Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Dec Tue 12, 2017 5:11 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 33 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Mon 06, 2017 5:46 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
Having read the Clubhouse thread about how hard it is to get a SSTrans, and previously having successfully combined the MW synthesizer with a Wenzel transmitter, I wondered if there was a simpler way. There is. It combines the stability and choice of frequency of the synthesizer with 100% modulation capability of a switching circuit.

This Class E is a switching circuit fed by a 50% duty cycle square wave provided by the frequency synthesizer. In the simplest configuration, the power is provided by the output of an LM386 amplifier. The problem with this configuration is that when there's no signal to the amplifier, the carrier drops very low, producing a signal so weak that AM radio noise is heard during silences or low volume audio.

The solution is to keep a carrier going and modulate that with the LM386. 12V is delivered through one winding of an audio transformer while the other winding is fed the output of the 386. Because of the winding resistance, the carrier is not as strong as when the coil tap is fed directly by 12V (unmodulated transmitter), but that's not an issue as this Class E is capable of 450 mW and power needs to be scaled back anyway.

I tested it with my Superradio and a sinewave. I can hear only up to about 16KHz, but that signal was coming through just fine with the Superradio set to wide band. Of course 16KHz disappears with the narrow band setting. The scope shows 100% modulation all the way up there. Even with the Superradio on wide bandwidth, AM doesn't sound like FM. It's a bit muffled for my tastes, and that bears more experimenting. I recall that my Wenzel circuit sounded almost as good as FM.

The first schematic is with no constant carrier. The second provides constant carrier.


Attachments:
class E.jpg
class E.jpg [ 14.55 KiB | Viewed 864 times ]
class E transformer.jpg
class E transformer.jpg [ 15.2 KiB | Viewed 864 times ]
Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Mon 06, 2017 12:13 pm 
Moderator

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19399
Location: Utica, NY 13502 (USA)
What is the difference between the lower circuit and an old-fashioned Class C, plate-modulated RF amplifier like I grew up with?

Dave


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Mon 06, 2017 5:07 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
From the internet:

Quote:
Classes E and F, much like class C, feature RF amplifier topologies that use LC tank circuits. Where class C amplifiers are widely used below 100 MHz, class E amps tend to fall into the VHF and microwave frequency ranges. The difference between class E and class C amps is the active device becoming a switch, rather than operating in the linear portion of its transfer characteristic.

I'm self taught, no authority on classes, so I can only surmise that it may boil down to the fact that the Mosfet is driven by a square wave. One enthusiast measured the efficiency at 98%.

I am curious why the waveform is asymmetrical. Could it be due to the way the LM386 output is biased?


Attachments:
scope.jpg
scope.jpg [ 44.66 KiB | Viewed 803 times ]
Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Mon 06, 2017 8:21 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 176
Location: Arlington, TX, USA
I'm no authority either, but what I understand about class E is the tuned circuit at the output is set up so that the current through the transistor is close to zero as it switches on, raising the efficiency of the circuit and lowering dissipation in the transistor. The capacitor to ground at the output is the main difference between class C and class E. Of course it has to be tuned correctly to work.

As far as the asymmetric waveform, my old Ramsey transmitter did the same thing and I never found out why. My guess is somehow the tuning is a bit off and one of the sidebands is lower than the other. I would need a spectrum analyzer to see what is going on... that's the next piece of test equipment on my list.

I took a different approach after never receiving the SSTRAN unit. I went with low level modulation and a low power linear amplifier and seem to get a nice looking wave on the scope as well as good sound quality. I'm sure the class E approach works as well, just need to do some refining of the design to get there.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Mon 06, 2017 8:42 pm 
Member

Joined: Aug Thu 06, 2015 2:20 am
Posts: 106
Any waveforrm that is not symmetrical above and below the X-axis will have a DC offset and significant 2nd harmonic content which apparently is an issue with Class E amplifiers which must be filtered out.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Tue 07, 2017 1:49 am 
Member

Joined: Apr Thu 14, 2016 8:25 pm
Posts: 282
Location: pensacola fl
Hi if your output tank or load if an antenna can have too hi a Q and this will narrow the frequency response. Class C as well as class E can have a full high fidelity response. We in the radio biz have both kinds as well as others. I suggest you check this angle out. Of course as you lower the Q the output will drop. There is a happy point in there that you can live with but do not expect high 90% + efficiency. As to symmetry look at the output of the modulator with and without audio. Look also at the match between modulator and modulated stage. This could be a factor but i am guessing as i am not sure of the version of the 386 or the current the final needs.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Tue 07, 2017 8:02 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 176
Location: Arlington, TX, USA
Like I said above, my Ramsey AM transmitter had that kind of output, but the unmodulated carrier waveform appeared normal viewed on an oscilloscope, so I don't believe it's excessive distortion of the carrier wave. I would expect such gross distortion to be readily apparent even on an oscilloscope if that were the case. That's why I suspect it's due to unequal sideband amplitudes, but without test equipment it's difficult to be sure. The Ramsey unit has a generic output filter in it with component values chosen based on a range of operating frequencies, so it's not optimized for a specific frequency, which makes me think it's just not tuned correctly.

