4CX350 On the HF Bands
I am looking for some info on how to drive 2 4CX-350 Tubes for the HF Bands they require only .3 watts of drive power for the grids i read in one of the old QST magazines not to run this tube in grounded grid it will decrease the life of the tube so the best thing i can come up with is to run it with a swamp grid of 50 to 75 ohm driveing it with a .005 or a .001 cap to the grids any help would be appreciated Thanks Robert KI4WAU 73s
You're on the right track . . .
The 4CX350 is a SSB class AB1 (NO grid current) variation of the 4CX250 class/series of tubes. The grid is slightly different for higher gain, but very fragile, and won't handle grid current over a few MICROamperes.
You'll have to do like many hams have done with the 4CX1000A, which is a big brother variation. Bias everything to class AB1, with normal grid bias and screen voltage (usually has to be stiffly regulated due to negative current at some times), and use the 50 ohm resistive load across the grid. This will have to be hefty enough to absorb the driver power, and only the VOLTAGE will drive the tube. One nice advantage is this effectively eliminates the need for neutralization.
Get some of the older W6SAI (editors & engineers) Radio Handbooks -- Bill had good design basics for tetrode amplifiers that you could pick up on and use.
A pair of those should be just fine for a 1 KW input amp -- three would do the legal limit without breaking a sweat. I seem to remember seening these tubes -- look just like a 250 except for a bigger finned plate.
If you revisit the old article it will probably say that you cannot simply tie the screen and control grids together as was popular with 813s 4-400s, 4-1000 etc.
You can still use your tubes in "grounded grid" but the grids will only be grounded for R.F. They will need control grid and screen grid bias. When the bias is set properly the tube should work without running into grid current.
Good advice about the screen supply not being all that simple.
The Orr book does indeed contain a nice example of a 4CX1000 in grounded screen. It was in the 1975 or so edition.
The use of the 4X150/4X250/4CX250/4CX300/4CX350 series of tetrodes in "standard" (control and screen at RF and DC ground) grounded grid service is emphatically discouraged by the manufacturers.
The December 1956 issue of QST actually HAS such a construction project featured, and the author admitted that the "design" was not that advisable, but on a low duty cycle, coupled by the low, low cost of mil. surplus and commercial 4X150 pulls, (abouit that of sweep tubes at the time) it was cost effective to purchase and keep routine replacements handy. the author even admitted that a standard two-tone test of the amp as an SSB linear would destroy the tubes almost instantly.
So much for tying the grids together and directly connecting them to ground...
Bill Orr, W6SAI had an EIMAC Amateur Service Newsletter (AS-1) that covered the topic of triode connected tetrodes in grounded grid service. (It's four pages long, so I can't post it here directly.) He does point out that cathode-driven operation IS quite feasible if the control and screen grids ARE "grounded" for RF with adequate bypasses, and supplied with normal DC control grid bias and screen potentials.
E-mail me if you would like a copy of AS-1. I'll wait a couple of days, to see if there are many requests, and e-mail limits may dictate one or at most two pages a day.
Originally Posted by KI4WAU
I don't know why amateur texts are so far behind on the proper way to calculate grid resistance value. Everyone seems to think you slap a 50 ohm resistor there and go.
The formula is:
First find the squareroot of ((drive power / desired input resistance) times 1.414). This will give you the peak input resistance current Ipk.
For 100 watts and a 50 ohm exciter it is 2 amperes peak at envelope crest.
Now that you know the peak current, you simply use:
Rg = Eb/Ipk to find the grid to ground resistance
Rs = Desired input resistance - Rg for the series input resistance.
For example a 27 volt bias tube grid (like the 4CX350A) 90 watts of exciter power would require 14 ohms of grid resistance to ground and 36 ohms of series resistance to the exciter.
This would give 50 ohms to the exciter, and allow the 4CX350A grid to remain in class AB1 at 90 watts PEP drive in any mode.
Second, some of Orr's super-cathode driven circuits are not good for stability. I would avoid them.
You can drive the tube's cathode with normal grid and screen voltages, but there is an inherent problem with that system. You have no control of the actual amount of RF drive power divided between the control grid and screen grid in the tube. With some tubes it works out OK, with some (like the 4CX1000A) it requires a fancy control grid system where cathode voltage is divided to reduce the drive applied to the control grid.
It is NOT, regardless of what anyone says, a simple thing like applying the correct voltages and just driving the tube. This is because the exciter now directly drives the screen and the control grid at the same time, unlike conventional grid drive which would only apply RF to the grid.
If the screen gets a disproportionate portion of the drive you can damage the screen, and the same is true for the control grid if the control grid gets a disproportionate portion of the drive. So the actual system needs looked at.
Another thing to consider are the IM products. You will absolutely need to regulate the screen grid voltage. You will need to limit the screen for current, and the screen has to be a low impedance shunt regulator system.
The method I use is to run the screen from a supply at least twice the desired voltage and use a shunt regulator that is set for the proper voltage with a series resistor to the supply source that sets the current at the absolute maximum allowed screen current.
Looking at TWO tubes the screen of the 350A is 400 volts, and the dissipation is 16 watts. That's 40 mA maximum screen current. Let's say you have a 1200 volt screen source taken from the center tap of a transfomer in a bridge supply for the anode. The series resistor to the shunt regulator would be 1200-400= 800 volts drop across the resistance at .04 amperes is 800/.04 = 20,000 ohms. That would take a 20k resistor capable of standing 32 watts dissipation. The shunt regulator would have to dissipate 16 watts. The screen dissipation could then never exceed 8 watts per tube no matter what you did. You could shut the HV off and not hurt the screens. You WOULD have to monitor screen current or voltage to be sure you are not pulling the regulator out of limit. The only way to tune a tetrode is to watch the screen carefully, it is the critical element when tuning.
With some tubes you might even need to add negative feedback by elevating the cathode from ground through an unbypassed resistor to reduce IM to acceptable limits.
I've used the above methods with tubes up to the 100 kW output category in commercial amps. Some 4CX350A sockets will not lend themselves to cathode driven circuits, which is another negative about using the tubes in cathode driven applications.
If you use the silly method of a 50 ohm resistor, the tube would drive with about 8 watts of drive power. If you use the method I outlined above the drive power would be 90 watts, and the tube would be MUCH more stable.
Of course you would need a 14 ohm low reactance 30 watt resistor at the grid, and a low reactance 36 ohm 60 watt series resistor to the input from the grid but that isn't hard to do at all. The 14 ohm resistor would be split between the two tubes at the grids, so that would be a 28 ohm 15 watt resistor at each grid. Piece of cake.
Last edited by W8JI; 11-07-2008 at 12:48 AM.
Reason: modified for two tubes
I'll make careful note of your comments, Tom. thanks for the input.
OK, sorry for the length.
Originally Posted by WA9SVD
I'm trying to point out the optimum grid resistor is often NOT 50 ohms.
As a matter of fact with a tube with low bias voltage like the 4CX350A, the drive power from using a 50 ohm resistor is about 8 watts!!!
Of course another solution would be to add a 50 ohm input attenuator of just over 10 dB loss, but for maximum stability it would be better to use an L attenuator pad right at the grid so the grid has the minimum impedance for an all resistor network.
Also you have to watch lifting the control grid from ground for stability reasons when using grounded grid. Super cathode drive is often bad news.