The AM Forum

What does the majority of folks believe the LOAD control is for in a typical transmitter's PI Network?

My understanding is a number of transmitters/transceivers have a variable output impedance range between 50 and 75 ohms. I understand there are transmitters that have a much lower and higher impedance range which includes the 50 to 75 ohm impedance range.

During tune up, after peaking drive and dipping plate current, then by adjusting the LOAD control varies the transmitter's output impedance between the load impedance range of 50 to 75 ohms so the transmitter's output impedance matches as close as possible to the load seen by the transmitter. If the impedance load presented to the transmitter is either higher or lower than the 50 to 75 ohm impedance range then a tuner would be required to bring the impedance load, presented to the transmitter, down so it's within this 50 to 75 ohm range.

73s
Mike

The load control does just what it says - varies the load placed on the final. A lighter load means less plate current when the plate/tune control is set for a dip in plate current. A heavy load means more plate current when the plate/tuned control is set for a dip in plate current. That's it.

Yes, impedances and impedance transformation are involved, but the range of input impedance that can still be properly tuned and loaded depends on the values of the components in the pi-network. Many modern rigs won't handle input impedances far from the preferred 50 Ohms. But older many older 50-60s vintage transmitters will handle impedance far different than 50 Ohms. I've tuned a Viking II into an antennas with 5:1 SWR, no problem.

The bottom line is that if you can tune and load the transmitter to the proper/desired plate current and you do not have any arcing or excessive heating in the pi-network (especially under modulation), you don't need a tuner.

Think of the final plate tank as a flywheel. Adjusting the load starts bleeding off some of the circulating power out to your antenna load. Max cap = light load & less power bled off of the flywheel, Min Cap = Heavy Load, takes more power from the tube to keep the flywheel spinning because your putting more energy into your load (Antenna).

At a normal level of grid drive, one then adjusts the output loading and plate tuning alternately until all the available power in the pi-network is being delivered. At that point one has adjusted the network to the point where its output impedance equals the load impedance. In other words, whenever the output loading is adjusted to deliver all the available power, the network output resistance equals the load resistance.

If there is reactance in the load it is reflected into the pi-network, but when the plate tuning is adjusted for the dip the reactance in the load is canceled by whatever change in plate tuning resulted in the dip in plate current, thus achieving resonance in the network. Since load reactance has been canceled the result is a conjugate match, because the remaining reactance in the network is equal to that in the load, but of opposite sign.

Walt, W2DU

Walt, is this why a transmitter with a separately adjustable tune cap, roller inductor, and load cap seems to be able to cater to the more unusual loads?

Patrick,
           In a poifikt woild that is nervana! It would give one the ability to match just about anthing, but................................. It also gives one too many possible choices, sometimes making it a difficult pain in the arse to find the sweet spot. Especially as you will find several possible scenarios, but some are better than others, as you can easily end up with a low Q and lots of "artifacts" and other junk as well.

I like to have a fixed L in the pi tank and just variable capz on the tune and load. (or the way old Viking 1&2s were where the cap and rollie duck are geared together.

                                                                       The Slab Bacon

Mike,

One terminology that throws most people in the beginning...  to load a transmitter more heavily (generally to produce more plate current and power) requires LESS C2 load capacitance. In other words the loading cap needs to be unmeshed to load the final more heavily.

And the opposite is true - more capacitance means lighter loading on the final.  It seems verbally logical to increase the loading capacitor's capacitance to load more heavily, but it's just the opposite.

(C1 = tuning cap..... C2 = loading cap in the pi-network tank circuit)

I usually mark the front panel of my rigs with an arrow showing the direction of less loading capacitance. When using a linear amplifier, unmeshing C2 to load the linear final heavier is important. This step is required to tune the linear amplifier for best IMD figures.

For a class C amp, C2 is basically used to adjust the final for desired operating current and power out. As Walt described above, this setting (along with C1 and L1)  is directly dependent on the antenna impedance for a particular freq.

T

How so?

Ok, so how does that improve IMD?

I knew Pi-L had better harmonic suppression since the HF cutoff is sharper (it's a higher order filter). Since IMD produces responses within a few kHz of the carrier frequency, I don't see how the Pi-L would make any difference here. Any IMD would be within the passband.

