Thread Tools
Jul 07, 2017, 07:44 AM
Radd Graduate '06
Mochaboy's Avatar
Quote:
Originally Posted by jjhdtv
All components except the LED's ordered for the Monkeyboard. Just noticed the specs on the LED's are not listed. Can someone let me know what LED's I should order?
Those boards look big enough to accomodate 5mm LED, but I'm a fan of 3mm LEDs for that approx form factor...gives you more options.

Here's a small sampling of LEDs from digikey
https://www.digikey.com/products/en/...=0&pageSize=25

3mm dome diffused, through hole, any color you like, but blue, red or green always work great as indicators...be careful about getting high millicandela leds...it's a fun idea until you stare directly at one and then see flashes of that led for the rest of the day every time you close your eyes.

or for 5 bucks just get an assortment

http://www.ebay.com/itm/150pcs-3mm-5...sAAOSwCypWpsoN
Sign up now
to remove ads between posts
Jul 07, 2017, 09:10 AM
Registered User
Hi all,
First of all, thanks a lot for your effort, Ben, for continuing development and keeping it free to use.
I've been reading a lot about ULRS in the last month and finally decided to buy a few RFM modules to use with some atmegas i have lying around. While thinking about the best place to put a vertical 30cm antenna i thought about the plane's tail. Would it be possible to put the RFM module attached to the tail with antenna wires directly soldered to it and have a cable all the way to the front, connecting with the atmega328?
I guess shielded cables, perhaps separating power and data, would be the best idea. If so, would it be better to solder the shield on the rfm side or on the atmega side?
Jul 07, 2017, 10:22 AM
Registered User
Thread OP
Quote:
Originally Posted by Wiky5
Hi all,
First of all, thanks a lot for your effort, Ben, for continuing development and keeping it free to use.
I've been reading a lot about ULRS in the last month and finally decided to buy a few RFM modules to use with some atmegas i have lying around. While thinking about the best place to put a vertical 30cm antenna i thought about the plane's tail. Would it be possible to put the RFM module attached to the tail with antenna wires directly soldered to it and have a cable all the way to the front, connecting with the atmega328?
I guess shielded cables, perhaps separating power and data, would be the best idea. If so, would it be better to solder the shield on the rfm side or on the atmega side?
Tail would be a good place for the antenna, I assume you'll use a dipole antenna ?

You can solder the wires directly to the RFM module, however the pads on the module itself are quite fragile, so maybe they'll get unsoldered in case of too much traction/torsion on the antenna wires. Using a short length of coax cable won't make any harm.

You can have some distance between the RFM module and the atmega, but remember they transport high speed digital data (400 kHz) so they can be sensitive to interferences. Use shielded or at least twisted cable. If using shielded cable, the shield should be connected to both rfm and atmega side. (You're maybe thinking about ground loops and the such, but honestly I've never seen anything like that with ULRS so I think that's overstated.)
Jul 07, 2017, 10:57 AM
Registered User

RFM23BP mod


I finally got around to modding a board. I started with a test board -- an old broken one I had.

If you are an expert with solder paste and hot air then you do not need my feedback or guidance! If like most people you are not then you will be pleased to know that this can be done with a very fine iron tip and very fine solder, if you have a steady hand. Even though I have hot air, I felt quite sure if I used it I would simply dislodge other components and do more harm.

I tried solder paste and decided, after I had trouble spreading it that it was not viable. You will see solder paste in the picture to the left. One other thing to note: I purchased solder paste from Banggood a while ago and left it in a drawer not realizing it has a shelf life and needs to be refrigerated. This why I had trouble spreading it. Also, the manufacturing date was 2014 and that means that even though I had this for a year, Banggood sold expired product! Quelle suprise.

You need very fine tweezers (the type used form working on cellphones), an illuminated magnifier (I used a $12 one I got from Bangood a while ago, a mini vice to hold the board, a needle point soldering iron (you really want a soldering station), and at least .5mm solder, if not thinner.

