1. Change the large blue, (2E225) from 2.2uF/250V to 1uf/400V. 
2. Change the smaller blue, (2E224) located close to the output connector, from 0.22uF/250V to 0.1uF/400V. 
3. Check that the 330k resistor, close to C26, the big capacitor, is 1/2W. It should be physically bigger than all other standard resistors. If not change it to 330k 1/2W.
4. Change the 220K 1/2W (in some models 1/4W) close to the output connector to 470k 1/2W.
   Important note: Recently, Paolo Masetti, found that this resistor replacement causes the appliance module to stutter under some conditions (reversed polarity of input power), so you may want to skip this step.
5. Change the 130V MOV in parallel with the input to 250V RMS. If you have 230 or 240V use 275V RMS.
6. If there is a resistor of 330k 1/4W going to the output connector, change this one to 470k 1/2W.

The built in appliance module is using a cam relay with a 110V coil. See appliance module modification for the coil.

Increasing the range

Open the unit and notice that the antenna is disconnected from the input circuitry.  There are 3 ways to increase the range, each progressively better.

• Solder a 22pF capacitor between the antenna base and the small brass plate inside the module.
• Solder  a 22pf cap in series with a 12uH inductor.
• Use a variable cap around that value and adjust until a max range is reached.

Warning: the brass plate is at high voltage, so make absolutely sure there is no galvanic (metal) short between it and the antenna. Else, the antenna will be at high voltage with all associated risk. The cap, if installed properly, does not pass the high voltage. Engineer this change properly, and of course, at your own risk.

A detailed description: from asimth@hereintown.net

Use a 22pF, 1kV capacitor. Start the modification by drilling a small hole in the antenna pocket until you get into the metal base. Cut the capacitor lead to fit, and put a bead of solder on the end. Stick it down in the hole and heated the lead to melt it to the plastic and to melt the solder to the metal. Clip the other lead and soldered it to the brass plate. Make sure to isolate the leads so they don't short. If you're really worried about safety, get caps that have UL Y1 spec (they fail in open state). The problem is that the smallest are 100pf, to the best of my knowledge. You can connect 3 or 4 of them in series to get to the 22pf ballpark.

Other factors affecting range:

1. Antenna length: set as close to 30cm as possible (use a ruler)
2. There is a report that height of antenna affects range: Peter Lawton writes:

I have noticed that placement of the transceiver greatly influenced the reception of my TM751.  Our house is a single story ranch, with a concrete slab.  When placed near the computers (3 of
them) and 1 or 2 feet off the floor, the transceiver had a marginal range again, around 50 ft.  Using an extension cord, I placed it near the 8 ft ceiling, on top of a bookcase.  The transceiver had a range of 172 ft.  I took both of these measurements after I made the TM751 mod.

I also moved the transceiver to a wall outlet, low and away from the computers, but in the same room.  It may have gotten a little better than the 50 ft. initial range.

You can also increase the range through optimized alignment of the module

Schematics
Credits: Mitch Orysh, personal correspondence

Finally, a great job detailing the RR501 schematics. PDF format (You may need to download acrobat reader). Notice of educational information: The information included herein is for educational purposes only. Oryshnet is not responsible for any use of this information.

RR501_Schematic vH.pdf

Modification to control All House Codes
Credits: Mitch Orysh, personal correspondence

And here is an advanced modification to the RR501 by Mitch Orysh that turns the RR501 to an All House Codes (AHC) transceiver. (You may need to download acrobat reader). The excellent description of this modification spans several pdf files:

• AHC Construction Instructions (long)
• AHC Theory of Operation brief
• AHC schematic

Also available are the JEDEC files for the programmable devices used in the AHC.  Folks who might want to build this design only need to have access to a device programmer and won’t need to recompile the source code.

JEDEC files are the files that can be downloaded to the device programmer to make each device for the AHC.  Most folks who know how to use a device programmer will know how to download these files.  They’re ASCII files that can be read by any editor.  Be aware that there are imbedded control characters in JEDEC files which must remain in the file to ensure it can be downloaded.

Files are:
“X10DEC.JED”: The AHC decoder PLD.  Goes with the source “X10DEC.PLD”
“X10MD.JED”:  The AHC Manchester Decoder PLD.  Goes with the source “X10MD.PLD”
“X10DEC.JED”: The AHC decoder PLD.  Goes with the source “X10DEC.PLD”
All are zipped in this file

In addition to the description, below are the source codes for the PALs:

• Common source file used for all PALs
• AHC 7-bit to 4-bit decoder source code
• AHC Manchester Decoder source code
• AHC Serial to Parallel Converter source code
  

The Quality of the RF stage

Comments by ND Lindsey, newsgroup post

Having held a Frequency Counter a Scope a Spectrum Analyzer and several RF Field Strength Meters next to many X10 transmitters of various flavors and varieties I found that the RF operating frequency on X10 devices is all over the band but in actuality is much closer to 290mhz than to the assigned freq. of 310mhz. These X10 transmitters are beyond the term "unstable" and I have to wonder how in h*ll the RF side of X10 works at all. We built better RF stuff than this in Jr. High School Electronics Classes. In the industry I've worked in for many years we were required to main a 100mhz RF signal within a tolerance of +/- 5 hertz.

Our long running argument regarding whether the transceiver antenna modifications actually improved RF performance has now become moot in my opinion after looking closer at the X10 RF circuitry. Virtually anything you do is bound to make some improvement because it certainly can't make it any worse.

Recognizing that the X10 RF designs and manufactured product were pretty poor, I once had hopes of tuning the transmitters and transceivers to one another just within my own home to improve performance. What I found when I started taking real world measurements on X10 devices is that there is simple nothing there to work with.

