80/40/20/6m Transmitter Design

[KI4YAN]

Hi guys, I'm about to start the design phase of a semi-coordinated attempt at a band-switching transmitter. I have a pair of NOS 826 transmitting triodes, and have found a published design for a 1KW rig, operating in the 6M and 2m bands. I'm not so sure I like the design, as it uses rather unorthodox methods of frequency generation. I have a sextet of 6BQ6, a TV sweep that traces out like a 6V6 on steroids, and have recently tested one up to 44W of CW output, good for three seconds key down till it melts.  I also have a pair of 6DQ5 sweeps, a powerful sweep capable of 75W of CW, but I'm not sure how much power can be made AM plate modulated. I have a large modulation transformer, multi-match, and will wind plate iron to match, if i can't find it. I've got plenty of iron...

I'm thinking use the sweeps for modulator tubes, and a single 826 for a plate modulated final. Whaddayouguys think? Any hints, tips, resources you have squirreled away?

[The Slab Bacon]

An 826 is a poor choice for a final for a couple reasons. First, the plate lead is a bottom pin instead of a plate cap. Second they are extreemly fragile mechanically. Third, they have a tendancy to lose vacuum. Electrically they are quite robust, I have run them with the plates glowing as bright as the filaments and not hurt them.

I used to have a pair of Gonset VHF amps that used them. I must have bought 25 or 30 hamfest specials to end up with a few good spares. Any small mechanical shock will break the filament and cause a filament to grid short. They will absolutely not tollerate any mechanical jarring or shock. (like dropping) And also they are getting kinda scarce.

If you are planning to build a HB rig, consider using fairly common and easy to get toobz. Dont build it around parts that are made of large amounts of unobtainum.

Also most sweep toobz are fairly low impedance (lower voltage / higher current) and manu of the more popular transmitting toobz are fairly high impedance (higher voltage / lower current) this could leave you in a situation needing a mod transformer with somewhat odd impedances and turns ratios, and may be somewhat hard to find. A pair of 6DQ5s would prolly kick ass modding a pair of 6146s. Both are common, easy to get, and cheap when found.

                                               The Slab Bacon

[Steve - WB3HUZ]

Here's info on sweep tubes in Class C. I'd adjust these rating by about 2/3 for plate modulation.

[Tom WA3KLR]

I looked up the pages in one of my RCA HB-3 databooks that has a separate list of the RCA transmitting tubes and the maximum frequency for full ratings and a maximum frequency at derated operation.  The pages are dated 1953.

For the 826 - 100 % to 250 MHz.,
Reduce plate input power and plate voltage to 80 % at 300 MHz.

It's glass envelope appears to be very similar to the popular VHF pentode 829.  No plate pins out the top however, as Frank said.

[KI4YAN]

As for price, you can't beat the 826. I just bought two NOS tubes for 17$ on ebay. As far as availability goes, you are quite right, Slab. They are getting hard to find.

So, 6DQ5 modulating 6146? How bout a quad of 6BQ6GTB modulating a pair of 6DQ5, you think that'd handle ok? A pair of 6DQ6GTB can easily handle 40W output in class AB1, and you aren't even pushing the rated plate or screen limits. Push a little harder and 50W would be trivial. From a quad, we're looking at a 2k2 primary impedance on the modulation transformer, and 80-90W AF output in class AB1. In the modulator alone, we're looking at under 6$ worth of power tubes...add in the pair of 6DQ5 RF finals, and we're at 30$, for new in box tubes. Not too shabby, methinks.

Now, How does one figure the secondary impedance needed for the modulation transformer? I'm puzzled by this one.

[Steve - WB3HUZ]

http://amfone.net/Amforum/index.php?topic=13025.0

Now, How does one figure the secondary impedance needed for the modulation transformer? I'm puzzled by this one.

[KI4YAN]

OK! so, for class C op, the modulated impedance is V = 2I * R, or rather, R = V/2I.

Sooooo...that means that at, lets say 450V and 180mA, modulated impedance is 1.25K. For two tubes, the plate current doubles, putting it at 360mA. Now I have a modulated impedance of only 625 ohms. Great! I actually have a modulation transformer that can handle that, and the AF power needed for full modulation.

