ADICON Support Forum Applied Digital, Inc.
Page 1 of 2 12>
Topic Options
#13707 - 12/31/03 10:24 PM Dave Houston's RF Support
Anonymous
Unregistered



Dave Houston has developed a way for the Ocelot/Leopard to directly control RF based X-10 camera mounts & celing fans.

http://www.mbx-usa.com/appnotes.htm

My thanks to Dave for sharing the information.

Dan

Top
#13708 - 12/31/03 10:28 PM Re: Dave Houston's RF Support
Anonymous
Unregistered



Copied from comp.home.automation:

From: embedded1 (embedded1@embed-x.com)
Subject: Ocelot control of Ninjas & Fans
This is the only article in this thread
View: Original Format
Newsgroups: comp.home.automation
Date: 2003-12-31 09:11:07 PST


ADI has a free utility that allows editing of Ocelot LIR (Learned IR)
files. It also allows importing CCF codes and converting them to LIR
codes.

In N. America, the Ninja camera mounts and _some_ Hunter ceiling fans
use 310MHz RF.

CodeGen will generate the codes used by the Ninja as CCF. These can be
converted to LIR format and the Ocelot IR output can be used to drive
a 310MHz RF transmitter module. mbx-usa.com stocks a 310MHz RF
transmitter module.

An IR booster like the one Guy Lavoie designed can be used to drive
both the IR emitter and the RF module.

Many other ceiling fans (and some fireplaces) use 303-304MHz RF.
CodeGen has codes for some of these. mbx-usa.com will stock a
303.825MHz RF transmitter module for these if there is sufficient
interest.

In Europe the X-10 camera mount uses 433.92MHz. There are many sources
for RF transmitter modules. The Ninja codes from CodeGen will also
work with the European camera mount.

Top
#13709 - 01/01/04 07:19 AM Re: Dave Houston's RF Support
Guy Lavoie Administrator Offline
Beyond All Hope
*****

Registered: 12/21/02
Posts: 6548
Loc: Montreal, QC, Canada
The circuit on Dave's page ends up doing the same thing as the one I referred to in the ceiling fan thread in September:

http://www.appdigsupport.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=3;t=000188;p=1

But with the difference that Dave is sending a modulated envelope (probably around 38 kHz) instead of just the envelope like I do (using a demodulator circuit made with a 555 timer), but it still seems to work ok... Also, he uses ready made RF oscillators while I ended up hacking into the fan's original remote to key its own oscillator. Whatever way you use, it shows that controlling fans and other RF controlled devices is indeed quite possible to do with an Ocelot/Leopard or SECU16IR output. I'll resend the schematic to Dan to post it here so that anyone interested in building one can use either that or Dave's circuit, whichever is more convenient for the project builder.
_________________________
"If you don't know what you're doing, do it neatly..."

Top
#13710 - 01/01/04 05:03 PM Re: Dave Houston's RF Support
Anonymous
Unregistered



Guy,
Thanks for the schematic.


Top
#13711 - 01/01/04 05:21 PM Re: Dave Houston's RF Support
Guy Lavoie Administrator Offline
Beyond All Hope
*****

Registered: 12/21/02
Posts: 6548
Loc: Montreal, QC, Canada
Thanks Dan. I have to explain a bit about the "inside remote" part of the diagram. This is for the "Mainstays" fan controller remote described in the thread I previously referred to. In that remote, the RF oscillator is keyed on and off by a +9 volt signal envelope (+9v = on, 0v = off). The circuit can be adapted as needed to other remote or even purpose built oscillator modules like the ones shown on Dave Houston's web page. I looked for ready made oscillator circuits but by the time shipping costs to Canada were factored in, it was actually cheaper to just buy another fan controller kit (about $20 US) just to get another remote, and get a bonus spare fan module too...

If anyone wants to use this module, I have learned the command button codes for all 16 possible fan addresses. Just ask and I'll email you the LIR file. You can then copy the codes you need into your own LIR file using IR-Max.
_________________________
"If you don't know what you're doing, do it neatly..."