The RF side of X10 is pure crap and not worth the time it would take to try and tweak it. The 16 or 17 units I've actually measured myself would not hold to any given frequency within +/- 10mhz at room temperature. It was so bad that it actually became a game to chase it across the bands and try to stay up with it as it while viewing it with a spectrum analyzer. Fortunately, it is of such low power that it doesn't seem to interact with other RF devices around the house.

Partial Description of the RR501 PIC 

(from John Payson)

Having examined the RR501 circuit, I find it odd that loading would affect
its behavior as you describe.  Although there seems to be some crosstalk
between the output-side "power-on" sensing and the X10 100KHz carrier pickup
such that receipt of X10 commands will cause glitches in the output-side
"power-on" input, I'm a bit puzzled as to why the firmware would care about
that.

BTW, in case anyone's curious:

PORTA:

    0=AC phase;
    1=X10 transmit;
    2=Relay out(*);
    3=RF input;
    4=X10 input(**)
PORTB:
      0-3=codewheen;
    4=button (pressed=low);
    5=line output sense;
    6=unused (grounded);
    7=1-8/9-15 switch (closed=1-8=low).

(*) To flip the relay, output a rising edge just after PORTA.0 goes high; this
    pin should be low at other times.  Delaying the output may make the relay
    quieter, but may affect reliability.

(**) To use this, you'll have to use the RTCC counter mode.  A bit of a pain
     since you won't have a timer any more, but since you only have to do that
     for a small part of the cycle it's easy to work around.

Differences between RR501 and TM751

Bringing a transceiver back from the dead 

Here is an account of restoring functionality to a presumed dead RR501 and TM751 modules. The same principles can be applied to resurrect other modules

source: Bala Chandar, personal correspondence

TM751

I found the output of the 78L05 IC to be be around 2 volts, whereas it should be 5 volts. But even when I soldered a new 78L05, the output was still around 2 volts.

Then I suspected some short in the other components receiving the supply from this IC.

I removed the LM358 IC (in the same small board) that amplifies the signal from the remote and feeds it to the microcontroller and tested the resistance between the supply pins. It was an unusual 50 ohms. When I replaced this IC, the module started working!

 The overheating of two resistors (220 ohm & 47 ohm, both 1 W)which were in close proximity to the 78L05 might have caused the problem in the first place.

RR501

Now that the TM751 had started working, I compared the voltages of the different pins of the microcontroller of the two modules. I found that the pin 4 (_MCLR) of the MUC in RR501 was low whereas in the TM751 it was high. Since that is the Reset pin, I understood that the MUC was not working with the Reset pin low. There were two NPN transistors (C 9014) connected to the pin 4. Desoldering both and testing them revealed that one transistor was not working. This is the component had got damaged with the spark coming from the 100 ohm resistor earlier. I replaced this transistor and the module started working!

Another account of fixing dead transceiver module

Credits: )) Sonic ((
http://siber-sonic.com/X10/X10world.html

There are any number of possible failure modes, so what i have listed below may easily NOT be what is wrong with your particular unit. I provide the information below here on this web page because i believe this to be one of the more common failures of the RR501 Transceiver.

If one or more of the RR501 functions is inoperative and pressing the On/Off button produces a loud staccato stuttering sound:

1. Unplug the transceiver and remove the two screws holding the cover together. Remove the PCB from the covers.
2. Inspect the unit for visibly damaged parts. In the failure i found, the PCB will show signs of overheating near the power supply components, otherwise no damage will be visible. Replace any visually damaged parts and retest before proceeding. (There is nothing to be done about the overheated PCB itself).
3. Locate a pair of large electrolytic capacitors along one edge. Value should be close to 1000µF at 25V. Replace them with a pair of the same value (slightly higher voltage and capacitance O.K. Try to maintain a match between the two). Carefully observe capacitor polarity!
4. Dig through your surplus parts for some washers and rubber grommets to hold the PCB securely against the front cover, using the original screw(s). Reinstall the PCB into the front cover, ensuring that the button and switches are correctly positioned.
5. Using a short length of extension cord, plug in the transceiver, and retest it.
6. If it works, disconnect power and reassemble (or leave as-is for alignment) - you’re done. If not, further conventional troubleshooting will be required.

It is very common for the main electrolytic filter capacitors in the power supplies of most electronic equipment to fail. It seems especially common in X-10 devices, given the high A.C. ripple, warm temperatures, and other demanding conditions the capacitors must endure. When in doubt, replacing any large (physically and value-wise) electrolytic capacitors (esp. 470µF or larger) may often cure problems. See also 'Fixing Stuttering Module' below

What has happened in this case is that both the positive and negative power supply main filter capacitors have lost most of their capacitance. The power supply voltages have dropped, and have a huge A.C. component across them. Here is what i measured:

Peak-to-peak values were read off an oscilloscope screen. D.C. and Average A.C. values were from a Fluke 77 multimeter.

Fixing Stuttering Module

Due to the low quality of X10 components, the modules sometimes 'stutters', or turns on and immediately off the local appliance module which is operated as unit code 1. This is due to the power supply not being able to move the relay armature against the return spring, and as mentioned above, one way is to replace the 470µF module with a larger one. Another way, perhaps simpler is to weaken the armature return spring.

Update: before trying the procedure below, try changing the polarity of the 220v power prongs. There's an account this helps (see modification to 220v above).

Procedure:

• Open the module. The area of interest is the module's armature, which is the part with the black rectangular plastic side on it. This is the relay.

• Near the electrical prongs, locate teh contact of the armature. This is the part that opens and closes. It is pulled down by a return spring.

• Gently lift it up about 1/2 an inch, streching the return spring. You might need to repeat this motion a few times, but don't overdue it. You have to just weaken the spring a tiny bit.

• Test the module, if it still stutters, repeat the procedure.