The AM Forum

 :)  Holes are about to be drilled.  I'm building the exciter, since I was able to score the coils, and the 813 Amp at the same time.  While the PA will consist of parts acquired for the unit, the exciter is another story.  

That is going to be a jonque box project based on the Heath AT-1.  It won't be a perfect clone because I'm not going to buy any parts regardless of what I see on the schematic and want to make some changes, especially in the power supply.  

There are three RFCs.  2 x 1.1 MH in the cathode of the 6AG7 oscellot and another in the plate of the 6L6.  The one that bugs me is the most is the 90uh in the plate circuit of the 6AG7. I don't have anything as low as a 90, but have loads of 2.5s.  May have a 1.Something or 3 floating around. Does it matter all that much?

Mike

The biggest enemies of RF chokes are usually high voltage and inter-turn capacitive coupling, series resonance, etc.   If you are running less than 400 volts, I can't see any reason why the 2.5 mH chokes will not work FB in all three circuits, especially replacing the 1.1 mH parts.  RF chokes in that service are normally not critical.

The only reason they may be using a single layer?  90 uH choke is because of 40M and above inter-winding capacitance concerns, but try the 2.5 mH choke there and see. The worst that can happen is it smokes and you replace it.

T

I had an AT-1 that wasn't in too good shape.  I reduced the thing to screws and parts.  The cabinet was the only thing that was in perfect condition.

this is a good excuse to install the chokes and sniff them out with a grid dipper .... check especially the ham bands for resonances ... as Tom sez usually not a problem

In addition to a grid dipper,  a vna / Mfj swr analyzer can be used to find any resonances.

I test all my Homebrew plate chokes with a '259.  Cheap and easy,  and a lot easier to remove a turn to five before the choke smoke,  rather than after.... 

--Shane
KD6VXI

Yes. The key here is to test the choke when installed in-circuit.  The added surrounding tube and stray chassis capacitance and inductance can shift a choke's resonance that is on the edge to in or out of the danger zone. Checking them out of circuit just gives a false sense of security.

Though I have always believed it is difficult to intentionally wind a choke (by chance) that is series resonant in a ham band. Something like the odds of rolling dice -  and getting six sevens in a row. But it does happen sometimes.  

The commercial chokes usually have an even better record or we hams would have heard about the suspect ones by now.

Out of perhaps fifty homebrew chokes I have wound for big rigs and linears, (and never pre-checked them)  I can remember maybe three that imploded - and who knows what caused it, series resonance or something else...   That's a 6% failure rate record, which is better than Russian roulette I suppose.. ;D

BUT... My rigs usually cover 160-40 or  75 - 20M.   Try making a choke that will cover 160 - 10M and it becomes a tougher feat with higher odds of hitting a ham band. Axe Tim about his homebrew 160-10M 4-1000A AM rig. He finally settled on adding a relay to switch in a second choke before taming it.

T

The 6AG7 is an excellent tube choice for a crystal oscillator.

Unfortunately, many oscillator/buffer combinations in many boat anchors are over coupled and cause frequency "pulling" of the oscillator.  

In this circuit, this is avoided by the combination of the 90 uH RFC and the 100 pF capacitor to reduce interstage coupling, which minimizes frequency pulling when the load on the osc. changes.

If you change the osc. RFC value you may want to reduce the following coupling capacitor's value to 27 to 47 pF in order to minimize frequency pulling.

I would also add a .005 uF ceramic cap close to the bottom end of any of the plate chokes to ground.

An alternative is to take a 1 mH RFC and remove windings until you reach about 100 uH.

Phil - AC0OB

That makes more sense.  Anytime there is an "optimized" value like that, someone probably spent the time to find it by trial and error, especially in an oscillator with stability goals.

In a pinch, good idea about stripping down a choke until 90 uH is found.

T