¿ªÔÆÌåÓý

  1. BITX20
  2. Messages
Expanded Topics Messages Polls
Date
Re: Follow-up Antuino question

 

H&P is not a phase locked loop in the sense that we usually consider it.??
There is no "lock" situation involved.? As Farhan mentioned, H&P continuously?
detects whether one edge of an HF cycle leads or follows an edge of an LF?
reference cycle.? A single D-type FF is used as a comparator to generate a steering?
voltage with a relatively long time constant until the next edge comparison happens.??
This occurs?at the LF rate.? The FF is always flipping back and forth at the LF
rate unless an edge match keeps it from flipping or flopping.? One of the "Q"?
outputs of the FF drives the steering voltage.? If it drives off frequency instead?
of on-frequency you only need to use the opposite 'Q' output to change the?
steering voltage so it tunes the other direction.? This fact that the correction is only?
applied at the?LF rate is what makes it also control the tuning step size.

There are several other ways to implement H&P.? I have not mentioned them?
here but they are covered in the QRP-labs web page section on H&P stabilizers.



It is also possible to count VFO frequency with a PIC or pre-scaled arduino input
and use the processor to generate scaled steering voltage for either stabilization?
and/or tuning of the VFO frequency.? While this is not H&P it is also relatively?
easy to implement if you can handle a bit of programming.? It is another way?
to get linear VFO output over the VFO tuning range, or over some translated?
range.? If you measure both BFO and VFO the microprocessor can be made?
to display actual frequency or actual frequency with calculated offsets for?
various transmission or receive modes.

Arv
_._


On Sat, Nov 14, 2020 at 7:08 PM Jerry Gaffke via <jgaffke=[email protected]> wrote:

This may be an alternate way of looking at what you are proposing.


I suspect you are describing a phase-locked-loop.
And that yes, you have tried it.
There's a phase-frequency-detector inside the Si5351.

The VCO in the Si5351 can be pulled anywhere between 600 and 900 mhz
(and quite a bit beyond, actually).
But if instead of that VCO we have a VFO that is first manually tuned very near the
target frequency, the error detection will have far less work to do.?
And the output will have far less jitter.

Jerry, KE7ER



On Sat, Nov 14, 2020 at 05:39 PM, Ashhar Farhan wrote:

I was expecting Arv to write in about H&P schemes. Between him and Hans, there is all we know about how this scheme works.
The challenge of Huff and Puff is that it needs a very stable oscillator to begin with. It should have very minimal wobble (short term drift). The drift between two correction pulses should be less than its step size. That is not difficult to obtain.
The inherent drawback of H&P is that the oscillator is always being pushed up or down from the central frequency between two cycles. This can have a disastrous effect on digital modes.
There is probably another way to fix this. It is to use two flip flops as parallel mixers, each is fed the clock pulse that is 90 degrees out of phase from the other (imagine it to be a phasing receiver). Noe, the combined DC output of the two will indicate if we have drifted above, below or stayed on the frequency. I haven't tried this, but I am guessing that this will work. In the professional literature it is called a phase frequency detector.
- f
Re: Follow-up Antuino question

 

If you have a uBitx and wish to experiment with sine waves instead of square,
perhaps add a low pass filter to each of the three Si5351 clocks.

The VFO at CLK2 would need to select from multiple filters if targeting?
more than one band.?

I'm assuming we select USB vs LSB by moving the BFO a few khz
instead of moving the second LO at CLK1 by 11+11=22mhz.

I'm not yet convinced that a sine wave will give significantly better
reception than a square wave

Jerry, KE7ER


On Sun, Nov 15, 2020 at 11:23 AM, Arv Evans wrote:
My comments were in response to someone's complaint about Si5351a outputs?
being square waves.? With H&P the VFO can be a linear output.??
?
Re: Follow-up Antuino question

 

Farhan

It is possible to reverse the H&P design by dividing the VFO down to LF and letting?
it be synched with a very stable HF reference (WWVB or GPS or TCO).? Result is?
the same when derived steering voltage is then applied to the HF part of the VFO.

My comments were in response to someone's complaint about Si5351a outputs?
being square waves.? With H&P the VFO can be a linear output.??

