Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Jun Sat 06, 2020 10:04 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 10 posts ] 
Author Message
 Post subject: Fringe howl and grid leak resistors.
PostPosted: Feb Sat 10, 2007 2:28 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 455
Location: Kalamazoo, MI USA
Had a problem with fringe howl yesterday. Electric Bill recommended increasing resistance of the grid leak resistor. Rocky, in his book recommends the same maneuver. In fact, he says that using a 5 M ohm resistor for the grid leak often improves sensitivity along with removing fringe howl.

My question is if using a 5 M ohm resistor for grid leak has no down side, why doesn't everyone do it? Why don't the plans for the sets we build spec a 5 M ohm resistor for the grid leak, rather than the 1.5 M ohm resistor that was called for in my particular set, for example?

Best Regards,

Rob


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 2:38 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20548
Location: Sandpoint, IDAHO 83864
You got to remember that the grid leak resistor along with the grid capacitor sets up an R/C time constant. The tube is detecting an audio signal from the RF signal and the time constant in the grid circuit has to be short enough to allow some audio fidelity to pass thru. I think the first thing you notice when the time constant is made too long, is the lack of high frequencies in the recovered audio. Everything in this world of regenerative sets is a trade-off.
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 2:51 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 455
Location: Kalamazoo, MI USA
Curt Reed wrote:
You got to remember that the grid leak resistor along with the grid capacitor sets up an R/C time constant. The tube is detecting an audio signal from the RF signal and the time constant in the grid circuit has to be short enough to allow some audio fidelity to pass thru. Everything in this world of regenerative sets is a trade-off. I think the first thing you notice when the time constant is made too long, is the lack of high frequencies in the recovered audio.
Curt


I did notice that while my fringe howl is gone, my audio is a little off. Is it possible that I'm at the point where my audio is being compromised? I have 3 M ohm resistance on the grid. See this post. http://www.antiqueradios.com/forums/vie ... hp?t=67305

Best Regards,

Rob


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 3:11 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20548
Location: Sandpoint, IDAHO 83864
Could be. Remember that this time constant can also be varied with the value of the capacitor being changed.

Another thing to keep in mind is that a lot of the older designs were intended for CW usage, and the audio quality was not paramount in the designer's eye. If you look at the circuit, you will see that the grid leak network connects to the coil and thus will load down the coil, reducing its Q. This is another reason the old timers used high values of the resistor, to minimize loading of the circuit which gave better selectivity. With a higher value of resistor, you would have to go lower in value on the capacitor to maintain the same time constant. Smaller values of capacity also reduce the coupling or loading of the tank circuit.

I read a rather lengthy treatise on this subject several years ago and it went into some pretty deep detail considering it was written around 1930 or so. I just wish I could remember where I read it at. Probably in an old issue of QST magazine.
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 3:31 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 455
Location: Kalamazoo, MI USA
Curt Reed wrote:
Could be. Remember that this time constant can also be varied with the value of the capacitor being changed.

Another thing to keep in mind is that a lot of the older designs were intended for CW usage, and the audio quality was not paramount in the designer's eye. If you look at the circuit, you will see that the grid leak network connects to the coil and thus will load down the coil, reducing its Q. This is another reason the old timers used high values of the resistor, to minimize loading of the circuit which gave better selectivity. With a higher value of resistor, you would have to go lower in value on the capacitor to maintain the same time constant. Smaller values of capacity also reduce the coupling or loading of the tank circuit.

I read a rather lengthy treatise on this subject several years ago and it went into some pretty deep detail considering it was written around 1930 or so. I just wish I could remember where I read it at. Probably in an old issue of QST magazine.
Curt


Yes, Rocky mentioned a smaller cap in his book. The set was built with a 100 pf cap at the grid. Since I doubled the grid resistance, should I halve the cap? Is it a 1:1 relationship?

Best Regards,

Rob


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 4:13 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20548
Location: Sandpoint, IDAHO 83864
T= RxC is the formula. Yes, it is a pretty linear relationship, and if you double one, you can cut the other one in half an still come up with the same thing.
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 6:27 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1576
Put a resistor across the audio transformer or choke. 220K is shown in Lindsay's books on chokes. I would just put my resistor box across it and see what works best. You can also put a resistor across the secondary of the transformer and shunt some audio to ground, this will fix it at the cost of some audio. It is easiar to use resistance coupling in the first place.

Also, how many tickler turns are you using? My SW coils have at most three or four turns. A coil that tunes 10 mhz will have one turn, spaced 1/8" form the main coil. An 80M coil might have three turns. The old books suggested way to many turns.

If you are runnig a triode make use about 20V on the plate. The set will work much better this way. Running it at high voltage is a mess.


