|
Hi all,
Does anyone try to QRV the CW mode by Raduino_v1.14.1 for BITX40?
Does it really works?
Here in my situation. The TX frequency does not down 800Hz when I select LSB mode.
Any suggestions welcome.
ja9mat hidehiko.
|
Hidehiko-san,
The frequency as shown on the display is always the TX carrier frequency.
In CW mode, the RX frequency is shifted by 800 Hz, not the TX frequency.
73, Allard PE1NWL
toggle quoted message
Show quoted text
On Wed, June 14, 2017 11:58, qrper72@... wrote: Hi all,
Does anyone try to QRV the CW mode by Raduino_v1.14.1 for BITX40?
Does it really works?
Here in my situation. The TX frequency does not down 800Hz when I select
LSB mode.
Any suggestions welcome.
ja9mat hidehiko.
|
TKS Allar,
OK but it is not a display issue.
800Hz frequency shift does not occur so I can not QSO with hi..
When I attach the voltmeter between "A0" and "GND" the problem is solved sometimes
but not stable. hi...
Maybe the "A0" PTT-SENSE signal line should be pull-up or pull-down?
ja9mat hidehiko.
|
I've put the 10k pull-down register at the "A0" and maybe the problem seems to be fixed now. This issue caused by the threshold problem?
ja9mat Hidehiko.
|
Hi Hidehiko-san
Yes you could be right.
Another possible cause could be poor grounding of BITX main board and
Raduino board.
If Raduino board is not grounded perfectly, some small potential
difference may exist between BITX board and Raduino. As a result, A0 input
will not be perfectly zero volts when PTT is not pressed.
73 Allard PE1NWL
toggle quoted message
Show quoted text
On Wed, June 14, 2017 15:33, qrper72@... wrote: I've put the 10k pull-down register at the "A0" and maybe the problem
seems to be fixed now. This issue caused by the threshold problem?
ja9mat Hidehiko.
|
Given your code, Allard, I'm surprised he needs a dropping resistor on that line.
Jack, W8TEE
From: Allard PE1NWL <pe1nwl@...>
To: [email protected]
Sent: Wednesday, June 14, 2017 10:51 AM
Subject: Re: [BITX20] CW Offset does NOT work on Raduino_v1.14.1 for BITX40
Hi Hidehiko-san Yes you could be right. Another possible cause could be poor grounding of BITX main board and Raduino board. If Raduino board is not grounded perfectly, some small potential difference may exist between BITX board and Raduino. As a result, A0 input will not be perfectly zero volts when PTT is not pressed. 73 Allard PE1NWL On Wed, June 14, 2017 15:33, qrper72@... wrote: > I've put the 10k pull-down register at the "A0" and maybe the problem > seems to be fixed now. This issue caused by the threshold problem? > > ja9mat Hidehiko. >
|
The code is doing a digitalRead(PTT_SENSE), I'd expect the threshold on the nano to be somewhere between 1 and 2 volts
The problem is that when the +12v TX line into the LM78L05 at U3 goes to zero volts, the output of U3 remains up around 5v because there is no place for the charge on C150 and C151 to go. ?The pulldown resistor will eventually discharge those caps. ?The RC time constant of 0.2uF*20k = 4 milliseconds, which should be fast enough. ?I'd put that new 10k resistor from U3-1 to ground, rather than from A0 to ground, the latter case would reduce the maximum voltage into the Nano A0 pin down from 5v to 2.5v. ?This also reduces the RC time constant to 2 milliseconds.
?
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 08:28 am, Jack Purdum wrote:
Given your code, Allard, I'm surprised he needs a dropping resistor on that line.
?
?
|
I must have misunderstood. If the internal pull-up resistors on a pin are activated, doesn't that impact the voltage seen on the pin? Also, if the code does a digitalRead(), the documentation says it returns HIGH or LOW. Is it true a digital read of a pin can return an "in-between" value with the internal pull-ups active??
