Jump to content

Mark all as read

Photo

Converting old transmitters to 2.4GHz using 'hack' modules


16 replies to this topic

#1 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 17 November 2012 - 01:01 PM

I have seen a few questions on the forums about converting vintage Acoms transmitters to 2.4Ghz using the Corona/Assan/Frsky DIY 'hack' modules that are available now. It has become a very popular mod for old aircraft and heli transmitters, but I've not seen very many vintage 2-channel surface transmitters that have been converted. As I have a few different makes and models of old transmitters I thought I'd look into it to see if it's possible.

Firstly, a bit of technical background as to how transmitters work. The electronics in a transmitter basically breaks down into 2 sections - the encoder section (which converts the stick positions into a waveform) and the RF section (the 'transmitter' part that takes the output of the encoder and uses it to modulate a Radio Frequency signal). The idea behind these DIY 'hack' modules is to take the output waveform from the encoder section and replace the existing 27 or 40Mhz RF section with a modern 2.4Ghz digital RF transmitter with all the advantages this brings.

The DIY modules normally require 3 connections from the existing transmitter circuitry - positive and negative power (which are usually easy to find) and the 'PPM' signal (the output waveform from the encoder section). Depending on how the encoder circuit is built in each radio, this can be easy or tricky to find. The aircraft and heli modellers have a distinct advanteage here as their transmitters usually have either plug-in RF modules and/or trainer ports on them, and the PPM signal is output to one of the pins of these ports.

Another thing to look at when installing a 2.4Ghz DIY module is that it's a good idea to disable the existing RF section, either completely or have the ability to switch the power between the old and new RF sections if you want the ability to keep using your old receivers. Removing the crystal will stop it transmiitting, but the circuit will still use some power (and may overheat or burn out components, which can happen if you remove the aerial as well). So, in addiiton to the PPM signal, you also ideally need to find and interrupt the power feed to the old RF section.

Next, a bit about the structure of the PPM waveform.

#2 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 17 November 2012 - 01:06 PM

PPM (Pulse Position Modulation) Basics
The waveform of a standard RC PPM signal consists of a 20 millisecond frame (meaning the stick
positions are scanned at 50 times a second). Each channel produces a pulse between 1ms and 2ms
long in the waveform, the length directly relating to the stick position (centre stick would be 1.5ms). At the end of the waveform there is an extra pulse which needs to be at least 4ms long, which is used to synchronise the receiver to the transmitter. The 20ms frame can therefore accommodate up to 8 channels ((8 x 2ms) + 4ms = 20ms) - if there are less channels fitted then no pulse appears for the unused channels and the sync pulse is extended. The picture below shows the PPM signal for a 4 channel system on the top line, with the 4 lines below showing how the positon of each channel relates to the time between each pulse in the PPM frame. The 'Frame space' is the sync pulse.
Posted Image
Most transmitters generate an 'active high' PPM signal (i,e, like the picture above, the waveform is normally low and pulses high), but there are some which are 'active low' (normally high and pulses to low).

Next, a bit about the various types of PPM encoder circuits.

#3 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 17 November 2012 - 01:14 PM

PPM Encoder Circuits
There are a number of methods that have been used over the years to build the encoder section in transmitters generate the PPM signal:

Transistor-based circuits (used in old radios, and the cheaper more modern transmitters)
These use a pair of transistors to form a multivibrator oscillator (to give the 20ms 'frame' timing, which is the RC standard PPM waveform length) with an additonal transistor circuit to generate the pulse length for each channel. for anything above 2 channels the PCB becomes quite large, which is why vintage aircraft transmitters are so big.

4017 CMOS counter-based circuits (found in old high-end multi-channel transmitters)
These use the same type of dual-transistor oscillator to provide the timing, but replaced the individual transistor cirsuit for each channel with a circuit based around a 4017 CMOS logic counter chip, which could generate the pulse length for up to 8 channels. This uses a lot less components and PCB space than a >2 channel transistor-based circuit.

'Encoder on a chip' based circuits
A number of manufacturers developed chips that contained the entire encoder circuit and outputted the PPM waveform, such as the Signetics CD8081/Philips NE5044 (capable of up to 7 channels) and the Oki L9362 (capable of up to 4 channels). These became quite popular with radio manufacturers as they reduced the component count and assembly costs dramatically. The Signetics/Philips chip was slighty more expensive and so was mainly used for aircraft transmitters where the 7 channels would be fully used (though Futaba did use it in some 2-channel radios). The Oki chip was used in 2 and 4 channel transmitters at the cheaper end of the scale. Unfortunately both devices have long been obsolete, so the radio manufacturers had to either develop their own custom encoder chips or go back to transistor circuits for the cheap 2 channel transmitters.

