MMMMM I can just smell that Memo-graph blue copy now....!!
There was the "smellevision" fad at movie houses. I'd like to resample those old solvent odors from the days of my youth and hand cranked copiers, but alas.. Like a bad but funny sci-fi movie where all household food, chemcials, and waste was pumped/transported through a plethora of dedicated pipes and tubing interconnecting residences with central plants, it might have been possible to share the aromas of seasoned old radios and MFP over the internet if infrastructure had evolved differently.
many guys wrote articles but few got the 4-1000A to be stable 160 through 10m.
This I don't doubt. It's always been a popular tube. The cited "memo-graph" article was included in the weighty tome concerning the transmitter, that I was allowed to copy. Because of that I have adhered to the writings therein. I admit that the article was sent to the builder, Colonel Tucker, because of his many problems getting rid of the instability caused by a 20" plate line (he was a BC man, not used to full HF or VHF) and his unfortunate reliance on a new but defective RF power meter.
speaking of the MB-40: single-page article in QST, February 1956, pg. 45.
It's here. I believe fair use of this single page applies in this case and would like to discuss it.
The neutralization circuit doesn't look right to me. I don't see the purpose of C1. Kinda a cross between "grid neutralization" with balanced grid circuit as used for triodes and "bridge neutralization" as described in the Eimiac literature for tetrodes. Maybe I'm missing something.
I had stability issues wit my 4-1000 until I used the bridge neutralized circuit from the Eimac book. I think the ARRL handbook has the same circuit, with a tab of metal next to the tube forming the neutralization capacitor. Mine is stable as a rock on 160-20M with no grid swamping resistor. It does not use the MB-40 however.
C1 is a value chosen to be the same value as the 4-1000 grid capacitance. Its purpose is to balance the tuned circuit against the grid to ground capacitance on the opposite side of the MB-40, which I believe we may consider to be a standard LC tank for this purpose but I could be wrong.
The neutralizing cap C2 is a metal rod of #10 wire passing up through the chassis. I believe it serves the same purpose as the metal tab. The swamping is only to improve the match at different frequencies without having to add extra coils as shown in the ARRL document and in a few other places. It is easily driven from 80-10 and does not seem to take too much power, not the 70W described in the article. The link works best fully engaged in this case and has been set to 4 turns.
I guess I have to pull it back out again and look. Maybe I need the proper CRT shield, I don't have one yet but how could the thing oscillate with a 3200 Ohm resistor & blocking cap from grid to ground, and there be little or no current drawn to make it so?
Didn't I tell you somewhere else that the filament leads going to the 4-1000 were too long? Remove them and make them twister pair. If necessary get some of the same size wire with modern insulation just a little bit longer and twist them together if the existing wires are too short to twist. Even if you use the grounded center tap scheme I would still twist the filament wires together. Just having the metal from the CRT next to them is probably causing a problem. Chances are that is the problem.
There is no way to shorten them.. I could twist them or run a 'zip cord' pair which ought to be the same. There is some 10 GA speaker wire around here. Considering the filament is grounded at one side and heavily bypassed, I don't see the point but it is one of the things I will try. What would be the real difference between twist and zip cord arrangements? That circuit is got to be at very low impedance.
How much RF current can really flow in the cathode of a 4-1000 at full bore? Peak maybe 6A? That across that fat GND tab should have eliminated other worries.
As for the CRT being too close to the socket, I'm not convinced unless the 'floating' electron gun structure is coupling RF elsewhere. The capacitance between any 4-1000 element and that entire CRT should be very low, only a few pF. I recently learned the CRT has a resistive helical post deflection accelerator and that it goes from about where the blue ceramic begins, all the way to the phosphor screen of the CRT. It is supposed to be spec at 200-1500 Meg Ohms.
(Thanks to a book on CRTs. "The cathode Ray Tube - Technology, History, and Applications" by Peter A. Keller., who was also kind enough to send me a copy of the data sheet on the tube in question. Did anyone suspect that Tektronix CRTs would have EIA registrations? I thought they were strictly in-house proprietary components. This one, from a 545, a "5ELP2")
This makes my request in "wanted" for a proper CRT shield all the more important to the project. I can ground a shield. I can't ground the stuff in the tube and the RF needs to be kept away from it even though the effect ought to be almost negligible.
Finally I have not measured the frequency of the oscillation. That ought to give a better clue as to where the parasite is. I should have time to do this next week as I am on vacation AUG 03-19 and have today completed almost the last chore except for starting and running two seldom used vehicles and removing the stale gasoline from the ATV.