On Mon, 6 Jun 2011 14:22:07 +1200, jeffers...@xtra.co.nz wrote: >The TR-4C does receive but, as the circuitry was designed using a Remote >Cutoff Pentode, is the use of a Sharp Cutoff Pentode degrading the performance >of the Rig ? >Should I replace the 12AU6 with a 12BA6 or leave well enough alone?
The following discussion of remote vs. sharp cutoff just came over the R390 mailing list. Although it specifically addresses the R-390 IF deck, the concepts are the same: <quote> I wrote: >The 6BA6 is a "remote cutoff" pentode, meaning that it takes a >relatively large negative grid bias to cut off plate current. The >6AU6 is a "sharp cutoff" pentode, meaning that its plate current cuts >off with a much smaller negative grid bias. Remote cutoff tubes are >generally used in stages with AGC control, to provide a proper linear >AGC action. > >In any case, all of the tubes on a common AGC bus should have the >same cutoff characteristic, so the IF gain is distributed properly >between the stages. To elaborate a bit: With 100 V on the plate and screen a 6AU6 is cut off with a grid voltage of -4.2 V (all tube parameters, voltages, and currents mentioned in this post are approximate). Cutoff is defined on the datasheet as a plate current of 10 uA or less. In typical operation, the 6AU6 with 100 V on plate and screen operates at 5 mA with a grid voltage of just over -1 V and a transconductance of 3900 umho. The transconductance does not change radically with grid voltage around this operating point, although it does change very rapidly between cutoff (-4.2 V) and the vicinity of the operating point (-1 V). A 6BA6 with 100 V on the plate and screen is cut off at a grid voltage of -20 V. Right away we see that it is intended for a different use: in this case, cutoff is defined as a transconductance of 40 umho, not as a particular plate current. In typical operation, the 6BA6 with 100 V plate and screen operates at 10.8 mA with the AGC wide open (or nearly so), also with a grid voltage of just over -1 V, and a transconductance of 4300 umho. However, it is designed to be used throughout the range of grid voltages, not just at one operating point. The transconductance of the 6BA6 changes smoothly with grid voltage, from around 5000 umho with a grid voltage approaching 0 to less than 10 umho with a grid voltage of -30 volts or so. Plate current never really does "cut off" -- the transconductance just gets lower and lower as the grid goes more negative. Because the amplification of a tube is directly related to its transconductance, we see that the 6AU6 is intended to maintain its amplification near a design value with reasonable changes about its nominal operating point, while the 6BA6 is intended to be a variable gain amplifier, with the gain input on the same element as the signal (i.e., the grid). (There are other possibilities -- gain can be varied by changing the screen voltage of a pentode, or one of the other grid potentials on a multi-grid tube, or changing the effective cathode resistance, for example by using a differential pair of tubes, or lots of other ways.) What does this mean for a radio's IF chain? The grid bias in AGC'd stages is set by the AGC line. Let's assume that a radio has typical AGC (for tube radios) that runs around -1 or -2 volts wide open on band noise, and reaches -15 V or more when it is hit with a lightning crash or your neighbor fires up a kilowatt to call the DX station you are listening to. Let's also assume you have your AGC set to Slow, just so we have time to analyze what is happening (it works the same with fast AGC, just ... well ... faster). Your neighbor just stopped transmitting, so the AGC line is recovering from -15 V or more. As it rises, the gain of the remote cutoff amplifiers increases smoothly, and the band noise (with that buried DX signal) also rises until the AGC stops it at a grid voltage of, say, -1 volts. Now replace one IF tube with a 6AU6. As the AGC line recovers from -15 V or more, the gain of all of the AGC stages except that one increases smoothly. When the AGC line reaches -6 volts, the 6BA6 stages are operating with a transconductance of 500 umho -- about 20 dB less than their maximum gain -- but the 6AU6 stage is still solidly cut off, and passing no signal. So, the IF as a whole is still doing nothing because no signal gets through the 6AU6 stage. By the time the AGC voltage reaches -4.2 V, the 6BA6 stages are up to a transconductance of 1000 umho -- within 12 dB of their maximum gain -- and the 6AU6 stage is just starting to pass signal (if that stage were another 6BA6, it would be operating within 12 dB of its maximum gain, too). So the IF as a whole is still doing essentially nothing. Then, over the next 2 V or so of AGC potential, the 6AU6 suddenly catches up and then passes the gain of the 6BA6 stages. The net result is that even with just one sharp cutoff tube, the useful AGC range has been compressed from around 20 volts to around 2 volts. The perceived results will vary according to where in the IF string the 6AU6 is located. If it is the last IF amp, the radio will be fairly quiet as the AGC recovers and then the band will rather suddenly pop in. If it is the first IF amp, the internal IF noise (but not the band noise) will increase steadily as the later IF amps increase in gain, then the band will rather suddenly pop in. As you replace more of the IF amps with sharp cutoff tubes, this "popping in" of the band will get more and more pronounced. The sharp cutoff stage will also have higher distortion. All tubes have harmonic distortion (largely second harmonic) because the transconductance varies with grid voltage, and thus the amplification varies from the negative signal peak to the positive signal peak. However, the much lower change in transconductance per incremental grid voltage of the remote cutoff tube (which goes from near zero to 4300 umho over a range of 30-odd volts of grid potential) compared to the sharp cutoff tube (which goes from near zero to 3900 umho over a range of only 2 or 3 volts of grid potential) means that the sharp cutoff tube will have much more distortion in the region where its transconductance is variable (which is where it will necessarily be operated in an AGC'd stage). This will not be audio distortion (at least not directly) -- rather, it will generate harmonics of the IF frequency, which may bleed into the RF and IF circuits to cause spurious responses as well as distortion at the detector. So, using sharp cutoff tubes in AGC stages is not a recommended practice. Best regards, Don </quote> 73 -Jim -- Ham Radio NU0C Lincoln, Nebraska, U.S.S.A. TR7/RV7/R7A/L7, TR6/RV6, T4XC/R4C/L4B, NCL2000, SB104A, R390A, GT550A/RV550A, HyGain 3750, IBM PS/2 - all vintage, all the time! "Give a man a URL, and he will learn for an hour; teach him to Google, and he will learn for a lifetime." HyGain 3750 User's Group - http://groups.yahoo.com/group/HyGain_3750/ http://incolor.inetnebr.com/jshorney http://www.nebraskaghosts.org _______________________________________________ Drakelist mailing list Drakelist@zerobeat.net http://mailman.zerobeat.net/mailman/listinfo/drakelist
