Reducing Power Toob Rigs

[ka4koe]

I've heard that one should NOT reduce power output on a classic AM transmitter by reducing the drive level. Can someone tell me "why"? I am aware of two methods to reduce power other than the VERBOTEN DRIVE! method....

1. Reduce B+ to the finals via a variac type circuit or equivalent.
2. Use an outboard attenuator.

I believe I am going to need to go with an amplifier with a pair of 3-500Zs to up my ERP. The AL82 seems to be the favorite for full 375W carrier. The SB220 is significantly cheaper, but due to cooling issues must be run at a lower constant carrier or it will go China Syndrome.

I am really enjoying myself, but tiring of the "Can you hear me now?" frame of mind.

Five months after buttoning it up, the Valiant is still chugging along.

Tnx.

FEELEEP

 

[steve_qix]

By a classic AM rig, I'm assuming you mean a class C RF amplifier that's modulated by something.... and if this is the case - you are correct - do not reduce the drive level to reduce the power output.

The reason for this is that reducing the drive level will change the conduction angle, may very possibly change the grid bias, and if the amplifier is a screen grid tube, will change the screen current.  Usually, transmitters are designed around a certain set of parameters and these all interact with the drive level.

The conduction angle for class C should be fairly small - less than 120 degrees and I've run less than 90 degrees.  The linearity and the efficiency of the class C stage is affected by the conduction angle and also the peak grid voltage both of which are affected by the drive....

So, the bottom line is:  Don't use drive to adjust power output from a class C stage.

Either reduce the plate voltage (and potentially the screen voltage if a screen grid tube) or load the stage somewhat more lightly.

[ka4koe]

Correct, plate modulated Class C transmitter, eg. Valiant.

Philip

[Pete, WA2CWA]

Use the Flex to drive the amplifier. Leave the Valiant for the "local" contacts. This way you don't have to screw around with the Valiant's electrical parameters.

[ka4koe]

Pete....

You have a way of convincing me to reduce the amount of projects on my workbench. If the Collins 32V2 did 160, I'd have one of those instead of the Valiant.

P

[WD5JKO]

a Valiant if modulated fully might do 500 watts PEP from a 125 watt carrier. The legal limit is 1500 watts PEP, so jumping from 500 watts to 1500 watts is about 4.8 db, or just under 1 'S' unit.

The secret to running lower power involves multiple things:

1.) A decent antenna
2.) Pick a time of day, and band when you can easily be 20 db above the noise. Here a 5 db drop in power from legal limit makes little difference. Perhaps 75m just before dark or just after sunrise for those stations within 300 miles.
3.) Modulate fully, and de-emphasize the lower bass frequencies below ~ 150 hz. Don't get rid of them, but knock them down about 10 db from the 1 Khz level. Then boost the stuff ~ 6 db going up from 1 Khz to to 3 Khz. A D-104 MC-320 element loaded to ~1 Meg-ohm load pretty much does this without an equalizer.
4.) If your voice is asymmetrical, then make sure your speech polarity is such that the positive modulation peaks are the greater of the two.
4.) Limit key down time to a minute or two at most. Sometimes the weakest guy in a QSO is the one that talks the longest. Then after a 5 minute transmission, a new QSO of SSB stations are right on the frequency.

Using some of these techniques, I've worked 150+ miles at 10AM running a Retro-75. The carrier was right at the other guys noise level, but the modulation was up there as described above.

73,
Jim
Wd5JKO

[N2DTS]

People report me being 10 over and/or having a good signal when I run 5 watts out of the argonaut 5.

A good antenna on an open band helps.

If you want to run 80 meters at prime time, you need big power and a good antenna, and it can be tough anyway.

[KZ5A]

With a plate modulated "driver" it is hard to beat a 6db attenuator.   It allows you to tune up the driver/amp combo with the driver in full power CW mode so the amp is tuned and loaded at max PEP as it should be for lowest distortion.  And then you switch to AM and switch in the 6db attenuator and every thing is happy.   

If your "driver" provides more drive in CW than your amp needs, use another attenuator to match up the overall drive levels.

6bd 75W Bird attenuators show up on ePay periodically and are perfect for the usual 100W AM rig application with an Amp that can use 100W of drive like the AL-82. 

72 Jack KZ5A

[steve_qix]

Use the Flex to drive the amplifier. Leave the Valiant for the "local" contacts. This way you don't have to screw around with the Valiant's electrical parameters.

Oh, Pete's got it!  If you have a Flex (I didn't know you had one) use THAT for the linear.  Nothing else you have will even come close with respect to audio, etc.  Sure, you could build a low power high level modulated transmitter to drive the linear and maybe it will sound better than the Flex, but it's a lot of work and what the heck.

[AB2EZ]

Using a ferrite core (or several ferrite cores) one can make a power divider that will send a portion of a transmitter's power to an ordinary dummy load, and the remainder to an amplifier's input.

The attached diagram shows how.

In the attached, the total load seen by the transmitter is the sum of: 50 ohms (the dummy load) and the input impedance of the amplifier. If the amplifier has a 50 ohm input impedance, then the transmitter will see a load impedance of 100 ohms, and 50% of the transmitter's power will go into the dummy load.

