I have no experience with the R7 but some experience with AGC loops in general. It is VERY difficult to design a good AGC and many radios get it wrong. For example, designing it for the user's eyes (vs ears) means paying too much attention to the S-meter indication. Using the loop feedback voltage to drive the meter in a semi-calibrated way may be convenient but it could easily result in making poor loop tradeoffs and that is what AGC design is all about, making those tough tradeoffs in an optimal manner. You are trying to adjust the signal levels in the chain to maximize linearity (in order to get good dynamic range), while at the same time leveling the avg overall gain for the benefit of your ears. Because it is a loop it can do all the nasty things that loops often do, such as overshoot, ring, and oscillate. When you change the AGC time constant cap, the loop dynamics and stability parameters change. A FAST setting means the loop attack has further amplitude to restore, meaning any tendency to overshoot and distort will get worse.
So speaking in general terms only (not the R7 in particular), the things one might try when overshoot and distortion on peaks is present could be the following: Lower the loop gain to reduce instability. For example, reduce IF gain. Some radios have a pot for this. This is why retarding the panel RF gain control works. You could try playing with the AGC threshold, but that will probably affect linearity/dynamic range adversely and will not change the feedback loop gain once all stages are engaged with feedback. Use a scope to determine the stage at which the distortion is being introduced and then change the distribution of AGC via adjusting the resistor values that determine how much AGC gets to each stage. If the distortion is occurring at the end of the chain in the audio section then changing AGC distribution won't help that. You would need to reduce overall gain. Sometimes the IF filter is the primary source of delay that leads to instability. I once put an aftermarket 250Hz crystal filter in place of a stock 500Hz. The group delay (i.e burst response) went from 0.5ms to 4ms, which caused a huge pop on the leading edge of CW in FAST and a lesser but still very annoying pop on SLOW. I would not expect a SSB width filter to exhibit this degree of delay, but maybe.... Try inserting a resistor in series with the AGC time constant cap and fiddling with the value. This will introduce a zero in the loop transfer function which could be adjusted to cancel a pole that may be exacerbating the loop instability. Some radios already have a pole-zero combo (series R and C) somewhere in the AGC loop just for this purpose, but not all radios. Bottom line is that designing good AGC systems is a tough job and anything we as users change within the AGC loop to compensate for original design deficiencies will interact with other loop characteristics and the meter readings. It's hard to come out ahead overall, which is why so many AGC mods to various radios end up with adverse tradeoffs. Just my humble two cents worth... Dennis AE6C On Sat, May 11, 2013 at 2:16 PM, Jim Shorney <jshor...@inebraska.com> wrote: > On Sat, 11 May 2013 22:52:57 +0200, Dieter Horst wrote: > > >@Jim, would you say that the service manual describes the right methode > >for aligning the AGC? > >I have had my doubts about it in the past, mostly because 'fast' would > >always create trouble. > > > Yes. In my discussions with John, he told me that the "ACG Bias" adjustment > R1127 (A.K.A. "AGC Pedestal") is basically not used, being left as set in > the > instructions (fully clockwise). The IF gain pot is actually used to set > the AGC > pedestal in this scenario. I've tried two or three alternate procedures, > both > of my own and other's devising, but once I understood what was going on the > "book" method seemed to work the best. There are two sections, coarse and > final > adjustment. If things are "pretty close", coarse probably does not need to > be > done. > > 73 > > -Jim > NU0C > > -- > The universe we're in will reach absolute zero in three hours. Safe is > relative. - Idris, "The Doctor's Wife" > > > > _______________________________________________ > Drakelist mailing list > Drakelist@zerobeat.net > http://mailman.zerobeat.net/mailman/listinfo/drakelist >
_______________________________________________ Drakelist mailing list Drakelist@zerobeat.net http://mailman.zerobeat.net/mailman/listinfo/drakelist
