Date   
Re: AGC circuit to try? new topic?

ajparent1/kb1gmx
 

Correct you cannot filter IMD with wide filters unless its outside the crystal filter bandwidth.
For  bitx that is about 2.5khz give or take and for ubitx that's 15khz (45mhz filer is for FM, wide).

However...  Limiting the audio bandwidth to the modulator helps.  A good compressor
can help.  Isolating the tx audio from the RX audio may help.  Why?   The RX preamp
Q70 even unpowered the base emitter junction can look like a diode across the mic amp
(Q6) output teh clipping level is about .5 to .6V P-P and it would be noticeable.  Inserting
an audio low pass filter between C222 and C63.  Then you get a free audio clipper with
lower IMD at that point.  Suggested cutoff is around 3.5Khz.

If the N6DT pop stopper is used that would isolate the RX and TX audio paths, worth a
try and it could help more than the pop.

The next step is insuring the drive from the modulator to the succeeding stages does not 
push the DBMs into compression.  Generally that means staying 10db or more below the
LO for each mixer.  That helps both inband (IMD) and out of band (spectrum).  You trade
that for less power out and lower average power.  You need very little gain from the
modulator to the next mixer and the amp plus filter should have a net gain of maybe
only a few DB otherwise the mixers are being hammered and will compress and
you will hear that.  

Now filtering from the last mixer to the power chain does help the out of band especially out
of ham band products.  The fewer undesired products you feed the amplifier chain with the 
better the result.  The key is only out power to the RF you want even weak spurs and splatter
is lost power.  

Then you have to deal with the amplifier chain itself.  Every stage must have enough current
to insure you do not hit even -1db compression.    Note the output matching may not be
optimal and can have the device in voltage or current limiting.  Underloading and overloading
can be troublesome. That's tricky as what may work well for 80-20M might be really poor at
10M.  If you do not need above 20M then optimize for that, you should be able to get a few DB
improvement.  For example the low pass filter could then have a 3db corner of 16mhz.  The
output low pass filter can be  then for split for 80 through 20M (or 40M). 

A 24 volt battery in the field... 2 8AH gells will do that nicely or 6 series LiPo.  Both cases give
more watt-hours with less current.

However for the ubitx that means protecting the most everything before the pre-driver from 
more than 12V. The TBA2822 seems fragile.

I do know that with the monobanders I have for 80/40/20/15/10/6M that they all perform 
better as they are optimum fot the band rather than trying to be everything for ever where.
Multiband with wide tuning is not easy ad requires some compromise to accomplish at that
uBitx does pretty well.


Allison

Re: W0EB/W2CTX Latest uBITX software release #ubitx

Jim Sheldon
 

Chris - If the KD8CEC software is working fine for you you should probably continue to use it. 

To use our latest software, you would have to either ADD a key jack and connect the tip terminal of that jack to the orange wire from the Raduino's P1 plug (PTT) line (you can easily just connect it to the PTT (ring) terminal on the MIC connector if you are using the provided 1/8" stereo jack for the microphone jack.  If not, connect the hand key jack's "TIP" contact to whatever contact on the microphone jack you connected to the orange PTT wire.  If you don't want to add another jack, simply wire your key with a plug to match your microphone plug.  The key would connect to the ring and round terminals on that plug with the TIP (microphone wire) left open. Simple as that.

Jim - W0EB

Re: W0EB/W2CTX Latest uBITX software release #ubitx

Jim Sheldon
 

The only place it's hosted presently is in the "Files" section of this group.

W0EB

Re: TDA2822 Audio problem #ubitx #tda2822

Jerry Gaffke
 

Sent an email to Oddwires through their contacts webpage, got a reply within minutes:
  
Hi Jerry, thanks for your inquiry. Board is configurable with a solder junction to 5 or 3.3V. Let me know if you need any more info. Thanks, Ian.


Very nice to have such a quick and concise response.
And he gave his direct line phone number with the response.
I'm sold.

Jerry





On Tue, Apr 10, 2018 at 10:06 am, Jerry Gaffke wrote:
The USB-to-UART cables sold for use by Arduino hobbyists generally have 0-5v logic levels,
not RS232.   That thing on Oddwires I suggested in a previous post:
    http://www.oddwires.com/cp2102-serial-adapter-module-usb-to-rs232-with-jumper-wires/
claims to be "USB to RS232", but I seriously doubt it.  All I see is the SiLabs chip on it,
no level translators.  Could be 5v logic levels, could be 3.3v, or could be configurable.
Claims to be suitable for use with both the Rasberry Pi (3.3v) or Arduino (5v).
These are nasty hardware details that everyone in the Arduino universe is oblivious to.

