From: Roy Lewallen on May 27, 6:49 pm

wrote:

I apologize for not being more precise in my nomenclature.

No problem to me...I fear I got off on a "lecture mode" again,
but was speaking in generalities to other readers about
receiver back-ends.

By "BFO" I mean the oscillator used for product detection when receiving
SSB and CW signals. No AM detector is involved. The AGC pickoff is of
course done from the IF preceding the product detector, and doesn't
intentionally use the BFO or product detector in any way. The problem I
was alluding to is that the BFO produces a large signal which is very
near the IF, and therefore can get into the AGC circuitry unless some
care is taken to prevent it. This produces a DC bias among other
problems, which can interfere with AGC circuit operation. I found it
necessary to completely shield the BFO, use a good doubly balanced
detector, and use differential amplifiers in the AGC chain in order to
reduce the BFO crosstalk to a tolerable level.

Sounds good to me. Separated, isolated detectors allow one to
concentrate on the particulars of each, makes it a lot easier to
work with.

For what it's worth on the audio-output part, I'm more fond of
rather high levels of IF into the detector to get around the
"square-law" response...looking for a better AM envelope
reproduction. While that results in better audio, it also makes
decoupling more difficult to avoid feeding the strong IF back
to the input. Different problem, same cuss-words on the bench,
though. :-)

I strongly suspect that a number of the complicated AGC circuits evolved
because a simpler AGC circuit was poorly designed and/or subject to
problems like crosstalk from the BFO. Instead of solving the fundamental
problems, increasingly complex circuits are developed until one
accidentally works correctly, then the improvement is credited to the
complex circuit rather than its accidental relative immunity to the
results of poor fundamental design. This isn't of course universally
true, but it happens pretty often.

I agree with you there. At least for voice-band detection
receivers (of which I've only built two in a half century from
my own design). Discounting copies of "All-American Five"
table-model cheapies using a single diode for both audio
rectification and (low-pass filtered) for AGC voltage to a
single controlled variable-mu amplifier. Ultimate simplicity
for reasons of price over the counter. One CAN put a BFO on
those (Hallicrafters did back in the late 40s) but the
performance is not the best.

Separating the "detectors" by function is best. The audio
"detector" (I still think of them as 'rectifiers') can be
optimized for best sound. The AGC detector can be optimized
for its action separately...and its response versus IF input
and overall receive chain amplification tailored for the
AGC control-loop "gain." Filtering-decoupling that follows
can be figured out to keep the low-frequency phase response
from upsetting the closed-loop AGC control.

Separate AGC and voice detectors lets one play around with
"attack" and "decay" time-constants with no more than a
single dual op-amp shaping circuit...multiple time-constants
under manual control if desired, that won't interfere with
the audio detection part. AGC detector input would have to
be the fastest-responding (to desired time-constant) with
a relatively simple op-amp doing the time-stretching.

Some folks might consider that op-amp addition "complicated."
Won't blame them if they do. From my experience, a
"complicated" AGC subsystem is having to AGC on a
1 uSec pulse with a time gate in the presence of other
assynchronous 1 uSec pulse sidebands located on 1 MHz
intervals (up to 3) on either side...with a decay to attack
time ratio of about 1000:1. :-) Did that for an R&D
airborne system at RCA...was somewhat too much but that
allowed a greater simplification for a following generation
of airborne equipment. A lesson there can be to "cover all
bases possible" the first time around, then investigate to
see what can be simplified for something less complicated.

AGC, in the basic consideration, should begin as a control
loop. From there on its a matter of choice of circuits.

Roy Lewallen wrote:
I strongly suspect that a number of the complicated AGC circuits evolved
because a simpler AGC circuit was poorly designed and/or subject to
problems like crosstalk from the BFO. Instead of solving the fundamental
problems, increasingly complex circuits are developed until one
accidentally works correctly, then the improvement is credited to the
complex circuit rather than its accidental relative immunity to the
results of poor fundamental design. This isn't of course universally
true, but it happens pretty often.

All too many software wannabees work this way too.

--
I miss my .signature.