By using different turns ratios, one can split the power in different ways.

If the secondary winding (that feeds the input to the amplifier) has twice as many turns as the primary winding (that feeds the dummy load), and if the amplifier's input impedance is 50 ohms.... then 20% of the power will go into the amplifier, and 80% of the power will go into the dummy load. The transmitter will see a load impedance of 62.5 ohms.

Stu

[flintstone mop]

I'll join the crowd and say use the Flex to input the linear. And keep the Valiant for a powerful barefoot signal. Good for quiet conditions and daytime use on 80 and 40.
If you're hell-bent to use the Valiant then you can lower power by reducing the screen voltage on the final PA tube. There was a reasonable circuit using a pot. and a transistor and minor circuitry to do it nicely. It was used in a Ranger to reduce power to drive a linear.
The Flex would be an easier approach.
Fred

[WD5JKO]

I am somewhat a newbie to Flex SDR technology, so my inexperience might be just that. What I have observed though is that a Flex on AM (I got a 3000) has two pitfalls:

1.) There is no provision to switch the speech polarity; it has to be done external to the rig. Any AM rig where we have a single OM, and Mic needs to get the speech polarity correct to optimize the audio peaks.
2.) I have not figured a way yet to get high positive peaks without making the occasional, or frequent wavelets of DSB. These wavelets are not distortion per say, but when received on a diode detector, the signal sounds horrid. Lots of Flex operators make wavelets on AM...sounds OK though with SAM detector.

I have concluded that I must do my processing external to the rig. I love (NOT) the front Mic RJ-45 input, or rear DB-9 Line input.

Back to the topic of this thread, I once was looking at a Grounded grid dual 304tl amplifier. As I recall, the low Mu of these tubes would lower the gain to about 5-6 db...just what we need here, and no attenuator required. There was an old Radio Handbook article (W6SAI) using a single 304TL in G-G. It took 200 watts drive, and of that 25 watts drove the tube, and 175 watts feed through to the output. Double those figures for two tubes. I was pondering ways to use two or three of them back in the DC input days where the DC input might be 1KW, and the RF output the same.

With an inefficient antenna close to the house, boosting the RF power 6 db will likely have a 20db increase in RFI complaints, and could possibly exceed safe signal strengths in the living areas, especially on the higher frequency bands.

Jim
Wd5JKO

[steve_qix]

Yes, you're right about external processing - welcome to the world of high fidelity audio    This is really very standard for high fidelity transmitters for a variety of reasons.  For instance, you don't generally see audio processing except for maybe the negative peak limiter in any broadcast transmitter or class E rig, etc.

Fred, I have finally purchased a Flex 5000a myself (I don't physically have it yet).  This is the first commercial rig I've *ever* bought that cost more than about $50.00.  I got mostly it to facilitate remote control operations, and for the fantastic receiver.

The first thing I'm going to do is to design a negative peak limiter that will work properly with the Flex to eliminate the "wavelets" problem that occurs with any transmitter that doesn't have a conventional AM modulator.  This is job 1 in my opinion for getting great audio from the Flex.  I will publish the circuit for sure when it's done.

The Flex 5000s I've heard on the air usually sound really good, but all suffer from the "wavelet" problem when "overmodulated" (overmodulation can't happen with a balanced modulator, hence the folding-back of the signal AKA "wavelets)".

Should be interesting !

[ka4koe]

I've seen those wavelets on my REA. Little globs and no flatlining of the overdriven waveform.

I'd be very interested to see what you come up with, cuz I'll likely do what you do.

Philip

[flintstone mop]

I have Jay's, N3WWL, first FLEX SDR 1000. I am not aware of wavelets. The audio reports are always pretty gratifying. The adjustments and settings and magical software give me 90% neg peaks and 120% pos. going into an AL1500 amplifier.
The mic is a Marshall condenser and the audio card is an outboard Edirol FA66. Balanced mic audio and FireWire into the 'puter.
And Steve, the RX and panadapter is really nice. It can turn the battleground of nearby SSB and hetrodynes into armchair copy. The SAM is as good as it can get without spending another $600 for a Sherwood unit. The SAM greatly reduces the selective fading distortion heard on shortwave, while listening to music. Many have said that the Sherwood Synchronous detector is the ultimate for music reception on shortwave.
And thanks for boosting the economy a little with your purchase........The SDR 5000 is not cheap!
Fred

[ka4koe]

Looks like I'll be getting an SB220 in March via a trade of my SGC SG500 (given to me by a relative). So, no money will be exchanged!

[w1vtp]

<SNIP>
I believe I am going to need to go with an amplifier with a pair of 3-500Zs to up my ERP. The AL82 seems to be the favorite for full 375W carrier. <SNIP>

Tnx.

FEELEEP

 

If you are going to use the classic 375 watt to determine "legal limit" you may be exceeding that PEP.  A better (and easier on the AL82) design goal is to use 300 watts of carrier and set a goal of 120% positive peaks.  And yes, the Flex 3000 would be a very painless approach for your legal limit plans.  You will need that polarity switch.  I don't know of any rig that has it built in.