Arv
_._


On Sat, Nov 14, 2020 at 6:39 PM Ashhar Farhan <farhanbox@...> wrote:
I was expecting Arv to write in about H&P schemes. Between him and Hans, there is all we know about how this scheme works.
The challenge of Huff and Puff is that it needs a very stable oscillator to begin with. It should have very minimal wobble (short term drift). The drift between two correction pulses should be less than its step size. That is not difficult to obtain.
The inherent drawback of H&P is that the oscillator is always being pushed up or down from the central frequency between two cycles. This can have a disastrous effect on digital modes.
There is probably another way to fix this. It is to use two flip flops as parallel mixers, each is fed the clock pulse that is 90 degrees out of phase from the other (imagine it to be a phasing receiver). Noe, the combined DC output of the two will indicate if we have drifted above, below or stayed on the frequency. I haven't tried this, but I am guessing that this will work. In the professional literature it is called a phase frequency detector.
- f


On Sun 15 Nov, 2020, 5:28 AM Jerry Gaffke via , <jgaffke=[email protected]> wrote:
This H&P jitter may not matter much if it isn't very frequent.
If it's a second or two between updates and the frequency slews
at a reasonable rate after an update, the "jitter" should be
no more objectionable than when manually making slight adjustments.

Jerry, KE7ER

On Sat, Nov 14, 2020 at 01:56 PM, Arv Evans wrote:
H&P stabilization is interesting because by default it sets the tuning?
step size as well as holding frequency drift to around 1 Hz, sometimes?
less.? However it does introduce its own type of slow jitter.?

Re: For Sale : uBITx 6.0 full kit

Jeff | VA2SS
 

Sold.?

73
--
Jean-Fran?ois M¨¦nard?
Jeff | VA2SS

Envoy¨¦ ¨¤ partir de mon appareil mobile.?
Sent from my mobile device.
Re: Follow-up Antuino question

 

There are lots of variants to doing a Huff and Puff.
EMRFD shows a very basic example on pages 4.5 and 4.7.
In this design, the frequency is constantly bouncing somewhere
between two frequencies that are 40hz apart.?

If the VFO is stable enough, one could slow down the loop filter between
the Huff and Puff detector and the varactor diode in the VFO, such that it
takes a minute or two (instead of under a second) before a change in
VFO frequency causes a significant change to the varactor correction voltage.
This would make the output frequency much more stable, but would take some time
to settle in after you manually tune the VFO.

The advantage of Huff and Puff is that you can make the VFO lock to
discrete steps without having to reprogram the dividers with each frequency change
as you would on a standard PLL.? Where a standard PLL will lock only to one specific?
frequency, a Huff and Puff might lock to any frequency that is an even multiple?
of 100 Hz. I doubt you could borrow a phase-frequency-detector directly
from a PLL design and have it work in a Huff and Puff.

Tom would almost certainly be better off with a standard PLL design
for his GPS stabilized crystal reference oscillator.

Huff and Puff isn't used a lot these days.? If you have a microcontroller in
the rig (such as the Nano on the Raduino), it's fairly easy to program the dividers
on a PLL device.? This will generally give a more stable output.? A PLL can allow
more resolution (much smaller steps) when choosing the output frequency.

Take the above with a large grain of salt.
I am no expert on either Huff and Puff or PLL's.

Jerry, KE7ER


On Sat, Nov 14, 2020 at 08:33 PM, Jerry Gaffke wrote:
It's been lots of years since I have looked at huff-and-puff oscillator stabilization.
What I wrote in that last post is incorrect, there's more going on here.

Jerry, KE7ER


Hide quoted text

?

On Sat, Nov 14, 2020 at 06:08 PM, Jerry Gaffke wrote:

This may be an alternate way of looking at what you are proposing.


I suspect you are describing a phase-locked-loop.
And that yes, you have tried it.
There's a phase-frequency-detector inside the Si5351.

The VCO in the Si5351 can be pulled anywhere between 600 and 900 mhz
(and quite a bit beyond, actually).
But if instead of that VCO we have a VFO that is first manually tuned very near the
target frequency, the error detection will have far less work to do.?
And the output will have far less jitter.

Jerry, KE7ER

?
For Sale : uBITx 6.0 full kit

Jeff | VA2SS
 

Hi,

I have for sale my uBITx 6.0 full kit for sale. I purchased it last spring. I just assembled it to make sure it was working, then it is now collecting dust on the shelf since that time.
I unfortunately not having enough time to spend with it that justify having one.