Thanks to all for the constructive and informative discussion. I wish I
could say that it has helped to solve my problem but I'm still wrestling
with it. The low frequency circuit analysis by Spice, Bode and Nyquist plot
are beyond my capacity but I have had some excellent help from Dave,
off-list, who eye-balled the DX-394 schematic and my mod. We identified
decoupling networks and what I assume to be AGC 'delay' networks and tackled
it on the basis that reducing the time constants of the larger ones should
reduce the low frequency phase shift that could be contributing to the
problem. The opposite occurred - increasing the 'delayed AGC' time constant
reduced the instability and effectively slowed the attack, especially on the
RF front-end. More or less the same effect was obtained by slowing the
attack in the attack network that affects all stages.

I believe 'delayed AGC' means a slower or delayed attack at the RF stages;
in the DX-394, there is a R-C network adding maybe 15 ms to the attack on
the AGC line affecting both the 1st mixer and the drain-source current of
the RF preamp and a second network adding maybe 10 ms on top of this
affecting the AGC gate of the RF preamp. I've doubled that first time
constant and doubled what I would like in my attack and release networks in
order to get the fastest stable speeds which I guess would be on the order
of 20-40 ms attack and 50-80 ms release.

My mod has a FET amp/buffer at IF driving the heck out of the diodes so that
should be fairly linear. I have adjustable gain at the output of the
detector/attack filter. I don't think BFO interference is an issue; I don't
notice any great difference in stability with it on or off - there are
separate envelope and product detectors.

I'd welcome any more input. The base DX-394 schematic is at
http://www.monitor.co.uk/radio-mods/dx-394/dx-394.htm and I'd be happy to
send anyone the schematic of my mod.

73, Tom

In message , Tom Holden
writes
Thanks to all for the constructive and informative discussion.

I believe 'delayed AGC' means a slower or delayed attack at the RF stages;
in the DX-394, there is a R-C network adding maybe 15 ms to the attack on
the AGC line affecting both the 1st mixer and the drain-source current of
the RF preamp and a second network adding maybe 10 ms on top of this
affecting the AGC gate of the RF preamp. I've doubled that first time
constant and doubled what I would like in my attack and release networks in
order to get the fastest stable speeds which I guess would be on the order
of 20-40 ms attack and 50-80 ms release.


73, Tom



Tom,
I haven't been following this thread. However, in my understanding,
'Delayed AGC' doesn't refer to a time delay. It normally means that the
AGC in the RF stage doesn't cut in until a certain level of signal is
reached. AGC is applied as 'normal' to the IF stages but the RF stage is
held at maximum gain until the input signal is higher. The effect is to
obtain a better signal-to-noise ratio with low-level signals.
Ian.
--

Thanks for the reference, Bill. I did learn something of value from it but
the devices are clearly intended for audio frequency although one might
actually support 0dB gain at 455kHz! However, they are a closed loop system
and it's not obvious that one could bring out the required control voltage
to drive the receiver AGC.

Regards,

Tom

"Netgeek" wrote in message
...
Hi Tom,

You might find some good info from reading the description and
data sheets for the Analog Devices SSM2165 and/or SSM2166
at www.analog.com. They address the issues of having a threshold
which can be adjusted and then varying the amount of compression
or limiting asymetrically. Perhaps you could modify your circuit to
emulate some of these features - or perhaps just use the devices
described?

Bill

wrote in message
oups.com...
This sounds like a classic negative feedback oscillation. You sense the
signal is too large, so you send a signal to kill the gain, and then
you sense the signal is too small, so you send a signal to increase the
gain.

Having different attack and release time means you have two different
time constants My guess is the quick attack leads to the instability,
since it is the lesser damped system. If this is true, then you should
concentrate on the attack time, i.e find how slow it has to be for the
sytem to be stable.

Of course this is really had to do without seeing the circuitry in
action.

