Using Linx RF modules ?

[Sam]

What is the data you want to send? How long are the messages?

My first thought was as follows, a PIC and TX module located at a CNC machine,
when the machine moves to specified locations a button closure would activate the PIC/TX to send a short message such as "Change Tool". And other sensors could initiate messages such as "motor stopped" or "coolant flow", ETC.,ETC. and this would be received and the message displayed on an LCD through a PIC on the receiver end.


Or, maybe the better way is to have the PIC/TX send a single character such "a" 5 times or so for one button and another like "b" to the RX/PIC and the RX PIC will generate the appropriate sentence to be displayed on the LCD.

So to answer your question, just short messages or single characters representing different button closures.

[dhouston]

...PIC and TX module located at a CNC machine...

Then, at minimum, you will need error detection. How radio-noisy is the environment? What's the distance between transmitter and receiver? Will this be the only CNC transmitter or are there multiple CNC machines where you might deploy this?

The shorter the message, the better.

I would use something like the NEC protocol for which I have an example on the page mackrackit referenced earlier. I would send each packet 3-5 times with a 10-20mS interval between them, discarding duplicates and any with errors received. This method allows you to handle other tasks within the main loop.

If you want to do it purely with RS232, use $FF,$FF,$FF,$FF,$00,$00,$55 as a preamble and have the receiver's PIC WAIT for the $55. The data (but not the preamble) should be manchester encoded. I would also send these packets 3 or more times with the above interval. With this method you will need to use a timeout if you need to handle other tasks. All in all, I consider it problematic for this application.

If range is not an issue, you might want to consider using the squelch circuit shown in Figure 11 of the Linx datasheet for the receiver. This would eliminate most of the noise (i.e. static) and simplify reception.

However, the best way to do this would be with an FSK transmitter and an FSK receiver with a Carrier Detect output.

Out of curiosity, what type of operation? I spent several years in the machine tool industry and most CNC controls either displayed or had serial outputs for this type of data.

[Sam]

There's allot more to this than I realized and while I love to learn new things, especially PICs and all electronics I think maybe I should go with encoder/decoder chips like THESE and then do the PIC LCD from that.

What do you think about going that route ?

Range only needs to be 50 to 70 feet at most.

And it's your site I bookmarked, sorry about that, very interesting site !

Regarding your question about my CNC setup, it's mostly a hobby setup. I converted a Seig milling machine to CNC and I use Mach3 to control it. I do think there are more outputs from Mach3 I can use but buttons and sensors will suit me fine for now. I do have output functions setup with it to turn the motor and the flood coolant pump on and off per the G-code via an optically isolated relay board I made for it.

Below is a picture of a pendant I made to do the basic functions of the CNC mill so I almost never have to touch the keyboard and a picture of the drivers and power supply in a PC tower case. The perfboard seen is a 12F675 that toggles the power on off via the OEM momentary push button on the front of the case. This picture was taken before I installed the dual optoisolated relay board in the case, also added a second large fan.

Thanks for your help

[dhouston]

...I think maybe I should go with encoder/decoder chips like THESE and then do the PIC LCD from that.

I think the Holtek chips are too limiting. However, Bruce Reynolds has posted some code here for receiving/decoding some Holtek codes.

I'm not familiar with the Mach3 package (and it's been about 25 years since I worked in the industry) but a quick scan of the introduction to its manual indicates it can handle most of what you want. I would study that before choosing a course.

It also seems you want a remote wireless readout so it may be that you will need to add to the built-in functions. The range you need is not a problem with a good antenna unless there are intervening walls. Most range estimates are line-of-sight and you need to figure about 20% of that for indoor use.

My preference is still for FSK transmitter/receiver (with Carrier Detect). Carrier detect with the FM signal along with manchester encoding makes for a very robust system and most FSK TX/RX pairs cost less than the AM modules you've chosen.

I would not do this with a PIC but would use one of the ZBasic chips - probably one of the native versions as it's very easy to create interrupt routines. The ZX328n costs $10, has a free compiler, and includes a serial downloader so you can reprogram it at will. Even if you decide on a PIC, ZBasic is great for developing things quickly which you can then rewrite in PBP.

[Sam]

In what way do you feel the Holtek chips are too limiting ?

Regarding Mach3, I would have to add another parallel port card to get anymore outputs than what I have and are using now and I don't want to go that route at this point even though they are cheap.

Why wouldn't THIS with the companion receiver (just as an example), from Rentron not work for me ?

I mean if this remote can turn LEDs at the receiver on and off why couldn't it activate a PIC programmed with the correct messages and send these corresponding messages to the LCD ?

I'll look into the FSK modules also, I may already have them ???

[dhouston]

I haven't kept up with the hardware so my views may be out of date - Bruce can correct me where necessary The Holtek (and clone) decoder chips have a fixed number of data lines (4 or 8) so you have a limited number of possible codes. These may be adequate for what you want to do - I just prefer more flexibility. Doing your own protocol allows more control. Now, if you want to skip the chips and just use Holtek's encoding scheme, you can have billions of codes.

The device you linked to may well work. I just think utilizing it represents more work than doing it from scratch. Of course, I'm biased as I've done it from scratch many, many times. Once you do it from scratch, you may come to share my bias.

Also, when you wrote CNC, I thought CONTROL which is why I emphasized error detection. But, if you are merely reporting status, error detection may be less of a priority.

[Sam]

Right, status reporting is all I'm wanting to achieve with this. I do prefer to do it from scratch, I just referenced that transmitter as an example. 8 lines will be more than enough for me, 4 will will actually do just fine.

So let's say I have the Linx TX unit and I only want it to report one single button closure on the machine to the receiver 30 feet away through one wall, and I'd like to keep it as simple as possible until I start to get a hold on this as I learn pretty fast once I get one basic thing working and can experiment from there. To keep it simple, lets forget about the LCD and just light an LED for now.

How would you recommend I do this ?

Much thanks.