Date   

APRS changes for the DV modes

Jonathan Naylor
 

Hi Folks

I'm revisiting the APRS support for the DV modes. To that end there are two changes, one has been done, and the other is to be done.

1. Instead of each gateway accessing an APRS-IS system such as aprs.fi individually, they now communicate via a local socket to a new gateway, called the APRS Gateway. This then sends the data unchanged to the configured APRS-IS system. Things like logging in to that system is also handled by the APRS Gateway. This simplifies each gateway that needs it, and hopefully reduces the load on the APRS servers, particularly aprs.fi.

The APRS Gateway is already uploaded to my GitHub and the modified gateways for D-Star, System Fusion, and NXDN are held in the APRS branches for each of them. Note that the configuration of each has changed to accommodate this new gateway. This work now compiles but has not been tested.

2. I will move away from my MobileGPS program to gpsd. It seems that some of the USB GPS pucks are a bit odd, and rather than spend my time trying to support them all, I'll let someone else do so. I haven't started work on this. I will probably create a suitable branch on each of the affected servers for this development work. It will be announced on here when I've finished.

Jonathan  G4KLX

 


Got FM working !

Frank C. D'Amato
 

I still need to do some fine tuning, but it does work.

MMDVM Repeater Test of C4FM - DMR - Analog FM


Re: Squelch tail

Frank C. D'Amato
 

When I use dmr or c4fm, I hear the digital noise also coming out to my Frequency on FM that has ctcss programmed in mmdvm and my radio.  How do I stop that from happening.  It appears the digital modes are also transmitting the ctcss from the FM side.


Re: Squelch tail

Jeff5bcb
 

Hi Geoffrey,
To avoid hearing the DMR modulation on an analog channel of the repeater, I put a CTCSS on transmission. In this way, the terminal only opens when the repeater is in FM mode.
When the repeater stops after sending its callsign when the hang timer expires, I hear a crackle for about 200 or 300 ms.
The repeater works with CTCSS + COS opening (in reality COS is the Carrierd Detect) to avoid false CTCSS detections.

73

Jeff, F5BCB


Re: Squelch tail

Geoffrey Merck
 

Hi Jeff,

Not sure I understand your issue. What kind of noise do you have ?

"Currently it works with CTCSS + COS" => You have CTCSS and COS activated ?

73
Geoffrey F4FXL - KC3FRA


Squelch tail

Jeff5bcb
 

Hi all,
I noticed a squelch tail phenomenon which produces an unpleasant noise when the repeater no longer receives a carrier.  Currently it works with CTCSS + COS
I do not know what parameter can act to solve this problem.
Have you noticed this behavior?
73
Jeff, F5BCB



FM/NXDN Mode and older MMDVM modems

Jonathan Naylor
 

Hi Folks

I see a lot of comments about the MMDVM not supporting various hardware with the new FM work. I think I need to point some things out:

1. The developers don't have access to all of the hardware that is out there. We can only support what we have, and even then it isn't easy. Ultimately if your hardware was created by G9BF, then it is G9BFs responsibility to support it. If we can make sure it runs correctly then we'll try to do so, but we make no guarantees. The most common quality modem hardware is provided by KI6ZUM, N3XCC (Repeater Builder), BI7JTA, and DF2ET. This list is not exhaustive, but these represent good quality boards and you can reasonably expect them to work. However see note 2 below.

2. With the additional work on the MMDVM, we are finding that some of the older MMDVM hardware no longer has enough processing power to run it. This first came up with NXDN mode, and this is now showing itself with the new FM mode. Geoffrey F4FXL had to do some amazing optimisations in order to get the standalone FM mode to run on the Arduino Due, but there is no way that this platform will support the use of external audio (i.e. Echolink). This comment also applies to other boards, typically ARM M4 boards with clock speeds lower than 180 MHz. This unfortunately applies to Arduino Due based interfaces, Repeater Builder V1 and V2 boards, SP6POG boards, and many no-name boards that are sourced from China via eBay.

If you are buying a new MMDVM modem board and you aren't sure what processor that it uses, ask the seller. If it's less than a 180 MHz ARM M4 and the seller claims that it supports FM or NXDN mode, don't believe them. Better, buy your board from a reputable seller, you may be paying more, but you'll be getting guaranteed hardware with a good chance that it will work with future upgrades.

