majek.sh – Page 7 – Marek Wodziński's home page

G-OSD mod – new input

Posted on Monday August 20th, 2012 by majekSunday November 18th, 2018

G-OSD schematic diagram and making new analog input to it.

I was very curious how everything is connected in G-OSD. It is similar to E-OSD but different – they added potientiometers, analog inputs and gps. And there was also one mystery: third pot that trims RSSI and do nothing in CL-OSD🙂

So, I made it:

G-OSD schematic diagram

Some conclusions from this schematic arrived:
– third pot really does nothing unless you want to measure 5V from voltage regulator 🙂
– acording to resistor values in voltage dividers, maximum voltage on both inputs could be as high as 23V. I don’t know at how high voltage is rated 10uF capacitor (probably it’s 50V) on battery input, so be carefull with this input. But on Voltage2 it is safe to apply up to 24V.
– video generation circuit is exactly the same as in E-OSD
– pin for RSSI haven’t filtering capacitor, so it could be used also as digital input (for example for measuring RSSI PWM directly from FrSky receiver)

And I used opportunity to make something based on information I learned from schematic and pcb investigation: new analog input! 🙂
So, I made useless pot useful 🙂

Warning: this input has some limitations:
– low input resistance (1kOhm), so it’s not good for current sensor as it require at least 4.7k. Some simple RSSI pwm converters based on passive elements could also work wrong with it.
– voltage from pot goes directly to processor, so putting voltage higher than 5V and turning around this pot without extreme care could end up with death of uC
If you want to measure voltage higher than 5V, put additional resistor in series with this input to be safe. For example for 3C battery, use 2k or 2.2k. If you want more, the rule is: 1kOhm (or more) for every additional 5V.

Now you know limitations of this new input, so let’s do this 🙂
This is my second mod (here is previous), so don’t be surprised by additional pins on pictures, it’s still G-OSD 🙂
At start (it’s obvious 🙂 ) you should remove shrinkwrap. Then locate third pot (RSSI) and capacitor below it. Please observe that between them there is also a via (small drill/hole). You must cut trace (under white paint) between capacitor and the via as shown on picture:

G-OSD – new input – where to cut trace

And then you have on opposite side of pcb quite large pad to solder new input:

G-OSD – new input pad

Do you want it look well and professional? Stay with me for another few pictures 🙂

Get new angle pinheader, remove unnecessary pins and cut middle two pins to the edge of plastic as on this picture:

G-OSD – new input – preparing new pinheader

Get CA and glue pinheader to pcb with pins in right position:

G-OSD – new input – glue new pinheader into place using CA

After few seconds everything is steady, so you could easily solder pins to pads below:

G-OSD – new input – solder connections

Now you have one pad connected to new input, second one you can connect by wire to some near gnd pad.
Now you can put new heat wrap, and that’s all.

But…
I don’t like make things twice, so I put few addidtional pins as RFU (and there will be nice to connect i2c and tx pins in future 🙂 )

G-OSD – more inputs 🙂

G-OSD – like a new 🙂

Voltage divider for FrSky

Posted on Monday August 20th, 2012 by majekSunday September 23rd, 2012

Very simple voltage divider for FrSky.

I use 3S battery, so I have maximum 12.6V while FrSky accepts only maximum 3.3V on its analog inputs.
This time without schematic because there is no need to draw two resistors and two connectors:-)
For my case 1/4 voltage divider is just enought. I made it from 1.5k and 4.7k resistors, so with 12.6V on input it will give about 3.05V which is fine for FrSky. I made is as small plug, so I could make connecting cable of any lenght and change it in future without resoldering of divider.

Pictures, pictures 🙂

Voltage divider inside

Voltage divider – ready

Just works 🙂

just to be clear: this divider isn’t designed to plug directly into FrSky receiver or battery plug! It will not work that way! It must be plugged to power harness similar to my power splitter and another plug should be used for connection with receiver.

Power splitter

Posted on Monday August 20th, 2012 by majekFriday August 31st, 2012

Nice and easy power splitter.
It’s not high current capable, but just to connect accesories with small current draw.

Have you ever wondered how to connect all sensors, osd, video TX etc to battery? Me too 🙂
This is simple one to four power spliter I made today.
Why four outputs?
1. Video TX and camera
2. OSD power and main voltage metering
3. Output to FrSky sensor
4. RFU 🙂

Because it’s primary Video TX power connector, so this connection I soldered permanently.