There are several websites which describe amateur radio class E amplifiers which explain the design and tuning pretty well, although at Part 15 AM power levels I don't think anything is going to go bang if out of adjustment.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Wed 08, 2017 5:33 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
Hi, Eric, similar to your experience, the unmodulated waveform is symmetrical with this circuit. I'll next try a different modulator.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Thu 09, 2017 4:39 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
On Class E websites, for an unmodulated transmitter, the usual coil configuration is as the first schematic. However the second schematic does away with a separate coil and produces similar output. Why is the first schematic the preferred configuration?

Also, I've been using a 6 foot antenna consisting of clip leads hooked together. When I add a clip lead to the last junction so it forms a Y, output increases 50% and so does current consumption. If I just add it to the end of the last lead, the output increases, but not as much as the Y configuration. Why?


Attachments:
loadingcoil.jpg
loadingcoil.jpg [ 13.42 KiB | Viewed 655 times ]
simplecoil.jpg
simplecoil.jpg [ 14.13 KiB | Viewed 655 times ]
Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Thu 09, 2017 4:45 am 
Moderator
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 22133
Location: Somers, CT
I would guess the bottom configuration is forcing your coil to act as an autotransformer. The tap going to B+ would
essentially short the that tap to earth for RF, assuming a low impedance power source. Thus the first half of the
coil acts like a primary winding, while the remaining coil acts like a secondary in an autotransformer configuration.

Looking at the inductances, there are less turns on the bottom half, thus the autotransformer is a voltage stepup.

Which gives the best field strength readings, say across the room? That is what I would go by.

Pete

_________________
Just because it can be done doesn't mean it should be done.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 10, 2017 3:33 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
Thanks, Peter. I got out my field strength meter and it agrees with scope peaking when I have the scope probe clipped onto the insulation of the first test lead attached to the antenna loading coil. The scope displays about 2V p-p when the probe is attached this way. I wonder how I might be able to calibrate that reading so I can calculate the actual voltage at the antenna.

When both the scope and the FSM, with unmodulated carrier the current draw is 87 mA. That's about one watt. Given that the transmitter is reportedly running at 98% efficiency, it would need to be scaled down for part 15 use. Maybe a 5V supply would be all that's needed.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 10, 2017 5:00 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 176
Location: Arlington, TX, USA
I did some playing around in LTSpice simulation with a textbook class E circuit and found the asymmetrical modulation is indeed due to excessive harmonic content (-20dB or so, or about 10% THD), but it's not enough to show up well on an oscilloscope with an unmodulated carrier. Adding a pi filter to the output in simulation gets the modulated waveform much more symmetrical, and I suspect that a higher order filter would work even better, but at the cost of reduced output power. If you're going for maximum range within Part 15 limits then the raw output might be good enough, but I personally would want the harmonics pushed down further for a cleaner signal. For anyone who does not have LTSpice, I highly recommend it; it's a free download from http://www.linear.com/designtools/software/ . This is the kind of thing that circuit simulation is real good at, and you can view FFT plots of the output showing the harmonics with and without the filter added.

Second, the classic class E circuit has another capacitor and inductor in series with the load, and there are several on line component calculators to find these values. The circuit in the original post omits these two components, so it might not be true class E operation, actually it looks like a class C amplifier with a capacitor to ground at the output instead of across the drain inductor, but in reality since B+ and ground are nearly the same at AC it's pretty much equivalent.

I've become intrigued by this setup and when I can get some suitable capacitors I'm going to try my hand at prototyping one of these circuits to run at 1500kHz since that is an unused frequency around here and see how it goes.

At 100mW input I would think the LM386 would work fine as a modulation amplifier since it is rated for 500mW (but I don't know what impedance without looking at the data sheet; it could be 4 ohm for all I know).

About the only drawback to this setup is the need to have everything optimized for one operating frequency only, not so good if the transmitter is meant to have its frequency changed easily, but OK when you know what frequency you're going to stick with (1500kHz in my case).


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 10, 2017 6:51 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
Erich Loepke wrote:
About the only drawback to this setup is the need to have everything optimized for one operating frequency only, not so good if the transmitter is meant to have its frequency changed easily, but OK when you know what frequency you're going to stick with (1500kHz in my case).

I had the same concern. But like you, I'd use it for primarily one frequency. For those who want a selection of frequencies, I have potentially good news.

I've found that different values of the tank coil and the cap to ground have very little effect on output power. Varying the cap does affect the waveform and allows me to peak it at one point, but that peak is negligible. In fact, the tank coil can be any value between 48-227 uH with no change in output. What DOES make a big difference in output is the loading coil. Tuning is very critical there. So the good news is there's a chance that multiple frequencies can be chosen with only needing to tune the loading coil.

I discovered why the broadcast sounded a bit muffled. It was the match between the mp3 player and the LM386 amp. I added a loading resistor to the MP3 player and now the highs are there.