Let the scheming begin!

[Steve - WB3HUZ]

Nope. It's just V/I or in your case 1250.



From the thread, read specifically this post.

http://amfone.net/Amforum/index.php?topic=13025.msg97294#msg97294

[AB2EZ]

There are a number of "rules of thumb" for the design of plate modulators, including the selection of the modulation transformer, that were created during a different era. As a result, since things have changed (see below) some of those rules of thumb can be a little misleading.

1. Designing a plate modulator in the era around 1960

Assumptions and objectives (valid in 1960):

a. You can design/order a modulation transformer with whatever turns ratio, rated output impedance, power handling capabilities, upper and lower 3dB roll off frequency specifications, allowable unbalanced secondary current, high voltage withstanding capability, etc. that you want.

b. The cost of the transformer is going to be essentially proportional to the sum of the amount of iron it contains (to accommodate the magnetic fields without excessive saturation) and the amount of copper it contains (to accommodate the current without excessive heating).

c. You want to design the modulator so as to achieve 100% modulation, with reasonable 3dB bandwidth (defined modestly as 300Hz - 3 kHz in most amateur applications), using an audio output stage with as little power output capability as possible, using as inexpensive a modulation transformer as possible.

Based on those assumptions and objectives, you would use the "rules of thumb" (derived from detailed engineering calculations) that were common in that era:

i. Design the modulator's audio output stage to be capable of delivering, into a matched load, 50% of the carrier input power of the r.f. stage being plate modulated; using whatever output tubes are most economical at that time.

ii. Design/purchase a modulation transformer whose turns ratio will transform the modulation-related plate impedance of the modulated r.f. stage (i.e., plate voltage/plate current = V/I) to the output impedance of the modulator's audio output stage (thus placing a matched load on the modulator's audio output stage)

iii. In addition, select the above modulation transformer so that its equivalent output impedance at the lowest audio frequency of interest (e.g., 300 Hz in this era) is at least as large as the modulation-related plate impedance of the modulated r.f. stage (V/I). This modulation transformer's equivalent output impedance, at any given frequency, is determined by the "magnetizing inductance" of the transformer. For example, if the transformer's specified "output impedance" is 6000 ohms, and the specified lower 3dB roll off frequency of the transformer is 100 Hz, then this means that the transformer's equivalent magnetizing inductance (which is an equivalent inductor across the output of the transformer) is 6000 ohms / [2pi x 100Hz] = 9.55H. Making the magnetizing inductance larger requires more turns on both the primary and the secondary (i.e., more cost)... but results in a modulator that can modulate the r.f. stage at lower audio frequencies.

[Note: one of the reasons why it is not best to lower the power output of a plate modulated transmitter by lowering the screen voltage, without reducing the plate voltage, is that doing so will increase the modulation-related impedance of the r.f. stage (i.e. V/I gets larger)... thereby (for typical modulator designs used in those transmitters) increasing the frequency at which the magnetizing inductance of the modulation transformer begins to roll off the low frequency response of the modulated r.f. stage. In other words, using this method to reduce the output power of a plate modulated transmitter, like a Johnson Ranger, has an unfavorable effect on the low frequency modulation performance of the transmitter]

2. Designing a plate modulator in the current era (2007/8]

Assumptions and objectives (valid today):

a. "Mod" transformers are hard to obtain, and we will be inclined to try to "make-do" with whatever mod transformer we can obtain. I.e., we may have to build our modulator to work with the turns ratio of the modulation transformer we have.

b. Given the above, it is not necessary to have the load on the modulator's audio output stage be matched to the output impedance of the audio output stage. Instead, it is only necessary for the modulator's audio output stage to be capable of delivering the required power (50% of the carrier level input power of the modulated r.f. amplifier) to the modulated r.f. amplifier, even if the actual load impedance on the audio output stage of the modulator is higher than the output impedance of the audio output stage of the modulator*.

