Topics

Compander AGC

Steven Weber
 

UPS came through and dropped off the compander chip this afternoon. For those of you who might be interested, the part number is NCY9100DW2G

 

It works great, but only one little problem. It wants to keep the output level at 2V p-p. That’s about  10X a comfortable level for headphones. It doesn’t like to drive headphones directly, but adding a series resistor cleans up the cross over the distortion.

 

So, one solution would be to just put a big enough series resistor to the headphones to get a 200 mV signal. But it takes a 200 mv p-p input signal before the AGC action kicks in. So, there would have to be a boat load of gain before the chip to arrive at the 200 mv input, just to knock it back down again.

 

It’s possible to reduce the gain of the AGC system, but not to increase it. Another possible solution could be to add an external gain stage to the loop. But I don’t have much real estate left to add more parts. The compander chip is in a 16 pin wide body package.

 

All in all, looked like a good idea, but as usual the devil is in the details. Good thing I only bought 10 of the chips.

 

Steve KD1JV

 

BTW, Happy New Year everyone. Better yet, have a Merry New Year !

 

Larry KA5T
 

NCY9100DWR2G maybe?

On Dec 30, 2019, at 7:22 PM, Steven Weber <steve.kd1jv@...> wrote:

NCY9100DW2G

Steven Weber
 

Yea, man I need new glasses 😊

 

 

NCY9100DWR2G  maybe?

 

> On Dec 30, 2019, at 7:22 PM, Steven Weber <steve.kd1jv@...> wrote:

>

> NCY9100DW2G

 

 

 

Steven Weber
 

Well, it seems I can make this compander chip do the job after all. I increased the gain up stream and put a 220 ohm resistor in series with the headphones. The AGC kicks in at about 50 uV, a S9 signal. The 220 ohm resistor limits the volume to a reasonable level.

 

So, it looks like I can eliminate the volume control all together and the LM386 amp.

 

Steve KD1JV

 

 

David Perry
 

Interesting Steve.  Takes me back to the old 'no volume' xtal  set days...

David G4YVM 

Joe Street
 

Hi Steve

AGC is always an interesting topic.  A couple of years ago I set out to build something of an improved audio derived technique.  I ended up using a vactrol with the LED portion driven by PWM via a PIC micro which was sampling a fast averaged signal level.  It required piecewise linear interpolation because the response of the LED/LDR is not linear but I was able to use a table lookup which was quite small and could tailor the AGC threshold, slope, attack and decay to anything I wanted and the great part was that the audio path was a simple resistor divider with zero impact on audio fidelity. The response characteristics of the LDR in the vactrol lend themselves well to the application of audio derived AGC but the technique gets around the typical limitations (ie too slow a response due to the need to integrate several cycles of audio rather than the time needed for several cycles of RF which causes a pop)  Perhaps the idea is more than you are looking for in a simple rig but I just thought I'd mention.

Happy New Year
Joe ve3vxo

On Mon, Dec 30, 2019 at 8:22 PM Steven Weber <steve.kd1jv@...> wrote:

UPS came through and dropped off the compander chip this afternoon. For those of you who might be interested, the part number is NCY9100DW2G

 

It works great, but only one little problem. It wants to keep the output level at 2V p-p. That’s about  10X a comfortable level for headphones. It doesn’t like to drive headphones directly, but adding a series resistor cleans up the cross over the distortion.

 

So, one solution would be to just put a big enough series resistor to the headphones to get a 200 mV signal. But it takes a 200 mv p-p input signal before the AGC action kicks in. So, there would have to be a boat load of gain before the chip to arrive at the 200 mv input, just to knock it back down again.

 

It’s possible to reduce the gain of the AGC system, but not to increase it. Another possible solution could be to add an external gain stage to the loop. But I don’t have much real estate left to add more parts. The compander chip is in a 16 pin wide body package.

 

All in all, looked like a good idea, but as usual the devil is in the details. Good thing I only bought 10 of the chips.