Items: uBITx 6.0 full kit (black enclosure), microphone, screen pencil, etc... see webpage for all included items



I would like to have 100$ USD + shipping via PayPal

Best 73.

--
Jean-Fran?ois M¨¦nard?
Jeff | VA2SS

Envoy¨¦ ¨¤ partir de mon appareil mobile.?
Sent from my mobile device.
Re: S-Meter Pin A7 Location V6

 

Mick,
I do not have a v6 kit, so not sure about the following:

A7 is the Purple wire coming off of the Raduino digital connector to goes to the encoder.? In one picture that I have seen of the encoder, the purple wire has been trimmed back.? You may need to pull that pin and insert another one crimped to a longer wire.

Here is a picture from HFSignals that shows the encoder connector:


A7 is the purple wire that is unconnected, wrapped around the other wires.

AGAIN, I do NOT have a v6, so you may want to wait for someone to post a picture of their uBITX.
73
Evan
AC9TU

S-Meter Pin A7 Location V6

 

Good afternoon all

I have just installed the kit-projects AGC board and read in the instructions to connect to pin A7 for the S-meter.
Please can someone confirm the location of pin A7?
Is it the pin 7 on the same header strip where I've connected the 4 wires that go the nextion display, or somewhere else?

An image showing pin A7 would be great if someone has one.

Many thanks

73

Mick M0GWD
Re: Raduino offset

 

Richard,
The name of the group is BITX20, not uBITX.? It is really for homebrewing as well as the kits that came out of that effort.? Feel free to continue to post if you so choose.? We are all interested in the type of experimentation that you are doing.? I can speak for myself that I learn with all of the questions and answers on bidirectional transceiver work that is being done.

Have fun with the build!
73
Evan
AC9TU
Re: Raduino offset

 

Evan, again thank you

I think this will be my last post on the subject as this is a uBITX group, not a Homebrew group.....

I think a *simple* fix involves both hardware and software.
Since I already have USB and LSB oscillators, I do not need the 57 / 33 MHz output of CLK1
I can change both these values to 27 MHz,? combine with the CLK2 in a mixer and put through a bandpass filter.
I will then have the correct IF? I need.
Since I will be using a Nextion, there will? be free pins on the Raduino / Nano.
I can put is some simple code to switch? on my USB or LSB oscillators.

I will also have to tweak the output filter code as I have 8 output filters, not 4.

Problem:? the rest of the CEC goodies will not be functional in my radio.? Thanks sadly OK
I may report back if successful, so others can play with their Raduinos as well.

Thanks a gain and best to all
Richard? VA3NDO
Re: My son's V6

Gary Rindfuss
 

I also used this on the antenna input for a while before I built my own filter. It worked ok. Definitly killed the AM interference. Their may have been some slight attenuation of the transmitted signal since its in both the Tx? and Rx path.? But I didn't notice it much.?



On Sun, Nov 15, 2020, 5:28 AM Gerard <kabupos@...> wrote:
Hello,
The other time when I did a search on E.. Y, with the keyword Ubitx, I came across this.
A filter, but I don¡¯t know any more...


cdt

?
Re: My son's V6

Gary Rindfuss
 

Kit-projects sells a uBitx broadcast filter kit on the web page where they sell their agc board. It requires you to wind 3 tiny torroids but it works well.?
There is another design you can build yourself that uses axial inductors it works well too. I have used one of each. The info on that is here:




On Sun, Nov 15, 2020, 5:28 AM Gerard <kabupos@...> wrote:
Hello,
The other time when I did a search on E.. Y, with the keyword Ubitx, I came across this.
A filter, but I don¡¯t know any more...


cdt

?
Re: Follow-up Antuino question

 

Bore da, Arv!

Sut dych chi?

(Dw i'n dysgu siarad Cymraeg!)

Related? There are teeming millions of we of the Evans clan!

On Sat, Nov 14, 2020 at 09:56 PM, Arv Evans wrote:
Gareth Evans

We have the same last name...are we related? 8-)
Re: My son's V6

 

Hello,
The other time when I did a search on E.. Y, with the keyword Ubitx, I came across this.
A filter, but I don¡¯t know any more...


cdt

?
Re: My son's V6

 

Bob,

I would agree with Gary's suggestion to add the Broadcast Band high pass filter.??

I had a similar experience to your son's one afternoon.? All of a sudden I had birdies all over the band.? Turned out that my wife had started our new high-efficiency washer at that time.? When the washer stopped (30+ minutes later), the noise went away.? It could be that there is an appliance being run in the area that is generating a lot of noise.? Switching to battery power did not eliminate the problem so it was not just power line related.

I would have him try again and see if the problem persists.

73
Evan
AC9TU
Re: No Rx from my V4.3

 

On Sat, Nov 14, 2020 at 04:58 PM, Patrick Peter Rosney wrote:
I uploaded V5 Firmware but this has made no difference
Patrick,
The v5 firmware uses a different SSB filter frequency.? That software will not work with a v4 board.? Here is a link to the KD8CEC software that has both the v4 and v5 board software.?


You want to use the uBITXV234 folder selecting the file that matches your display.? If it is the stock v4 1602 display then the file is?UBITX_CEC_V1.200_16P.hex.?The hex files are loaded into the Nano with Xloader.

If you want to go back to the stock v4 software, here is the link:


I would suggest the KD8CEC software, as that has more features and options for CW operation that the stock does not.

Once you have the correct software I would then start the troubleshooting as Jerry suggested.

Have fun with the rig.
73
Evan
AC9TU
Re: Follow-up Antuino question

 

It's been lots of years since I have looked at huff-and-puff oscillator stabilization.
What I wrote in that last post is incorrect, there's more going on here.

Jerry, KE7ER


On Sat, Nov 14, 2020 at 06:08 PM, Jerry Gaffke wrote:

This may be an alternate way of looking at what you are proposing.


I suspect you are describing a phase-locked-loop.
And that yes, you have tried it.
There's a phase-frequency-detector inside the Si5351.

The VCO in the Si5351 can be pulled anywhere between 600 and 900 mhz
(and quite a bit beyond, actually).
But if instead of that VCO we have a VFO that is first manually tuned very near the
target frequency, the error detection will have far less work to do.?
And the output will have far less jitter.

Jerry, KE7ER

?
Re: Follow-up Antuino question

 

On Sat, Nov 14, 2020 at 06:08 PM, Jerry Gaffke wrote:
But if instead of that VCO we have a VFO that is first manually tuned very near the
target frequency,
So it seems the Huff and Puff method is for providing the final little bit of frequency stabilization to a VFO that is already adjusted to oscillate very close to the desired frequency. Would the Huff and Puff method be one of the preferred ways to fine adjust the frequency of a 10mhz crystal oscillator to make a DIY GPS disciplined frequency reference?

Tom, wb6b
Re: Follow-up Antuino question

 

This may be an alternate way of looking at what you are proposing.


I suspect you are describing a phase-locked-loop.
And that yes, you have tried it.
There's a phase-frequency-detector inside the Si5351.

The VCO in the Si5351 can be pulled anywhere between 600 and 900 mhz
(and quite a bit beyond, actually).
But if instead of that VCO we have a VFO that is first manually tuned very near the
target frequency, the error detection will have far less work to do.?
And the output will have far less jitter.

Jerry, KE7ER



On Sat, Nov 14, 2020 at 05:39 PM, Ashhar Farhan wrote:

I was expecting Arv to write in about H&P schemes. Between him and Hans, there is all we know about how this scheme works.
The challenge of Huff and Puff is that it needs a very stable oscillator to begin with. It should have very minimal wobble (short term drift). The drift between two correction pulses should be less than its step size. That is not difficult to obtain.
The inherent drawback of H&P is that the oscillator is always being pushed up or down from the central frequency between two cycles. This can have a disastrous effect on digital modes.
There is probably another way to fix this. It is to use two flip flops as parallel mixers, each is fed the clock pulse that is 90 degrees out of phase from the other (imagine it to be a phasing receiver). Noe, the combined DC output of the two will indicate if we have drifted above, below or stayed on the frequency. I haven't tried this, but I am guessing that this will work. In the professional literature it is called a phase frequency detector.
- f
Re: Follow-up Antuino question

 

On Sat, Nov 14, 2020 at 05:33 PM, Tom, wb6b wrote:
using the ATTiny481 chip
That was a typo. The chip is the ATTiny841.

Tom, wb6b
34081 - 34100 of 116787