I have struggled with this in the past
It is a function of the behaviour of the servo loop at low freq and I
dont have the theory to analyse it properly.
However, the most successful AGC system I used in a DC receiver had a T
attenuator as the control element, consisting of resistors on the
horizontal arms of the T with a Darlington pair to ground as the control
element. This was driven by Opamp - rectifier (BE junction of a 2N3904)
driving an emitter follwer driving a conventional RC circuit. It seemed
that when you got rid of any DC shift in the system this fixed the
problem. As a bonus you got a surprisingly accurate log detector(over
about 60bB range) for an S meter. The difficult part in all systems
seems to be the control element.

Richard


Tom Holden wrote:
I'm looking for some advice/guidance on the design of AGC detection and
timing circuits, prompted by some level of frustration with a modification I
have been doing to a DX-394 SW radio. My questions, though, probably apply
to receiver design generally. I have a problem with stability - the receiver
gain oscillates at medium and fast release speeds.

Previously I had done a mod that pretty successfully provided 3 release
speeds for the DX-394 but fell short of what I thought was the ideal: an
attack time of ~1 millisecond, independent of the release time. That was
based on a survey of receivers from which I concluded that the attack should
be less than 13 ms and that 1 ms seemed to be the goal. Release speeds
should probably be on the order of 30ms, 300ms and 3 seconds, for fast,
medium and slow, respectively, although there seems to be lots of scope for
subjective preference. My mod required a rather large capacitor for Slow
release so my Slow was more like 1.2 seconds and the attack was slowed to
maybe 50-70 ms for the slow release..

The objectives of the enhanced mod are to:
a) improve the attack speed to better less than 13ms for all release speeds
b) extend the Slow release using smaller cap
c) reduce the loading of the AGC detector on the output of the 2nd IF amp
and also possible distortion due to the AGC and AM/Product detectors fed in
parallel

I used a JFET to buffer between the IF amp and the diode detector and an
emitter follower between the attack R-C circuit and the release R-C circuit,
dc coupled to the stock AGC amplifier. On the release side, about 1/10 the
capacitance vs the earlier mod is required for slow release and the attack
does seem to be similarly less affected by the release network.

However, at the fast and medium release settings, the receiver gain
literally oscillates at a rate that seems to be a function of attack and
release time constants, manual RF/IF gain setting, AGC gain setting and
signal strength. The depth of this gain modulation is affected by AGC and RF
gain. In order to get stability, it seems that I have to slow down the
attack (and/or release) time constant and carefully tweak the AGC gain
between the onset of oscillation and receiver peak distortion caused by not
enough gain reduction.

Have I completely misunderstood the meaning of attack/release speeds? My
'ideal' attack circuit has a R-C time constant of 1 ms, which means it will
even respond substantially to 1kHz modulation. That seems high. The R-C time
constant for my target fast release of 30 ms means that it will
substantially follow a 30Hz signal. I have had to pad these out to ~20ms
attack, 50ms release for stability or tolerably low gain oscillation depth
at medium and lower signal strengths. With this slower attack, stability is
much improved with the 500ms medium release speed.

The target attack/release of 1ms/30ms is not good for AM reception anyway as
it causes considerable distortion on heavy bass modulation - it is for data
services on steady carriers, e.g., PSK, FSK, DRM. But if the AGC causes
oscillation, then that's interference of another kind that would adversely
affect error rates. Several, including myself, have noted that DRM SNR is
improved by defeating AGC, on a wide variety of receivers.

Is this a typical problem for receiver design? Would 'hang' AGC stabilise
the AGC loop? Are my design objectives reasonable?

Comments from experienced radio designers/builders/experimenters much
appreciated.

Tom

From: Richard Hosking on Tues 31 May 2005 20:05

I have struggled with this in the past
It is a function of the behaviour of the servo loop at low freq and I
dont have the theory to analyse it properly.

The "theory" part should be evident to anyone who has made
a negative-feedback amplifier, single transistor to op-amp.
Getting to know op-amp responses both open-loop and closed-
loop (with negative feedback) can be helpful. Note that
op-amp designers actually build in open-loop phase shifts
at high frequencies to avoid oscillation with feedback.