Jonathan  G4KLX


Re: Latest MMDVM firmware compatibility with STM32F104

Razvan M
 

Hi Geoffrey,

RaspberryPi 3B / Stretch / Raspbian GNU/Linux 9 / Kernel 4.19.66-v7+
I don't think it's a platform problem since I have the same hardware/software combo since last year and I've tested a lot of firmwares

Cheers!
Razvan / YO6NAM


Re: Latest MMDVM firmware compatibility with STM32F104

Geoffrey Merck
 

Hi,

What platform did you compile on ?
I needed to compile under Debian 9. Compiling under debian 10 resulted in w non working Firmware.

73
Geoffrey F4FXL - KC3FRA


Re: Issues with COS - Low output from MOSFET

f5vmr23
 

Yes of course - I see that now. Have to consider that when conversion time comes.


On 22 May 2020, at 20:07, Dominic <dominic@...> wrote:

It is beneficial when using the new FM Analogue mode on the MMDVM. Not needed for digital modes.

Dom G7NPW

On 22 May 2020, at 6:36 pm, f5vmr23 <f5vmr@...> wrote:

Running 3 DMR repeaters I have not used COS in any of them, and they all use different modem cards Zum-USB, Zum-Radio Pi-Hat & RepeaterBuilders Hat, I presume this is for other modes?

Regards
Chris F5VMR/G4NAB

On 22 May 2020, at 18:02, dominic@... wrote:

I pulled mine up high with a 10k pull up resistor on my F0DEI board then set the Tait to be active high when busy. And applied this signal between the resistor and the board input. 

You have the option to invert the sense in the software now and also the radio so have you tried it the opposite way around?



--
Dominic G7NPW

On 22 May 2020, at 3:23 pm, Matthew 2E0SIP <groups.io@...> wrote:



Hi All,

I am experiencing some issues with COS which I believe are being caused by a low voltage from the MOSFET (~1.6V) not being high enough to register as "high" on the STM32F446RE.

I am using a Tait TM81XX with 3.3v logic connected to the RADIO_SQL pin. When the output from the radio goes HIGH, only ~1.6V is output at the SQL pin. This configuration with a 
2N7000 MOSFET appears to be common amongst MMDVM boards. Jeff (F5BCB) and Toufik (F0DEI) have kindly made suggestions and I've swapped out the 2N7000 and swapped the 47K resistor for 4.7K, but the issue remains.

Unfortunately discreet electronics isn't my forte. Does anyone have any suggestions? Is this configuration working for others?

Any thoughts appreciated,
Matthew

2E0SIP

 


<Screen Shot 2020-05-22 at 15.07.51.png>




Re: Issues with COS - Low output from MOSFET

Dominic
 

It is beneficial when using the new FM Analogue mode on the MMDVM. Not needed for digital modes.

Dom G7NPW

On 22 May 2020, at 6:36 pm, f5vmr23 <f5vmr@...> wrote:

Running 3 DMR repeaters I have not used COS in any of them, and they all use different modem cards Zum-USB, Zum-Radio Pi-Hat & RepeaterBuilders Hat, I presume this is for other modes?

Regards
Chris F5VMR/G4NAB

On 22 May 2020, at 18:02, dominic@... wrote:

I pulled mine up high with a 10k pull up resistor on my F0DEI board then set the Tait to be active high when busy. And applied this signal between the resistor and the board input. 

You have the option to invert the sense in the software now and also the radio so have you tried it the opposite way around?



--
Dominic G7NPW

On 22 May 2020, at 3:23 pm, Matthew 2E0SIP <groups.io@...> wrote:



Hi All,

I am experiencing some issues with COS which I believe are being caused by a low voltage from the MOSFET (~1.6V) not being high enough to register as "high" on the STM32F446RE.

I am using a Tait TM81XX with 3.3v logic connected to the RADIO_SQL pin. When the output from the radio goes HIGH, only ~1.6V is output at the SQL pin. This configuration with a 
2N7000 MOSFET appears to be common amongst MMDVM boards. Jeff (F5BCB) and Toufik (F0DEI) have kindly made suggestions and I've swapped out the 2N7000 and swapped the 47K resistor for 4.7K, but the issue remains.

Unfortunately discreet electronics isn't my forte. Does anyone have any suggestions? Is this configuration working for others?