Microcontroller-based circuits
When the encoder chips went obsolete, programmable 'system on a chip' microcontrollers (PIC chips, for example) started to become available at affordable prices. Rather than develop their own expensive custom chips, most of the radio manufacturers started to develop encoders based on these devices - they had the required analogue and digital inputs, digital outputs and timing citcuity buit in, and the functionality could be developed in firmware. As they became more powerful, LCD screens, multiple model memories, onboard mixing, programmable endpoints, digital trims and so on could be added by just modifying the firmware. The encoders in most modern medium and high-end transmitters are built around these devices.

When it comes to finding a PPM output signal to use with a 2.4Ghz DIY module without the presence of an RF module socket or trainer port, the order of difficulty of the 4 types of circuits above is (from easiest to hardest):
'Encoder on a chip' < Transistor < 4017 CMOS < Microcontroller

In reality, the last 3 are about the same difficulty, as they all rely on having the circuit diagrams of the transmitter available to study (or some educated guessmork and poking about with an oscilloscope). The transistor-based designs all use very similar circuits , the difficulty is translating the design into how it is laid out on the PCB. The 'Encoder on a chip' types are a lot easier to work with as they were commercial chips and the datasheets and pinouts for the chips are available to download.

Next, a survey of my radios to see what types of encoder circuits they have.

#4 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 17 November 2012 - 01:41 PM

I have a number of different transmitters, so as part of my investigation I opened them up to identify the encoder type and try to find the PPM output.to see how easy or hard they would be to convert:

1) Acoms AP-227 MkII
Posted Image
Probably the definitive vintage Tamiya radio (here in the UK at least). Given it's age and lack of features, it is an all transistor-based design as you would expect.

2) Acoms Techniplus AP-27 MkV
Posted Image
This classic chrome-plated unit from the early 90s. The circuit design is almost identical to the ancient MkII (basically the only difference is the addition of the servo reverse switches).

3) Acoms Techniplus Alpha AP-201
Posted Image
Acoms (or JR) finally updated their design - this is based on the Oki L9362 chip (this means that if you use it as the basis for a 2.4Ghz mod, you could also potentially add the 2 extra channels the chip is capable of).

4) Acoms Techniplus AP-202 (rebadged JR Propo Beat Sonic)
Posted Image
A backwards step. The Oki chip had gone obsolete by this time this was made, so JR/Acoms went back to the all-transistor design of the AP-27 MkV, but converted to surface-mount components and with an additional PCB to provide the LED battery meter.

5) Conrad Razor X4 (rebadged Acoms Techniplus 4 AP-401)
Posted Image
This is a more modern 4 channel radio, and the encoder is based on a Mitsubishi microcontroller. On closer inspection of the PCB, it had some unused switch connections, pads for a trainer socket to be fitted, and the silkscreen printing on the PCB indicates it was actually designed by JR. Armed with this information, I did a search on the US FCC database and managed to find the schematics for the JR Quattro transmitter. On comparing the schematic to the AP-401 PCB, it appears to be a cut down version of the JR radio. It uses the same version of the Mitsubishi microcontroller, but the PCB layout has been simplified to remove the tracking and pads for channels 5 and 6 that were present on the JR Quattro.

A bit more research showed that the radio was originally sold in the USA as the 'Expert 4' by Horizon Hobby (the US JR distributor). Expert was a high-end american radio manufacturer (similar to Fleet Systems in the UK) who folded and the name was bought by Horizon. It looks like they asked JR to design an entry-level aircraft radio (hence the presence of the trainer port) to be sold under the 'Expert' brand, and JR produced a cost-reduced version of the Quattro as a result. JR subsequently licenced this design to both Acoms and Conrad as a 40Mhz surface radio (minus the trainer port) to be rebadged under their names. The schematic for the 'Expert 4' is also available from the FCC database and matches this radio almost exactly (apart from the omitted trainer port components)

6) Sanwa Dash Saber
Posted Image
Dates from around the same time as the AP-202. With Sanwa being one of the better names, you would think they would have an 'encoder on a chip' or microcontroller-based design, but it's obviously been built down to a price and so it's another transistor based encoder. Very similar to the AP-202 design, but with a much more confusing PCB layout.