I'm not so concerned about harmonics for two reasons. First, the waveform looks pure on the scope. Compared to when I first experimented, checking with a shortwave radio, the second harmonic is way down, although the first harmonic does show up. Second, being part 15, the power is so low that the harmonics would probably be problematic only to SWLs next door. Tonight I took a receiver outside and I could hear it 400 feet up the street. The nightly DX was interfering at that point, but on my property there was full quieting. I think this circuit has potential for a kit that would be easy to assemble. An on board indicator could be as simple as a milliameter. The circuit draws the most power at resonance.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Tue 21, 2017 4:09 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 176
Location: Arlington, TX, USA
Hey, any more news on your circuit? I put something together myself and it worked somewhat, but all I had on hand was an IRF640 which is not a very good device for this kind of circuit because it has far too much gate capacitance. It runs really hot. I was getting just over a watt into 68 ohms with an input of about 1.8 watts. The MOSFET runs quite hot because of all that internal capacitance. The IRF510 has about 10% the capacitance and is much more suited to this use.

The real class E circuits have a series LC in series with the load, and the tuning is critical to get true zero crossing switching that makes it class E. What I see in other's designs is the C is a variable capacitor which is tuned to get that zero crossing. The big problem is these capacitors quickly get expensive when high power is involved, and even at Part 15 levels the voltages are still high, measuring about 50V or more.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Tue 21, 2017 7:19 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
Hi, Eric, I'm working on it now. With unmodulated carrier, I get about a watt with the IRF510 running at room temperature. I estimate the power by current draw assuming 98% efficiency.

Instead of the LM386 modulator, I'm experimenting with PWM. Both sections of a TL082 comprise a square wave and integrator that produces a clean triangle wave that goes into one input of an LM311 comparator.

The other comparator input is the audio input signal that needs be 0-2 volts to fit the lowest and highest peaks of the triangle wave. Thus the LM311 outputs a 0-100% duty cycle PWM at 27,000 Hz, which according to Nyquist should be high enough for AM radio bandwidths. I like the lower PWM frequency because waveforms are sharper than at higher frequencies. One capacitor change will bring it to 48,000 or 96,000 Hz, if necessary. I just figure that harmonics into RF are less likely to be a problem if the PWM frequency is low to begin with.

The PWM output is fed into one input of a discrete two diode AND gate. The other AND gate input is the synthesizer output. The AND output goes to the gate of the IRF510. Thus the synthesizer output appears at the gate only during the duty cycle highs of the PWM square waves.

Where I am now is that I'm ready to connect the AND gate to the IRF510 and associated circuitry. Hope it works!


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Wed 22, 2017 3:20 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 17961
Location: Warner Robins, GA
That would definitely be nice if it does indeed work.

A PWM part 15 transmitter.

Possibly the first of its kind.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 24, 2017 1:54 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
UPDATE: My PWM scheme does work, after a fashion. I barely had time to work on it before Thanksgiving preparations, but I did get to transmit a sine wave, albeit noisy. My sine generator has a high impedance output, and I'll next put an amplifier between it and the input to the comparator.

I've not actually connected the output of the AND gate to the IRF10 section of the transmitter. So the output I was listening to was not optimum. And the sine wave max amplitude was half of what is needed to provide full modulation. But I could hear the sine wave clearly.

The scoped 27000 Hz PWM peaks contained the 1690 MHz component as expected. I'm really wondering what the waveform will look like when connected to the IRF510 section.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 24, 2017 4:07 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 176
Location: Arlington, TX, USA
If it works out, please post some scope screen shots of the circuit if you can. I'll do the same and start a thread about my low level modulation transmitter as soon as I get it working again; I had a mishap with the Arduino falling off the solderless breadboard which damaged the on board voltage regulator which overvolted the DDS chip and killed it. I'm still waiting on replacements from China; they should be here in about another week since it's been two weeks already. This time I bought two evaluation boards in case something like this happens again. The way I see it, there are two potential possibilities for quality antique radio Part 15 transmitters: My design which will give high quality audio at the expense of some range (I was getting 50-100' with a loading coil at this time at 1500kHz), and a Class E design for maximum range but still decent audio. If we can come up with some working designs I'd be willing to do some PC board layouts for free, just need to pay a board house to have them made. These will be double sided boards with ground planes.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 24, 2017 2:44 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 17961
Location: Warner Robins, GA
Would be nice to have a part 15 transmitter that transmits farther while staying within the legal limits.


Top
 Profile  
 
 Post subject: Re: Class E transmitter with $12 MW Frequency Synthesizer
PostPosted: Nov Fri 24, 2017 4:29 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3143
Tube Radio wrote:
Would be nice to have a part 15 transmitter that transmits farther while staying within the legal limits.


With Part 15 it's mostly about the antenna/ground system and optimum matching. I could hear my Wenzel transmitter a mile away when I mounted my antenna on the roof and used radials. Some say that roof mounting is impermissible under Part 15 rules, others say not, YMMV depending on the FCC inspector. Personally, I can't imagine FCC denying Part 15 use to an apartment dweller, but what's my opinion worth?

In any event, an efficient antenna within Part 15 rules usually will get you several hundred feet range, and at least 100 feet of strong signal.


Top
 Profile  
 
Post New Topic Post Reply  [ 33 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: Southerner and 6 guests



Search for:
Jump to:  


















Privacy Policy :: Powered by phpBB