 

Steve KD1JV

 

BTW, Happy New Year everyone. Better yet, have a Merry New Year !

 

Steven Weber
 

Joe,

 

Some time ago I made a similar AGC with LED and LDR, but all analog. The photo resistor controlled the gain of an op amp which also drove the LED. It worked pretty good. I haven’t looked for photo resistors in some time, wonder if they are still available?

 

KD1JV

 

 

Hi Steve

AGC is always an interesting topic.  A couple of years ago I set out to build something of an improved audio derived technique.  I ended up using a vactrol with the LED portion driven by PWM via a PIC micro which was sampling a fast averaged signal level.  It required piecewise linear interpolation because the response of the LED/LDR is not linear but I was able to use a table lookup which was quite small and could tailor the AGC threshold, slope, attack and decay to anything I wanted and the great part was that the audio path was a simple resistor divider with zero impact on audio fidelity. The response characteristics of the LDR in the vactrol lend themselves well to the application of audio derived AGC but the technique gets around the typical limitations (ie too slow a response due to the need to integrate several cycles of audio rather than the time needed for several cycles of RF which causes a pop)  Perhaps the idea is more than you are looking for in a simple rig but I just thought I'd mention.

Happy New Year

Joe ve3vxo

 

On Mon, Dec 30, 2019 at 8:22 PM Steven Weber <steve.kd1jv@...> wrote:

UPS came through and dropped off the compander chip this afternoon. For those of you who might be interested, the part number is NCY9100DW2G

 

It works great, but only one little problem. It wants to keep the output level at 2V p-p. That’s about  10X a comfortable level for headphones. It doesn’t like to drive headphones directly, but adding a series resistor cleans up the cross over the distortion.

 

So, one solution would be to just put a big enough series resistor to the headphones to get a 200 mV signal. But it takes a 200 mv p-p input signal before the AGC action kicks in. So, there would have to be a boat load of gain before the chip to arrive at the 200 mv input, just to knock it back down again.

 

It’s possible to reduce the gain of the AGC system, but not to increase it. Another possible solution could be to add an external gain stage to the loop. But I don’t have much real estate left to add more parts. The compander chip is in a 16 pin wide body package.

 

All in all, looked like a good idea, but as usual the devil is in the details. Good thing I only bought 10 of the chips.

 

Steve KD1JV

 

BTW, Happy New Year everyone. Better yet, have a Merry New Year !

 

 

Joe Street
 

With ROHS it is likely that cadmium selenide resistors would face some hurdles, but on the other hand some grad students in my lab recently started a business selling high brightness LED displays based on CdSe quantum dots which are very hazardous to manufacture so who knows.  I'm pretty sure the CdS LDR's will be available surplus for many years to come.  I have a few discontinued VTL5C3's which is what I used for that AGC but I have also rolled my own using a peice of tubing, some black heat shink and a green LED and CdS cell.  They are also handy for optimally remote terminarting beverage antennas. :)

Cheers
Joe

On Tue, Dec 31, 2019 at 10:10 AM Steven Weber <steve.kd1jv@...> wrote:

Joe,

 

Some time ago I made a similar AGC with LED and LDR, but all analog. The photo resistor controlled the gain of an op amp which also drove the LED. It worked pretty good. I haven’t looked for photo resistors in some time, wonder if they are still available?

 

KD1JV


Steven Weber
 

I looked at ebay and there are all kinds available there, cheap too. Mouser also has one, Adafruit branded.