7) Graupner MC10
Posted Image
'Computer Radio' built by JR for Graupner's German/European market with all the modular features that market expects (though the MC10 is the bottom of the 'MC' range). As it's name implies, it's microcontroller-based, though thanks to it's modular design and being sold in both air and surface frequwncies there is a PCB connector for a trainer socket so the PPM signal is easily accessible.

8) Graupner C4-X (rebadged JR Propo Beat Gear)
Posted Image
Early 90s bottom of the range JR 2-channel unit (the predecessor to the Beat Sonic/Acoms AP202), rebadged for the European market by Graupner. Like the Sanwa Dash Saber, you would hope JR would be above using transistor-based encoders, but unfortunately not. The circuit is very similar to the Acoms AP-27 MkV (which was probably derived from this JR design).

After opening them up, studying the PCBs and poking about with the oscilloscope, I managed to find whereabout on the PCBs the PPM signal could be accessed. That will be in the next update.

#5 speedy_w_beans

speedy_w_beans

    Member

  • Members
  • PipPipPip
  • 1572 posts
  • Gender:Male
  • Location:North Carolina

Posted 17 November 2012 - 03:10 PM

Great thread, keep it up! The background information about the signal encoding is very nice to know.

-Paul

#6 PaPeRo

PaPeRo

    Member

  • Members
  • PipPipPip
  • 732 posts
  • Gender:Male

Posted 17 November 2012 - 08:52 PM

Very informative thread. I've converted my Futaba 2PL FM using the Corona DIY kit.

#7 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 17 November 2012 - 09:29 PM

Very informative thread. I've converted my Futaba 2PL FM using the Corona DIY kit.

Yes, the 2PL is very easily converted, having a separate PCB for the RF section so the PPM and power connections are easily accessible. Unfortunately Acoms were not so considerate ;)

#8 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 17 November 2012 - 10:19 PM

I'll start at the oldest radio, the AP227 Mk2. My particular radio is a bit battered (the orginal aerial, T-handle and battery meter have disappeared at some point in it's past) but it still works. Unlike more recent radios, it uses 6 AAs, meaning the supply voltage (Vcc) will be around 9V, depending on the type and state of the batteries. My example is fitted with a PCB that is marked 'A42 - 602 - 0' in the copper on upper right corner on the solder side (just below the throttle stick).

Any advice below is provided for educational purposes only - any damage, explosions, house fires, injuries or visits from the FCC/DTI as a result of modifying your radio are your responsibility! :)

Here's a picture of inside the radio, the left-hand side of the solder side of the PCB:

Posted Image

The PPM output appears on the emitter of transistor Q4 (the output transistor for the PPM section), and can be picked up at the solder points on the pad outlined in yellow. In this radio, the PPM signal is active high, and pulses up to Vcc (9V), which may be a problem depending on which DIY module you choose (some expect 'logic level' inputs of 3.3V or 5V for the PPM signal). The ground connection for the module can be picked up from the large round pad in the upper centre of the picture. To get power for the module, I would suggest sodering direct to the switched power (the bottom of the 3 rectangular pads for the switch just underneath the aerial).

There are 2 separate power feeds to the RF section in this design (one to the RF output amplifier and one ot the crystal oscillator). To disable the feed to the output amplifier, I would suggest cutting the track at the yellow line between the switched power pad and the large circular pad to it's left (use the back edge of a scalpel blade to scratch through the copper in a straight line repatedly until the PCB material can be seen). To disable the power for the crystal oscillator, heat the solder on the right-hand pad of resistor R5 (the side connected to Vcc) and use tweezers or pliers to lift the leg from the hole.

The above solution would leave the RF section permanently disabled. If you wish to have it switched between 27MHz and the 2.4GHz module, a SPDT (Single Pole Double Throw) switch could be used. The centre connection would be connected to the suggested power point (the powr switch) from the PCB. One of the outer connections would be back to the PCB (the other side of the cut) to power the 27MHz RF amplifier, and to the lifted leg of R5 to power the oscillator. The other connection would be the power for the 2.4GHz module.

The only other observation is that it may be a challenge to fit a 2.4GHz module in the case if you want the option of keeping the 27MHz RF section operational. This radio is a bit smaller than later ones (mostly due to using 2 less AAs) and the only place to fit a DIY module would be the back half of the case in the area behind the stick mechanisms, or if the existing aerial is removed the space between the sticks could be used.

#9 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 18 November 2012 - 12:57 PM

Next in chronological order, it's the AP27 MkV Techniplus. This design has moved on to using 8 AAs, meaning the supply voltage (Vcc) will be around 10-12V, depending on the batteries. My example is fitted with a PCB that is marked '1771-500' in the ground plane copper on the right-hand edge of the solder side, and also sikscreened on the component side above the crystal socket.

