The double peak detector that I have attached will output between 0 and 7 volts. You can incorporate it in your scheme. Barry K3bo
George,
An AGC detector must be placed after the the desired signal has been separated out
with the narrowest bandwidth filter in the radio, so after the 11mhz crystal filter is correct.
Unfortunately, when trying to detect that 11mhz signal on the uBitx,
you will likely see nothing but the very strong BFO, also at 11mhz.
It's possible to do, but you will need a carefully designed stage for isolation
between the crystal filter and the demodulator, it is currently a direct connection.
And likely need a well shielded BFO and demodulator.
There's a reason all the various AGC implementations for the uBitx
detect the audio signal, not the IF signal.
###########
Building an amplifier to go from mV DC to Volts is fairly straightforward..
I would probably do it with two op amps configured as inverting amplifiers,
one after the other.? Use three if you really want an inverted output.
Trying to do it all in one op amp stage might be possible, but that much gain
approaches the limits of what typical op amps can do? ? And while not
terribly important in this case, it would drastically reduce the possible bandwidth.
Lots of gain in a single stage can be unstable.
Op amps are cheap, 2 or even 4 in a single package for well under a dollar.
This could also be done with transistors instead of op amps.
Jerry, KE7ER
On Fri, Oct 30, 2020 at 09:51 AM, George Metaxas wrote:
Hello, I am a uBitx v5 owner and modder. I am opening this thread again as I am considering placing a p-channel JFET in parallel to the signal entering the 1st IF amplifier as a means to control the RF gain. I am currently trying to figure out a circuit to be added right before the product detector and perform envelope detection on the 2nd IF signal, then use the DC voltage to do AGC.
I have found out by simulation - not built any test jig yet - that the gate voltage to the JFET should range between zero and 5 V in order to control the RF gain as such. However, the envelope detector produces a voltage that is both much smaller and ranges over much smaller values, of the order of mV.
What do you think of that approach? Could that be a good way to have IF-derived AGC on uBitx? How can I use the output of an envelope detector as a way to produce a 0...5 V AGC voltage signal? I'm really stuck on that.
73, George, SV1DEQ
|
On Fri, Oct 30, 2020 at 09:29 PM, Jerry Gaffke wrote:
As with the true meaning of AVC, we may have language issues here,
or at least a cultural difference.? I sincerely mean that.
no cultural difference, it is extremely difficult to translate Italian sarcasm into English, just to clarify English is certainly not the master of culture ... don't you think? ... try reading Dante in the original, then we'll talk about it!
|
Loris,
Of course, this part was sarcastic:
> Some of us also talk about how we intend to exercise more
> and make a change from our current diet of mostly beer and cheetos.
At least somewhat self deprecating, note that I say "we", not "you".
Humorous in some circles, at least for those of us that can laugh at ourselves.
As with the true meaning of AVC, we may have language issues here,
or at least a cultural difference.? I sincerely mean that.
Best Regards,
Jerry, KE7ER
toggle quoted message
Show quoted text
On Fri, Oct 30, 2020 at 01:16 PM, Jerry Gaffke wrote:
Loris,
Sorry, not being sarcastic.
|
Loris,
Sorry, not being sarcastic.
Many people posting here have added AGC or are planning to do so.
Many are interested in other technical projects involving the uBitx.
Many are interested in operating.
Many are simply interested.
I honestly feel that considerably less than 50% have carried out an AGC hack.
I seem to have rubbed you the wrong way when I disagreed you previously.
In the future, I'll try to stay out of your way and avoid answering your posts.
Reading them too.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Fri, Oct 30, 2020 at 12:28 PM, IW4AJR Loris wrote:
Hi Jerry ...?
always sarcastic and perched on your shining ivory tower?
Relax !!! many more than you think like to modify his radio and like experiment new solution,
don't need to be a professor for to be a good OM!
|
On Fri, Oct 30, 2020 at 07:42 PM, Jerry Gaffke wrote:
I'd be very surprised if a majority of uBitx's have AGC added.
Perceptions may be skewed from reading posts in the forum
because it is so much fun to talk about the various ways to add AGC.
Some of us also talk about how we intend to exercise more
and make a change from our current diet of mostly beer and cheetos.
Jerry
Hi Jerry ...?
always sarcastic and perched on your shining ivory tower?
Relax !!! many more than you think like to modify his radio and like experiment new solution, don't need to be a professor for to be a good OM!
|
I'd be very surprised if a majority of uBitx's have AGC added.
Perceptions may be skewed from reading posts in the forum
because it is so much fun to talk about the various ways to add AGC.
Some of us also talk about how we intend to exercise more
and make a change from our current diet of mostly beer and cheetos.