The only difficult part is in MEASURING the PHASE at low
frequencies in the 0.1 to 10 Hz range...especially that of
the AGC control-line (feedback) circuitry. If not, some
dog-work on analyzing the magnitude and phase response of
that circuit will show that. [ability to handle complex
quantities is preferred there]

If the phase response is 0/360 degrees between AGC control-
line input and output (to the gain-controlled stages), AND
the closed-loop gain of the system is greater than unity,
there be troubles there! :-(

To get a view into AGC behavior with any general receiver,
disconnect the AGC control-line from the gain-controlled
stage and substitute a small variable DC source for the
AGC control-line input to that gain-controlled stage.
Using a reasonably-well-calibrated RF source, pick some
RF levels over the expected dynamic range of receiver input.
At each input level, adjust the DC control-line substitute
to be the same as the value of the disconnected AGC control-
line. Measure the DC value of both the input and output of
that AGC control-line circuitry.

The reason for doing that is to remove any phase effects
at low frequencies. That's a baseline value set that
SIMULATES the closed-loop control range of the AGC.

With enough RF input signal levels, the characteristic curve
of the closed-loop AGC action at DC can be seen...from no
AGC (maximum receiver gain) to high values of AGC control
(essentially minimum receiver gain). That will show the
delta of tiny AGC-control line variations which is the
equivalent of the "feedback percentage" of a negative-
feedback amplifier simple formula.

Alternately, one can do an open-loop gain measurement using
a series of AGC control-line value increments from minimum
to estimated maximum. Setting the simple DC supply (substitute
for the AGC control-line) to those increments will do it nicely.
Log the RF input level for those DC increments and measure the
input to the AGC control-line feedback circuitry (even though
it is disconnected from the controlled stages). That will
result in the same input signal characteristic curve.

Either way will result in "seeing" what the RF input signal
characteristics are, allow one to use the AGC line for things
like an S-Meter indicating circuit, squelch control, etc.
Note: The alternate method can also be done analytically on
paper if the controlled stages' gain v. control line is known.
A "gain budget" can be tabulated of the total receiver
sensitivity to various RF input levels that produce various
AGC control-line values.

That takes part of an afternoon's bench data logging, dreary
though that may be. It will establish THE characteristics of
that receiver, valuable reference for later work on it. That
curve will be no different than that of a single amplifier
stage with varying amounts of negative feedback.

Next is to either measure or calculate the low-frequency
magnitude and phase characteristics of the AGC control-line
circuitry (ALL of it, even to bypass caps at the controlled
stage input connection). Magnitude alone will yield the
feedback percentage of the equivalent negative-feedback
amplifier. The phase response at various low frequencies
has to be compared to the "attack" and "decay" times as
desired. THAT is not intuitive but must be examined to see
if the low-frequency AGC control characteristics will result
in a negative-feedback or positive-feedback (oscillatory,
motorboating) amplifier equivalent.

A stable receiver WITH AGC should ALWAYS have some error.

If actual low-frequency oscillation occurs, one cure is to
attenuate the AGC control-line range. A voltage divider if
the AGC control is through voltage does that. Such
attenuation works on the magnitude of the AGC control but
will also affect the phase.

I hope this simplified explanation is a help for all who
aren't familiar with Control System theory. Control Systems
aren't as intuitive as many instructors on the subject claim
so there isn't a lot of literature on it in popular
publications for hobbyists. Those usually present some
very simple analogue such as the ball governor valve on a
steam engine of old and let it go at that. :-(

wrote in message
oups.com...
From: Richard Hosking on Tues 31 May 2005 20:05

I have struggled with this in the past
It is a function of the behaviour of the servo loop at low freq and I
dont have the theory to analyse it properly.

The "theory" part should be evident to anyone who has made
a negative-feedback amplifier, single transistor to op-amp.
Getting to know op-amp responses both open-loop and closed-
loop (with negative feedback) can be helpful. Note that
op-amp designers actually build in open-loop phase shifts
at high frequencies to avoid oscillation with feedback.