In my post, just to clarify -  what I meant by IMD improvement was the ability of the loading capacitor to swing the final tube's operating point (in a linear amplifier) into the straighter part of its transfer characteristic curve. (Into the more linear part of its curve by loading the amplifier more heavily.)

However, I believe that the tank circuit bandwidth itself has virtually no effect on IMD because it is so broadly tuned (low Q) it will have little effect on improving 3rd, 5th, etc., IMD products that are say within  +-  10kc away.

But maybe there is some other IMD benefit of a pi-L I am missing?

Certainly the pi-L will improve harmonic rejection/filtering  performance X2, X3 of the fundamental.

T

Thanks guys, I appreciate all the responses.

The responses stated what I believed to be what the LOAD control does and what it is for.

The reason I asked the question was because I read on another ham forum where a ham would tune his transmitter up for maximum power output of 100 watts. Once tuning was completed he would then rotate the LOAD control to reduce the power output down to 40 watts. I disagreed with that for the reasons stated plus because that's not what the LOAD control is for.

Again thanks guys.

73s
Mike

Hi Mike,

In a sense he is correct, though he also needs to reoptimize the other parameters that the loading control will change from a big power reduction like that. (C1, screen current, grid current, etc)  You didn't mention if it was a linear or class C amplifier, but here's the deal....   The pi-network is simply an adjustable impedance matching circuit to go from a high impedance tube's plate output to a low impedance antenna.  When running class C, it is perfectly acceptable (in a limited range) to adjust your power output using the loading control as long as the other tube parameters are kept optimized too. With my 4X1 class C plate modulated rig, there are days when I feel like loading it up to low power or running full strap. I can do it with the loading control (and repeak the tuning cap, etc) or do it with a combination of higher and lower plate voltage too.   However you may find there is a particular "sweet spot" that the rig likes where the mod transformer ratio, tank L/C ratio efficiency and other parameters all work together best, so we may choose to run it at those parameters all the time.

Not to be cornfused... it is a common practice using linear amps to peak c1 and c2 for maximum power output, then unmesh C2 (loading cap) until the power drops off about 5-7%. This will load the tube more heavily and produce a cleaner signal as I described in a post above. I use a low duty cycle (10% pulse)  "chirper" to accomplish this without putting a big strain on the amp at full carrier power.

So, the loading control is part of a matching circuit and in addition can be used to control power output in some limited range.

Hope this helps.

T

Good analogy, and the load control is like the clutch on a manual transmission.  Notice that with the clutch disengaged, you can "blip" the throttle and the engine speed will increase quickly, but little power to the rear wheels.  This is analogous to the sharp dip you can get with a lightly loaded final.  Now engage the clutch (ie load the engine) and the vehicle will not respond nearly as quickly to a throttle input (analogous to the broad dip you get with a heavily loaded final) but more power to the wheels.

Question: I wonder if we could say that the broad and sharp dips you described are the result of the tank Q changing as a result of the higher and lower load placed on the tank by the final and antenna? Just like when we see the Q drop as we add lower and lower resistors across a parallel resonant circuit.  ;)   ie, Does the effective Q of the tank change from say, 12 to 10 as we load an amplifier more heavily? From the sharp and broad tuning, it seems to act that way...

When I load my little Henry linear runinng at only 1500V with HUGE loading capacitance for maximum linearity, the dip is so broad I can barely see it. I need to use the output power meter to find it. This is the way I create a super clean driver. I found these settings by trial and error watching the IMD readings.

T

Linear amps class AB are different animals.  The sharper dip with the "dip and load" rigs is because they are class C PAs.  It's important for the original questioner to understand that plate current meters show average current.   A typical AM plate modulated tube rig has a class C PA so the tube is in conduction for only a tiny part of each cycle.  So when the tank is resonant, the current dips because you are seeing the actual low average current flow due to the tube being cut off most of the time.  

When you load the plate by reducing the load cap you take the tank off resonance a bit and the average plate current seems to rise because the tank flywheel is "hiding" the tube's small conduction time.

