[R-390] AGC Slow - Moment of silence
David Wise
David_Wise at Phoenix.com
Fri May 13 16:12:08 EDT 2016
If you have refurbished your C551, here is an AETMOS variant that takes advantage of it. Call it "AETMOS-51".
1. Refurbish C551
2. Disconnect and tape off the wire (ground) at S107-9
3. Connect C591, a 0.15uF 400V cap, from S107-8 to S107-9
4. Connect R590, 10M, from S107-7 to S107-8
SLOW: Unchanged (V506A with Miller feedback).
MED: Instead of using C551 passively, we operate like SLOW, only with a smaller cap.
C590 (0.15uF) in series with C551 gives about 1.9uF of effective capacitance.
R590 (10M) keeps the high end of C551 near V506A's plate voltage, and keeps C591 discharged.
FAST: Unchanged. R590 (10M) keeps the high end of C551 near V506A's plate voltage.
==================================
And here is an improved AETMOS write-up, which replaces yesterday's write-up.
I thought of a mod that's halfway between Larry's simple mod
(20uF in series with S107-8 plus 10Meg to ground)
and my "Ending The Moment Of Silence" article.
Like Larry's mod, and ETMOS, it affects only the front panel.
THE PROBLEM
The problem with Larry's mod is the extremely long time to charge C590 (20uF) through R590 (10Meg).
The time constant is 200 seconds. It takes about five TC's to get near full charge. That's 17 minutes!
If you are listening to a weak signal, all is well.
If you switch from SLOW to MED, all is well.
If you switch from MED to SLOW on a stable signal, all is well.
If you switch from MED to SLOW on a signal that recently got stronger, you will get a Moment Of Loudness.
If you switch from MED to SLOW on a signal that recently got weaker, you will get a Moment Of Quiet.
You will get a MOQ, not a full MOS, because the jump is divided by V506A's gain.
The cause behind MOQ and MOL is that V506A's plate voltage changed, taking the low
end of C590 with it, and R590 hasn't had time to pull it back to ground.
It will happen once in a while, seemingly out of the blue, depending on your habits.
During the 17-minute window after a fade, you will get a MOQ/MOL every time you go SLOW/MED.
The depth and length decrease with time and number of switches.
The following mod shortens the window from 17 minutes to 1.7 minutes.
Plus, you don't need a 20uF cap.
It can fit on the same five-position terminal strip that's pictured in ETMOS.
THE MOD
Refer to the IF Module schematic, Figure 5-22, e.g. Y2KR3 Chapter 5 page 43.
1. Disconnect the wire at S107-8 (V506A plate)
2. Connect the wire (V506A plate) to two new caps, 2uF and 0.15uF, C590 and C591
3. Connect the other end of C590 to S107-8 (SLOW contact)
4. Disconnect the wire (ground) at S107-9 (MED contact)
5. Connect a 10Meg resistor, R590, between S107-8 and ground
6. Connect the other end of C591 to S107-9
7. Connect a 10Meg resistor, R591, between S107-9 and ground
8. Disconnect and tape off the wire at S107-7 (wiper)
9. Connect S107-7 to S102-5 (MGC contact, and AGC output to TB102-3)
Now the TC is only 20 seconds instead of 200.
You only get one MOQ, instead of over and over when you switch back and forth.
The MOQ is the same depth and length, but the window of opportunity is shortened.
To close it entirely, you have to do ETMOS, or Charles' all-passive mod.
I think I will call this mod "Almost Ending The Moment Of Silence", AETMOS. :)
PRINCIPLES OF OPERATION
C551 is abandoned in place.
V506A's grid goes directly to S107-7.
V506A's plate goes to two caps, which in turn go to S107-8 (SLOW) and S107-9 (MED).
Each cap is also returned to ground through a resistor.
SLOW: Unchanged (V506A with Miller feedback), except the cap is on the front panel.
MED: Instead of using C551 passively, we operate like SLOW, only with a smaller cap.
0.15uF gives about 2uF of effective capacitance.
FAST: Unchanged.
In MED and FAST, R590 pulls the low end of C590 to ground.
If it isn't at the AGC voltage when you go SLOW, AGC will jump then
taper off. The effect is proportional to how far behind C590 is,
i.e., the change in signal strength just before SLOW.
