Tom ..there are the two filament leads downstream of the filament choke....the input is coupled to each filament lead thru an .o1cap.... one each on each filament lead...I have always used HV mica caps for this purpose...How important is it that these two caps be matched? I have always just found a couple from the junk box that looked alike and had the same (.o1) value on them and never gave it any more thought....But after reading the recent post by Don KYV on the testing of old mica caps I see that there can be quite a difference in actual capacity vs what the cap has printed on it.....I have also noticed that some GG schematics show a value of .1 and others use .o1......
I have started gathering parts for the deuce 4x1 GG....I have the tubes, 2 eimac air sockets , a 60 amp fil trans, 1 chimney, a blower, a 10 x 17 chassis, meters, plate choke...... I have 1 vac variable for the tune side of the out put and I have might have to use a breadslicer for the load side..Probably use plug in coils on the output since I have lots of them. ..... What values of caps and coil do you use for your tuned input? I probably won't get this thing built this winter as I have other projects going but the accumulation of parts is going well.... Steve
Hi Steve,
You brought up some good points...
Over the years, I've heard MANY rigs with hum on the carrier as a result of filament imbalance problems. There's many ways it can happen. With indirectly heated cathodes, like the 8877, etc, there needs to be a TRI-filar wound filament choke. Using a standard bi-filar causes hum.
With directly heated cathodes, like the 4X1, 3-500Z, etc, balance is also important, though a bifilar choke is OK. When I select the filament bypass caps, I consider the input impedance. When doing something real low, like the quad 4X1's, I usually use something like .05's. Any input up over 50 ohms can usually get away with .01's. The math is to calculate the reactance of the cap on the lowest band, like 160M, and be sure it is at least 10 times lower than the input impedance, or more. Otherwise, you may see amplifier instabilities. The fil choke should also be 10:1 in impedance ratio - 10 times higher (or more) than the input impedance of the tube. (at the lowest operating frequency) If it takes more drive on the lower bands for a given power output, this may be a clue that your fil choke needs more inductance. I've also seen the amp get unstable as a result, too.
Being on the cold side of the choke lessens this requirement somewhat, just like the bypass cap on the cold side of the plate choke. But might as well do it right.
As far as balance of the bypass caps... I usually just check the caps on the cap checker before use. It's not hard to find a pair at .01 that are real close. It might not be a big deal to get them closely matched - just a matter of degree. As long as there is not a big mistake in imbalance of the overall structure, your hum will be canceled out pretty well. One side of the fil to chassis, like I did last night, is a BIG mistake and makes the amp inoperable from severe hum.
** I would imagine that a LEAKING bypass cap could simulate a partial short to chassis and be a tough thang to troubleshoot. So maybe replace suspect bypass caps when you have eliminated everything else for hum problems.
As for the L/C input to the amp filament.... you can use a pi-network or a T network to match most any situation. If the impedance is close to 50 ohms already (like a pair of 4X1's) a simple L/C parallel resonant circuit to ground will work FB. Esp if you are driving it with a tube driver with a pi-network itself.
I would like to hear more from the guys here about filament balance and problems they've had with hum. I know Frank/GFZ had a long standing problem with his 4X3 amplifier 'til he realized the bifilar choke should not be wound on two separate rods. Hum problems like these can drive ya nuts, cuz you check all your other supplies for hum, audio leads, etc and the hum is still there. Hum induced by filaments in high power amplifiers is probably more common than most hams realize... :-)
Your dual 4X1 project sounds great. Oughta be a beauty.
Later -
Tom, K1JJ