I've attached the calculations for an Excel calculator I ginned up to make my point.

Al

[w1vtp]

A better design goal would be for a 250 watt carrier at 120% positive peaks for the SB220.  It's doesn't have the cooling for higher power / percent mod levels.

Al

[N2DTS]

I really liked the flex 5000, the 3000 not so much.

The 5000 has all the nice ins and outs, audio and antenna, its quiet in operation, and the receiver is very good.
I liked the one button record and playback of people over the air with no change in the audio, nice for signal reports.

I think the old software was better though, back when the AM carrier adjustment was 0 to 100.
You could adjust it for 150% positive modulation without over modulating, and I do not think the new software works or sounds as good (on AM). It does have less latency though.

Part of why the old flex 1000 sounds so good may be the older software versions they run.
I never had much luck running the built in audio processing stuff, its nothing like good equipment..

[wa3dsp]

I do the power reduction on my Ranger with a pot mounted behind the crystal access socket. It varies the bias to the clamp tube and reduces screen voltage. Note this is also in combination with the screen/plate modulation ratio circuit shown in the modulator schematic. This is an easy mod and when the pot is at full power it is exactly the same as stock. You basically replace the final grid bias resistor with an equivalent value pot, wiper to the clamp tube grid. Load the rig up and adjust the pot for the desired power out and retouch tuning.

I run about 150 watts carrier with an SB200 and get good reports. When you reduce the screen voltage watch the grid current. It will change and needs to be readjusted.

http://www.crompton.com/hamradio/JohnsonRanger/RangerModsSchematic1_2-5-2013.pdf

http://www.crompton.com/hamradio/JohnsonRanger/RangerModulatorUpdateWA3DSP.pdf

http://www.crompton.com/hamradio/JohnsonRanger/RangerRestoration.html

A better way to do this is to reduce the plate and screen voltages but it would involve a more extensive mod especially if you also wanted to use the rig barefoot.

[AB3L]

Pete....

You have a way of convincing me to reduce the amount of projects on my workbench. If the Collins 32V2 did 160, I'd have one of those instead of the Valiant.

P

I'm pretty sure there was an article in an old QST that detailed the conversion of a V-1 or a V-2. It's here but can't find it for the year.

[ka4koe]

For essentially no monetary outlay, I think its a good deal.

[flintstone mop]

I really liked the flex 5000, the 3000 not so much.

The 5000 has all the nice ins and outs, audio and antenna, its quiet in operation, and the receiver is very good.
I liked the one button record and playback of people over the air with no change in the audio, nice for signal reports.

I think the old software was better though, back when the AM carrier adjustment was 0 to 100.
You could adjust it for 150% positive modulation without over modulating, and I do not think the new software works or sounds as good (on AM). It does have less latency though.

Part of why the old flex 1000 sounds so good may be the older software versions they run.
I never had much luck running the built in audio processing stuff, its nothing like good equipment..

I'm using Flex V2.3.5 on my SDR 1000. I do not know what version they up to now.

[Pete, WA2CWA]

I really liked the flex 5000, the 3000 not so much.

The 5000 has all the nice ins and outs, audio and antenna, its quiet in operation, and the receiver is very good.
I liked the one button record and playback of people over the air with no change in the audio, nice for signal reports.

I think the old software was better though, back when the AM carrier adjustment was 0 to 100.
You could adjust it for 150% positive modulation without over modulating, and I do not think the new software works or sounds as good (on AM). It does have less latency though.

Part of why the old flex 1000 sounds so good may be the older software versions they run.
I never had much luck running the built in audio processing stuff, its nothing like good equipment..

I'm using Flex V2.3.5 on my SDR 1000. I do not know what version they up to now.

From the Flex web site:
"Starting with PowerSDR v2.6.x, FlexRadio Systems formally announces the "End of Software Support" for the SDR-1000 transceiver. The last RECOMMENDED stable version of PowerSDR for the SDR-1000 is v1.18.6."

Currently PowerSDR is at Version 2.7.2.

[WU2D]

The 400 lbs of nothing drops the power by reducing the screen voltage on the 814's so that the plate current is 1/2. This is approximately 1/4 Power out. This gives a low power CW mode as well as a control grid modulated mode for AM. That's right: control grid modulation; very popular in the 1930's designs. The 814 has no suppressor grid. In control grid modulation you need plenty of drive so you sure do not want to reduce power with the driver. They achieved the excess drive by using an 807 as a driver. Typically you want to swamp the driver tank (grids of the final) and waste about half of the drive power in that resistor. Of course this broadens the response killing Q making tunning less critical, but more importantly it tames the wildly varying grid impedance as you bias modulate.

In another control grid modulated rig, the famous British Wireless Set 19 XCVR, they cleverly apply AGC with a 6H6 detector to the very hot EF-50 super-pentode driver to linearize the control grid modulation of the 807.

Finally, with any efficiency modulator (Screen, Suppressor or Control grids), a global feedback system utilizing the detected RF envelope sent back to a voltage amplifier close to the mic and combined in antiphase, will linearize the kinks and make tuning and loading less critical.