The effect is also biased toward MOL. With constant signal
strength, you will get a 10% MOL that you may not notice.
(This asymmetry is not present in Larry's mod, because we connect
to AGC a capacitor terminal that was previously returned to ground.
ETMOS has the same asymmetry and I never notice it.)
In SLOW and FAST, R591 pulls the low end of C591 to ground.
If it isn't at the AGC voltage when you go MED, AGC will jump
then taper off, but it's too brief to notice.
In SLOW and MED, R590 or R591 loads the AGC line.
Since the line is 390K, this perturbs it by less than 4%.
The radio will play a tiny bit louder; you won't notice,
especially since it's probably coming off a MOL.
Regards,
Dave Wise
May 2016
-----Original Message-----
From: David Wise
Sent: Wednesday, May 11, 2016 10:11 AM
To: R-390 Forum
Subject: RE: [R-390] AGC Slow - Moment of silence
Thank you, Larry. Your sketch is correct, but I think it's important to show all of the wire to S102-5, so it's clear that the destination is V506A/G, and we are rerouting S107-7 around C551. It's a neater markup if you use Y2KR3's Figure 5-22 instead of TM 11-856A.
One can also disconnect and tape off the original S107-7 wire, which leaves C551 open instead of shorted. The result is the same, and the next guy will have easier understanding in case the documentation is lost. I'm editing it into the copied post below.
The problem I saw with your mod is a corner case which only manifests if you have certain habits. It happens to be something I'm likely to do.
I think this is the best we can do with the original S107. ETMOS substitutes a 2-pole switch, which makes a separate path to charge the off-duty cap quickly, closing the window almost entirely.
You can see drawings and pictures showing how I mounted the parts in the ETMOS article. (AETMOS can use the same hardware.) And if you figure out a better way, let me know!
Regards,
Dave Wise
-----Original Message-----
From: R-390 [mailto:r-390-bounces at mailman.qth.net] On Behalf Of Larry H
Sent: Tuesday, May 10, 2016 8:14 PM
To: R-390 Forum
Subject: Re: [R-390] AGC Slow - Moment of silence
David, I like this one that you made a lot. Although I haven't seen the problems you correctly point out with my simple one, I think this new one is a keeper. I through a quicky schematic together for what I believe you are suggesting - is this right? http://s32.postimg.org/aziojxup1/IMG_7216s.jpg Yours is not that much more work than mine and is much better - I'm going to try it.
Regards, Larry
From: David Wise <David_Wise at Phoenix.com>
To: R-390 Forum <r-390 at mailman.qth.net>
Sent: Tuesday, May 10, 2016 6:01 PM
Subject: Re: [R-390] AGC Slow - Moment of silence
I thought of a mod that's halfway between Larry's simple mod and my "Ending The Moment Of Silence" article.
Like Larry's (and ETMOS), it's front-panel only.
THE PROBLEM
The problem with Larry's mod is the extremely long time it takes to charge C590 (20uF) through R590 (10Meg).
The time constant is 200 seconds. It takes about five TC's to get near full charge. That's 17 minutes!
If you are listening to a weak signal, all will be well.
If you switch from SLOW to MED or FAST, all will be well.
If you switch from SLOW to MED or FAST then right back to SLOW, all will be well.
If you receive a strong signal for a while on FAST or MED, then switch to SLOW, all will be well.
BUT: If you receive strong for a while, *and then the signal fades out*, and then you switch to SLOW, you will get the same old Moment Of Silence you were trying to fix. (It's because V506A's plate dropped when the signal faded.) It will happen once in a while, seemingly out of the blue, depending on your habits. The intensity decreases the longer you stay no-signal before switching, finally petering out around the 15 minute mark.
The following mod makes it 10x better. Plus, you don't need a 20uF cap.
It can fit on the same five-position terminal strip that's pictured in ETMOS.
THE MOD
Refer to the IF Module schematic, Figure 5-22, e.g. Y2KR3 Chapter 5 page 43.
