RF project, designing 433MHz digital low power walkie-talkie

[steviemidnight]

manchester pic rf project

for my 3rd year project at uni i have the task of designing and building 433MHz digital low power walkie-talkies. i am looking for circuit designs in this area.

i will be using qfmt5/qfmr5 fm radio modules. these will be connected to coders/decoders and then to the adc/dac. i am stuck though.

can somebody please help me. has anyone done a project of this type etc.

 

[feedbackkiller]

Hello.

I don´t exactly see the problem you have.
Where are you stuck in the design process ?
No idea for the basic system design ?
Unexpected results during simulation ?
Problems with the PCB layout ?

Please provide a little more information.

 

[steviemidnight]

ok i have purchased a transmiter and reciever from rfsolutions.co.uk qfmt5 and qfmr5 with data rates upto 128kbit/s. i am looking for some schematics that i could attach these modules to.

the design so far is to attach these modules to holtek coders/decoders and then attach these to 8bit adc/dacs with a 8000Hz sample rate. i need a low passs filter after the mic to limit the incoming freq to my cut of freq of 4kHz with a nyquist 8kHz. so i need help with the a good quality filter design either low pass or band pass(300Hz to 3kHz). also how do i filter the out going analogue sig from the dac to the speaker?

this description is my plan for a basic one way working model. i would like to make a two way model with several channels which is where i need the majority of your help

basically has anyone attempted a project like this? has anybody got and schematics that i could modify? any good links would be helpful.

plus can somebody advise me on the design and testing? aslo has anbody got ISIS libary files which contain holtek pics HT12e/ht12d and adc0804lcn?

please help if you can

 

[steviemidnight]

ok i have purchased a tx/rx from rfsolutions.co.uk qfmt5 and qfmr5 with data rates upto 128kbit/s.
the design so far is to attach these modules to holtek coders/decoders and then attach these to 8bit adc/dacs with a 8000Hz sample rate. i need a low passs filter after the mic to limit the incoming freq to my cut of freq of 4kHz with a nyquist 8kHz. so i need help with the a good quality filter design either low pass or band pass(300Hz to 3kHz). also how do i filter the out going analogue sig from the dac to the speaker?
this description is my plan for a basic one way working model. i would like to make a two way model with several channels which is where i need the majority of your help

basically has anyone attempted a project like this? has anybody got and schematics that i could modify? any good links would be helpful.

plus can somebody advise me on the design and testing? aslo has anbody got ISIS libary files which contain holtek pics HT12e/ht12d and adc0804lcn?

 

[Encrypted]

I don't see why you need the Holtek chips. These are encoders/decoders for remote control applications, they will not help you with you with your project.

 

[Pim]

I might give you a small, but hopefully good tip: Regular A/D and D/A converter chips are usually linear in the sence that a double digital value corresponds to a double analogue value. There exists A/D and D/A chips that do not have this linear transfer characteristic. Such are especially aimed for the A/D and D/A conversion of voice and usually have a resolution of 6-8 bits, as far as I can recall from memory.

As you know the dynamic (amplitude-wise) of the human voice could be very high and using such non-linear transfer characteristics will improve the voice quality. Otherwise you run the risk of the sound being heavily distorted if speaking silently.

Another approach could be to use an AGC (Automatic Gain Control) circuit incorporated into the A/D and D/A circuits. But this solution seems to me to be more complicated than necessary.

/Pim

 

[yellowtooth]

Must U do the design of filter ?
If so , digital or analog ?
U can use design software .
Good luck.

 

[feedbackkiller]

Hi !
I´m quite busy right now, but I think I can help with the antialiasing-filters.
I will try to get it done during the weekend and then will post again or PM you, if that´s O.K.

 

[steviemidnight]

Holtek chips i thought all though coders and decoders would enable the siganal to be coded and so would give some security when the info travels throught the airwaves? i was thinking that once i get the basics working, to make a more complex coding that will reduces the chances of the signal being recivered elsewhere, thus increses the security. the holtek coders alllthough i know not designed for this purpose may workfor this process.