Jack, W8TEE
From: Jerry Gaffke via Groups.Io <jgaffke@...>
To: [email protected]
Sent: Wednesday, June 14, 2017 12:07 PM
Subject: Re: [BITX20] CW Offset does NOT work on Raduino_v1.14.1 for BITX40
The code is doing a digitalRead(PTT_SENSE), I'd expect the threshold on the nano to be somewhere between 1 and 2 volts The problem is that when the +12v TX line into the LM78L05 at U3 goes to zero volts, the output of U3 remains up around 5v because there is no place for the charge on C150 and C151 to go. ?The pulldown resistor will eventually discharge those caps. ?The RC time constant of 0.2uF*20k = 4 milliseconds, which should be fast enough. ?I'd put that new 10k resistor from U3-1 to ground, rather than from A0 to ground, the latter case would reduce the maximum voltage into the Nano A0 pin down from 5v to 2.5v. ?This also reduces the RC time constant to 2 milliseconds. ? On Wed, Jun 14, 2017 at 08:28 am, Jack Purdum wrote:
Given your code, Allard, I'm surprised he needs a dropping resistor on that line.
?
?
|
Hmm, yes you may be right that in some cases the 7805 output hasn't
dropped sufficiently to represent a digital LOW.
Another way to circumvent this situation might be to define A0 as an
analog input and define a higher threshold for a digital LOW/HIGH. What do
you think?
73 Allard PE1NWL
toggle quoted message
Show quoted text
On Wed, June 14, 2017 18:07, Jerry Gaffke via Groups.Io wrote: The code is doing a digitalRead(PTT_SENSE), I'd expect the threshold on
the nano to be somewhere between 1 and 2 volts
The problem is that when the +12v TX line into the LM78L05 at U3 goes to
zero volts, the output of U3 remains up around 5v because there is no
place for the charge on C150 and C151 to go. ??The pulldown resistor will
eventually discharge those caps. ??The RC time constant of 0.2uF*20k = 4
milliseconds, which should be fast enough. ??I'd put that new 10k resistor
from U3-1 to ground, rather than from A0 to ground, the latter case would
reduce the maximum voltage into the Nano A0 pin down from 5v to 2.5v.
??This also reduces the RC time constant to 2 milliseconds.
??
On Wed, Jun 14, 2017 at 08:28 am, Jack Purdum wrote:
Given your code, Allard, I'm surprised he needs a dropping resistor on
that line.
??
|
Funny you should ask.. As I was reading Jack's post I was thinking, it's
an analog line, use it like one.
Vince.
toggle quoted message
Show quoted text
Hmm, yes you may be right that in some cases the 7805 output hasn't
dropped sufficiently to represent a digital LOW.
Another way to circumvent this situation might be to define A0 as an
analog input and define a higher threshold for a digital LOW/HIGH. What do
you think?
73 Allard PE1NWL
On Wed, June 14, 2017 18:07, Jerry Gaffke via Groups.Io wrote:
The code is doing a digitalRead(PTT_SENSE), I'd expect the threshold on
the nano to be somewhere between 1 and 2 volts
The problem is that when the +12v TX line into the LM78L05 at U3 goes to
zero volts, the output of U3 remains up around 5v because there is no
place for the charge on C150 and C151 to go. ??The pulldown resistor
will
eventually discharge those caps. ??The RC time constant of 0.2uF*20k = 4
milliseconds, which should be fast enough. ??I'd put that new 10k
resistor
from U3-1 to ground, rather than from A0 to ground, the latter case
would
reduce the maximum voltage into the Nano A0 pin down from 5v to 2.5v.
??This also reduces the RC time constant to 2 milliseconds.
??
On Wed, Jun 14, 2017 at 08:28 am, Jack Purdum wrote:
Given your code, Allard, I'm surprised he needs a dropping resistor on
that line.
??
--
Michigan VHF Corp.
|
Allard:
I'd give this a try.
