The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: W8ACR on April 14, 2018, 04:12:16 PM



Title: More grid bias stuff
Post by: W8ACR on April 14, 2018, 04:12:16 PM
I didn't want to hijack the other thread. I would like a critique of the grid bias circuit that I use in my big rig which is a Taylor 822 modulated by 811's. It has similarities to the circuits discussed in the other thread. The RF exciter is a Johnson Viking II. The grid coil is a B&W JVC. The bias transformer provides about 110VDC of bias voltage. The 822 plate B+ (2000VDC) is always on. With no RF drive, the tube is 100% fixed bias from the bias supply, and is biased well beyond cutoff (which is about -80VDC). When RF drive is applied, the bias becomes 100% grid leak - if I understand it correctly.

I can adjust the grid bias voltage to whatever I want by adjusting the RF drive and the rheostat. FWIW, the circuit seems to work just fine.

I guess my main questions would be:

1. is the 1N5408 diode good, bad, or indifferent?
2. what is the effect of the voltmeter, if any, on the rest of the circuit?
3 is there any reason that I should not use a rheostat as a grid leak?

Thanks, Ron







Title: Re: More grid bias stuff
Post by: DMOD on April 14, 2018, 04:52:09 PM
Here are the problems I see with in this circuit:

1) When the Pot wiper goes to zero ohms it shorts out the Bias Supply.

3) Anytime the grid goes more positive than the bias supply, the diode conducts bringing back the voltage down to the bias supply voltage with it's accompanying switching transients (at half the RF drive frequency) and the grid sees a very low impedance situation, with no degree of freedom.

Personally I like to see the grid voltage with no modulation so I can actually see what it is.

Below is my suggestion for this situation.


Phil - AC0OB


Title: Re: More grid bias stuff
Post by: W8ACR on April 14, 2018, 05:20:53 PM
OK Phil,

Thanks, your circuit makes good sense to me. I will say that the rheostat would never be at short unless by mistake.

Also, if the grid voltage went too positive, wouldn't that affect the voltmeter reading, even at RF frequencies? The voltmeter reading in operation is rock solid. It never flinches even a bit.

Lastly, can one calculate the positive peak grid voltage, or does it have to be measured?

Ron


Title: Re: More grid bias stuff
Post by: DMOD on April 14, 2018, 06:04:38 PM
The grid voltage meter is damped and would never be fast enough to show 1/2*F pulsations.

I just noticed that there is a .001 uF RF shunt cap at the bottom of the grid RFC so that should attenuate some RF.

My view is the diode is just an additional noise source.

Phil 


Title: Re: More grid bias stuff
Post by: N1BCG on April 15, 2018, 10:05:58 AM
When the Pot wiper goes to zero ohms it shorts out the Bias Supply.

Yeah, I'm having a hard time with this also and it could be easily remedied by making the wiper the bias supply output to the mA meter and removing the strap on the rheostat. Your 822 will sleep better at night.

A 10H choke in the bias supply...impressive!


Title: Re: More grid bias stuff
Post by: W8ACR on April 15, 2018, 04:49:32 PM
OK Phil et al,

You have convinced me that I should rearrange the bias circuit as you have outlined. Thank you for taking the time to respond to my questions.

I have two RF decks that use the same power supply and modulator. On one, the grid parameters are -200V operating bias and 50mA grid current. The other uses different tubes in push pull, and requires -400V operating bias and 100mA grid current. Both decks therefore require a grid leak of around 4000 ohms, but of course the push pull deck will require a much heftier grid leak resistor. I was thinking of splitting the difference and putting a 2000 ohm/50W WW resistor in the circuit between the ammeter and the grid coil and then adjusting the rheostat for proper voltage. Does this sound reasonable, or would you choose a different amount of fixed resistance?

Also, if you refer back to my original schematic, you see that I have an RFC in the circuit. What do you think about eliminating that RFC and just using a bypass cap on the ground side of the fixed resistor?

Thanks again, Ron



Title: Re: More grid bias stuff
Post by: N1BCG on April 15, 2018, 05:53:13 PM
Also, if you refer back to my original schematic, you see that I have an RFC in the circuit. What do you think about eliminating that RFC and just using a bypass cap on the ground side of the fixed resistor?

Actually, I believe you could take out the RFC, making the center tap of the input coil RF grounded. That still leaves the output and neutralization point hot.


Title: Re: More grid bias stuff
Post by: DMOD on April 17, 2018, 07:30:09 PM
Hi again Ron,

The purposes of the grid choke is to 1) at the top of the choke provide a high RF voltage at the grid driving node over a wide range offrequencies, 2) at the bottom, a high impedance to RF over a wide range of frequencies while keeping RF off of any external supplies and ground, 3) offering a DC path for external or protective bias.

I would keep it.

As far as the differing grid leak values, you might consider the following:


Phil


Title: Re: More grid bias stuff
Post by: Opcom on April 20, 2018, 12:50:50 AM
Better bias regulation should be had by ditching the cap right after the bias rectifier making it a choke input. More AC volts would be needed from the transformer.
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands