ad: vanity

Grounding the Grids vs. Cap/RFC Grounding

Discussion in 'Amateur Radio Amplifiers' started by W4LAC, Jun 13, 2019.

ad: L-HROutlet
ad: l-rl
ad: L-Geochron
ad: MessiPaoloni-1
ad: Left-2
ad: HRDLLC-2
ad: L-MFJ
ad: abrind-2
ad: Left-3
  1. G0HZU

    G0HZU QRZ Member

    I'm afraid I again did a very quick guess of the grid inductance based on a casual glance at photos and dimensions of the 3-500Z tube. Some 3-500Z tube drawings look different to others but It looks to me like the length of each 'narrow' grid pin is typically about 0.5" long and I guessed the diameter at 0.1". So this gave about 6nH. There is also another thicker section of grid pin and I guessed this as being about 1" long and 0.2" wide. This added another 12nH. This gives 18nH per pin. I rounded it up to 21nH to allow for chunky/ideal socket connections and placed three of these in parallel. I then added another 5nH for the top section. This could be completely wrong because the effective inductance of the top section will be hard to guess. I think the effective inductance of the internal grid 'mesh' structure itself will be lower than expected because it is tightly distributed inside the cathode and the anode structures. I guessed it at 5nH. It will obviously be more than this to the top of the grid mesh but I just need a 'working' value for all of the internal grid mesh. However, when I use all these values together as a total grid L of 12nH with a tube gm of 12mA/V the tube is stable with the RLC suppressor with a reasonable margin. I've also used the factory data for the internal capacitances here.

    I don't think that minor changes in the grid inductance will have a lot of effect on the frequency of oscillation. It does have a big effect on the amount of negative resistance generated by the tube at VHF though. For me at least, this is the critical area of the design and I'm sure the designers of the 3-500Z amplifiers in the 1960s would have used a slide rule and a smith chart and the datasheet to estimate the impact of all this. They would also have a much better idea of how to quantify the overall grid inductance as they would have had access to real hardware to measure and play with.

    I think what impacts the typical frequency of oscillation is the tube output capacitance and the overall inductance of the tank circuit when measured at VHF. I'd expect this to be somewhere around 15onH and this is the total inductance starting inside the tube at the anode and working through the tank interconnection wiring and into the plate capacitor which will look like a low impedance at VHF. So it can be treated a bit like a short circuit. Hence the overall guess at 150nH. If the tube looks like 4.5pF (with maybe another 1pF stray) and the tank looks like 150nH then the system will go unstable (without a suppressor) at the resonant frequency of 5.5pF and 150nH = 175MHz. This is because the tube can generate a fair bit of negative resistance up at 175MHz. The equivalent series resistance of the suppressor at 175MHz is there to cancel this negative resistance in the tube and ensure system stability.
    Last edited: Jul 16, 2019
    W1BR, K9AXN and KM1H like this.
  2. K9AXN

    K9AXN Premium Subscriber QRZ Page

    Very nicely done!
    Regards Jim
  3. SWL37632

    SWL37632 QRZ Member

    Spent a lot of time reading and understanding this thread and learned much. It appears that using floating grids bypassed by ( 150 to 220 pf) caps add to the stability, linearity, neutralization, but, Bottom line, was the original question answered ?? And if so, what were the trade-offs ??

Share This Page