Jack, W8TEE
From: Allard PE1NWL <pe1nwl@...>
To: [email protected]
Sent: Wednesday, June 14, 2017 1:09 PM
Subject: Re: [BITX20] CW Offset does NOT work on Raduino_v1.14.1 for BITX40
Hmm, yes you may be right that in some cases the 7805 output hasn't
dropped sufficiently to represent a digital LOW.
Another way to circumvent this situation might be to define A0 as an
analog input and define a higher threshold for a digital LOW/HIGH. What do
you think?
73 Allard PE1NWL
On Wed, June 14, 2017 18:07, Jerry Gaffke via Groups.Io wrote:
> The code is doing a digitalRead(PTT_SENSE), I'd expect the threshold on
> the nano to be somewhere between 1 and 2 volts
> The problem is that when the +12v TX line into the LM78L05 at U3 goes to
> zero volts, the output of U3 remains up around 5v because there is no
> place for the charge on C150 and C151 to go. ??The pulldown resistor will
> eventually discharge those caps. ??The RC time constant of 0.2uF*20k = 4
> milliseconds, which should be fast enough. ??I'd put that new 10k resistor
> from U3-1 to ground, rather than from A0 to ground, the latter case would
> reduce the maximum voltage into the Nano A0 pin down from 5v to 2.5v.
> ??This also reduces the RC time constant to 2 milliseconds.
> ??
>
> On Wed, Jun 14, 2017 at 08:28 am, Jack Purdum wrote:
>
>>
>> Given your code, Allard, I'm surprised he needs a dropping resistor on
>> that line.
>> ??
>>
>
|
Allard:
> Another way to circumvent this situation might be to define A0 as an analog input?
I think I'd add a pulldown resistor somehow, make the behavior vaguely defined.
As you have found, there are significant differences between rigs on how fast that charge leaks away.
Using the extra complication of an ADC to read a digital input pains my soul, and is not guaranteed to fix the problem.
Jack:
> If the internal pull-up resistors on a pin are activated, doesn't that impact the voltage seen on the pin?
I haven't looked at the Nano documentation for the effective resistance of their pullup/pulldown scheme,
but it's most likely well under 100k.
You could program the Nano to give A0 a pulldown and use that instead of my $0.01 10k resistor.
Would probably work, and would probably happen within 10ms or so.
> ?Also, if the code does a?digitalRead(), the documentation says it returns HIGH or LOW.
> ?Is it true a digital read of a pin can return an "in-between" value with the internal pull-ups active??
You either get a 0 or a 1 back when you read a pin with digitalRead(). ?If the voltage at the pin happens to be midrange the result is indeterminate, but the result will be either a 1 or a 0.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 10:08 am, Allard PE1NWL wrote:
Another way to circumvent this situation might be to define A0 as an
analog input and define a higher threshold for a digital LOW/HIGH. What do
you think?
?
|
OK, now I have checked the ATmega328P documentation.
The digital IO pins have a pull-up only, guaranteed to be between 30k and 60k, which can be disabled.
The ATmega328P does not have an option to make it a pull-down like some processors do.
So me, I'd add an extra 10k resistor from the LM78L05 output to ground, and continue to use A0 as a digital input.
If you scope the LM78L05 output and find that when 12v TX is shut down the A0 line immediately falls from 5v to 3v, and then slowly decays toward zero, perhaps the analog read trick would work. ? But not something I would do.
John Smith:
> I already have right channel PTT circuit in the form of the Fldigi CW keying circuit
If all I wanted was a simple way to drive the Bitx40 for CW, I'd add an extra 10k resistor and skip all the FLdigi stuff.
If you already have the FLdigi stuff somehow dealing with this PTT thing, that's great.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 10:45 am, Jerry Gaffke wrote:
I haven't looked at the Nano documentation for the effective resistance of their pullup/pulldown scheme,
but it's most likely well under 100k.
?
|
Thanks Allard, PE1NWL.
OK all, Yes because my situation is still in experiment ballkack.
So the perfect grounding and casing to avoid the stray capacity will be good.
ja9mat Hidehiko.
|
Now Jerry, how'd you know I was going to come over here and read this thread about CW?