Jerry
toggle quoted message
Show quoted text
On Fri, Oct 30, 2020 at 10:38 AM, Curt wrote:
one can wonder what % of ubitx have agc installed but it may be a vast minority
|
Gerard,
There are only 3 regions per this map:
I believe that you are in Region 1, though that is through your reference to the French language.
I have never found a good explanation for all of the values on the Memory Manager.? Pieces here and there and then trial and error.? Here is one place that kind of describes one of the earlier version operations including a video at the end.
73
Evan
AC9TU
|
George,
An AGC detector must be placed after the the desired signal has been separated out
with the narrowest bandwidth filter in the radio, so after the 11mhz crystal filter is correct.
Unfortunately, when trying to detect that 11mhz signal on the uBitx,
you will likely see nothing but the very strong BFO, also at 11mhz.
It's possible to do, but you will need a carefully designed stage for isolation
between the crystal filter and the demodulator, it is currently a direct connection.
And likely need a well shielded BFO and demodulator.
There's a reason all the various AGC implementations for the uBitx
detect the audio signal, not the IF signal.
###########
Building an amplifier to go from mV DC to Volts is fairly straightforward..
I would probably do it with two op amps configured as inverting amplifiers,
one after the other.? Use three if you really want an inverted output.
Trying to do it all in one op amp stage might be possible, but that much gain
approaches the limits of what typical op amps can do? ? And while not
terribly important in this case, it would drastically reduce the possible bandwidth.
Lots of gain in a single stage can be unstable.
Op amps are cheap, 2 or even 4 in a single package for well under a dollar.
This could also be done with transistors instead of op amps.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Fri, Oct 30, 2020 at 09:51 AM, George Metaxas wrote:
Hello, I am a uBitx v5 owner and modder. I am opening this thread again as I am considering placing a p-channel JFET in parallel to the signal entering the 1st IF amplifier as a means to control the RF gain. I am currently trying to figure out a circuit to be added right before the product detector and perform envelope detection on the 2nd IF signal, then use the DC voltage to do AGC.
I have found out by simulation - not built any test jig yet - that the gate voltage to the JFET should range between zero and 5 V in order to control the RF gain as such. However, the envelope detector produces a voltage that is both much smaller and ranges over much smaller values, of the order of mV.
What do you think of that approach? Could that be a good way to have IF-derived AGC on uBitx? How can I use the output of an envelope detector as a way to produce a 0...5 V AGC voltage signal? I'm really stuck on that.
73, George, SV1DEQ
|
Thanks Curt
Darrell Lee
Advanced Data Systems, Inc. dlee@... 864-230-9626
On October 30, 2020 1:38:23 PM "Curt via groups.io" <wb8yyy@...> wrote:
Darrell I don't know - the special ingredient here is needing a high resistance on the output side with no signal and a low one with no signal (but not to low to eliminate the signal completely).? I won't claim its a superior solution - but it was the first one I got to work in my v4.? one can wonder what % of ubitx have agc installed but it may be a vast minority - and the types of AGC are spread out - with maybe the ND6T design the most popular - except it did not work well in my v4.?
73 curt
|
Darrell I don't know - the special ingredient here is needing a high resistance on the output side with no signal and a low one with no signal (but not to low to eliminate the signal completely).? I won't claim its a superior solution - but it was the first one I got to work in my v4.? one can wonder what % of ubitx have agc installed but it may be a vast minority - and the types of AGC are spread out - with maybe the ND6T design the most popular - except it did not work well in my v4.?
73 curt
|
Evan,
I found the map on this site
https://hamwaves.com/index/en/index.html
Frankly, there is something to read!!
It¡¯s hard to understand everything.
It takes hours and hours to be an informed Ham.
cdt
|
Hello, I am a uBitx v5 owner and modder. I am opening this thread again as I am considering placing a p-channel JFET in parallel to the signal entering the 1st IF amplifier as a means to control the RF gain. I am currently trying to figure out a circuit to be added right before the product detector and perform envelope detection on the 2nd IF signal, then use the DC voltage to do AGC.
I have found out by simulation - not built any test jig yet - that the gate voltage to the JFET should range between zero and 5 V in order to control the RF gain as such. However, the envelope detector produces a voltage that is both much smaller and ranges over much smaller values, of the order of mV.
What do you think of that approach? Could that be a good way to have IF-derived AGC on uBitx? How can I use the output of an envelope detector as a way to produce a 0...5 V AGC voltage signal? I'm really stuck on that.
73, George, SV1DEQ
|
Evan,
I don¡¯t think done like that.I had modified directly in the displayed bands.
