Hello, It's my first time here. I have no idea in signal processing math but reasonable C programming, and need to send data (minimum 2400bps though more is better) thru the audio (baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru 8KHz linear PCM. In the transmitter after my connection point there’s 300-3000Hz bandpass filter. There’s no space for an external dedicated modem chip and thus need to implement it within the MCU of the main app (without FPU, though it can be 32 bit). Will H-CPM for example be optimal and feasible for this BW, and what’s the bit rate? Of course C source code would be appreciated. Thanks Rony
CPM
Started by ●June 14, 2013
Reply by ●June 14, 20132013-06-14
On Fri, 14 Jun 2013 16:15:19 -0500, "Ro_ny" <95309@dsprelated> wrote:>Hello, > >It's my first time here. > >I have no idea in signal processing math but reasonable C programming, and >need to send data (minimum 2400bps though more is better) thru the audio >(baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru >8KHz linear PCM. In the transmitter after my connection point there’s >300-3000Hz bandpass filter. > >There’s no space for an external dedicated modem chip and thus need to >implement it within the MCU of the main app (without FPU, though it can be >32 bit). > >Will H-CPM for example be optimal and feasible for this BW, and what’s >the bit rate? Of course C source code would be appreciated. > >Thanks > >RonyDoes your 8kHz PCM stream go through the FM mod/demod? That'll make a difference. Eric Jacobsen Anchor Hill Communications http://www.anchorhill.com
Reply by ●June 14, 20132013-06-14
>On Fri, 14 Jun 2013 16:15:19 -0500, "Ro_ny" <95309@dsprelated> wrote: > >>Hello, >> >>It's my first time here. >> >>I have no idea in signal processing math but reasonable C programming,and>>need to send data (minimum 2400bps though more is better) thru the audio >>(baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru >>8KHz linear PCM. In the transmitter after my connection point there’s >>300-3000Hz bandpass filter. >> >>There’s no space for an external dedicated modem chip and thus need to >>implement it within the MCU of the main app (without FPU, though it canbe>>32 bit). >> >>Will H-CPM for example be optimal and feasible for this BW, and what’s >>the bit rate? Of course C source code would be appreciated. >> >>Thanks >> >>Rony > >Does your 8kHz PCM stream go through the FM mod/demod? That'll make >a difference.I doubt that it connects directly to the modulator but this rate refers to the transmitter, while the receiver has same BPF (bandpass filter) but its PCM rate is over twice (not sure what it gives as it has same BPF)> > >Eric Jacobsen >Anchor Hill Communications >http://www.anchorhill.com >
Reply by ●June 14, 20132013-06-14
"Ro_ny" <95309@dsprelated> writes:> Hello, > > It's my first time here. > > I have no idea in signal processing math but reasonable C programming, and > need to send data (minimum 2400bps though more is better) thru the audio > (baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru > 8KHz linear PCM. In the transmitter after my connection point there’s > 300-3000Hz bandpass filter. > > There’s no space for an external dedicated modem chip and thus need to > implement it within the MCU of the main app (without FPU, though it can be > 32 bit). > > Will H-CPM for example be optimal and feasible for this BW, and what’s > the bit rate? Of course C source code would be appreciated. > > Thanks > > Ronyv.34? Implementing such a modem is a gigantic task. It appears you can purchase the software from companies like this: http://www.vocal.com/data-modem/v-34-modem-software/ Good luck. -- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
Reply by ●June 14, 20132013-06-14
On 6/14/2013 4:15 PM, Ro_ny wrote:> Hello, > > It's my first time here. > > I have no idea in signal processing math but reasonable C programming, and > need to send data (minimum 2400bps though more is better) thru the audio > (baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru > 8KHz linear PCM. In the transmitter after my connection point thereтАЩs > 300-3000Hz bandpass filter. > > ThereтАЩs no space for an external dedicated modem chip and thus need to > implement it within the MCU of the main app (without FPU, though it can be > 32 bit). > > Will H-CPM for example be optimal and feasible for this BW, and whatтАЩs > the bit rate? Of course C source code would be appreciated.This is well known task. Simplest solution is DBPSK or DQPSK over audio carrier. An 8bit/1MIPS MCU is good enough. Project for one weekend. Could do C code for you, however this is going to be business. Contact info is at web site below. Vladimir Vassilevsky DSP and Mixed Signal Designs www.abvolt.com
Reply by ●June 14, 20132013-06-14
On 6/14/2013 7:46 PM, Randy Yates wrote:> "Ro_ny" <95309@dsprelated> writes:> v.34? Implementing such a modem is a gigantic task.No problem. I did it. VLV
Reply by ●June 14, 20132013-06-14