Re: W0EB/W2CTX Latest uBITX software release #ubitx

youainti@...
 


 Is this software hosted somewhere such as github or bitbucket?

Will   KG7YQB

Re: AGC circuit to try?

Jerry Gaffke
 

By this:
>  Yes it would seem that way but it does work very well.

sounds like you mean that a PIN diode bridge attenuator should have no problem 
with the strong RF signals it might encounter toward the front end of the uBitx.
That's good to know.

I do think your gain control scheme using the 1n400x parts would work fine.
Though one BAP64Q gives more dynamic range, and can be all in one spot.

And is just plain cool.
It's puzzling to me that they can maintain a 50 ohm impedance so well.

Jerry


On Tue, Apr 10, 2018 at 11:16 am, ajparent1 wrote:
Yes it would seem that way but it does work very well.

Re: AGC circuit to try?

ajparent1/kb1gmx
 

Yes it would seem that way but it does work very well.

Using the bridged T design from Avago apnote with common 1n4007 worked very well for me
at 6M and the environment was "must still work with 100mW from the antenna".    It is what
you get with a very local ham that can't run a repeater contact at less than 1KW.  Never
minding BC stations 4 miles away, and 7 miles the other way the local Boston TV nexus. 
Bridged T has the feature of near constant 50 ohms and even with 1n4007 50DB range
was easily had.  That went in my 6M SSB contest radio that I called Crunchproof.

Another part that works well is the DBM,  Apply small DC (less than 20mA) to the IF port and
the loss though it is small. Reduce the applied current and 40-60DB of attenuation is had
(depends on balance).

The closer to the antenna the better the result, however it has to be out of the TX path.

With the uBitx/Bitx that easiest to do in the amps. as they are split for RX and TX.  The
switching is already done.

Allison

Re: AGC circuit to try?

ajparent1/kb1gmx
 

That's why the CDS cell and the #49 bulb  worked so well the attack and decay are limited
by filament heating and cooling.

I'm aware of all the various opto isolator, that parts bin has about 15 different parts.
The Jfet based parts might work well.  I've used the 4n25 and 26 parts in attenuators as well.

FYI a common Jfet works well too, generally as a series device.

Allison

Re: AGC circuit to try?

Jerry Gaffke
 

Any an attenuator used up front at the uBitx antenna port 
will be dealing with a whole world's worth of RF, 
so any non-linearity in the attenuator could create IMD products 
due to a strong local AM or FM or TV or hairdryer signal.

The situation gets slightly better after the 30mhz low pass filter at L1,2,3,4
Much better after the 45mhz crystal filter.
Much much better after the 12mhz filter.

Though as Tim has pointed out, the 12mhz filter might
be especially sensitive to variations in impedance.

Pick your poison.

Jerry


On Tue, Apr 10, 2018 at 10:36 am, Jerry Gaffke wrote:
The half-life of the charge carriers in a PIN diode is only around a microsecond,
so won't work at all at audio frequencies, must be prior to D5 pin 3.

Most of the AGC schemes we see here suppress the audio after the audio pre-amp.
And they work well enough.
But we should get more dynamic range (be able to receive strong signals) if the AGC
attenuator is somewhere between C50 and the antenna port.

Re: AGC circuit to try?

Jerry Gaffke
 

The half-life of the charge carriers in a PIN diode is only around a microsecond,
so won't work at all at audio frequencies, must be prior to D5 pin 3.

Most of the AGC schemes we see here suppress the audio after the audio pre-amp.
And they work well enough.
But we should get more dynamic range (be able to receive strong signals) if the AGC
attenuator is somewhere between C50 and the antenna port.



On Tue, Apr 10, 2018 at 10:19 am, Tim Gorman wrote:
So you think the BAP64 would be better placed before the audio pre-amp
rather than in the IF? Before C50 maybe?

Re: AGC circuit to try?

Tim Gorman
 

So you think the BAP64 would be better placed before the audio pre-amp
rather than in the IF? Before C50 maybe?

tim ab0wr

On Tue, 10 Apr 2018 09:23:34 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

Yup.
Getting AGC right however you do it is not trivial.
And not something I am able to fully analyze.
For example, Henning Weddig pointed out some time ago that
delays through the crystal filter can cause instability when you
close the loop.

He also pointed out that it is our audio pre-amp that limits the 
total dynamic range of the rig, and thus an attenuator prior to 
that pre-amp is preferred if you might be dealing with really strong
signals.

On Tue, Apr 10, 2018 at 09:10 am, Sandy T wrote:


At least the BAP64Q at Q70 would only cost $0.56 but proper biasing
to work "as expected AGC" might be a trick.