You can heat and flick off R2 with a soldering iron tip, quite easily. To solder R25, you really need to add a dab of solder to the pads, press with tweezers and get one leg soldered (even partially) and then solder the others, including the one used to pin it initially. My R25 is bigger than the OEM one and the pads did not align properly.

R2 is tricky. It is really tiny. Basically, I tinned it with fresh solder (by placing it on a part of the RF board with no copper), re-tinned the pads, and held in in place with a finger and eventually got one side soldered. Then I did the other. With some practice it can be done.

If you have a sacrificial board to practice with, use it.

I will try and do a pair of working Dolphins in the next few days.
Last edited by Marc Dornan; Jul 07, 2017 at 10:34 PM.
Jul 07, 2017, 12:38 PM
Radd Graduate '06
Mochaboy's Avatar
For my SMD rework - especially removing fine components...this stuff has paid off in spades:

http://www.ebay.com/itm/Fast-and-Qui...1%26rkt%3D1%26

so easy to use to too that anyone can do this.
Jul 07, 2017, 03:01 PM
Registered User
Quote:
Originally Posted by Mochaboy
For my SMD rework - especially removing fine components...this stuff has paid off in spades:

http://www.ebay.com/itm/Fast-and-Qui...1%26rkt%3D1%26

so easy to use to too that anyone can do this.
Never even knew this existed! I guess that is how you easily remove RFM23s? (Prob not needed or useful for resistors and such).
Jul 07, 2017, 03:24 PM
Radd Graduate '06
Mochaboy's Avatar
Quote:
Originally Posted by Marc Dornan
Never even knew this existed! I guess that is how you easily remove RFM23s? (Prob not needed or useful for resistors and such).
Marc,

You can use it for anything. It's a special formulation of solder that stays liquidous for 3-5 seconds depending on how much heat you set your iron to. In most cases - that's just enough time to circle an atmega 328p (for instance) then literally watch it slide off a board..it's pretty cool stuff.

This is not the best example of how it's used but it should give you an idea of how it works.

Header Removal (1 min 47 sec)


I just finished recording an RFM32bp repair video so you'll see it in action a little better there.
Jul 07, 2017, 07:18 PM
Registered User
Quote:
Originally Posted by Mochaboy
I just finished recording an RFM32bp repair video so you'll see it in action a little better there.
I would really like to see this video
Thank you for your posts!
Jul 07, 2017, 10:24 PM
Registered User

Dolphin Mod Pt 2


I realized that the second vreg on the board is actually a 5V (not 3.3V) one, so I added a separate 3.3V board that is fed from the LM317 output of 6V. This feeds the modded board at 3.3v as per Narpat's suggestions.

One nice consequence of this is that the RFM board is not powered when the USB is connected, and so you cannot fry your board while flashing without an antenna.

It works! I am getting a solid link. Brief bench testing has shown that neither the R25s or the 5.1k R2 (in my last post I incorrectly said R9) get hot at all -- this is better than it was. I still need more conclusive testing. When I mod my other boards I will pass serial data and keep a PPM link for a few hours.

So, regardless of whether you are using Open LRSng or ULRS you should do this mod. The extra R25 transistor is quite easy. The resistor is a bit trickier but also quite doable by anyone without superhuman skills.

I think that all the boards should be revised to also supply 3.3V as an option-- this could be done with another 3.3V reg on the Dolphin. It could be an optional component to allow 3.3V for the RFM for those that want to do the mod. An extra jumper could select 3.3V for the RFM. If Artemen is still reading maybe he will want to revise the Dolphin (and get rid of the linear LM317). This could probably safely produce 1.5-2W with a different resistor and bumping the LM317 up a little.

I think I will next mod 2 boards to output 1.5-2W. It is a little annoying to have that little resistor as a weak point or we could add an electronic potentiometer to the LM317 reference pin and have output power controlled by firmware with a trivial firmware mod. Or at very least a jumper. Maybe there is an optimum value that will allow 0.5-1.5W output and we could select output power by jumper.