The only difficult part is in MEASURING the PHASE at low
frequencies in the 0.1 to 10 Hz range...especially that of
the AGC control-line (feedback) circuitry. If not, some
dog-work on analyzing the magnitude and phase response of
that circuit will show that. [ability to handle complex
quantities is preferred there]

Lacking a calibrated RF source and much other critical equipment, I do have
a 45 year old Eico scope that once belonged to the famous Bach pianist Glenn
Gould, and could cobble together a variable dc source and a low freq
oscillator. To observe phase response of the open loop system, I'm thinking
that the loop could be broken between the AGC detector and the AGC time
constant/buffer. Drive the latter and the X input of the scope with the dc
supply and superposed low freq signal, feed the receiver with steady state
RF carrier and take the output of the AGC detector to the scope's Y input.
The variation of the input to the AGC system will cause variation in the
receiver gain and the output of the AGC detector. If in phase, the scope
would show a line with positive slope; if antiphase, a line with negative
slope; if in-between, an ellipse or some open shape subject to time
constants and non-linearities. This arrangement would leave the receiver's
RF gain control intact and its effect on time constant and phase observable;
it appears to modify the discharge resistance seen by a 1uF cap at the RF
and 1st Mixer in addition to pulling down the AGC voltage applied to them.

Does that seem to be a practical approach, Len?

Tom

From: "Tom Holden" on Tues 31 May 2005 22:08

wrote in message
roups.com...
From: Richard Hosking on Tues 31 May 2005 20:05

Lacking a calibrated RF source and much other critical equipment, I do have
a 45 year old Eico scope that once belonged to the famous Bach pianist Glenn
Gould, and could cobble together a variable dc source and a low freq
oscillator. To observe phase response of the open loop system, I'm thinking
that the loop could be broken between the AGC detector and the AGC time
constant/buffer. Drive the latter and the X input of the scope with the dc
supply and superposed low freq signal, feed the receiver with steady state
RF carrier and take the output of the AGC detector to the scope's Y input.
The variation of the input to the AGC system will cause variation in the
receiver gain and the output of the AGC detector. If in phase, the scope
would show a line with positive slope; if antiphase, a line with negative
slope; if in-between, an ellipse or some open shape subject to time
constants and non-linearities. This arrangement would leave the receiver's
RF gain control intact and its effect on time constant and phase observable;
it appears to modify the discharge resistance seen by a 1uF cap at the RF
and 1st Mixer in addition to pulling down the AGC voltage applied to them.

Does that seem to be a practical approach, Len?

If that tells you what you want to know, it is practical.

However, the phase information from that Lissajous display is
rather gross. If, with a closed-loop condition, there is
marginal stability, then a better handle on phase response
would be necessary...or just reducing the AGC control-line
magnitude (which would offer less AGC action).

I'll have to presume the Eico scope doesn't have a slow sweep
rate. If that scope has a DC input on both horizontal and
vertical, then the cobbled-together low-frequency source could
be built with a ramp output that would act as the horizontal
sweep; the display would then be just one cycle but that would
indicate the phase difference. Suggestion for source: Exar
XR-8038 DIP which has both square-wave and sine outputs.

A "bounce-less" switch circuit can be put together out of two
NAND gates connected as an R-S flip-flop, an SPDT switch
grounding/earthing one input on each NAND gate. That simulates
a very extreme "attack" situation to check the response of the
AGC control-line circuit. It's a bit much to infer anything
of numerical value out of that, though, since the amount of
analysis of the waveform out of the AGC control-line is lengthy
and probably more time than it's worth.

I'll have to remind all that a reasonably-calibrated RF signal
source is also necessary. That will yield both the open-loop
gain and the closed-loop gain...which can then be applied to
a standard negative-feedback amplifier formula. Even with a
"cheap" RF signal source, an RF output voltage meter circuit
(even if a 1N34 diode rectifier is used, good to ~ 30 MHz) will
provide a maximum RF output level. Resistor Tee or Pi pads
built on DPDT switches (cheap slide switches work out best due
to least internal inductance) external to the RF generator are
effective although not to the wideband accuracy of the
waveguide-below-cutoff type used in older commercial RF
generators. A sequence of 1, 2, 3, 5, 10, 20, 40 etc db pads
would do well enough. If needs be to make the pads the most
accurate, a spoiler pad of around 10 db at the start of this
chain of pads would insure a good source impedance. While
not of greatest metrology quality, those would be better than
nothing at all.