Again, any advice below is provided for educational purposes only.

Here's a picture of inside the radio, the left-hand side of the solder side of the PCB:
Posted Image

Again, the PPM output appears on the emitter of transistor Q4 (the circuit is so similar to the Mk2 that most of the component names are the same), and can be picked up at the solder points on the pad outlined in blue. As wih the Mk2, the PPM signal is active high, and pulses up to Vcc (12V in this case), again check this is suitable for the module you choose. The ground connection for the module can be picked up from anywhere around the large ground plane that occupies all the unused space arond the board (or from the negative battery terminal). To get power for the module, I would suggest soldering direct to the switched power - the top pair of contacts on the switch, or the square pad on the PCB that the red wire from the switch is connected to just below the aerial bracket (hidden under the brown/yellow pair of wires in this picture).

As in the Mk2, there are 2 separate power feeds to the RF section. To disable the feed to the crystal ocsillator, I would suggest cutting the track at the yellow line on the Vcc track. To disable the power for the RF amplifier, heat the solder on the right-hand pad of inductor L5 (the side connected to Vcc) and use tweezers or pliers to lift the leg from the hole.

This would disable the RF section permanently. If you wish to switch between 27MHz and 2.4GHz, use a SPDT switch. The centre connection would be connected to one of the solder points on the 'Vcc' pad in the picture above.. One of the outer connections would be back to the PCB (one of the 2 solder points on the other side of the cut) to power the crystal oscillator, and to the lifted leg of L5 to power the RF amplifier. The other connection on the switch would be the power line for the 2.4GHz module.

Again, there is not a lot of room inside the case - although the 2 extra cells means it's wider than the Mk2, it's not as deep and the stick mechanisms reach almost to the back of the case. If you want to keep the 27Mhz output usable, the only space I can see is to the upper left of the steering stick. If you just want to convert to 2.4GHz, the space between the sticks where the aerial was becomes available.

#10 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 18 November 2012 - 05:54 PM

Now onto the AP201 Techniplus Alpha. This transmitter is built around the Oki L9362 'Encoder on a chip', and is totally different to any of the other Acoms designs (which makes me wonder if they either copied or licenced a circuit design from Futaba or JR). It runs on the now standard 8AAs, meaning Vcc will be around 12V. Another difference in this design is that it uses 2 PCBs. The main encoder/transmitter board and a daughter board that contains the LED voltmeter and power switch. The battery and charge port connections are carried from the main PCB to the daughter board via a 4-way ribbon cable, with the Vcc supply for the main board being returned through one of the wires in the ribbon cable.

As usual, any advice below is provided for educational purposes only.

Here's a picture of inside the radio, the right-hand side of the solder side of the PCB:
Posted Image

Now, this is very different from the previous two PCBs, and is a lot easier to convert. The L9362 chip handles the encoding - it is a 14-pin device, and the PPM signal is output on pin 9. This can be picked up on the small PCB pad outlined in blue connected to that pin.

Power to the RF section is taken from the Vcc track (the inverted 'T' shaped track coming from the ribbon cable connection) via a wire jumper, J2, on the component side of the PCB. Disconnecting this jumper should be enough to leave the RF section unpowered.

If you wish to switch between 27MHz and 2.4GHz, you could wire the two pads at the ends of jumper J2 to a switch, which could then be used to switch Vcc between the 27MHz and 2.4GHz RF sections.

Again, there is not a lot of spare room inside the case - it uses the same deep stick mechanisms as the AP27 MkV, so pretty much the only place to mount a DIY module would be in the space between the sticks occupied by the existing aerial.

#11 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 18 November 2012 - 11:08 PM

Last of my Acoms 2-channel sets is the AP202 Techniplus. Back to an all-transistor design, but now most of the components have been changed for surface-mount equivalents (except for the larger capacitors and inductors) which are mounted on the solder side of the PCB. The PCB is marked '3511-500-1-91' on the component side, which would suggest it was the next PCB to be designed after the AP201.

Another interesting detail is the PCB is also marked 'JR CH7D'. This would indicate it is actually a design licenced or made by JR, like the AP401 Techniplus 4 and the Technisport Hayabusa (which was a rebadged JR Cobra). A bit more research shows that this exact model was also sold as the JR Propo Beat Sonic in Japan (http://www.jrpropo.c...69&db_flg=jp_dl), which was the replacement for the Beat Gear at the bottom of the JR range. It also makes me wonder if the Casio takeover had reduced Asahi/Acoms to an offshore manufacturing and rebadging/reboxing company with no design input.

