4cx250B

[exces3]

is it possible to get correct grid bias with passive parts.
ie, resistors or zeners?

I assume correct bias would be -50 to -100 Volts
on the grid

poorly written.
...............
is it possible to get correct grid bias,
something between -50 an -100 volts,
using only resistors and zeners,
without an active power supply...
.........no transformer....

 

[Crusher]

The transformer I was using for grid bias and screen is a 110v primary, 250vac @30mA and 50vac @30ma secondaries. Single primary and dual secondaries.

 

[KM3F]

With operating voltage so far out of the norm and no metering, how do you expect to see what's going on?
This tube requires a closed chassis on the bottom side with a forced air blower to cool the anode fin assembley up through the socket.
Do you have that kind of blower and not just a fan?
What is the resting plate current?
What is the number 1 grid current under drive? It will not take more than a few watts of grid 1 dissapation.
What is the screen current?
When you lower the plate voltage the screen will try to take more current.
The negitive bias is dependent on the screen and anode voltage and the mode of operation.
You just can't fly blind on all this and expect the operation to be right.
The tube's max anode dissapation is only 250 watts.
Sorry to sound a bit harsh but you need understand what your doing.
I have an amplifier that had a 250 in it for 6 meters and ran a max of 325 watts flat out.
It has 2400 volts on the anode and 350 on the screen.
The amplifier now has a 4cx400/Russian GS36B tube and makes 600 watts at 400 ma anode current and no more than 2 ma grid 1 current on single side band.

 

[Shockwave]

Now im getting somewhere. I changed bias, so that i have -68V at all times. The bias circuit turns on when plate V is turned on now. When i first turn it on, amp stays nice and cool. But when i keyit up and unkey, amp stays in full conduction. I am thinking i need more bias? Probably around -100V for closer to full cut off. -68V just aint gonna cut it i guess. Part of the learning curve. So looks like i will have to install a second bias transformer and run them in series to give me about -140v. Then switch in a zener regulator or something more suitable to handle regulation so that amp will conduct when amp is keyed and drop it down to a more workable voltage.

Don't get discouraged, be careful the advice you follow does not have you chasing your tail in unrelated areas and listen to what your amp is telling you. All you need to know to diagnose this problem is contained in these two sentences "When I first turn it on, amp stays nice and cool. But when I key it up and unkey, amp stays in full conduction."

If the negative bias voltage is great enough to hold the tube in cutoff so it does not produce noticeable heat before RF is applied, it's enough to return the same condition once RF is removed. Without a current meter, you are working blind and I highly suggest you install one for testing even if it's only your VOM inline with the plate return line.

Your symptom is one that is classic with improper neutralization of the tetrode. Once RF is applied, the amp goes into self oscillation and does not return to the cutoff current as a result of the current drawn by the RF produced in the self oscillation. One other common mistake that can contribute to this is poor RF bypassing of the screen grid. You should have a low inductance bypass cap to ground off of every screen pin on the socket.

Many people shy away from experimenting with a proper negative feedback loop to provide good neutralization and will try anything including swamping the heck out of the control grid with non inductive resistors. This can stabilize the amp but at the expense of gain since more drive power is wasted heating the resistors but the reduced gain can prevent self oscillation. Sometimes the 4CX250B can be stabilized with as little as a 100K 2 watt carbon resistor swamping the control grid.

If you decide to go all out and neutralize this amp, it's easier then you may think and can be done without risk to you or the tube in a cold condition with no voltage applied. It really is just forming a very low capacitance loop from after the plate blocking cap, back to the control grid input. You then connect a low level signal generator (the MFJ analyzer can be used) to the output of the amp. Connect a RF detector (field strength meter can be used) to the input of the amp.

Now engage the RF relay so the connectors of the amp are tied to the input and output circuits. This can be done manually or by energizing just the relay coil. You then adjust the capacitor value so that power applied at the output of the amp is nulled out at the input of the amp. The value of the typical neutralizing cap is very small in pf value. Usually between 1.5 and 10 pf.

The most important thing here is that you choose one well capable of withstanding at least the plate voltage. Even though it is tied into the plate circuit after the plate blocking cap and will not have DC across it, nearly all of the RF voltage is dropped across this cap since only a very tiny portion makes it through the cap and arrives at the control grid. Hope this helps.

 

[Crusher]

KM3F, yes the amp has a blower. I put a larger blower on it than originally. The bottom deck is sealed. I will have to run external meters as this amp on bottom does not even have enough room to put a small chinese meter on it. I did not build amp, merely trying to upgrade so that it runs proper. Before it had 2400V on the plate and would not drop. Meaning its a solid supply. Really need 3 meters not 1.

 

[Crusher]

Well I put HV the way it was originally in amp, now have 2400vdc on plate. 300vdc on screen. Have to change up bias with a properly fixed zener, but now it will key and unkey properly. Cuts off when amp unkeys. No more conduction issues. Load cap needs to be padded some. I installed 100K 5W resistor from grid to ground. Need to do more testing. Need to replace batteries in bird meter so I can see what it does peak, but on average carrier is 50W and swings up to 175 average. Think I may need to pad the input slightly more. Think when plate V was so low, screen might have been trying to compensate. Also I should mention, I pulled out 250A and installed a 350A. I have a small dropping resistor in series with filament so I had to make minor adjustments to get it within spec I had a .2 ohm 5W and had to change it to a .1ohm 5 W. This has me at 6.1V. Drops slightly under load, but I still sit right at 6V.

 

[Crusher]

I took the 100K back out. Its not needed. Think the problem was it didn't have the plate volts to begin with. Seems to have been cured with higher plate volts. At idle it is in complete cut off now. Not sure if I'm gonna put a 250a back in, its working now.