Any thoughts appreciated,
Matthew

2E0SIP

 


<Screen Shot 2020-05-22 at 15.07.51.png>



Re: Building new firmware and mmdvm host

pd1acf
 

Hi Scott and all,

I did program the firmware, before I noticed the message about 'not load the firmware' version.
So far no real issues observed with the 'older' modes, except for the fact that opening the FM repeater with configured CTCSS code results in these log lines:

M: 2020-05-22 17:37:43.926 Mode set to FM
M: 2020-05-22 17:38:34.235 Mode set to FM
E: 2020-05-22 17:38:34.486 MMDVM DAC levels have overflowed
E: 2020-05-22 17:38:34.738 MMDVM DAC levels have overflowed
E: 2020-05-22 17:38:34.989 MMDVM DAC levels have overflowed
E: 2020-05-22 17:38:35.241 MMDVM DAC levels have overflowed

If there is any way I can help testing FM firmwares let me know!

73'

Ron


Re: Issues with COS - Low output from MOSFET

f5vmr23
 

Running 3 DMR repeaters I have not used COS in any of them, and they all use different modem cards Zum-USB, Zum-Radio Pi-Hat & RepeaterBuilders Hat, I presume this is for other modes?

Regards
Chris F5VMR/G4NAB

On 22 May 2020, at 18:02, dominic@... wrote:

I pulled mine up high with a 10k pull up resistor on my F0DEI board then set the Tait to be active high when busy. And applied this signal between the resistor and the board input. 

You have the option to invert the sense in the software now and also the radio so have you tried it the opposite way around?



--
Dominic G7NPW

On 22 May 2020, at 3:23 pm, Matthew 2E0SIP <groups.io@...> wrote:



Hi All,

I am experiencing some issues with COS which I believe are being caused by a low voltage from the MOSFET (~1.6V) not being high enough to register as "high" on the STM32F446RE.

I am using a Tait TM81XX with 3.3v logic connected to the RADIO_SQL pin. When the output from the radio goes HIGH, only ~1.6V is output at the SQL pin. This configuration with a 
2N7000 MOSFET appears to be common amongst MMDVM boards. Jeff (F5BCB) and Toufik (F0DEI) have kindly made suggestions and I've swapped out the 2N7000 and swapped the 47K resistor for 4.7K, but the issue remains.

Unfortunately discreet electronics isn't my forte. Does anyone have any suggestions? Is this configuration working for others?

Any thoughts appreciated,
Matthew

2E0SIP

 


<Screen Shot 2020-05-22 at 15.07.51.png>



Re: Issues with COS - Low output from MOSFET

Dominic
 

Hi Matthew,

This is what I did on mine and it works 100%

Repeater pinout might be different to yours

Dom G7NPW






















On 22/05/2020 17:19, Matthew 2E0SIP wrote:

Thanks for the responses.

Brian - The power comes from a Raspberry Pi over USB to a Nucleo Board where I guess it's dropped down to 3.3v. All grounds are commoned. I've confirmed there's 3.3V at the Drain. 

Dominic - it sounds like we're running a very similar setup. I have tried fiddling with the inversion. As far as I can tell it only works when the MMDVM is detecting the signal as LOW and in that case it's permanent and is just falling back to CTCSS detection. When it's detecting HIGH the repeater doesn't open at all. Have you omitted the MOSFET entirely? 

Thanks again
Matthew



Re: Issues with COS - Low output from MOSFET

Brian Morrison
 

On Fri, 2020-05-22 at 09:19 -0700, Matthew 2E0SIP wrote:

Thanks for the responses.

Brian - The power comes from a Raspberry Pi over USB to a Nucleo Board where I guess it's dropped down to 3.3v. All grounds are commoned. I've confirmed there's 3.3V at the Drain. 

Dominic - it sounds like we're running a very similar setup. I have tried fiddling with the inversion. As far as I can tell it only works when the MMDVM is detecting the signal as LOW and in that case it's permanent and is just falling back to CTCSS detection. When it's detecting HIGH the repeater doesn't open at all. Have you omitted the MOSFET entirely? 

Thanks again
Matthew

The 2N7000 is an n-channel device, so you are using it as a source follower with a fairly high output impedance (that 47K). Note that the device itself has a minimum on-resistance of about 5 ohms. A better way is to add in a second P-channel device with its source connected to 3V and the drain of the 2N7000 connected to its gate (disconnect 2N7000 drain from 3V), then use the p-channel device drain to drive the STM32F446 input with a pull-down resistor of a reasonable value to ground. This should give you active high on the squelch input and a good drive capability which you won't get with the existing circuit.