I have not added the capacitors that Narpat recommended - I was waiting to see if this worked. I will order them and add them as well.

So while I really wish that I had not spent so much time on Bluetooth woes and now this, at least the end result is something much better. Only time will tell if it is truly reliable, but it should be.
Last edited by Marc Dornan; Jul 07, 2017 at 10:38 PM.
Jul 08, 2017, 08:15 AM
Radd Graduate '06
Mochaboy's Avatar
Quote:
Originally Posted by Marc Dornan
I realized that the second vreg on the board is actually a 5V (not 3.3V) one, so I added a separate 3.3V board that is fed from the LM317 output of 6V. This feeds the modded board at 3.3v as per Narpat's suggestions.

One nice consequence of this is that the RFM board is not powered when the USB is connected, and so you cannot fry your board while flashing without an antenna.

It works! I am getting a solid link. Brief bench testing has shown that neither the R25s or the 5.1k R2 (in my last post I incorrectly said R9) get hot at all -- this is better than it was. I still need more conclusive testing. When I mod my other boards I will pass serial data and keep a PPM link for a few hours.

So, regardless of whether you are using Open LRSng or ULRS you should do this mod. The extra R25 transistor is quite easy. The resistor is a bit trickier but also quite doable by anyone without superhuman skills.

I think that all the boards should be revised to also supply 3.3V as an option-- this could be done with another 3.3V reg on the Dolphin. It could be an optional component to allow 3.3V for the RFM for those that want to do the mod. An extra jumper could select 3.3V for the RFM. If Artemen is still reading maybe he will want to revise the Dolphin (and get rid of the linear LM317). This could probably safely produce 1.5-2W with a different resistor and bumping the LM317 up a little.

I think I will next mod 2 boards to output 1.5-2W. It is a little annoying to have that little resistor as a weak point or we could add an electronic potentiometer to the LM317 reference pin and have output power controlled by firmware with a trivial firmware mod. Or at very least a jumper. Maybe there is an optimum value that will allow 0.5-1.5W output and we could select output power by jumper.

I have not added the capacitors that Narpat recommended - I was waiting to see if this worked. I will order them and add them as well.

So while I really wish that I had not spent so much time on Bluetooth woes and now this, at least the end result is something much better. Only time will tell if it is truly reliable, but it should be.
Man I love what you did here...good lord so much to catch up on but getting there slowly...
Jul 08, 2017, 08:32 AM
Registered User
@Mochaboy -- all this is down to Narpat, the RF engineer on this thread. I think you will find these mods pretty easy to do actually, when you get to that point. Ideally, the boards designs would be tweaked a little with this new knowledge. But you will not have much trouble adapting the boards you have now.

My thoughts are to have a 1.5-2W on the Antenna Tracker Yagi and the slightly lower power on on the aircraft - since the Yagi is so sensitive.
Jul 08, 2017, 01:54 PM
Registered User
Quote:
Originally Posted by Marc Dornan
I realized that the second vreg on the board is actually a 5V (not 3.3V) one, so I added a separate 3.3V board that is fed from the LM317 output of 6V. This feeds the modded board at 3.3v as per Narpat's suggestions.

One nice consequence of this is that the RFM board is not powered when the USB is connected, and so you cannot fry your board while flashing without an antenna.

It works! I am getting a solid link. Brief bench testing has shown that neither the R25s or the 5.1k R2 (in my last post I incorrectly said R9) get hot at all -- this is better than it was. I still need more conclusive testing. When I mod my other boards I will pass serial data and keep a PPM link for a few hours.

So, regardless of whether you are using Open LRSng or ULRS you should do this mod. The extra R25 transistor is quite easy. The resistor is a bit trickier but also quite doable by anyone without superhuman skills.