When you speak " region", does that correspond to that?
Good for now this is not the most important. I will do a test "region" later. Thanks for the info.
It¡¯s a shame that it doesn¡¯t have documentation on all the manager¡¯s parameters. I looked, not found
good week
|
Could you replace the led/ldr with an opto-isolator ic chip?? Just wondering.....
Darrell Lee
Advanced Data Systems, Inc. dlee@... 864-230-9626
On October 30, 2020 10:44:54 AM "Curt via groups.io" <wb8yyy@...> wrote:
Bob
interesting discussion.? first the variable resistor needs a total range that is higher than the gain control potentiometer for small signals, and adequately below it for large signals -- something like perhaps > 10 k ohms to < 200 ohms in very rough numbers.? second, the mechanism for providing the bias to this variable resistor needs to be highly isolated from the audio chain - so its flow does not also impact the receive signal level.? two aspects of the circuit work nicely for this - first the 27k sense resistor and second the inherent isolation of the LED to LDR light path.? I see where the mosfet potentially can provide high isolation, but the question then becomes is its variable resistance behavior based upon the available bias obtained by the simple diode detector.? ND6T and others have provided a mosfet variable resistance design -- I confess I tried 2 or 3 versions of this but could not get desired behavior in my v4? - but the idea is similar.? the LED-LDR only won the competition in my case because it is the solution I got to work.? also this other design is installed differently - not upon the volume control but in the RF chain as I remember.? whether it has suitable variable resistance to install across the volume control pot is to be determined if someone wants to try it.?
I remember someone in Australia was going to kit a version of this circuit -- I don't know the availability of suitable LDRs as new stock components - mine was found in my parts stash likely acquired it a lot of components bought at a hamfest flea market.?
73 Curt
|
Re: A strange thing about calibration!!
Gerard,
Good information.
With your findings not sure it is worth the time to find out the beating that is causing the 16KHz tone.
Thank you,
73
Evan
AC9TU
|
Bob
interesting discussion.? first the variable resistor needs a total range that is higher than the gain control potentiometer for small signals, and adequately below it for large signals -- something like perhaps > 10 k ohms to < 200 ohms in very rough numbers.? second, the mechanism for providing the bias to this variable resistor needs to be highly isolated from the audio chain - so its flow does not also impact the receive signal level.? two aspects of the circuit work nicely for this - first the 27k sense resistor and second the inherent isolation of the LED to LDR light path.? I see where the mosfet potentially can provide high isolation, but the question then becomes is its variable resistance behavior based upon the available bias obtained by the simple diode detector.? ND6T and others have provided a mosfet variable resistance design -- I confess I tried 2 or 3 versions of this but could not get desired behavior in my v4? - but the idea is similar.? the LED-LDR only won the competition in my case because it is the solution I got to work.? also this other design is installed differently - not upon the volume control but in the RF chain as I remember.? whether it has suitable variable resistance to install across the volume control pot is to be determined if someone wants to try it.?
I remember someone in Australia was going to kit a version of this circuit -- I don't know the availability of suitable LDRs as new stock components - mine was found in my parts stash likely acquired it a lot of components bought at a hamfest flea market.?
73 Curt
|
Re: A strange thing about calibration!!
You say "Another is that I have two different audio amps in the rigs."
Evan,
No, I don¡¯t think it¡¯s audio-related.
Signal 16 khz, is already present at the entrance of the amp.
There is a combination of things that we do not yet control that makes this effect.
The SSB filter at 12mhz, is can help a number too "Round".
The main thing is that it disappears with a filter 11.059.
The result is, you can calibrate in all directions, as soon as it is validated and the reboot is done, the 16khz comes back.
cdt
|
Gerard,
You are correct in that there are at least two different sizes for the ILI 9341.? That is because the ILI 9341 really represents the controller interface, not the screen.
To get the pinout, look at this schematic from HFSignals:
or
73
Evan
AC9TU
|
Re: A strange thing about calibration!!
Evan,
Yes, very interesting what you say.
So, there is a problem if we use 12mhz quartzs.
I¡¯m glad you looked at that.
Like I said, the ear does the audio filter.
That confirms my theory.
I got quartz 11.059.
so I¡¯ll switch later to V5
I need to test them for the best filter.
Thank you for watching
cdt
|
Evan,
You can tell a little more on this screen.
on amazon.com, if I look at Il9341, they¡¯re just little 2.4- or 2.8-inch shells.
It¡¯s not as clear about the connection.
There are also 4 wires, ground, +5v, Tx, RX????
Thank you for given us more information on this. Screen bigger, connection.
Thank you
|
Gerard
George Metaxas (SV1DEQ)