Here are some of the preliminary answers and hints that I have received. Plus some comments and questions from my own experience with that tube on HF. A lot of it is common sense to some and may be new to others so please bear with me. - Many have tried and failed to tame the 4-1000 on 6M, particularly at outputs exceedind 800W or so. - Only one out of many could get the ARRL 1985 era Handbook circuit to work. - The input circuit must be completely isolated from the output. This means full shielding with complete bypassing of all non RF leads below deck. The 6M drive should come into the below deck pressurized compartment via a BNC connector or similar. This also means a real blower and chimney, not some muffin fan and open chassis jumble. - Use a real filament choke, not some totally unknown hunk of ferrite from a junked TV or whatever. Wind a R73-50-400 ferrite rod full of bifilar #10 Formvar. I sell these parts but so do many others. - At the input side of the filament choke use .01 and .001 disc ceramic bypass caps in parallel. I only use .01 or .02 with 3-500Z's on 6M but it certainly cant hurt to try. Whatever works. - Own, buy, borrow or steal a Grid Dip Meter. It is your greatest weapon for discovering reasonance problems. I use old tube type units such as the Millen and the Measurements Corp. Model 59. With this type of circuitry just about anything including solid state should work. - Run the filaments at exactly 7.5VAC RMS at the socket or even 5% below. This might be a very good idea for new or like new pulls with lots of gain. - Never use an untuned input. An L network or Pi-net is required. Keep Q under 3. The 4-1000 has an input impedence of roughly 100 Ohms in GG service. Use a healthy sized .01 or .02mf disc ceramic for coupling; voltage ratings of 2KV or so will stay stable with 100W of drive. - Never use the old BC style block mica capacitors at 6M for any purpose. They are highly inductive and can really aggravate a marginal problem. - In GG operation use 2 or 3 1Kv ceramic discs in parallel to bypass the grid. Use a total of .005 to .01 mf with very short leads. Use shielded wire to the metering circuit. ( I always liked to put the series and shunt metering resistors, carbon composition only, right at the socket as another method of lowering any inductive Q and then run the lead out thru shielded wire and filtering.) - Use single point grounding for grid and screen in GG. This means a wide copper strap from pins 2 and 4 to ground, absolutely minimum length. Grid components from pin 1 should go to this strap with "zero" leads and NOT to a separate ground. Some designers say to ground the strap at both pins 2 and 4 ends....I've always leaned to the single point idea but lets get comments. - In grid driven service l do prefer the 2 point ground for the screen for convenience plus the tube is being utilized much differently. Again I welcome comments here. About 2000pf total should suffice. For voltages <500V a pair of 500pf dipped silver micas in parallel at each pin are fine. For higher voltages I suggest a 1000pf 5KV RF rated doorknob at each pin. NOT a TV capacitor. A one piece strap should be used to go across the socket from cap to cap and pins 2,4. - In grid driven service install a 3 to 10 ohm 1W carbon comp resistor right at the grid pin and in series with the RF. In the screen use a 100 Ohm 1W similarly. Both will help to suppress parasitics and also act as current limiters/fuses. - In extreme cases in grid driven service it may be necessary to actually resonate the screen to effectively neutralize the tube. This should not be necessary if the grid is sufficiently swamped. Anyway, Eimac recommends a 25-50pf variable from screen to ground to tune out the screen inductance. They also recommend the 2 screen leads be tied together with copper strap but they then suggest using the center point of the strap as the take-off point to the cap. This makes the 3rd suggestion as to proper RF grounding. I'm referncing here to an old Eimac spec sheet that I just dug up...dated 1966. Tnx to W3OTC for waking up the brain cells! - In grid driven service the grid should be swamped with about 200 Ohms or so for drive in the 75-100W range. This almost totally precludes any requirement for neutralization...note I said "almost'. A few responses suggested a 4:1 transformer as a simple way and others said to still terminate it in a tuned circuit. Personally, I dont like broadband stuff at VHF...garbage in=garbage out. The swamped grid is excellent....I use to drive the NCL-2000 on 6M....and can actually be tailored within reason to fit individual drive requirements. Input networks can be almost anything; L : Pi-net: link coupled; tapped coil, etc. Its whatever that makes your exciter happy and does an efficient transformation to the tube grid...whatever is in the path. - Output circuit: The plate choke should be between 8-10 uh and have no reasonances near 6M. I test all of mine for reasonance above at least 70MHz. Here is where the GDO (Grid Dip Oscillator) is important. A 1000pf 6KV disc ceramic is fine for plate voltages up to about 4500VDC as a choke bypass since the choke impedence to 6M is very high and it sees very little RF with the above choke. Any indication of heat in the bypass is a sign of something wrong in the choke/tank circuit. - Dont skimp on the plate blockig capacitor. Use 750 to 1500pf. DONT use TV caps or micas. At plate voltages of <3500V a pair of 5kv RF rated ceramic doorknobs in parallel are fine. At higher voltages use one or two 15KV units such as the 857 series. Ideally you want very little heat disappation in the capacitors since that causes capacitance changes and tuning drift. - In GG service the 4-1000 needs lots of voltage to deliver decent levels of RF. Decent meaning over 1KW since it takes 150W just to light the filament. 4200 to 5500VDC is required for the 1500W level depending upon tube condx, drive level and your circuit. In an efficient circuit, up to about 150W drive is acceptable with decent >30dB IMD. Dont expect much over 10dB stage gain with the average "pull" tube. - GG service usually requires more plate voltage than grid driven for equal output since stage gain is lower. This simplifies the power supply but can add a big expense to the plate circuit components. - Grid driven stage gains are higher than GG so less plate voltage is usually required for similar output. The plusses are a much simplified tank circuit using readily available air variables, etc. 1500W possible with <4KV on the plate. But now you need a screen voltage of up to 1Kv plus a negative bias voltage of -50 to -130 depending upon AB1 or AB2 operation. The minus is really more attention to details; higher gain = more stability potential. With todays technology the extra voltages required for grid driven are miniscule and very cost effective to the serious builder. - In grid driven service the output C of a 4-1000 only about 8pf. Why any sane person would try and use the C of the tube plus squeeze a coil to tune is beyond me. I cant imagine the output C is too much worse in GG.....answers please. A high tank Q of over 16 or so is actually practible and beneficial on 6M. - I'll limit this part to basics. A plate circuit input Pi-net C should be variable from roughly 7-25pf so that reasonance occurs at 50-70% open. Output C is roughly a 250pf at maximum. Those criteria give sufficient range for 50 or 75 Ohm antennas and plenty of tuning range for SWR problems. Always use an RF choke across the output just like at RF. On 6M a RFC-50/Z50 or similar is all thats needed. On 6M a RFC-50/Z50 or similar is all thats needed. - I missed something above.... In GG svc at >4KV a vacuum variable for the input C is almost mandatory. Most with a 7.5KV rating and 30pf or so are fine; the minimums re in the 3-5pf area. Check it before you own it. - I've probably missed a bunch of good input and the response scattered but I've been intermittently typing this over 2 days as time/family permit. Excuse the grammar and typos please. - I'm really hoping for lots of involvement from experienced hams and also from the totally non-experienced and all in between. Amplifiers are still one of the items that are home buildable at low cost if you know the procedures....and it dont take a Phd (Piled Higher & Deeper) to learn them. Its getting late..... I really hope that some others comment or contribute to this. Lets get info sources available too. Its time to kick ass and generate RF! 73..........Carl......KM1H