"Ro_ny" <95309@dsprelated> writes:>>On Fri, 14 Jun 2013 16:15:19 -0500, "Ro_ny" <95309@dsprelated> wrote: >> >>>Hello, >>> >>>It's my first time here. >>> >>>I have no idea in signal processing math but reasonable C programming, > and >>>need to send data (minimum 2400bps though more is better) thru the audio >>>(baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru >>>8KHz linear PCM. In the transmitter after my connection point there’s >>>300-3000Hz bandpass filter. >>> >>>There’s no space for an external dedicated modem chip and thus need to >>>implement it within the MCU of the main app (without FPU, though it can > be >>>32 bit). >>> >>>Will H-CPM for example be optimal and feasible for this BW, and what’s >>>the bit rate? Of course C source code would be appreciated. >>> >>>Thanks >>> >>>Rony >> >>Does your 8kHz PCM stream go through the FM mod/demod? That'll make >>a difference. > > I doubt that it connects directly to the modulator but this rate refers to > the transmitter, while the receiver has same BPF (bandpass filter) but its > PCM rate is over twice (not sure what it gives as it has same BPF)This "clarification" is confusing to me. I'm assuming you mean that you want the analog input sampled at 8 kHz linear PCM at the beginning of the Rx chain, and the analog output generated from linear PCM with an 8 kHz D/A at the end of the Tx chain, and that you want the demodulation and modulation algorithms to operate at this sample rate. I further assume you chose this sample rate based on the ~3 kHz bandwidth required. --Randy> >> >> >>Eric Jacobsen >>Anchor Hill Communications >>http://www.anchorhill.com >>-- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
Reply by ●June 15, 20132013-06-15
On Fri, 14 Jun 2013 18:37:04 -0500, "Ro_ny" <95309@dsprelated> wrote:>>On Fri, 14 Jun 2013 16:15:19 -0500, "Ro_ny" <95309@dsprelated> wrote: >> >>>Hello, >>> >>>It's my first time here. >>> >>>I have no idea in signal processing math but reasonable C programming, >and >>>need to send data (minimum 2400bps though more is better) thru the audio >>>(baseband) section of FM VHF/UHF radio with 12.5KHz channel spacing thru >>>8KHz linear PCM. In the transmitter after my connection point there’s >>>300-3000Hz bandpass filter. >>> >>>There’s no space for an external dedicated modem chip and thus need to >>>implement it within the MCU of the main app (without FPU, though it can >be >>>32 bit). >>> >>>Will H-CPM for example be optimal and feasible for this BW, and what’s >>>the bit rate? Of course C source code would be appreciated. >>> >>>Thanks >>> >>>Rony >> >>Does your 8kHz PCM stream go through the FM mod/demod? That'll make >>a difference. > >I doubt that it connects directly to the modulator but this rate refers to >the transmitter,It's important to know whether the FM modulator can be, or is intended to be, bypassed or not. When you say "thru the audio (baseband) section" it suggest that the signal is sent in place of the audio signal through the modulator, but then you reference H-CPM, which doesn't go through the modulator in the Project 25 standard. The architecture of the system may dictate what is possible to do or not, and without knowing that one doesn't really know how to answer the question. Is the PA linear? That'll matter as well, as some of the possible solutions could be eliminated as not possible if the PA isn't linear. H-CPM was proposed specifically because the PA in the handset isn't linear in order to improve power efficiency. Likewise, if the FM discriminator/demodulator can't be bypassed in the receiver, the same restrictions may apply. FWIW, H-CPM supports, as I understand it, 12kbps at the air interface, with a sustainable rate in a single link necessarily quite a bit below that. Depending on the number of simultaneous users, I could see it providing potentially on the order of 2400bps.> while the receiver has same BPF (bandpass filter) but its >PCM rate is over twice (not sure what it gives as it has same BPF)A higher rate in the demodulator can be expected in order to provide some processing gain for SNR improvement and to aid in synchronization. Eric Jacobsen Anchor Hill Communications http://www.anchorhill.com
Reply by ●June 15, 20132013-06-15
eric.jacobsen@ieee.org (Eric Jacobsen) writes:> FWIW, H-CPM supports, as I understand it, 12kbps at the air interface, > with a sustainable rate in a single link necessarily quite a bit below > that. Depending on the number of simultaneous users, I could see it > providing potentially on the order of 2400bps.Does H-CPM stand for "Harmonized Continuous Phase Modulation?" If so, where is this defined? I have not heard of this modulation format. I did some googling and could only find references to it and not an actual definition. -- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
Reply by ●June 15, 20132013-06-15
Randy Yates <yates@digitalsignallabs.com> writes:> eric.jacobsen@ieee.org (Eric Jacobsen) writes: > >> FWIW, H-CPM supports, as I understand it, 12kbps at the air interface, >> with a sustainable rate in a single link necessarily quite a bit below >> that. Depending on the number of simultaneous users, I could see it >> providing potentially on the order of 2400bps. > > Does H-CPM stand for "Harmonized Continuous Phase Modulation?" If so, > where is this defined? I have not heard of this modulation format. I did > some googling and could only find references to it and not an actual > definition.PS: I also don't want to shuck out money to Project 25, the TIA (or anyone else) for the standard. -- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