Like the AP201, there is a small daughter board for the LED meter. However, in this design it is just the for the meter, the switch and power control circuitry are now back on the main PCB.

As usual, any advice below is provided for educational purposes only.

A picture of the solder side of the PCB:
Posted Image

Although the change to surface-mount components makes it look very different, the circuit is very similar to the AP27 MkV (though it in now a bit more tricky to modify). The PPM output appears on the emitter of the transistor outlined in blue (there are no component names now), and can be picked up at the solder points on the pad outlined in yellow. This pad is now quite small, so rather than soldering to one of the tiny SMD solder pads I'd suggest carefully scraping the lacquer off in the middle of the pad to get to the copper, and soldering to that. As wih the AP27 MkV, the PPM signal is active high and pulses up to Vcc (12V),

The ground connection for the module can be picked up from anywhere around the large ground plane (or from the solder point marked 'BK' that has the black wire attached). To get power for the module, I would suggest soldering direct to the switched power - either the leftmost rectangular pad for the switch, or the solder pad further up the track where the red wire is attached.

As in the previous transistor circuits, there are 2 separate power feeds to the RF section. To disable the feed to the crystal oscillator, I would suggest carefully removing the zero-ohm resistor (outlined in red, and marked with the value '000' - zero-ohm resistors are used in surface-mount circuits as equivalents of a wire jumper connection). To disable the power for the RF amplifier, carefully unsolder the leg of inductor L4 from the pad where it connects to the switched power track.

This disables the RF section permanently. If you wish to switch between 27MHz and 2.4GHz, use a SPDT switch. The centre connection would be connected to one of the solder points on the 'Vcc' pad in the picture. One of the outer connections would be back to the PCB to the tiny solder pad on the 'crystal' side of the removed zero-ohm resistor to power the crystal oscillator. and to the lifted leg of L4 to power the RF amplifier. The other connection on the switch would be the power line for the 2.4GHz module.

There is a bit more room inside this case, as it's a bit deeper. There should be just enough room to fit a DIY module in the centre of the rear half of the case behing the existing aerial if you want to keep the 27MHz RF section functional.

#12 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 19 November 2012 - 08:26 AM

After a bit more research, it turns out there is a relationship between JR and Acoms/Asahi Corp/Casio Creative Products. JR's lower-end radios are built in Asahi's factories in Malaysia and Thailand. It looks like as part of this deal Asahi were allowed to rebox/rebadge these JR designs and sell them under their Acoms brand (which would explain the similarities in the circuit designs and 'JR' markings appearing on the PCBs of the later models).

My guess is that that Asahi were only allowed to export the Acoms-branded JR designs to countries where the JR importer does not take the lower-end gear. This would explain Acoms' ubiquity in the UK market, as MacGregor (JR's UK importer) only deal in the higher-end JR sets. Germany is an interesting situation, as Graupner import and rebadge JR gear (including the low-end 'Beat' 2-channel sets, which are rebadged as Graupner C4/C4X), and Carson import Acoms gear, so you end up with the same designs being sold by 2 competing companies.

Horizon are JR's partner for US market (JR also design and build the Spektrum sets for them too), and also they import the lower-end pistol-grip JR sets (and rebadge some of the JR-designed Acoms gear under thier own brands). It looks like this restricts Acoms from having much of a presence in the US.

I wonder how the relationship works now, after Casio sold Asahi Corp/CCP to Bandai in 2006?

#13 waterbok

waterbok

    Member

  • Members
  • PipPipPip
  • 761 posts
  • Gender:Male
  • Location:Dutch living in Frankfurt, Germany

Posted 19 November 2012 - 08:43 PM

this is a very clear and interresting read, as you seem to know your way around Tx and Rx, could you consider doing an explanation thread on antenna lengths? both old style AM and FM and 2.4 especialy the receiver wires, who always seem to get caught and ripped out. :wacko:
58035swb 58035lwb 58039 58048 50087 58154 58161 58192 58204
58242 58344 58405 58445 58452 57602 57604 57606 57743 ax90014

#14 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 19 November 2012 - 09:20 PM

While I'm looking at the connection between Acoms and JR, I thought I'd document the Graupner C4-X next, which is a rebadged version of the JR Propo Beat Gear.