For my digital upgrade I will be using serial PTT rig control in Fldigi. A year ago I made the CW keying circuit found in the Fldigi Manual and added a relay module for solid sounding keying of the radio. And I hope this new control scheme work equally as well. Check out this link for the VOX style CW keying circuit. It does require a particular balance of combined volume controls (side tone on radio out, received signal, output tone on right....).
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 10:58 am, Jerry Gaffke wrote:
John Smith:
> I already have right channel PTT circuit in the form of the Fldigi CW keying circuit
If all I wanted was a simple way to drive the Bitx40 for CW, I'd add an extra 10k resistor and skip all the FLdigi stuff.
If you already have the FLdigi stuff somehow dealing with this PTT thing, that's great.
Jerry, KE7ER
?
|
All, many thanks for all of your thoughts and comments. I tend to agree with Jerry that probably the easiest and most reliable way to improve the situation would be to just add a 10K resistor at the ouput of the 7805. I believe the most convenient place for the resistor would be to solder it across C150. I haven't tried this yet, what do you think? Couldn't it have some effect on the IRF150 bias setting? Do we perhaps need to check/realign the bias after adding the 10K resistor?  73 Allard PE1NWL
|
Adding a 10k load to the 78L05 will not affect bias to the IRF510 gate.
The 78L05 can supply between 0 and 100ma while maintaining a constant 5v at the output.
That 10k resistor will draw an extra ?5v/10k = 0.5ma.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 03:37 pm, Allard PE1NWL wrote:
Couldn't it have some effect on the IRF150 bias setting? Do we perhaps need to check/realign the bias after adding the 10K resistor?
?
|
The reason to go with the output of the 78L05 was because it swings between 0 and 5v, which is the range acceptable to the Nano.
But looks like we want two resistors to do that properly.
Perhaps the 10k series resistor ?from 78L05 to the Nano is not absolutely needed,
but we've seen enough blown Nano IO pins here that I would definitely include it in case
the 12v TX line goes high but the Raduino is not powered up for some reason.
Might be more straightforward to instead sense the 12v TX line using a resistive divider of 20k and 10k in series,?
with 12v TX into the top of the 20k, bottom of the 10k to ground.
The join between the 20k and 10k goes to the Nano's A0 input.
The 78L05 is not involved.
So the Nano sees 12v * 10k/(20k+10k) = 4v when TX is at 12v, otherwise 0v.
Your choice, either this or the two 10k's on the 78L05, either should work just fine.
?
Jerry
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 03:53 pm, Jerry Gaffke wrote:
Adding a 10k load to the 78L05 will not affect bias to the IRF510 gate.
The 78L05 can supply between 0 and 100ma while maintaining a constant 5v at the output.
That 10k resistor will draw an extra ?5v/10k = 0.5ma.
Jerry, KE7ER
?
|
Yes, sensing the 12V from the TX/RX line is also an option.
But I'm still a bit reluctant as I'm not completely convinced yet that we
anyway need a two-resistor solution.
The elegantly simple one-resistor method has worked fine in all cases so
far (at least that I know of). So far only one builder (ja9mat Hidehiko)
reported that in his case it didn't work reliably.
As he mentioned that his radio was still in an experimental setup
situation (perhaps with long or loose wires, poor ground connections,
etc.), there is still a possibility that his problem will disappear once
everything is properly built together and wired correctly.
So before overreacting and advising everyone to add extra resistors I'd
like to hear if there are any other builders having similar issues, and if
Hidehiko-san's problem disappeared after he has properly wired everything.
73 Allard PE1NWL
|
Sounds good.
The reason the Bix40 is so popular is that it gives great performance with?
very few parts. ?Your approach is in line with that philosophy.
Keep it simple wherever we can.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jun 14, 2017 at 04:36 pm, Allard PE1NWL wrote:
So before overreacting and advising everyone to add extra resistors I'd
like to hear if there are any other builders having similar issues
?
|
Hidehiko Komachi - JA9MAT
Allard PE1NWL