Re: AGC circuit to try?

Tim Gorman
 

Jerry,

I've ordered four BAP64Q units for when my second ubitx gets here. I'm
going to stick it in the rf path somewhere as we discussed earlier.
Maybe between R17 and C12?

tim ab0wr

On Tue, 10 Apr 2018 08:05:57 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

It was just a thought, getting an AGC scheme to work well is not
trivial. I'm still thinking a BAP64Q at R17 might be best.

Base of Q70 could work.  It might do better on the other side of C50,
where that weak incoming audio is centered on ground. 
Though as we have seen here, have to think about what happens when
you have strong audio from the mike amp during transmit, potentially
going through the intrinsic diode in that FET. 

You're on your own!

Jerry

On Tue, Apr 10, 2018 at 07:32 am, Tim Gorman wrote:


I have mine connected in the receive line before the first mixer.
Your idea might work better. I assume you would put it at the base
of Q70 in the ubitx?

Re: Variation on Ian's KD8CEC uBitx software (based on his 1.04 release for now) and ATU sketch. #ubitx

Tim Gorman
 

I think you have it nailed. It's why the mjkdz backpacks need a
different software initialization. Based on what I have read the mjkdz
backpacks wire to the lcd pinouts differently than the other backpacks.
The software apparently needs to know about that difference.

tim ab0wre

On Tue, 10 Apr 2018 09:02:23 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

I think that list of pins describes the interface between the PCF8574
i2c chip on the backpack into the parallel interface of the LCD
display. Nothing to do with pins on our Nano, all the Nano sees is
the SDA and SCL pins for the i2c bus that is shared between the i2c
display and the si5351.

The parallel interface into an lcd requires 4 parallel data lines,
and they are used for both read and write.

On Tue, Apr 10, 2018 at 08:56 am, Jack Purdum wrote:


It appears that most of those extra pins are for read/write
capability, and most of the BITX displays don't use software that
reads the display. Also, since most of those use pins 2 and 3, they
would kill the ability to use external interrupts on the rig. Since
I use external interrupts for tuning, such an I2C display for my
projects would be a bad choice. They probably are fine for projects
that don't use external interrupts. Personally, I'd look for
displays that only require the device address and the row/column
specs.

Re: TDA2822 Audio problem #ubitx #tda2822

Jerry Gaffke
 

The USB-to-UART cables sold for use by Arduino hobbyists generally have 0-5v logic levels,
not RS232.   That thing on Oddwires I suggested in a previous post:
    http://www.oddwires.com/cp2102-serial-adapter-module-usb-to-rs232-with-jumper-wires/
claims to be "USB to RS232", but I seriously doubt it.  All I see is the SiLabs chip on it,
no level translators.  Could be 5v logic levels, could be 3.3v, or could be configurable.
Claims to be suitable for use with both the Rasberry Pi (3.3v) or Arduino (5v).
These are nasty hardware details that everyone in the Arduino universe is oblivious to.

I have an old TTL-232R-3V3 FTDI cable I use,
    http://www.ftdichip.com/Support/Documents/DataSheets/Cables/DS_TTL-232R_CABLES.pdf
since 3.3v and less is what most of electronics have gone for the last couple decades.
For use with the Nano, we should have the 5v logic level version, the TTL-232R-5v
FTDI is kind of the industry standard, note that the true Arduinos have the same FTDI chips.

Better yet, here's a newer FTDI cable that can be configured for either 5v or 3.3v:
    http://www.ftdichip.com/Support/Documents/DataSheets/Cables/DS_TTL-232RG_CABLES.pdf

Those FTDI cables are around $20.
FTDI clones and alternatives are now out there for $2 or $3.
If that's a real SiLabs chip on the Oddwires device it probably works fine,
though Oddwires hasn't bothered to fully describe what the logic levels are.    

Jerry, KE7ER


On Tue, Apr 10, 2018 at 09:18 am, Arv Evans wrote:
Beauty of using the built-in UART and built-in USB interface on an Arduino NANO is
that you don't have to mess with voltage level conversion between RS-232 and TTL.
This is also convenient because many modern PC do not have built-in RS-232 ports.

Re: Variation on Ian's KD8CEC uBitx software (based on his 1.04 release for now) and ATU sketch. #ubitx

Tim Gorman
 

According to your second link the higher priced unit from oddwires uses
lcd(0x20, 4, 5, 6, 0, 1, 2, 3, 7, NEGATIVE) as it is an "mjkdz"unit.