 

Steve

 

 

With ROHS it is likely that cadmium selenide resistors would face some hurdles, but on the other hand some grad students in my lab recently started a business selling high brightness LED displays based on CdSe quantum dots which are very hazardous to manufacture so who knows.  I'm pretty sure the CdS LDR's will be available surplus for many years to come.  I have a few discontinued VTL5C3's which is what I used for that AGC but I have also rolled my own using a peice of tubing, some black heat shink and a green LED and CdS cell.  They are also handy for optimally remote terminarting beverage antennas. :)

Cheers

Joe

 

On Tue, Dec 31, 2019 at 10:10 AM Steven Weber <steve.kd1jv@...> wrote:

Joe,

 

Some time ago I made a similar AGC with LED and LDR, but all analog. The photo resistor controlled the gain of an op amp which also drove the LED. It worked pretty good. I haven’t looked for photo resistors in some time, wonder if they are still available?

 

KD1JV

 

 

WA0ITP
 

Awesome news!! Great design feature imho.
72 WAØITP
I love this radio stuff.
www.wa0itp.com
www.4sqrp.com
On 12/30/2019 10:33 PM, Steven Weber wrote:

Well, it seems I can make this compander chip do the job after all. I increased the gain up stream and put a 220 ohm resistor in series with the headphones. The AGC kicks in at about 50 uV, a S9 signal. The 220 ohm resistor limits the volume to a reasonable level.

 

So, it looks like I can eliminate the volume control all together and the LM386 amp.

 

Steve KD1JV

 

 


dl2arl@...
 

Hello Steven, Hello Group,
this posting is GREAT NEWS! But please be aware that for a kit to fit them all out there you might imperatively need some impedance setting device to fit for the headphone used and also for the desired AGC Kick-In Point to acommodate loud or quiet environements . There are many types of headphones out there and it is not said that the expensive ones are better, but they differ very much in what their impedance is concerned. So adapting the AGC for one specific headset will inevitably lead to malfunction with another one having an impedance ten-fold bigger or smaller.

These are only details! I think the sourcing of this AGC chip is a big step forward and rigs made for rough portable  use absolutely need one. They need one so desperately, that I seriously think of retro-fitting some of my MTR and ATS rigs with such an AGC.

On the other hand, I am happy to own one fabulous Rig, the SST, that has one of the simplest and most genial audio solution that I can think of, No! I do not ignore the difference between a limiter (that is what the SST has) and a full grown-up AGC, but I try to humbly express my enthusiasm for the simple and efficient solution the SST has. I also do not ignore the difference between an AGC solution working in the audio frequency stage (much more prone to "plopping") and one looping back as far as the Intermediate Frequency or even the front end. But let us keep it simple.

Wish you all a Happy New Year!
73 de Razvan DL2ARL

Reed White
 

Earphones have different sensitivities, so if using with an AGC, it would be good to have a tiny screw-adjusted pot on the circuit board to match the earphones.

... Reed K7FLY

--
Reed White, reed@..., 719-695-0880

steve_wg0at
 
Edited

Still have my SST ...it's one of my favorite mono band SOTA rigs
...just love the sound of it (AGC?) and it's sharp xtal filter!
Great field OPs radio! ...73, Steve/wGOAT

Joe Street
 

Hey Razvan

I agree with your comments.  The AGC subject is always a good one and the configuration of it for different situations is really nice if the equipment allows you that level of control, otherwise it might be better riding the gain control by hand!   In the absence of that, on a simple CW rig a limiter followed by a hypermite type filter is a very good alternative.  The filter removes the nasty harmonics from the limiter and the result is a pleasing audio even when the signal is limited.

Joe

On Fri, Jan 3, 2020 at 7:57 PM <dl2arl@...> wrote:
Hello Steven, Hello Group,
this posting is GREAT NEWS! But please be aware that for a kit to fit them all out there you might imperatively need some impedance setting device to fit for the headphone used and also for the desired AGC Kick-In Point to acommodate loud or quiet environements . There are many types of headphones out there and it is not said that the expensive ones are better, but they differ very much in what their impedance is concerned. So adapting the AGC for one specific headset will inevitably lead to malfunction with another one having an impedance ten-fold bigger or smaller.

These are only details! I think the sourcing of this AGC chip is a big step forward and rigs made for rough portable  use absolutely need one. They need one so desperately, that I seriously think of retro-fitting some of my MTR and ATS rigs with such an AGC.

On the other hand, I am happy to own one fabulous Rig, the SST, that has one of the simplest and most genial audio solution that I can think of, No! I do not ignore the difference between a limiter (that is what the SST has) and a full grown-up AGC, but I try to humbly express my enthusiasm for the simple and efficient solution the SST has. I also do not ignore the difference between an AGC solution working in the audio frequency stage (much more prone to "plopping") and one looping back as far as the Intermediate Frequency or even the front end. But let us keep it simple.

Wish you all a Happy New Year!
73 de Razvan DL2ARL

Brien Pepperdine
 

I realise (in theory) that what is being discussed and wanted is internal  AGC /  limiting kind of stuff, but for those interested, or who might want to incorporate into a somewhat larger 'kit'.. this might be of interest. For interest's sake or for practical application.

FWIW - from Alan Wolke W2AEW, who knows a few things. From there.. to eternity ( I guess ).

https://www.youtube.com/watch?v=pUii_pDQxg8

Brian

---------- Original Message ----------
From: dl2arl@...
Date: January 3, 2020 at 7:57 PM

Hello Steven, Hello Group,
this posting is GREAT NEWS! But please be aware that for a kit to fit them all out there you might imperatively need some impedance setting device to fit for the headphone used and also for the desired AGC Kick-In Point to acommodate loud or quiet environements . There are many types of headphones out there and it is not said that the expensive ones are better, but they differ very much in what their impedance is concerned. So adapting the AGC for one specific headset will inevitably lead to malfunction with another one having an impedance ten-fold bigger or smaller.

These are only details! I think the sourcing of this AGC chip is a big step forward and rigs made for rough portable  use absolutely need one. They need one so desperately, that I seriously think of retro-fitting some of my MTR and ATS rigs with such an AGC.

On the other hand, I am happy to own one fabulous Rig, the SST, that has one of the simplest and most genial audio solution that I can think of, No! I do not ignore the difference between a limiter (that is what the SST has) and a full grown-up AGC, but I try to humbly express my enthusiasm for the simple and efficient solution the SST has. I also do not ignore the difference between an AGC solution working in the audio frequency stage (much more prone to "plopping") and one looping back as far as the Intermediate Frequency or even the front end. But let us keep it simple.

Wish you all a Happy New Year!
73 de Razvan DL2ARL


 

Joe Street
 

Hi Brian

Your circuit is very clever and elegant in its simplicity using the current through the series diodes to modulate the on state resistance (a voltage divider essentially).  However the peak detector rectifies the audio waveform in order to drive the level control diodes and therefore is still subject to the time delay needed to integrate several cycles of the audio signal (several milliseconds) to generate the control voltage for the system.  I wish you could do another demonstration video but this time using CW.  This will show that in between words or maybe even characters (depending on the recovery time of the detector) the level control signal will change and there will be a disturbing pop or click on the next strong CW signal (especially annoying with headphone use).  If the peak detector was rectifying RF or IF those 3 to 5 cycles necessary to settle the drive signal are RF cycles and therefore happen so fast (microseconds) that the ear is not bombarded during the transition.  This is the real problem of any audio derived AGC and the reason I developed the technique I described earlier using the PIC micro which worked but ended up being rather complex.

Joe


On Sat, Jan 4, 2020 at 7:59 AM Brien Pepperdine <brianpepperdine@...> wrote:

I realise (in theory) that what is being discussed and wanted is internal  AGC /  limiting kind of stuff, but for those interested, or who might want to incorporate into a somewhat larger 'kit'.. this might be of interest. For interest's sake or for practical application.

FWIW - from Alan Wolke W2AEW, who knows a few things. From there.. to eternity ( I guess ).

https://www.youtube.com/watch?v=pUii_pDQxg8

Brian