Top
#13712 - 04/02/04 04:22 PM Re: Dave Houston's RF Support
Dave Matthews Offline
newbie


Registered: 01/25/03
Posts: 19
Loc: Tennessee
Thanks for the 555 circuit Guy, and thanks to Dave Houston as well for CODEGEN.

I'm building a rf fan control for my 6 Hampton Bay fans to be controlled by an Ocelot. The 555 circuit works perfectly, and use it to drive a RF transmitter module I bought from Computronics (303.875 mHz). Dave Houston's CODEGEN program was vital. It still took a while to get my circuit to work, it now works reliably only if I specify 5 code repeats from CODEGEN. The graph function in IrMax shows 6 repeats and some junk bits, but it works fine so-far.

I'm going to test the range next, but this sure is fun...at three hundered and four million cycles per second.

Dave

Top
#13713 - 04/03/04 02:42 AM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
This circuit is probably easier to build. If you want only RF, you can skip the lower section.

http://www.mbx-usa.com/ocelot-rf.htm

In CodeGen™ 1 initial + 5 repeats = 6 copies.

This URL...

http://www.mbx-usa.com/rf-noise.htm

shows why multiple copies are needed. The Hampton Bay codes have no lead-in so extra copies are needed to set the receiver AGC.

Top
#13714 - 04/03/04 05:14 AM Re: Dave Houston's RF Support
Dave Matthews Offline
newbie


Registered: 01/25/03
Posts: 19
Loc: Tennessee
Thanks for the information about the receiver AGC, that explains it. So I just repeat the code...

I made an error in my previous post wrt repeats: I don't know if I'm doing something wrong with CodeGen and IRMax, but when I ask for 5 repeats, the IRMax graphing functions shows 7 groups with two odd pulses at the end.

I considered the op-amp version, and I appreciate the advice. I went with the 555 circuit since I had a bunch of 555s laying around . I was put off a bit by the notice associated with Dave's demodulator about the potential need to edit the .lir code. I am a complete novice when it comes to this rf stuff and even the IR stuff, so I am very grateful to you for your reply!

Dave

Top
#13715 - 04/03/04 06:02 AM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
Most IR receivers will tolerate +/-10% variations in pulse width so you're unlikely to need to edit the codes because of the slight stretching. In fact, it may be more severe with the 555 circuit.

I've never looked at IRMax but it sounds like it may be adding the noise. All CodeGen™ does is string 6 copies of the basic code together. You can see what a single copy looks like by setting it for 0 repeats.

You might also improve things by manually editing the single code to reduce the gap between codes. Then, string 3-4 together and change the 3rd word to reflect the total number of burst pairs.

If you click on any word in the CodeGen™ output window, a popup window explains that word. You can see the number of burst pairs by clicking on the third hex word . Multiply this by the number of copies and replace 000D with the appropriate hex value. The final hex word in the single code is the gap (in numbers of carrier cycles). Reducing it will help the AGC and may reduce the number of copies you need to send.

You'll get the best range with an antenna that is about 9 inches long. A piece of solid hook-up wire is adequate. Experiment with horizontal vs. vertical orientation. The antenna in the fan receiver is not very efficient.

The RF transmitter module you're using can handle 3-12V for its supply. Range is proportional to supply voltage.

Top
#13716 - 04/03/04 06:22 AM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
If you do need to edit the pulses/spaces, subtract one cycle from each pulse and add it to the space that follows. Pulses and spaces alternate. The first pulse is word #5.

With the Hampton Bay codes, the first pulse is 25 cycles of carrier. 1 cycle is 4%. Any error from the filter will be constant rather than proportional to the pulse width so shifting one cycle from pulses to spaces will do the trick.

With the CCF format, resolution is limited to 1 cycle of carrier which is about the same amount (at 38kHz) that the filter will add to pulse widths. In other words, it's unlikely that the slight stretching of pulses (and commensurate shrinking of spaces) will present any problem.

A future CodeGen™ release will allow you to adjust pulses/spaces individually or throughout the entire code. But, it's simple enough to do the same thing using Search & Replace with most text editors.