The lower priced one probably uses the more common
lcd(0x20,2,1,0,4,5,6,7,3,POSITIVE).

This site also has a nice discussion:
forum.arduino.cc/index.php?topic=157817.0

I'm going to check the backpacks I threw away to see if they are mjkdz
units!

I agree we should standarize on the cheaper ones! They seem to be more
standard everywhere.

It looks as if the LiquidCrystal_I2C lcd(x,y,z.....) function is to
create an object (whatever that means). The lcd.begin(16,2) is to
initialize the lcd itself.

So the lines are doing two different things.

The w0eb et al. software does just that:
LiquidCrystal_I2C lcd(0x27, 2, 1, 0, 4, 5, 6, 7, 3, POSITIVE);
lcd.begin(16,2);

I can't see where the test software from john does that. Maybe I'm
missing something.

tim ab0wr


On Tue, 10 Apr 2018 08:45:56 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

You might email or call Oddwires, see what they say is different
about the two. I think the more expensive one just has a less common
parallel pin-out from PCF8574 i2c chip into the lcd display.
They have to explicitly set the pins, whereas the cheap one uses the
default pins.

Enough of these being sold that there are now clones of clones of
clones. Choosing one because the board layout looks similar to one
that failed is probably not a good plan.  Better to just choose a
vendor you trust.

There are lots of "how I done it" webpages regarding these i2c
displays, but nothing I've found really describes all the gotcha's in
one place. Weird that the libraries don't include a decent Readme.

Here's an interesting page that better describes all those args
to some of the Liquid_Crystal_I2C libraries: 
    http://masteringarduino.blogspot.com/2013/12/lcdi2c.html
He never mentions using a scanner to figure out what the i2c address
is. He shows two primary ways those parallel pins might get stirred,
the fact that backlight could be POSITIVE or NEGATIVE, and two of the
16 or so i2c device addresses you might run into:

###### Some example of initialization codes (try them to find one
that suitable for your device), instantiated on object *lcd* :

* LiquidCrystal_I2C *lcd* (0x20,2,1,0,4,5,6,7,3,POSITIVE);
* LiquidCrystal_I2C *lcd* (0x27,2,1,0,4,5,6,7,3,POSITIVE);
* LiquidCrystal_I2C *lcd* (0x20,4,5,6,0,1,2,3,7,NEGATIVE);

This page shows a variant of the first entry above, the last two args
giving the backlight pin and polarity are not included:
    // addr, en,rw,rs,d4,d5,d6,d7 LiquidCrystal_I2C lcd ( 0x27 , 2 ,
1 , 0 , 4 , 5 , 6 , 7 ); // Set the LCD I2C address Anyways, stirring
those parallel pins is something to try for those with apparently
dead i2c displays. Hopefully the Arduino world is slowly coalescing
into one version of parallel pinout, since users with i2c backpacks
that don't conform to the majority of the documentation will be
wanting their money back.

If you run into an i2c lcd with the "wrong" polarity on the
backlight, it will probably just stay lit by default if you don't
mess with the backlight in your code.

Here's a wiki page describing some of the various lcd interfaces you
can run into that are supported by the Malpartida library, including
a brief mention of the i2c interface:
https://bitbucket.org/fmalpartida/new-liquidcrystal/wiki/schematics#!configurable-i2c-address-modules

And this old post describes confusion regarding whether i2c device
addresses are 7 or 8 bits: https://groups.io/g/BITX20/message/39966

Hopefully we can all settle on one backpack type and one library.
For US customers at least, I'd tend toward the cheaper backpack from
oddwires, which also comes mounted on an LCD if you wish.

Jerry

On Tue, Apr 10, 2018 at 07:19 am, Tim Gorman wrote:


The $2 backpack looks like the ones I have. I might give the $5 one
a try and see how it works. The code they give for it uses the pin
assignments although with a negative type instead of positive
type.

Re: Variation on Ian's KD8CEC uBitx software (based on his 1.04 release for now) and ATU sketch. #ubitx

Jack, W8TEE
 

I don't know, as I've never had one that used more than 3 arguments in the constructor.

Jack, W8TEE


On Tuesday, April 10, 2018, 12:52:45 PM EDT, K9HZ <bill@...> wrote:


I believe this is correct, or at least what my digging into the libraries shows. 


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Apr 10, 2018, at 11:02 AM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

I think that list of pins describes the interface between the PCF8574 i2c chip on the backpack
into the parallel interface of the LCD display.
Nothing to do with pins on our Nano, all the Nano sees is the SDA and SCL pins
for the i2c bus that is shared between the i2c display and the si5351.

The parallel interface into an lcd requires 4 parallel data lines, and they are used for both read and write.