Power splitter – just solder wires to pinheader 🙂

It’s very easy to do, so I won’t write much about it. One thing is worth to notice: before heating shrinkwrap put connectors on all pins, so after shrinking it become nice formed.

Power splitter – heat shrinkwrap with connectors to make nice shape

Additionally I added ferrites on both ends to filter EMI from video TX and main power source.

Power splitter – ready to use

And now I can connect bunch of stuff to my current sensor 🙂
For example my small voltage divider for measuring main voltage by FrSky receiver..

Current sensor (30A)

Posted on Sunday August 19th, 2012 by majekSunday August 26th, 2012

Current sensor for OSD or FrSky elemetry up to 30A.

Just another useful brick in telemetry equipment is current sensor. With right OSD or FrSky software it allows to show not only current draw but also remaining battery capacity, and this feature is VERY helpful in FPV.
I checked market for possible solutions and easiest (but not very cheap) way of doing this is to use Allego MicroSystems hall chips. I chose ACS713 chip because it is available at reasonable price, is available :-), and have current range sufficient for me (I don’t expect to go beyond this in my current models). It has also linear output proportional to passing current and moreover: it has separation between sensor and output, so one can use it to sense on high or low side.

So, chips arrived and there was time to design sensor. Agrhhh, it requires 5V of supply voltage so I had to add voltage regulator. 2 chips, let’s make schematic:

Current sensor 30A – schematic diagram

I set filtering capacitor to 100uF to average current, so few redings per second should be accurate with it. If not, larger capacitor will be needed.

PCB size is designed to fit XT-60 connectors on both sides:

Current sensor 30A – PCB

So, let’s make PCB:-)

Current sensor 30A – making PCB step 1
Traces printed on paper and laminate ready to operation.

Current sensor 30A – making PCB step 2
Toner transferred from paper to laminate.

Current sensor 30A – making PCB step 3
Etching pcb.

Laminate etched. It looks like it was a little too long in bath, but it will be tinned anyway, so no problem this time.

Current sensor 30A – PCB ready to solder parts

Soldering time:-) Because copper traces could be too thin to pass such high current, I added additional copper wires on its path.

Current sensor 30A – soldering parts and wires

Current sensor 30A – almost ready

Current sensor is also excellent place to branch power supply for other sensors, OSD and Video transmitter, so I added small JST connector.

Current sensor 30A – adding additional power connector for powering OSD and other stuff.

With heatshrink wrap cover sensor is ready.

Current sensor 30A ready to use:-)

Current sensor 30A – sensor output

Tested and it (of course 🙂 ) works.

Frequency meter up to 100MHz

Posted on Saturday August 18th, 2012 by majekSaturday March 17th, 2018

Short story: I needed frequency meter, so I made it 🙂

Why my own? I found few frequency meters over the internet, but most of them was based on PIC microcontroller (and I hadn’t experience with this type of UC), had range up to 40-50MHz and not so accurate. But making frequency meterer based on well known AVR was difficult because maximum frequency that this chip could count directly is only half of CPU frequency which is 20MHz maximum. So, I needed divider or prescaler. Looking for right prescaler was difficult, all good looking solution was unavailable for purchase or very expensive. At the end I looked into my stock of ‘nice to have’ parts and I surprisingly found 74HC393 chip which have maximum frequency of 100MHz! And from that point everything became easy:-)
From my junk I also digged small, cheap LCD display with serial (SPI) interface, made prototype, and surprisingly everything worked 🙂

Schematic:

Frequency meter – schematic diagram

Prototype:

Frequency meter – prototype

Design of PCB:

Frequency meter – PCB

Making PCB:

Frequency meter – PCB after termotransfer

Frequency meter – making PCB in high speed 🙂

Frequency meter – PCB almost ready

Everything assembled:

Frequency meter – assembled, bottom side

Frequency meter – assembled, top side

Because I had problems with analog input not working as expected I left this part without elements, so I have only TTL/CMOS input. As you can see below, it’s sufficient to me:-)

Testing:

Frequency meter – measuring RC controller frequency

Software:
It was quite easy to write first working version. But then appeared that even measuing itself, there was an error. After some checking I found a reason: delay between reading value from divider and internal counter, and clearing it. Error was exactly equal to number of clock cycles between these two operations. I added subroutine to correct this and it now works much more accurate. There is a possible hazard situation between reading and clearing counters as it couldn’t be atomic operation, so it requires also some fix, but maybe next time 🙂
Right now it just works for me:-)

And one important additional feature I almost forgot to mention: it outputs current frequency value also via serial interface 🙂

Eagle files and assembler source code you can download here.