--

Brian G8SEZ


Re: Issues with COS - Low output from MOSFET

Matthew 2E0SIP
 

Thanks for the responses.

Brian - The power comes from a Raspberry Pi over USB to a Nucleo Board where I guess it's dropped down to 3.3v. All grounds are commoned. I've confirmed there's 3.3V at the Drain. 

Dominic - it sounds like we're running a very similar setup. I have tried fiddling with the inversion. As far as I can tell it only works when the MMDVM is detecting the signal as LOW and in that case it's permanent and is just falling back to CTCSS detection. When it's detecting HIGH the repeater doesn't open at all. Have you omitted the MOSFET entirely? 

Thanks again
Matthew


Re: Issues with COS - Low output from MOSFET

Dominic
 

I pulled mine up high with a 10k pull up resistor on my F0DEI board then set the Tait to be active high when busy. And applied this signal between the resistor and the board input. 

You have the option to invert the sense in the software now and also the radio so have you tried it the opposite way around?



--
Dominic G7NPW

On 22 May 2020, at 3:23 pm, Matthew 2E0SIP <groups.io@...> wrote:



Hi All,

I am experiencing some issues with COS which I believe are being caused by a low voltage from the MOSFET (~1.6V) not being high enough to register as "high" on the STM32F446RE.

I am using a Tait TM81XX with 3.3v logic connected to the RADIO_SQL pin. When the output from the radio goes HIGH, only ~1.6V is output at the SQL pin. This configuration with a 
2N7000 MOSFET appears to be common amongst MMDVM boards. Jeff (F5BCB) and Toufik (F0DEI) have kindly made suggestions and I've swapped out the 2N7000 and swapped the 47K resistor for 4.7K, but the issue remains.

Unfortunately discreet electronics isn't my forte. Does anyone have any suggestions? Is this configuration working for others?

Any thoughts appreciated,
Matthew

2E0SIP

 


<Screen Shot 2020-05-22 at 15.07.51.png>


Re: Issues with COS - Low output from MOSFET

Brian Morrison
 

On Fri, 2020-05-22 at 07:23 -0700, Matthew 2E0SIP wrote:

Hi All,

I am experiencing some issues with COS which I believe are being caused by a low voltage from the MOSFET (~1.6V) not being high enough to register as "high" on the STM32F446RE.

I am using a Tait TM81XX with 3.3v logic connected to the RADIO_SQL pin. When the output from the radio goes HIGH, only ~1.6V is output at the SQL pin. This configuration with a 
2N7000 MOSFET appears to be common amongst MMDVM boards. Jeff (F5BCB) and Toufik (F0DEI) have kindly made suggestions and I've swapped out the 2N7000 and swapped the 47K resistor for 4.7K, but the issue remains.

Unfortunately discreet electronics isn't my forte. Does anyone have any suggestions? Is this configuration working for others?

Any thoughts appreciated,
Matthew

2E0SIP

 



What does
What is the input circuit to the STM32F446? What supply does that chip run from?

--

Brian G8SEZ


Issues with COS - Low output from MOSFET

Matthew 2E0SIP
 

Hi All,

I am experiencing some issues with COS which I believe are being caused by a low voltage from the MOSFET (~1.6V) not being high enough to register as "high" on the STM32F446RE.

I am using a Tait TM81XX with 3.3v logic connected to the RADIO_SQL pin. When the output from the radio goes HIGH, only ~1.6V is output at the SQL pin. This configuration with a 
2N7000 MOSFET appears to be common amongst MMDVM boards. Jeff (F5BCB) and Toufik (F0DEI) have kindly made suggestions and I've swapped out the 2N7000 and swapped the 47K resistor for 4.7K, but the issue remains.

Unfortunately discreet electronics isn't my forte. Does anyone have any suggestions? Is this configuration working for others?

Any thoughts appreciated,
Matthew

2E0SIP

 



Latest MMDVM firmware compatibility with STM32F104

Razvan M
 

I compiled and uploaded the latest version on a STM32F104 board and noticed it won't start (no led sequence, only two lit when powered on).
The version (prior to FM) that's working is https://github.com/g4klx/MMDVM/commit/b4647d8

Can someone confirm if the latest FW works on boards based on STM32F1xx?

Thanks

521 - 540 of 852