I think that all the boards should be revised to also supply 3.3V as an option-- this could be done with another 3.3V reg on the Dolphin. It could be an optional component to allow 3.3V for the RFM for those that want to do the mod. An extra jumper could select 3.3V for the RFM. If Artemen is still reading maybe he will want to revise the Dolphin (and get rid of the linear LM317). This could probably safely produce 1.5-2W with a different resistor and bumping the LM317 up a little.

I think I will next mod 2 boards to output 1.5-2W. It is a little annoying to have that little resistor as a weak point or we could add an electronic potentiometer to the LM317 reference pin and have output power controlled by firmware with a trivial firmware mod. Or at very least a jumper. Maybe there is an optimum value that will allow 0.5-1.5W output and we could select output power by jumper.

I have not added the capacitors that Narpat recommended - I was waiting to see if this worked. I will order them and add them as well.

So while I really wish that I had not spent so much time on Bluetooth woes and now this, at least the end result is something much better. Only time will tell if it is truly reliable, but it should be.
This implementation is very important to our knowledge, thanks to Narpat and to you who put everything to work.
Jul 08, 2017, 02:17 PM
Registered User

Ultimate LRS Penguin Board Beta


Hi Everyone.
I am putting the new Ultimate LRS board which I mentioned last month I am working on. Well the design of the board went through usual changes mid way, like any other project, specially due to report of RFM23BP board heating issue with higher voltage and LM 317T LDO heating issue with 3S and higher Lipo supply. So below are salient and hated features of this board :

1. Onboard Switching BEC based on RT8293B 3A, 23V, 1.2MHz Synchronous Step-Down Converter. I chose this chip over other cheaper chips which operate at 100Khz . The Schematic values is set for a 6V out put and you can choose the output voltage by changing R7 and R8 values, most suitable for the desired power from RFM module you want.
http://www.richtek.com/assets/produc...DS8293B-03.pdf

2. Small Foot Print of under 1"x1.8" . The AVR ISP headers in the middle to save safe real estate. You can use Pogo Pins / normal 2.54 Pins headers pressed against the Pad to burn the Boot loader which is anyway one time affair. You can use the space later to put a heat sink to cool down the 5V LDO or the Switching Chip.

3. No USB, PWM output. Full fledged Board are already there and this is for people who would be using it strictly with FC like F3/F4 based MultiWii boards and Pixhawk/APM .

4, Most hated DF13 / Molex Picoblade Connectors used in current Pixhawk. Reason : the Crimped connectors are cheaply and widely available and fits best with the desired form factors. If you hate DF13, you could easily solder the wires on the board on the pad for the connectors.

5. I am using a 10uf Cap for Analog RSSI instead of 1uf in the original schematic by Ben. This is the value I have used to convert Digital RSSI from many FrSky Rx and worked well. If people feel this is a wrong change, I will revert it immediately.

I am attaching the Eagle Schematic, Brd, BOM in the Zip file as well as the render of the Board.

I would need the review of the Design for correctness or any other errata from the active people here, specially like Ben, Marc, Narpat etc.

It was a great opportunity to brush up my PCB skills as well as on some new technology.

Thanks in advance for all the eyeballs and reviews.
Last edited by ujjwaana; Jul 09, 2017 at 07:51 AM.
Jul 08, 2017, 03:57 PM
Registered User
Quote:
Originally Posted by ujjwaana
Hi Everyone.
I am putting the new Ultimate LRS board which I mentioned last month I am working on. Well the design of the board went through usual changes mid way, like any other project, specially due to report of RFM23BP board heating issue with higher voltage and LM 317T LDO heating issue with 3S and higher Lipo supply. So below are salient and hated features of this board :

1. Onboard Switching BEC based on RT8293B 3A, 23V, 1.2MHz Synchronous Step-Down Converter. I chose this chip over other cheaper chips which operate at 100Khz . The Schematic values is set for a 6V out put and you can choose the output voltage by changing R7 and R8 values, most suitable for the desired power from RFM module you want.
http://www.richtek.com/assets/produc...DS8293B-03.pdf

2. Small Foot Print of under 1"x1.8" . The AVR ISP headers in the middle to save safe real estate. You can use Pogo Pins / normal 2.54 Pins headers pressed against the Pad to burn the Boot loader which is anyway one time affair. You can use the space later to put a heat sink to cool down the 5V LDO or the Switching Chip.

