The AM Forum

Hello, I have built an 2x 813 amplifier for HF. The bias is AB2.
When I give modulation in AM mode (with the driver) I receive some distortion in the final signal.
Possible a consequence of the AB2 stage?
Has someone a suggestion?
73'Paul PE1PAL

I'd push the bias up to run class B.  Make sure you don't exceed the tubes total power rating of around 140 Watt output for the pair as a linear amp.

dg

Is it GG or grid driven?

If grid driven, put a DC VOM on the grid, at the DC point of the RF choke and watch the grid bias level as you apply signal. It should be regulated and not fluctuate. 

A better way for bias is to put back biased diodes or zeners in the fil center tap for a solid bias.


** Also, make sure the final is loaded heavily enough. Peak the plate tuning and loading for maximum output, then unmesh the loading cap until power drops off about 7%. Do this with full drive.  If the final is loaded too lightly, this will cause good efficiency, but with distorsion.  A heavily loaded amp has reduced efficiency - that's the trade-off.


Look at your RF waveform (100% modulated) at both the input and output. You will probably see where the sine wave starts to deteriorate.

Tom, K1JJ

You can be pumping the bias as Tom Suggests but another possible problem is the Q of your input tuning. AB2 input Z drops as the grid is driven positive reducing the input Z. If you are grounded grid triode connected add some tuning on the input to add a flywheel effect on the drive. If this is your problem chsnging to class b will just shift the problem to a diffeent power level. Watch your dissipation. Also a low final Tank q will compress the peaks due to light loading.

Paul,

What are you driving it with?  Since class AB2 develops grid current it does require some power from the exciter and presents a varying load to the exciter.  If your exciter has plenty of drive power capability, you might try a swamping network to reduce the impact of the varying grid load.

Are you staying well within the power capabilities of your amplifier?  With modulation peaks it needs to remain linear at 4X unmodulated carrier output level and if the HV supply has poor dynamic regulation this can create an issue.

Rodger WQ9E

Hi Rodger and all the others, thanks for the information!
I run the amplifier with a 200 ohm dummy resistor on the grid connected.
The transceiver is coupled with unun 1:4
SWR is optimised with the internal tuner of the transceiver.
One site of the 200ohm resistor is connected with -90V power supply directly from supply rectifier, so there is enough current available.
The G2 voltage is around 725V with Anode 2500V / 150ma (2 tubes).
Output around 125W in AM.
At these moment I use a pi filter with 2x T200-2  38 windings, I see at 160m the need for a big capacitance of 3000pf The first capacitance is then around 100pf.
Next listing I place some pictures.

 

Make sure you bypass the low side of the 200 ohm resistor to ground with a disc cap. .01 in parallel with .001. It needs to be at RF ground.