Homebrew 3 Channel Pure SIN Generator ( 0 ... 60MHz )

[Yucca]

I'm starting this thread to document my build of a 3 channel pure SIN generator with the following

specs:
==============================
3 Channels (hopefully phase locked).

0 ... 60MHz with 1Hz resolution.

4V p2p into 50 Ohm.

Microcontroller controlled.

3D sweeping capability (sweep width and resolution adjustable on each channel)

When sweeping samples analog input signal in order to hunt for optimum power out. 8bit sampling taken through digital adjustable attenuator (16bit) so that clipping is avoided and true maxima is found. This effectively increases sampling resolution to 24bits.

Ability to log scan results into PC over USB cable for analysis on PC in excel etc.
==============================

I already have a 2nd hand tektronix FG504 sig genny which is great and handy but the output, like most bench gennys, is like a wailing drunk karaoke singer when compared to a true radio quality signal required for experiments in intermodulation.

The hard work I will buy premade in the form of 3 x DDS60 modules, designed and used by amateur radio enthusiasts. The design has evolved over quite a few years and is very stable:

http://www.amqrp.org/kits/dds60/index.html

The modules will be under the control of an ATMega Microcontroller (under the screen in the photo).

I am thinking about and working on the user interface at the moment. The user can either use the small joystick switch you can see in the photo just to the right of the screen, or they can attach a rotary shaft encoder and spin a wheel to intuitively sweep through frequencies having first selected the channel and decade to adjust.

I'll post more as the build and programming progresses.

I posted in TPU section because 3 chan sin is often spoken of in theories of operation. But I would prefer to post in an "Equipment & Measurement" section. That section could include reviews of existing equipment, which would attract lots of search hits, measurement techniques and other tips on how to build and use stuff or how to improvise etc.

Yucca.

[dankie]

Does the TPU require AC or a sine waveform .

At higher frequencies the square wave will be rounded off a bit , a vacuum tube looks like it rounds off the edges also .

I think you are overcomplicating the task at hand , controlling the "phase" of each signal will be hard with a sine wave generator . It can be done more easily with logic lvl pulsing . Then amplified to be smooter .

[Yucca]

Well, I'm commited now; Just purchased the 3 x DDS60 modules. Will now have to explain to my better half exactly what the hell they are lol!

[Yucca]

Does the TPU require AC or a sine waveform .

I dunno, better ask SM. As you know pure variable sin is not easy to do, variable square is easy.

What do you mean AC or SIN? pure AC is pure SIN.

At higher frequencies the square wave will be rounded off a bit , a vacuum tube looks like it rounds off the edges also .

Rounded squares aren't pure freq, square waves are composed of odd integer harmonics of the fundamental. In experiments regarding intermodulation this is tantamount to cluttering noise.

It is easy to turn square into sin by using a lowpass filter set to roloff at fundamental but every time you change freq you have to alter component values. This genny will give very pure sin at any freq rock solid at 1Hz resolution. A nice bit of kit however you look at things.

I think you are overcomplicating the task at hand , controlling the "phase" of each signal will be hard with a sine wave generator . It can be done more easily with logic lvl pulsing . Then amplified to be smooter .

As I said I would prefer to post in an eqpuipment section. I want a 2 chan pure sin gen anyway for stuff other than TPU research, I just thought I'd add a 3rd chan to explore Bruces ideas. The equipment will not be solely for TPU research. I'm just thinking cover lots of combinations of various RF interference phenomena in various coil configs and see if I spot any anomolies worthy of further investigation. I'm just figuring it'll please me more than watching telly because I enjoy programming micros and building things.  To buy a piece of equipment with these specs would cost a few grand.

With regard to phase the chips are digital (they are a complex logic circuit with a final DAC stage) and all driven by xtal clocks, these clocks can be phase locked and so the derived frequencies can also be phase locked and adjusted with some effort in interlinking and programming.

[Peterae]

Go for it Yucca

A great bit of kit for the bench, and future TPU exploration work.

Not sure if the feedback idea will work , as SM has mentioned you can expect to find a hole in your bench if you find the correct frequency's, might be best a put it in a flame proof box