It dates from around the same era as the AP27 MkV Techniplus, and once you open the case you find that they have a lot of things in common - the same stick mechanisms, and the PCBs are virtually identical, with just some minor component position changes. The PCB is numbered '1771-503-0-91', which fits into the same numbering scheme as the other JR/Acoms designs (the Techniplus MkV is '1771-500'). However, instead of the mechanical voltmeter, it has a simple 3-LED meter.

Again, any advice below is provided for educational purposes only.

Here's a picture of the PCB:
Posted Image

Compare it with the MkV PCB earlier in the thread - it's virtually identical. Just as in the MkV, the PPM output appears on the emitter of transistor Q4 (the component names are the same on both PCBs), and can be picked up at the solder points on the pad outlined in blue. As wih the MkV, the PPM signal is active high, and pulses up to Vcc (12V). The ground connection for the module can be picked up from anywhere on the ground plane, or from the negative battery terminal. To get power for the module, I would suggest soldering direct to the switched power - the top pair of contacts on the switch, or the large square solder point just under the aerial bracket that the pinkish-red wire from the switch is connected to.

Just like the MkV, there are 2 separate power feeds to the RF section. To disable the feed to the crystal ocsillator, I would suggest cutting the track at the yellow line on the Vcc track. To disable the power for the RF amplifier, heat the solder on the right-hand pad of inductor L5 (the side connected to Vcc) and use tweezers or pliers to lift the leg from the hole.

If you wish to switch between 27MHz and 2.4GHz, use a SPDT switch. The centre connection would be connected to one of the solder points on the 'Vcc' pad in the picture above.. One of the outer connections would be back to the PCB (one of the 2 solder points on the other side of the cut) to power the crystal oscillator, and to the lifted leg of L5 to power the RF amplifier. The other connection on the switch would be the power line for the 2.4GHz module.

Although the case is a bit chunkier than the MkV's, there is only the same amount of room inside (the extra depth is due to a thicker bezel around the sticks). If you want to keep the 27Mhz output usable, the only space I can see is to the upper left of the steering stick. If you just want to convert to 2.4GHz, the space between the sticks where the aerial was becomes available.

#15 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 19 November 2012 - 10:26 PM

this is a very clear and interresting read, as you seem to know your way around Tx and Rx, could you consider doing an explanation thread on antenna lengths? both old style AM and FM and 2.4 especialy the receiver wires, who always seem to get caught and ripped out. :wacko:


Thanks for the comment. RX antenna wire lengths is a bit of a minefield. Antenna theory would have it that for best reception, an antenna length based on a fraction of the wavelength (1/2, 1/4 or 1/8) would give best results, with quarter-wave being the optimum. However, for 27MHz a quarter-wave antenna would be approximately 2.8 metres long, and for 40Mhz it would be 1.9 metres long (as the frequency increases, the wavelength decreases). These antennas would be too long for fransmitters, never mind receivers (though a transmitter could get away with an eighth-wave antenna at half those lengths with reduced performance).

However, to get around the need for a long antenna you can design the transmitter and receiver circuits to have an 'impedance matching' circuit in line with the antenna. This allows you to tune the circuit to give good perfomance with a much shorter antenna (especially important for receivers). The ideal antenna length for the tuned circuit is decided at the design stage by the manufacturer, but from a quick survey of my 27MHz receivers it seems around 50cm seems to be a common length (the only 40MHz gear I have is JR Graupner, and they seem to tune their receivers to use much longer antennas).

If your antenna gets damaged, it's probably safest to either get a replacement wire from the manufacturer/distributor (they are usually available as spare parts) or find someone with the same model and ask them to measure theirs so you can replicate it (I usually measure the antenna of any new type of RX I get before it's installed and make a note of it in the manual, just in case). If you have no idea, and having the optimum range or signal to noise ratio isn't too critical, then 50cm would seem to be as good a guess as any for a 27MHz receiver antenna.
Out of interest, this page is a good practical study of changing the length of the antenna wire and the effect it has on signal strength:
http://www.rc-cam.com/ant_exp.htm

The antennas for 2.4GHz systems are much more sensitive to length changes. as the higher frequencies mean the antennas will be much shorter and so exact length becomes more critical. A quarter-wave antenna for 2.4GHz comes out as about 3.2cm, which is about the same length as the little stub of wire you get on most 2.4GHz surface or park flyer receivers.

Multi-channel 2.4GHz receivers designed for large aircraft and helicopters usualy come with 2 antennas, made from short lengths of coax with a length of bare wire at the end. The 2 antennas form a dipole, which provides better directional reception as the model moves in 3 dimensions relative to the transmitter (this is known as 'antenna diversity'). The length of bare wire at the end of the coax is actually the antenna, which again is usually about 3.2cm long.