Top
#13717 - 04/03/04 06:29 PM Re: Dave Houston's RF Support
Dave Matthews Offline
newbie


Registered: 01/25/03
Posts: 19
Loc: Tennessee
Thanks for all of the information. I'll try to make some comparisons between the op-amp circuit and the 555 circuit before I hard wire anything. I have been able to edit the .lir code to shorten the time between pulse repeats, and I just need to establish a few benchmarks to see where I am in order to make evaluations of performance.

I'm still struggling with understanding the codes represented in pronto and .lir, but I'm getting there. I believe the code is converted properly, including repeats, but I may be using the graph function of IRmax incorrectly. I'm still not so sure I'm doing everything correctly. (I don't get any explanitory popup window when I click on a word in the CodeGen output window either.)

I plan to use a 12V power supply with the final unit, and I am using the 9" antenna, coiled up.

Do you know what the drive requirement to the code input of the TWS-303 is? I'm presently coupling the output of the 555 to the code input via a 1K resistor...I seem to be getting the full voltage swing at the code input, just wondering.

Thanks again for the efforts you've made explaining this to me, I greatly appreciate it.

Dave

Top
#13718 - 04/04/04 06:11 AM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
I don't have any data beyond what's given on the Computronics website which indicates the maximum current on the data line is 100ľA.

Most of my experience has been with a later, 4-pin transmitter from the same manufacturer. I think it needs about 2V to turn the transmitter on.

Top
#13719 - 04/04/04 06:39 PM Re: Dave Houston's RF Support
Dave Matthews Offline
newbie


Registered: 01/25/03
Posts: 19
Loc: Tennessee
Ouch, I should have seen that! Thanks again. I see that the input high voltage minimum is Vcc-0.5V, I'm ok there. I've got to check the low voltage from the 555 output, at this current it should be less than the 0.3V maximum specified for an off state.

I actually looked at the module, saw and then measured a 47K resistor from the input pin to a transistor. So I guessed that my 1K series resistance wasn't affecting anything. So I replaced it with a wire, and suddenly the performance improved...kinda funny because putting my scope probe on the input made it work better too. I think I have some serious lead-dress issues, I'm doing all this on a simple plastic plug-in breadboard and I wonder if at these rf frequencies I shouldn't be a bit more realistic about connections! Any tips regarding module connections?

The Dave Houston information included the antenna instruction of winding a 9" length of 22ga wire into a tight helical shape. Could you advise me further regarding the antenna? Is a straight wire as good, better? I'm using the helix currently.

Thanks for your help,
Dave

Top
#13720 - 04/04/04 07:29 PM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
I use breadboards all the time with no apparent ill effects. Maybe you just have iffy connections.

A straight wire will give maximum range. The helix gives minimum size.

Top
#13721 - 04/11/04 01:31 PM Re: Dave Houston's RF Support
Dave Matthews Offline
newbie


Registered: 01/25/03
Posts: 19
Loc: Tennessee
rfguru, everything seems to be working well, but I'd like to get a little more reliable distance. I'm presently getting about 40'. I am going to a 12V supply when I hardwire this, and that will help some, but I wonder a bit more about the antenna.

Would the receiver antenna MBX310ANT used as a transmitting antenna give me any advantage over the straight wire? Would it make positioning less critical? I am transmitting to an L shaped room with 4 of the fans on X and 2 on Y axes. Would a 1/2 wavelenth straight wire antenna give any better performance than the 1/4 wavelength wire?

Would increasing the fan rf receiver module's antenna length help at all? I have to install a jbox at each fan and I could extend the receiver antenna to 1/2 wavelength or so.

Thanks again,
Dave

Top
#13722 - 04/11/04 04:51 PM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
40' is about what I would expect with 5V supply and the dinky receiver antenna. 12V will likely increase the range by a factor of 2 or 3. However, this is hard to predict because most houses are not all that hospitable to RF. You may be getting reflections, refractions, multi-path reception, etc.