[*Typically, we want the actual load on the audio output stage of the modulator to be higher than the output impedance of the modulator's audio output stage... so that the modulator will behave as a source whose source impedance is lower than the modulation-related impedance (V/I) of the modulated r.f. amplifier... thus causing the modulated voltage on the r.f. output stage to more closely follow the desired modulating waveform]

c. We want the equivalent inductance across the output of the modulation transformer (i.e., the modulation transformer's magnetizing inductance) to be large enough so that the impedance associated with that inductance is at least equal the modulation-related plate impedance of the modulated r.f. amplifier at the lowest audio frequency of interest. If we desire a lower frequency 3dB roll off point of 50 Hz, instead of the 300 Hz target typically used in the 1960's, then we need a modulation transformer with a higher magnetizing inductance (more turns on the primary and the secondary) than we would have required  in the 1960's era. In other words, if we have an r.f. amplifier whose modulation-related plate impedance (V/I) is 5000 ohms, and we want to achieve a low frequency 3dB modulation cutoff frequency of 50 Hz, then we need a transformer whose output impedance at 50Hz is at least 5000 ohms. If we consider a transformer whose specified output impedance is 2000 ohms, and whose specified frequency response is 3dB down at 20Hz, then its output impedance at 50Hz is at least 2000 x 50/20 = 5000 ohms... so this transformer would meet our needs in this regard.

Therefore we obtain the following new "rules of thumb"

i. Find a modulation transformer whose output impedance at the lowest desired audio frequency of interest is at least equal to the modulation-related impedance of the modulated r.f. amplifier at that lowest audio frequency of interest. Naturally, it has to be capable of handling the high voltages that will be present between the primary and the secondary, between the primary and/or the secondary and the case, etc. (with lots of margin), it has to be capable of handling the heating associated with currents that will flow through the primary and through the secondary (sufficient wire gauge), and it has to have enough magnetic material to avoid magnetic saturation from the combination of average current and audio current flowing through the windings. If you are using the transformer at input (source) and output (load) impedances that are "in the ballpark" of its specified input and output impedances, and if the transformer is rated for the power level that is going to be delivered to the modulated r.f. amplifier (the load), then it will probably be able to handle the associated currents without excessive heating and without saturation. If the load impedance on the transformer's output (i.e., the modulation-related plate impedance of the r..f. amplifier) is lower than the specified output impedance of the transformer, then the power-handling capability of the transformer will be reduced (because the primary and secondary currents will be higher, for a given output power).

2. Given the turns ratio of the modulation transformer, calculate the impedance of the load on the audio amplifier output stage, as seen from the primary of the transformer. Design the audio output stage of the modulator to be capable of delivering at least 50% of the carrier level input power of the modulated r.f. amplifier into this load. It is desirable to design the audio output stage to be capable of delivering at least 50% more power into this load (i.e., 75% of the carrier level input power of the modulated r.f. amplifier) in order to have adequate margins for achieving 100% modulation.

Happy Holidays!
Stu

[ka3zlr]

It would be interesting from a nostalgic stand point but duty cycle looks a little questionable.

[nu2b]

This push-pull modulator spreadsheet might be of additional use

http://www.qsl.net/nu2b/xls/PPMod20070324.xls

Regards,
BobbyT

[KI4YAN]

Well, my grandiose plans have fallen a bit, I lucked into a bunch of free NOS 6CU6 sweep tubes. Like fifty of the buggers...

And, when I blow all those up, then a 6DQ6, or 6BQ6 will replace them easily.

SO, I've lowered my expectations a bit, and decided to start a little smaller. A pair of 6CU6 in parallel, modulated by class B 6CU6. 1963 ARRL handbook has a 25W modulator using 6BQ6, (almost exactly the same, in fact 6CU6 is the direct equal. Mine all have larger plate assemblies, looking like a 6DQ5.) With a 4K load, a pair in PP class AB1 will produce 25-35W of audio, capable of modulating 50-70W of DC input. In class B, a quite a bit more is easily attainable.

Can I still use the much larger modulation transformer I have, as long as I can get the correct impedance match? (I'm going to assume that it's just like when i'm winding an audio output transformer, too big is about right, ridiculously huge is better.)