Thanks everyone for your help so far. i will looking into the linearity of adc & dca converters. i wil be working all weeekend on the project so ill keep you posted on my progress. im designing the circuit in ISIS. can you recomend a better program as i would like to test the circuit before construction(pspice will probably be used for this)

cheers

 

[Encrypted]

Forget the Holtek chips, they can't pass enough data and they're not fast enough.

The filters will be the least of your problems. You do realise that you can't just hook the ADC output straight into the TX part. You will need a microprocessor to add your start, stop and parity bits, and do the encoding, you can add your security here in the firmware. You'll then end up with a data stream over 128Kbits so you'll need to look into compression.

 

[steviemidnight]

then look at DS305(sorry if this takes awhilt the link is low i know. i think i have underestimated this task.

microprocessor? you mean i need to program a pic? could you please explain more on the parity bits, stop and start stuff.

i have done some stuff on compression so ill go home and read up on the different coding methods.

i want to get a basic setup working first then sort out a more complex version. please keep posting your thoughts

i think i will need all the help i can get

 

[feedbackkiller]

Hello again. I´m terribly sorry, I just had to alter my plans for the weekend, so I won´t have enough time to help you.
Just so much: I did a little reading and research on the web and agree with Encrypted that the filters are the smallest of your problems.
For the antialiasing filter at the output:
Try a simple lowpass solution first, some basic R-C-filter found in any student textbook should work.
You said you want to realize a two-way communication in the end. Then why don´t you use some ISDN telephone IC that does it all ? You end up with a 64KBit data stream that just needs to be XORed with a pseudo-noise data stream to ensure encryption.
But I also agree with Encrypted, that you will need a microprocessor for all this.

 

[Encrypted]

I'd already looked at the datasheets!

The output from your ADC will be a string of 0 and 1's at 64Kbis/s, presuming 8-bit resolution and 8K samples per second. If you feed this straight into the tx, then you have two problems:

1. How will the rx/DAC know where the start of each sample is?

2. You will need to maintain synchronization; if you send a long string of 1's (or 0's) and one clock is running a little slower than the other then you will loose bits.

So, you'll need a way of identifying the start of each sample and a way of sending the clock with the data (look up manchester or NRZ encoding).

You'll definately be programing a micro, a DSP may be better if your encryping and compressing!

I'd be interested to know how you get on so keep us posted.

 

[steviemidnight]

i know about compression, manchesterd and NRZ encoding but how can i implement this into the circuit. i have been looking for cheap dsp chips. i would like to program a pic which would be cheaper and look better in my report. programmed in C then saved to hex i believe so if somebody has any good DSP code or knows where to find them.

sorry if i sound stupid but i have forgotten alot of this. thanks for pushing me in the right direction

the filter is the easy bit i know but could you please remind me of what filter after the dac(DSP) does. thanxs

ill keep you all posted

 

[Encrypted]

The output from your DAC will be a staircase waveform, the LPF removes the fast edges and restores the signal.

 

[Encrypted]

I just found this page for you, it's an RS232 link using similar modules:

https://members.tm.net/mevans/index.html

 

[ms1234]

For walkie-talkie's application, I think it is not need to use ADC/DAC, only CVSD is good enough! That will save the CPU, communcation port, and software.

 

[Encrypted]

A CVSD would definately make life a lot easier, I was just following your original design.

Check out: **broken link removed**

You'll still need to manchester encode to recover the clock at the receiver but you can do this with logic (or a small PIC).

 

[steviemidnight]

how dumb must you lot think i am
i know this:
The output from your DAC will be a staircase waveform, the LPF removes the fast edges and restores the signal
It just completely slipped my mind, thanks though for getting my brain up and runnning again.

those links are good, them modules are made by the same manufacture as mine Quasar(well now rf solution have taken them over).
i will study that page some more.

thanks for sticking with me guys

 

[yellowtooth]

I am modest.
why do not U ?
they do not know if U R familar with DSP.