Top
 Profile  
 
 Post subject:
PostPosted: Feb Sat 10, 2007 7:58 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 455
Location: Kalamazoo, MI USA
Scot Armstrong wrote:
Put a resistor across the audio transformer or choke. 220K is shown in Lindsay's books on chokes. I would just put my resistor box across it and see what works best. You can also put a resistor across the secondary of the transformer and shunt some audio to ground, this will fix it at the cost of some audio. It is easiar to use resistance coupling in the first place.

Also, how many tickler turns are you using? My SW coils have at most three or four turns. A coil that tunes 10 mhz will have one turn, spaced 1/8" form the main coil. An 80M coil might have three turns. The old books suggested way to many turns.


Scot,

I don't have an audio transformer on this set as I'm using headphones.

You are right about the tickler. Rocky recommended the same thing in his book. I don't think that I have any more than three turns. I built the coils using his specs.

Best Regards,

Rob


Top
 Profile  
 
 Post subject:
PostPosted: Feb Wed 14, 2007 7:28 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 251
Location: Scotch Plains, NJ, USA
I restored an old one-tube homebrew regen set with a "variable" grid leak control (fixed mica cap and an open carbon disk pot). Very archaic design, but it works...

I am glad that I read this post, because I never really understood why they put in a variable grid leak control. You always seem to learn something on this site!

Thanks.


Top
 Profile  
 
 Post subject: Re: Fringe howl and grid leak resistors.
PostPosted: Jul Thu 18, 2019 8:18 pm 
Member

Joined: Nov Mon 06, 2017 2:35 pm
Posts: 120
Location: Texas, U.S.A.
1. The reason that all grid leak detector designs do not use very high resistance grid leak resistors is that the grid leak resistance is not the cause of fringe howl (also known as threshold howl).

2. In small signal operation of the grid leak detector the function of the grid leak resistor is to set the grid bias, in other words to set the operating point on the grid voltage vs. grid current curve. The demodulated audio is developed in series with the grid resistance (Rg). The audio frequency response at the grid is that of a low pass filter. The -3 dB point is determined by the capacitance of the grid condenser and the grid resistance Rg. Increasing the grid leak resistance increases Rg. The grid leak resistance is also part of the audio frequency load impedance on the grid.
(F. E. Terman, "Grid-leak Grid-condenser Detection", Radio Broadcast, Mar. 1929)
(F. E. Terman, "Some Principles of Grid-leak Grid-condenser Detection", Proc. Institute of Radio Engineers, Oct. 1928)
(W. L. Everitt, Communication Engineering 2nd ed., 1937, pp. 418 - 421)

3. Fringe howl is an audio frequency oscillation ocurring just as the amplifier starts into radio frequency (RF) oscillation. It is more likely to be encountered in detectors using an audio choke plate load or audio transformer plate load.

Possible causes of fringe howl:
1. Excessive tickler inductance (too many turns) and excessive coupling (mutual inductance) from tickler to grid coil (as mentioned by previous commenters).
(S. Kruse, "The Receiving Experimenter" QST, Jul. 1924)

2. Interaction between an inductive audio plate load (i.e. choke or transformer primary) and change in plate resistance when RF oscillation starts. This is where changing the grid leak resistance may move the plate resistance to a point at which howl does not occur, but increasing the leak resistance will likely have deleterious effect on the audio frequency response and increase susceptibility to external noise fields. Reducing the leak resistance will reduce the demodulated audio amplitude for a given RF signal level. Instead, placing a resistance across an inductive plate load (on primary side of a transformer) limits the maximum audio plate load impedance to essentially the value of the resistance, while retaining the advantage of the choke or transformer primary in establishing the dc plate voltage when only relatively low available B+ supply voltage is available (e.g., 45 volt battery). Zepler suggests 10,000 to 20,000 ohms but higher resistance may be found sufficient. (E. E. Zepler, The Technique of Radio Design, 1943, pp. 231-232)
If around 150 volts of B+ or more is available, a resistor is preferred as the plate load for a screen grid or pentode detector.

3. Amplitude modulation of the RF input to the detector caused by coupling of the detector grid circuit by common impedance to an audio stage of the receiver. The common impedance may be internal resistance of a battery or wiring resistance and inductance. Providing a couple of microfarads of bypass capacitance between ground and B+ near each tube may be required.
(E. E. Zepler, The Technique of Radio Design, 1943, pp. 212-213, 231-232)
(J. M. Thomson, "Motorboating and Howling" QST, Nov. 1927)

4. An RF choke between the tickler and inductive audio frequency plate load must be effective at the operating radio frequency of the detector. I would choose an RF choke that will be operating below its self resonant frequency and that will provide reactance of at least ten thousand ohms at the operating radio frequency.


Eric LaGess
WB5HDF


Top
 Profile  
 
Post New Topic Post Reply  [ 10 posts ] 

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 2 guests



Search for:
Jump to:  


































-->


Privacy Policy :: Powered by phpBB