Amps conduct for much more of the cycle being linear so the dip is not so obvious because the average current is higher anyway at resonance.  They should still be loaded but you won't see as big a difference in Ip or to put it another way, you don't see as much of a dip.  Making it even worse (or better depending on how you look at it) are commercial tube amps that are made to cover 160 to 10 with as few parts as possible so on say 160 the tank may not be screaming as much as it could if it were a HB amp where the builder optimized the tank for the low bands.

I have never built any amp or rig in my life so I don't know s**t about this stuff compared to you guys; I just happen to know a little bit about this one thing because last year I totally did not get the dip and load thing and why you increase loading by decreasing the load cap (which sounded totally unintuitive to me) so I did some reading to try to figure it out.  Making it more confusing is some mfrs. label the load cap in ways that say things like "more loading ---->" and when you look inside the plates are meshing but others are unmeshing.  

Didn't we go through this rather extensively just a little while ago?

The amplifier operates Class AB1 in a sideband transceiver.

Again, thanks guys for the informative discussion.

73s
Mike

Yes. There is no improvement in IMD when using a Pi-L network. Any amp can be loaded more heavily to the point where IMD is minimized. But the Pi-L provides no documented improvement.

AM broadcast uses Pi-L for 2nd harmonic reduction improvement. (unless you were just refering to IMDs!)


 Pi-L and Pi output Networks  (http://www.smeter.net/amplifiers/pi-l-and-pi-amplifier-outputs.php)

 


        I used PI-L output networks in my old contest single band finals 'cause it
        made it possable to use a smaller value output (loading) cap...

        No need for multi section BC tuning and switched padding caps in the
       160-80 meter units, etc...

       I also had a regulated switchable bias supply that went from AB thru to C
       so the finals were usable in SSB and CW as well as AM if I stuck the modulator
       in the HV line...

    

I guess you just made the assumption it was a class C amp.  ;D

That's OK -

He called it a "transmitter" in his first post and didn't mention it was driver/linear until his third post.  I think most AMers think of a transmitter as being class C  vs: talking about a ricebox /driver with a linear.

But no big deal -  cuz not being sure, we all addressed both class C and linear operation related to loading early on the first page for him...



Other news:  Gawd... it's 102 degrees right now looking at our porch thermometer in the shade - new record for this day... :o


T

It makes ya wonder, T I just came in off the Roof of the Barn Man it's Freaking Hot.

73

Jack.

To me, "transmitter" is a self contained unit that has all the necessary circuitry to generate an RF signal on some selected frequency and modulated in some way with a tool provided by the user.  The tool might be a switch such as a cw key, or equipment to provide some audio signal source.  And that's all it does.  It doesn't receive.  If it does then it is not a transmitter, it is a transceiver.  So, an amp is not a transmitter.  A transceiver is not a transmitter.  The two together is not a transmitter, even though they can transmit.  So, then the guy wrote "transmitter" I pictured a Ranger, DX100, BC610...something like that and I bet everyone else here did too (except maybe the solid state guys  :) ).


Yikes, reminds me of the summer of '95 when we had about 6 weeks of daily highs in at least the 90s, nightly lows only down to the low 80s and many days over 100.   We hit a record 106 on one of those days.  I felt like I was back in southern Alabama.   Only time up here I have taken a cold shower and it felt good.  I was younger and dumber and I went outside one afternoon when it was around 100 and started to mow the grass with a reel lawn mower, the kind with the drum of blades you push around.  After about 15 minutes I started to feel very strange and I didn't like it.  I dropped the mower and went directly inside where I took off my clothes and got in the shower with it on cold and stood there for about 20 minutes.  To this day I think that kept me out of the hospital.   A lot of elderly people died that summer from the heat.  From that the City of Chicago set up a heat emergency system with cooling centers for people. 

R. 

Was doing some errands and was driving back home around 1:30 PM today and the temp gauge that reads outside temps, said 105. It think Newark, NJ and New York City both hit 103 today. At 9:27 PM tonight, temp in central NJ is 90 degrees with a feels like of 96 degrees.

Please provide some documentation comparing the Pi and Pi-L networks in identical setups for HF operation and the prodcedures for finding the sweet spot. If requesting such valuable information is causing trouble, so be it.   Even troublemakers want to tune their amps properly.  ;)

Yes, IMD. I've never seen any documentation that the Pi-L improves IMD. I'm sure Brian will post some here soon.

Cool ;D