#16 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 20 November 2012 - 09:17 PM

Before I go on with any more radio details, I thought it might be a good idea to discuss some of the various 2.4GHz DIY modules that are available.

Jeti (from the Czech Republic) and Xtreme Power Systems (XPS) were amongst the first manufacturers to offer a 'DIY' 2.4GHz upgrade module that could be used on any transmitter (Jeti even designed installation kits for various popular transmitters). XPS got a lot of bad publicity early on as the 'frequency hopping' algorithm of their system did not work very well with interference and the receivers would lock up (a big problem, as aircraft modellers were their target market). Both systems were very expensive, which was to be expected in the early days of the technology (but they could also could work with the most complex high-end transmitters, such as those that used faster scanning and longer PPM frames to allow up to 16 channels to be used).

Once radio manufacturers started bringing out plug-in 2.4GHz RF modules for their transmitters (and some started offering them for other manufacturer's transmitters), people started looking at them as a cheaper way of modifying existing radios that either didn't have module sockets or did not have a 2.4GHz option available. Once this started becoming a worthwhile market, various far-eastern manufacturers started making their own 2.4GHz systems, As of now, there are numerous manufacturers in the 'DIY Module' area, such as Assan, Corona, Flysky, FrSky, Flydream and WFly.

When it came to choosing which system I wanted to go for, I read a lot of reviews and opinions of people on forums. Originally I was going to go with the Corona V2 system, but in the end I settled on the FrSky system. My main reasons for choosing FrSky were
1) They are fully CE and FCC approved.
2) They have a UK distributor (GiantShark) and are readily-available.
3) They have implemented a full-featured frequency hopping algorithm with a fast 'lost signal' recovery time and their receivers have a proper failsafe function that can be programmed for each channel.
4) There is a wide user base for surface models (a lot of the members on Oople have converted their old high-end car radios with FrSky modules and 4-channel receivers)
5) FrSky maintain a technical presence on forums like RCGroups and Giantshark's user forums and seem to be very receptive to customer suggestions
6) Although they are not the cheapest, they are still competitively priced (currently £13 for the non-telemetry TX module, and £13 for the 4-channel receiver)
7) The PPM input of the module is tolerant of a wide range of input voltages (I have tested it with PPM 'high' pulse levels of 0.6V up to 12V) and will accept both positive and negative-pulsing PPM signals. Some other modules will only accept positive-pulsing PPM signals of at least logic-level (3.3V), and so require additional circuitry to invert negative-pulsing PPM signals or boost lower-voltage PPM signals.

The transmitter module I ended up purchasing was the V8HT DIY module (this is the non-telemetry version) and a couple of receivers - the V8R4-II 4 channel unit, and the V8R7-II. These 'Version II' receivers are compatible with both the non-telemetry and telemetry versions of the TX modules, but do not send telemetry back (the 'D'-series receivers are the ones with telemetry capability). This could be important if you intend to use it for racing, as telemetry may be against the rules (it is for BRCA-sanctioned events in the UK), and the presence of a telemetry-capable receiver may cause trouble with the scruitineers.

One other thing to remember about 2.4GHz systems is that unlike 27MHz and 40MHz systems, by and large no two manufacturer's products are compatible with each other, though there are a a couple of exceptions. Some JR and Spektrum receivers and transmitters will work with each other (not surprising, as JR make both) and there are now some third-party JR/Spektrum and Futaba compatible receivers on the market. For all other systems, you have to buy transmitters and receivers from the same manufacturer. The reason for this is because each manfacturer has developed proprietary spread-spectrum implementations and frequency-hopping algorithms, and don't want to give this information away to their competitors. However, if the user base is large enough (as is the case for Spektrum and Futaba) then the other companies will find it worth the effort to reverse engineer the big name's systems, so they can get a slice of their receiver market.

Another important thing to remember is that hardly any 2.4GHz receivers have BEC circuitry built into them. Some manufacturers make receivers that will work with supply voltages up to 13V (these are usually marked 'HV', or something similar), but the regular receivers will only take a supply voltage of 4.8V-6V. You either need an ESC that has a built-in BEC or an external BEC module between the ESC and receiver, even for 'HV' receivers (otherwise they would end up supplying up to 13V out to the servos, which wouldn't be very good for them).

#17 fastfordrc

fastfordrc

    Member

  • Members
  • PipPipPip
  • 173 posts
  • Location:1274

Posted 24 November 2012 - 06:08 PM

Well, my package has arrived from GiantShark, so it's time to stop waffling on about theory and actually do a conversion.

