatShark. Are there any cheap micro-display sources? It seems the compact photo camera spare screens could be used, but need to get some knowledge on interfacing.
Sign up now
to remove ads between posts
Dec 27, 2016, 12:21 PM
Registered User
Thread OP
More detailed info on Diversity board is on the photos attached.
You can find both sides with 2x RX5808 etc.
The MCU compares RSSI levels from both receivers and switches video/audio output to the receiver with more powerful signal.
Every part is marked directly on the photos.
Please find that:
1) Audio outputs goes all the way throught switches to the board connector but not used after that point.
2) Video from DVR output connected by the long connector not used on the Diversity board itself but just routed to the smaller connector to pass alltogether with V1 to the Display board.
3) There is voltage sensing resistor divider that allows MCU to check battery state and show that "Lo" on the 2-digit LED. But "Lo" is way early starting from 7.2v. That divider can be easily modified to set another "Lo" level.
Buzzer. Just did not install it back after mod.
The power scheme is :
- The 2xLiIon voltage (ideally 6...8.4V but the connecting wire and thebattery connector itself are so weak that 8.4 hardly could be measured on the board input) - will call it "7.4V" for shot - (a) directly powers StepDown with 5V output (b) passes to the Display board (c) divided and monitored by the MCU.
- 5V StepDown output powers up: (1) MCU (through the diode - cannot understand what it for) (2) 2-digit LED with shift register (again the dode on the way) (3) both of RX5808 (4) passes to the DVR board
- RX5808: 5V input is lowered down to 3.3V by Richtek LDO .

Heat/Power calculations:

Single RX5808 consume ~160mA @3.3V and the same 160mA @5V
Two RX5808 LDOs produce excess heat 2*(800-530)=540mW (and current is 320mA @5V
The MCU with 2-digit LED consume ~40mA @5V (can be wrong as was not systematic during measurements)
Then, there are 2 video repeaters powered by 5V with output voltage in range 1.5-3.0V and loads are (150R parallel to 75R at the destination input) - ~90mA to add.
(320+40+90)*5 = 2250mW after StepDown or ~2560mW (2250/efficiency) before - that means 300...420mA @6...8.4V

The, the DVR is powered from the same StepDown and consuming upto 240mA @5V (during recording): 240*5=1200mW, 1200/0.88=1360mW from 2xLiIon or 160...220mA @6...8.4V.

The last power hungry guy is the Display Board. To be continued...
Last edited by Batlapin; Dec 27, 2016 at 12:31 PM.
Dec 27, 2016, 01:57 PM
tHe n00b on YouTube
Great work!
Dec 28, 2016, 07:02 PM
Registered User
thank you very much!
Dec 28, 2016, 08:53 PM
Registered User
Excellent. I'll be following this and trying some of mods out.
Dec 29, 2016, 06:29 AM
Registered User
Thanks for the excellent analysis!
I received my goggles yesterday and I am already planning to do some improvements. I am thinking of powering my VR D2 from regular 3s or 4s lipos, as I would like to reduce the number of batteries to be charged and things to be remembered before heading out for FPV flying. Also it seems that the original 18650 goggle battery would not be enough for a full day of flying anyway. I have some cheap chinese Step-Down circuits based on LM2596 at hand and was planning to used these to power the goggles. I think I could do something similar to what you did and power each component directly from the step-down unit at 5V, but do you think the ripple from a switching regulator cause problems with the receiver modules or the display? Or should I just adjust it to ~7.5 V and power eveything through both LM2596 and their own, in-built linear regulators? The latter option wouldn't be as efficient, but I think that the linear regulators might help in filtering out some of the noise from the switching regulator.
Jan 19, 2017, 05:32 AM
Registered User
Thread OP
Quote:
Originally Posted by Repe
Thanks for the excellent analysis!
I received my goggles yesterday and I am already planning to do some improvements. I am thinking of powering my VR D2 from regular 3s or 4s lipos, as I would like to reduce the number of batteries to be charged and things to be remembered before heading out for FPV flying. Also it seems that the original 18650 goggle battery would not be enough for a full day of flying anyway. I have some cheap chinese Step-Down circuits based on LM2596 at hand and was planning to used these to power the goggles. I think I could do something similar to what you did and power each component directly from the step-down unit at 5V, but do you think the ripple from a switching regulator cause problems with the receiver modules or the display? Or should I just adjust it to ~7.5 V and power eveything through both LM2596 and their own, in-built linear regulators? The latter option wouldn't be as efficient, but I think that the linear regulators might help in filtering out some of the noise from the switching regulator.
In your case it is easier to StepDown to ~7.5V. There are not so much troubles powering modules from 5V as the display board at first steps-down the battery voltage to 5V and after that there is the LDO to 3.3V - if you use external step-down to 5V you can power the LDO directly. The DVR is powered from 5V provided by diversity board step-down. And the diversity is powered by the same 5V.
Jan 31, 2017, 05:08 PM
Registered User
alex_007's Avatar
Excellent work! Thanks a loooot. Hope to read more from you mod soon! Big thanks
Feb 07, 2017, 03:42 PM
Registered User
flybytim's Avatar
Quote:
Originally Posted by Batlapin
More detailed info on Diversity board is on the photos attached.
You can find both sides with 2x RX5808 etc.
The MCU compares RSSI levels from both receivers and switches video/audio output to the receiver with more powerful signal.
Every part is marked directly on the photos.
Please find that:
1) Audio outputs goes all the way throught switches to the board connector but not used after that point.
2) Video from DVR output connected by the long connector not used on the Diversity board itself but just routed to the smaller connector to pass alltogether with V1 to the Display board.
3) There is voltage sensing resistor divider that allows MCU to check battery state and show that "Lo" on the 2-digit LED. But "Lo" is way early starting from 7.2v. That divider can be easily modified to set another "Lo" level.
Buzzer. Just did not install it back after mod.
The power scheme is :
- The 2xLiIon voltage (ideally 6...8.4V but the connecting wire and thebattery connector itself are so weak that 8.4 hardly could be measured on the board input) - will call it "7.4V" for shot - (a) directly powers StepDown with 5V output (b) passes to the Display board (c) divided and monitored by the MCU.
- 5V StepDown output powers up: (1) MCU (through the diode - cannot understand what it for) (2) 2-digit LED with shift register (again the dode on the way) (3) both of RX5808 (4) passes to the DVR board
- RX5808: 5V input is lowered down to 3.3V by Richtek LDO .