1. Disconnect the wire at S107-8 (V506A plate)
2. Connect the wire (V506A plate) to two new caps, 2uF and 0.15uF, C590 and C591
3. Connect the other end of C590 to S107-8 (SLOW contact)
4. Disconnect the wire (ground) at S107-9 (MED contact)
5. Connect a 10Meg resistor, R590, between S107-8 and ground
6. Connect the other end of C591 to S107-9
7. Connect a 10Meg resistor, R591, between S107-9 and ground
8. Disconnect and tape off the wire at S107-7 (wiper)
9. Connect S107-7 to S102-5 (MGC contact, and AGC output to TB102-3)
Now the TC is only 20 seconds instead of 200. You can still get a MOS - but the vulnerable window is 10x smaller. To close it entirely, you have to do ETMOS, or Charles' all-passive mod.
I think I will call this mod "Almost Ending The Moment Of Silence", AETMOS. :)
PRINCIPLES OF OPERATION
C551 is abandoned in place.
V506A's grid goes directly to S107-7.
V506A's plate goes to two caps, which in turn go to S107-8 (SLOW) and S107-9 (MED).
SLOW: Unchanged (V506A with Miller feedback), except the cap is on the front panel.
MED: Instead of using C551 passively, we operate like SLOW, only with a smaller cap. 0.15uF gives about 2uF of effective capacitance.
FAST: Unchanged.
In MED and FAST, R590 charges C590. Since it pulls C590 to ground, not to AGC, if you switch to SLOW on a strong signal, you will get a brief MOL (Moment Of Loudness).
In SLOW and FAST, R591 charges C591. Since it pulls C591 to ground, not to AGC, if you switch to MED on a strong signal, you will get a very brief MOL (Moment Of Loudness).
In SLOW and MED, R590 or R591 loads the AGC line. Since the line is 390K, this perturbs it by less than 4%. The radio will play a tiny bit louder; you won't notice, especially since it's probably coming off a MOL.
Regards,
Dave Wise
-----Original Message-----
From: R-390 [mailto:r-390-bounces at mailman.qth.net] On Behalf Of David Wise
Sent: Monday, May 09, 2016 11:35 AM
To: R-390 Forum
Subject: Re: [R-390] AGC Slow - Moment of silence
I would not notice a 20% reduction in the SLOW time constant.
The symptom appears when you switch to MEDium, but the situation got set up during SLOW. Your solution eliminates the voltage jump at the positive end of C551.
On paper, none of the commonly available 22uF/450V aluminum electrolytic caps has acceptable leakage, but I expect that actual parts beat spec by a wide margin. You can also use a film cap (a motor-run type can be less than $10) but it can't go behind the front panel.
While the low end of C590 averages 0V, you'll get a bump if you go to SLOW during an excursion or during the first minute or so after power-on.
I attacked the problem differently, running the Miller integrator on MED as well as SLOW. See "Ending The Moment Of Silence" at www.r-390a.net .
Larry's mod is easier and still pretty good, and I encourage the rest of you to try it first.
Regards,
Dave Wise
-----Original Message-----
From: R-390 [mailto:r-390-bounces at mailman.qth.net] On Behalf Of Larry H
Sent: Sunday, May 08, 2016 6:16 PM
To: R-390 Forum
Subject: [R-390] AGC Slow - Moment of silence
It looks like the problem with the circuit is the charge time of C551 when switching the AGC into the slow position. This causes the agc line to be pulled + for a second while C551 charges. Even if it had zero leakage, one end is switched from ground to B+ potential, causing the problem.
I came up with a fairy simple solution to this problem. Add a 10 meg to 20 meg resistor to the AGC switch contact 8 to ground. Remove the wire that goes to that same contact and add a 20 mfd cap (call it C590) in series with that wire to contact 8. This will maintain a charge on C590 all the time and will have very minimal affect on the operation of the AGC circuit. I know that this will reduce the effectiveness of the slow position a little, but I don't think you will notice it. Use the largest res that will overcome the leakage in C590 (should be very low), but I don't think it should be less than 10 meg. See this schematic.
http://s32.postimg.org/6310y6av9/IMG_7198s.jpg
I've been using this for a while and have not noticed any issue with it.
Regards, Larry
____
______________________________________________________________
R-390 mailing list
Home: http://mailman.qth.net/mailman/listinfo/r-390
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:R-390 at mailman.qth.net
This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
______________________________________________________________
R-390 mailing list
Home: http://mailman.qth.net/mailman/listinfo/r-390
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:R-390 at mailman.qth.net
This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
More information about the R-390
mailing list