Note on the above: The RF signal generator meter would
determine the signal level into the attenuator chain. The
chain's output would then be that value minus the total db
of the attenuators switched-in. Making the attenuator-
switch mountings in-line in an outboard long metal box
having 1:2 ratio of width to height will reduce most of the
RF feed-around across switched-in attenuators; if that is
1 x 2 inches it is roughly high C-Band waveguide size and a
maximum of 30 MHz RF input would certainly be below cutoff
frequency of that "waveguide." Attenuation through that
long metal box would be a linear relationship of db v.
length. I did just that with an old Heathkit RF generator
(meter calibration set against lab equipment) and outboard
switched attenuators...until I lucked-out and obtained a
pair of HP 355 step attenuators (wideband to 500 MHz,
easier to use).

wrote in message
oups.com...
[snip]
However, the phase information from that Lissajous display is
rather gross. If, with a closed-loop condition, there is
marginal stability, then a better handle on phase response
would be necessary...or just reducing the AGC control-line
magnitude (which would offer less AGC action).

I thought that phase errors of a few degrees would not be an issue.


I'll have to presume the Eico scope doesn't have a slow sweep
rate. If that scope has a DC input on both horizontal and
vertical, then the cobbled-together low-frequency source could
be built with a ramp output that would act as the horizontal
sweep; the display would then be just one cycle but that would
indicate the phase difference. Suggestion for source: Exar
XR-8038 DIP which has both square-wave and sine outputs.

The scope does not go below 10 Hz sweep. I have a simple gen board that can
be pushed to 3 Hz and maybe lower with mods.

A "bounce-less" switch circuit can be put together out of two
NAND gates connected as an R-S flip-flop, an SPDT switch
grounding/earthing one input on each NAND gate. That simulates
a very extreme "attack" situation to check the response of the
AGC control-line circuit. It's a bit much to infer anything
of numerical value out of that, though, since the amount of
analysis of the waveform out of the AGC control-line is lengthy
and probably more time than it's worth.

I'm hoping that the qualititative observation would get me headed in the
right direction


I'll have to remind all that a reasonably-calibrated RF signal
source is also necessary. That will yield both the open-loop
gain and the closed-loop gain...which can then be applied to
a standard negative-feedback amplifier formula. Even with a
"cheap" RF signal source, an RF output voltage meter circuit
(even if a 1N34 diode rectifier is used, good to ~ 30 MHz) will
provide a maximum RF output level. Resistor Tee or Pi pads
built on DPDT switches (cheap slide switches work out best due
to least internal inductance) external to the RF generator are
effective although not to the wideband accuracy of the
waveguide-below-cutoff type used in older commercial RF
generators. A sequence of 1, 2, 3, 5, 10, 20, 40 etc db pads
would do well enough. If needs be to make the pads the most
accurate, a spoiler pad of around 10 db at the start of this
chain of pads would insure a good source impedance. While
not of greatest metrology quality, those would be better than
nothing at all.

Been meaning to build something like that. I need some basic, low cost gear
for RF/IF testing.

Note on the above: The RF signal generator meter would
determine the signal level into the attenuator chain. The
chain's output would then be that value minus the total db
of the attenuators switched-in. Making the attenuator-
switch mountings in-line in an outboard long metal box
having 1:2 ratio of width to height will reduce most of the
RF feed-around across switched-in attenuators; if that is
1 x 2 inches it is roughly high C-Band waveguide size and a
maximum of 30 MHz RF input would certainly be below cutoff
frequency of that "waveguide." Attenuation through that
long metal box would be a linear relationship of db v.
length. I did just that with an old Heathkit RF generator
(meter calibration set against lab equipment) and outboard
switched attenuators...until I lucked-out and obtained a
pair of HP 355 step attenuators (wideband to 500 MHz,
easier to use).




You're a wealth of info, Len. Because of impending holiday, I'm going to
have to shelve this for a few weeks. Hope to get back to it in July...

73, Tom