The AM Forum

I am designing a grid overcurrent protection circuit for my 8877 amplifier.
What should be the maximum grid current trip point be?
I see everything from 75 to 225Ma. on various sites for overcurrent trip levels.
The normal grid current with 38 watts of drive producing 1500 watts out is around 40Ma.
Anyone out here have any insights on this?
Thanks and have a happy 4th of July!
Don W9BHI

Maybe whatever current makes the 25W grid dissipation value? Will depend on plate voltage and 'single tone' power output but for example using eimac AB2 specs at 2500V, assuming the driving Z at the cathode is 50 Ohms:

10mA, drive 57W, 53V drive (if z=50 Ohms), grid input is 0.5W no sweat. The highest grid current in the datasheet is 74mA but that's not the whole story by a long shot.

maybe opening cans of worms beyond my understanding.. The current goes up at lower plate volts, and also changes with loading.

Hi Don,

As Patrick says, grid current will vary widely depending upon plate voltage, drive, loading, class of operation, etc.

I have a HB 8877 amplifier too and made up circuits to trip the plate current and grid current when each came near to "abusive conditions."

I determined the trip levels more seat of the pants. Always monitor grid and plate current during this setup procedure:   I first loaded the amplifier up to maximum power. I use an SDR spectrum analyzer to determine the cleanest and most linear parameters vs: maximum power. I then adjusted things for worst-case plate and grid current conditions. I set the trips so that they stayed on during full power optimum conditions that I normally run and  tripped off when I overdrove the grid or put the plate too far out of resonance.

Play around with these parameters for a while and you will soon determine where to set the trips for normal operation and when they should trip out.

Be sure the grid current always stays within data sheet standards and you will be FB.  Besides, the less grid current you run, the cleaner the amplifer will be. In my class A RF linear chain, I draw zero grid current on the 8877 grid for exquisite cleanliness... :-)

T

Thanks for the replies,
I played around with the amp and driving it to 2kw showed less than 50Ma. of grid current.
I guess that if I set a trip current of 80 Ma. it would give a good safety margin since it would never operate anywhere near that.
Don W9BHI

Don,

   If this is a home brew Amp, consider this..

Many 8877 amplifier's for HF will ground one side of the filament, and just drive the cathode alone. Then when in receive, a relay inserts a resistor of about 50K (cathode to ground) to bias the tube off. This works FB until you hear a snap crackle pop while the amplifier is just sitting there while you are receiving..

What happens is when (not IF) there is a touch of heater cathode leakage, the tube protective bias is lost, and the bias zener (if used) is bypassed to ground. The tube turns on with high idle current, and at best you will notice your fan come on high speed (if a thermostat is used). More likely, the tube will zoom off in a destructive VHF parasitic and will self destruct. I had this happen 3 times here...Two 8877's with an open filament (after the snap crackle pop), and the 3rd (pulse version) got a dead short heater to cathode. This last one never went into a parasitic, but the fan just went to high speed as I was in listen mode...

The cure to prevent this scenario is to float the filament from ground. Then make your cathode drive include the filament, much like is done at VHF.

How many "used" 8877's have a touch of heater cathode leakage? My used pulse version showed it from the beginning during power up when the time delay relay was locking out a PTT command. About 30 seconds after power up, I saw about 50ma of cathode current for 2-3 seconds, and then it was gone. Worked that way for about a year before the heater to cathode became a dead short. Tube still works FB with a floating filament..

Jim
Wd5JKO

Hi Don,

Sounds like a plan.   Setting the grid trip is almost like setting spark gaps on mod iron.  Talk normally and watch the 2KW meter swing up to pin.  Then sock a few atomic yallos (with the compression backed off) and you will arc the gaps or trip the grid protection.   Nothing better than real world settings.



Hi Jim:  Is your suggestion the equivalent of using a "tri-filar" filament choke?  That's what I have always used here for the indirectly heated cathode tubes I often use for linear service.   

Here's a sample schematic:

http://wc6w.50webs.com/wc6wamps/index.html?fr40.html

T

  Don, Just curious what you will do with the grid over protection circuit? Perhaps just to make an LED flicker on voice peaks (SSB), or perhaps if the limit is exceeded for ~1 second then the trip circuit deactivates the PTT line? A latched trip, or momentary trip?

   Tom, the circuit you linked has the same problem since one side of the filament is at chassis ground (or nearly so...when the FET is on). Floating that circuit needs more thought. Might work to remove the ground, and replace the "HTR OFF" transistor with an opto-coupler.

Jim
Wd5JKO

Hi Don..... I've attached the data sheet incase you ain't got one.

Absolute maximum grid dissipation is 25 watts.

Hi Jim,

Yes, his complex filament circuit eventually went to ground... didn't see that.

The same trifilar choke with a floating fil transformer would be a better choice.... and simpler.

T  

Jim,
In normal operation in AM or SSB I barely see any grid current.
I wanted something to shut down the keyline if the grid current goes over 80Ma. or so.
Like if the amp is accidently over driven, on the wrong band or the antenna is not connected. (you know, the dumb things that happen).
The circuit that I am using causes a relay to latch and break the keyline and has to be manually reset.
Thanks,
Don W9BHI