The contents of the FrSky V8HT DIY Module package looks like this:
Posted Image

There are 3 sets of leads coming out of the module - the aerial coax cable (with the gold-plated fitting on the end), the Bind button/Status LED lead (with the small PCB on the end) and the power and PPM input. This is the important one, and of you look closely one of the 3 wires in this lead has a heat-shrinked section at the end. This is the PPM input, and the heatshrink is covering a zener diode which has been soldered onto the end of the lead. This was an important modification suggested by the users on the RCGroups and GiantShark forums that protects the module's circuitry from being damaged by the PPM output generated by certain designs of transmitters.

To install the module, you need to find somewhere to put the module, aerial, Bind/Status PCB and then solder up the 3 connections. The transmitter I chose was the Techniplus MkV, as it's my favourite one to use, and in my opinion probably the nicest-looking of the Acoms units I have.

Firstly i looked for a location for the module. In my previous post about the MkV I said there wasn't much room in the case, and I found this to be true. A few temporary fittings using sticky tape showed that the best place to fit it where it didn't interfere with the stick mechanisms and gave easy cable routing was on the far right of the rear case, snugged up against the top of the battery compartment:
Posted Image

Notice the heatshrink on the yellow wire - I have moved the zener diode further up the PPM input wire. I needed the connections to the PCB to be flexible, which wouldn't have been the case with the diode at the end.

Next, I looked for a position for the aerial. As I wasn't keeping the 27MHz RF section functional, I decided to use the existing position, but this meant the case needed modification. The aerial support boss on top of the case is moulded as part of the front half of the case, but there is a thin section under the base which is split in half between the front and back of the case. I cut the boss off flush with the top of the case, then added a half-ring of ABS underneath to reinforce it and give a level surface for the aerial fitting to be bolted up to (this required the thin section on the rear half of the case cutting away to fit as well):
Posted Image

The next thing was to find a position for the bind/status PCB. Some people mount these on the back of the transmitter case, but I wanted the LED somewhere where it was visible in use. As I wasn't keeping the 27MHz RF section functional, that meant the hole for the crystal socket was available. Again, a few trial fittings showed that this needed a bit of modifying. If the bind PCB was sandwiched between the hole and the radio PCB, it meant the radio PCB sat about 2mm too high and the case wouldn't fit back together, and the hole was too deep to use the bind button. I ended up cutting 2mm off the back of the hole for the crystal, which meant the bind button would end up flush with the front:
Posted Image
The reinforcing webs were also trimmed to make room for the cable and to let the PCB sit squarely.

The original PCB now fitted back in place, but when the button was pressed the bind/status PCB would push inwards as well. The solution to this was to put a blob of hot glue on the back of the PCB. One it cooled, it meant the crystal socket on the original PCB was pressing against the back of the bind/status PCB, holding it firmly in place:
Posted Image

The 4-way cable fits through the gap between the original PCB and the bottom on the case, then folds upwards and runs in the gap between the solder side of the PCB and the battery box in the rear half of the case. it can be seen at the top of this picture:
Posted Image

The power and PPM input cable can be seen above, running diagonally down to the PCB. This also folds at the bottom when the case is reassembled, which is why I didn't want to leave the zener diode at the end. The connections and cut tracks can be seen in more detail below:
Posted Image

The black wire is the Ground, which was soldered to a convenient pad on the ground plane. The red wire is the power supply, and is soldered to the track that supplied power to the existing RF section. The yellow wire is the PPM input, which is soldered to the output of transistor Q4, as identified in my previous post. The 2 cuts isolate power from the existing RF section. In my original post on this radio I suggested 1 cut, and disconnecting one of the legs of inductor L4. The upper cut in the picture above does the same job as lifting the leg of L4, as I was not keeping the original RF section switchable.

The aerial connector was then bolted in place and the 2 halves of the case were carefully put back together, taking care not to snag any of the new wiring. The transmitter was then powered up, and the flickering LED indicated that a valid PPM signal was being received by the module. I then connected up the 4-channel receiver to a battery pack and servo, and went through the bind procedure (very simple - power on the TX with the 'bind' button pressed, them power on the RX with the 'bind' button pressed, though you have to be careful not to have them too close together while doing so). Everything is now working OK:
Posted Image

I have another TX module to use, which is for my Graupner MC10. That should be easier, as it has a trainer port connection on the PCB so the power and PPM signals are easily accessible.



Reply to this topic



  


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users