3. No USB, PWM output. Full fledged Board are already there and this is for people who would be using it strictly with FC like F3/F4 based MultiWii boards and Pixhawk/APM .

4, Most hated DF13 / Molex Picoblade Connectors used in current Pixhawk. Reason : the Crimped connectors are cheaply and widely available and fits best with the desired form factors. If you hate DF13, you could easily solder the wires on the board on the pad for the connectors.

5. I am using a 10uf Cap for Analog RSSI instead of 1uf in the original schematic by Ben. This is the value I have used to convert Digital RSSI from many FrSky Rx and worked well. If people feel this is a wrong change, I will revert it immediately.

I am attaching the Eagle Schematic, Brd, BOM in the Zip file as well as the render of the Board.

I would need the review of the Design for correctness or any other errata from the active people here, specially like Ben, Marc, Narpat etc.

It was a great opportunity to brush up my PCB skills as well as on some new technology.

Thanks in advance for all the eyeballs and reviews.
Are you feeding the RFM board at 3.3v with your design? Do you have a pad to attach the jumper wire for RM the output stage higher voltage? In order to do Narpats mod you do need a 3.3v supply.

Sorry if you covered this in other posts.

Oh and while you are still hot why not do a version with pads for molex click mate connectors
Last edited by Marc Dornan; Jul 08, 2017 at 04:03 PM.
Jul 08, 2017, 04:35 PM
Registered User
Thread OP
Quote:
Originally Posted by ujjwaana
Hi Everyone.
I am putting the new Ultimate LRS board which I mentioned last month I am working on. Well the design of the board went through usual changes mid way, like any other project, specially due to report of RFM23BP board heating issue with higher voltage and LM 317T LDO heating issue with 3S and higher Lipo supply. So below are salient and hated features of this board :

1. Onboard Switching BEC based on RT8293B 3A, 23V, 1.2MHz Synchronous Step-Down Converter. I chose this chip over other cheaper chips which operate at 100Khz . The Schematic values is set for a 6V out put and you can choose the output voltage by changing R7 and R8 values, most suitable for the desired power from RFM module you want.
http://www.richtek.com/assets/produc...DS8293B-03.pdf

2. Small Foot Print of under 1"x1.8" . The AVR ISP headers in the middle to save safe real estate. You can use Pogo Pins / normal 2.54 Pins headers pressed against the Pad to burn the Boot loader which is anyway one time affair. You can use the space later to put a heat sink to cool down the 5V LDO or the Switching Chip.

3. No USB, PWM output. Full fledged Board are already there and this is for people who would be using it strictly with FC like F3/F4 based MultiWii boards and Pixhawk/APM .

4, Most hated DF13 / Molex Picoblade Connectors used in current Pixhawk. Reason : the Crimped connectors are cheaply and widely available and fits best with the desired form factors. If you hate DF13, you could easily solder the wires on the board on the pad for the connectors.

5. I am using a 10uf Cap for Analog RSSI instead of 1uf in the original schematic by Ben. This is the value I have used to convert Digital RSSI from many FrSky Rx and worked well. If people feel this is a wrong change, I will revert it immediately.

I am attaching the Eagle Schematic, Brd, BOM in the Zip file as well as the render of the Board.

I would need the review of the Design for correctness or any other errata from the active people here, specially like Ben, Marc, Narpat etc.

It was a great opportunity to brush up my PCB skills as well as on some new technology.

Thanks in advance for all the eyeballs and reviews.
Thanks for your work, I'm ready to add it on the site, let me know when you'll have the PCB order links.


Quick Reply
Message:

Thread Tools