1/2 wavelength probably won't help much but you can always cut it back to 1/4. If you lengthen the receiving antenna, I suggest soldering a new wire in place of the short one they use as opposed to splicing it.

I would go step-by-step. Try 12V. If you need more range, replace the receiver antenna with a ~9" wire. If you need still more range, the MBX310ANT would probably give you maximum (but may be too much).

Top
#13723 - 04/12/04 06:57 AM Re: Dave Houston's RF Support
Robert Jenkins Offline
active contributor


Registered: 07/08/03
Posts: 122
Loc: Worksop, UK
Hi,
one thing that may help is to use another 1/4 wave wire (9.25") as a counterpoise; just connect it directly to the receiver ground nearest the aerial point & arrange it so it's aimed away from the existing aerial wire.
In effect, you are putting the receiver in the centre of a half wave dipole.

The same should work on the transmitter.

Helically wound aerials are not as good in open, line-of-site conditions but are far less affected by proximity to other objects & people etc., so can be better in some circumstances.
_________________________
Robert Jenkins.

Top
#13724 - 04/12/04 07:24 AM Re: Dave Houston's RF Support
rfguru Offline
junior


Registered: 03/19/04
Posts: 33
In theory, a 1/4 wave monopole is equivalent to a 1/2 wave dipole.

Anyway, it's not particularly easy to add the ground half of a dipole to the ceiling fan. You probably need to see the receiver that mounts at the fan to appreciate this.

Adding a large ground plane underneath the transmitter will probably help. It need not even be connected electrically to the transmitter ground - it just needs to be physically close to it. A large, flat piece of metal will do - even a piece of aluminum foil.

But, I expect going to a 12V supply will be more than adequate.

Top
#13725 - 04/12/04 01:44 PM Re: Dave Houston's RF Support
Dave Matthews Offline
newbie


Registered: 01/25/03
Posts: 19
Loc: Tennessee
Thanks gents for the replies. The six fans that I want to control are in a swimming pool house that has an aluminum/foam/aluminum sandwich ceiling/roof, and the fans are right up close to the ceiling. So maybe I have that large ground plane!

I realized after my last post that maybe I haven't done enough research myself on this matter, it has been so convenient just to ask here. Thanks again for ya'lls help, and I'll do some research and experiments (12V) before I post again.

Talk of a counterpoise reminded me that I've got a 1958 Radio Amateur's Handbook around here somewhere....

Dave

Top
#13726 - 04/12/04 02:22 PM Re: Dave Houston's RF Support
Robert Jenkins Offline
active contributor


Registered: 07/08/03
Posts: 122
Loc: Worksop, UK
They are very Un-equal... a 1/2 wave Dipole has 2db gain relative to a 1/4 wave ground plane monopole.

There will be considerably more effect when converting a 'floating' (i.e. no ground plane) antenna to a dipole.

Even if the grounded section has to be poked up through the ceiling or whatever, it will give a significant improvement. Just try and keep the 'free' ends of the antenna wires away from objects - any distance is better than being in contact with something.

If it's possible at both the transmit and receive ends, you will probably get double (or more) operating range, before any increases due to transmitter voltage. One end only should give you around 50% improvement.

The dipoles will be more directional, though. This is the trade-off for antenna gain. The best signal will be side on; if you can arrange them near vertically, they should work from any direction.

Alternatively, if the transmitter is going to be in a fixed location, you could connect it to a piece of coax and put an antenna near the fans?
_________________________
Robert Jenkins.

Top
Page 1 of 2 12>


Moderator:  Dan Smith, Monte G, ADI Tech Support, Guy Lavoie 
Hop to:
Who's Online
0 registered and 49 anonymous users online.
Recent Posts
Ocelot not working
by JPO
11/08/17 06:02 AM
Shout Box

Newest Members
brigiel, vevevie, zhutree, 416, saiqul
3003 Registered Users
Forum Stats
3003 Members
19 Forums
4004 Topics
23442 Posts

Max Online: 132 @ 11/13/16 10:07 AM
November
Su M Tu W Th F Sa
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30