Heat/Power calculations:

Single RX5808 consume ~160mA @3.3V and the same 160mA @5V
Two RX5808 LDOs produce excess heat 2*(800-530)=540mW (and current is 320mA @5V
The MCU with 2-digit LED consume ~40mA @5V (can be wrong as was not systematic during measurements)
Then, there are 2 video repeaters powered by 5V with output voltage in range 1.5-3.0V and loads are (150R parallel to 75R at the destination input) - ~90mA to add.
(320+40+90)*5 = 2250mW after StepDown or ~2560mW (2250/efficiency) before - that means 300...420mA @6...8.4V

The, the DVR is powered from the same StepDown and consuming upto 240mA @5V (during recording): 240*5=1200mW, 1200/0.88=1360mW from 2xLiIon or 160...220mA @6...8.4V.

The last power hungry guy is the Display Board. To be continued...
Whats doing the video switching from the receivers? MCU?
Last edited by flybytim; Feb 21, 2017 at 04:18 PM.
Feb 09, 2017, 03:29 PM
Registered User

Strips of one of the channels.


Did anyone what is causing the strips from one of the channels?

Faulty VR-D2 goggles from Banggood II. (1 min 1 sec)
Mar 02, 2017, 07:16 AM
Quads/electronics enthusiast
Do you have any schematic for the video repeater? I want to split the signal out from a Realacc 5808 diversity module, to source video for both display and dvr without causing to much problems.
Mar 02, 2017, 11:25 AM
Registered User
flybytim's Avatar
Quote:
Originally Posted by jaukur
Do you have any schematic for the video repeater? I want to split the signal out from a Realacc 5808 diversity module, to source video for both display and dvr without causing to much problems.
Sorry can you go over that one more time, I Dont understand what you want.
Mar 02, 2017, 01:06 PM
Quads/electronics enthusiast
"Then, there are 2 video repeaters powered by 5V with output voltage in range 1.5-3.0V and loads are (150R parallel to 75R at the destination input)"

Do you have the schematic for that part? I have had trouble with just splitting my video signal in two and have my DVR and screen in parallel
Mar 19, 2017, 12:14 PM
Registered User
I love your thread sir, thanks so much...
May 28, 2017, 02:29 AM
Registered User
Hey Batlapin and any others interested. I replaced the microcontroller in the goggles with an Arduino Pro Mini loaded with a modified version of rx5808-pro-diversity. Here are some instructions I've written up.

https://www.rcgroups.com/forums/show...nd-Scanner-Mod

Cheers


Quick Reply
Message:

Thread Tools