On Tue, Apr 10, 2018 at 08:56 am, Jack Purdum wrote:
It appears that most of those extra pins are for read/write capability, and most of the BITX displays don't use software that reads the display. Also, since most of those use pins 2 and 3, they would kill the ability to use external interrupts on the rig. Since I use external interrupts for tuning, such an I2C display for my projects would be a bad choice. They probably are fine for projects that don't use external interrupts. Personally, I'd look for displays that only require the device address and the row/column specs.

Re: Variation on Ian's KD8CEC uBitx software (based on his 1.04 release for now) and ATU sketch. #ubitx

K9HZ <bill@...>
 

I believe this is correct, or at least what my digging into the libraries shows. 


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Apr 10, 2018, at 11:02 AM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

I think that list of pins describes the interface between the PCF8574 i2c chip on the backpack
into the parallel interface of the LCD display.
Nothing to do with pins on our Nano, all the Nano sees is the SDA and SCL pins
for the i2c bus that is shared between the i2c display and the si5351.

The parallel interface into an lcd requires 4 parallel data lines, and they are used for both read and write.



On Tue, Apr 10, 2018 at 08:56 am, Jack Purdum wrote:
It appears that most of those extra pins are for read/write capability, and most of the BITX displays don't use software that reads the display. Also, since most of those use pins 2 and 3, they would kill the ability to use external interrupts on the rig. Since I use external interrupts for tuning, such an I2C display for my projects would be a bad choice. They probably are fine for projects that don't use external interrupts. Personally, I'd look for displays that only require the device address and the row/column specs.

Re: Raduino no longer recognized by PC

Jack, W8TEE
 

As a general rule, the real thing has "Arduino Nano" on the body of the device. (See the Mouser ad.) The clones do not. I buy Nano clones 10 at a time and have always been able to get them to work, even though occasionally you have to work around an old bootloader or FDTI chip.

Jack, W8TEE


On Tuesday, April 10, 2018, 12:05:25 PM EDT, Mike Bryce <prosolar@...> wrote:


Jack and the group,

how does one tell the difference between a Chinese clone and the real McCoy? I mean other than buying four of them for $6 on eBay.

and where does one fine the genuine article?

Inquiring minds and all.



On Apr 10, 2018, at 11:30 AM, Jack Purdum via Groups.Io <jjpurdum@...> wrote:


Another common problem is a faulty bootloader. Sometimes you need to change the processor selection (Tools --> Processor --> ATmega328P (old bootloader)). Also, many of the clones use the CH340G chip for the FTDI. Usually you can google "Arduino CH340G driver download" and find one for your op system:


Re: Raduino no longer recognized by PC

Arv Evans
 

Mike, WB8VGE

Good question.  Major difference is related to the serial-TTL/USB conversion IC.  In original
NANO boards this was an FT240RL device.  In subsequent built-in-China (and Philippines,
Japan, Mexico, etc.) boards this is the much less expensive CH340 device.  Functionally
there is no difference between the FT240RL and the CH340 devices, although some
DOS/Windows users have reported driver program problems when switching between
NANO boards that have the FT240RL versus the CH340.  Linux users will find that their USB
ports automatically work with either original or clone boards.


The so-called clones have been sold and used by many with apparently no compromise in
performance  The Arduino IDE works well with either original or clone boards.  Cost is the
main differentiator between original and clone purchases. 

Arv  K7HKL
_._


On Tue, Apr 10, 2018 at 10:05 AM, Mike Bryce <prosolar@...> wrote:
Jack and the group,

how does one tell the difference between a Chinese clone and the real McCoy? I mean other than buying four of them for $6 on eBay.

and where does one fine the genuine article?

Inquiring minds and all.



On Apr 10, 2018, at 11:30 AM, Jack Purdum via Groups.Io <jjpurdum@...> wrote:


Another common problem is a faulty bootloader. Sometimes you need to change the processor selection (Tools --> Processor --> ATmega328P (old bootloader)). Also, many of the clones use the CH340G chip for the FTDI. Usually you can google "Arduino CH340G driver download" and find one for your op system:



Re: Fw: Re: [BITX20] AGC circuit to try?

Jack, W8TEE
 

OK...I'm a software guy and should learn to keep my mouth shut on the EE side of things!

Jack, W8TEE

On Tuesday, April 10, 2018, 12:21:18 PM EDT, Sandy T <windy10605@...> wrote:


Jack,

Phototransistors are great/cheap for isolation and switching applications ......as with your keyer. Not so much for symmetrical and variable output resistance, tracking an AC input.

73 Kees K5BCQ