EMRFD Message Archive 11583

Message Date From Subject
11583 2015-09-05 07:25:51 ravimiranda Source Coupled JFET - Oscillator
Hi
Apologies to those who may have seen the document below.

I'm trying to build an antenna analyser, the oscillator section (2 x J310), Buffer (J310), Amplifier 2N5109
detector board, and the Instrumentation Amplifier fed back to the OSC section via a BS170

What would be the best way to have a distortion free circuit, and avoid any stray parasitics? Ideally this should work from 1Mhz - 30Mhz (50?)

You can access this file at the URL:
https://groups.yahoo.com/neo/groups/emrfd/files/2E0RVI/New-Oscillator-Spice.PNG

 

Thank you for your time.

Regards

Ravi


11584 2015-09-05 08:11:54 Thomas S. Knutsen Re: Source Coupled JFET - Oscillator
Are you sure you want to try to build a distortion free oscilator that spans a 30:1 or 50:1 range?  Doing that without any bandswitching and filtering would be quite hard. 
I'd suggest you take a look at one of the great classics from HP, 8601 is a sweep generator coverring the range 1MHz to 110MHz. This is done by building a VHF source, and mixing with a fixed oscilator.

In order to cover the range you want, you could build a 100MHz fixed oscilator (crystal) and a 100-130MHz VCO. The VCO would then cover a range of 1.3:1, quite doable. You should be able to use the oscilator you have, by adding a tuning element like a varicap, or there are a lot of ready made modules that could work, Mini-circuits POS150 comes to mind.  A lowpass filter and perhaps a amplifier and you are off to a great start.

In addition, there is a lot of ready made oscilator modules available, just observe that most of them have square wave outputs, and as such you would probably need filtering.

Just observe that while building a directional coupler with good directivity is easy at HF, it quick gets harder as you get onto VHF. 
How do you plan to do the measuring in your analyzer project?


73 de Thomas LA3PNA.


2015-09-05 16:25 GMT+02:00 ravimiranda@gmail.com [emrfd] <emrfd@yahoogroups.com>:
 

Hi
Apologies to those who may have seen the document below.

I'm trying to build an antenna analyser, the oscillator section (2 x J310), Buffer (J310), Amplifier 2N5109
detector board, and the Instrumentation Amplifier fed back to the OSC section via a BS170

What would be the best way to have a distortion free circuit, and avoid any stray parasitics? Ideally this should work from 1Mhz - 30Mhz (50?)

You can access this file at the URL:
https://groups.yahoo.com/neo/groups/emrfd/files/2E0RVI/New-Oscillator-Spice.PNG

 

Thank you for your time.

Regards

Ravi


11585 2015-09-05 09:10:05 Ravi Miranda Re: Source Coupled JFET - Oscillator
Hi Thomas

Since I had to break the circuit down into a smaller chunk as I've built the whole circuit on bits of stripboard (just as a Proof of Concept). The original build failed and so I resorted to looking at things one section at a time.

I've planned for 6 bands (switched coils) and this I feel should be adequate for this design. Thank you for the suggestion around HP/8601 and hope to try that once I make this successful. The detector board is based around a bridge 50ohm and 3 Ge-diodes.

Kind regards

Ravi


11586 2015-09-05 10:12:23 Rod Re: Source Coupled JFET - Oscillator
Ravi,

I have built such a device. I took the view that for antenna purposes,
and for light-duty
bench purposes, I made some greatly simplifying decisions.

I used an AD9850 dds board from ebay, an arduino mini-pro, and an AD8307 log
amp to measure RF levels. The output of the AD9850 is amplified by a MAV11,
followed by a compromise 16MHz LPF.

The AD8307 circuit is from the W7ZOI power meter article. The DC output
is fed into
the ATmega328 a/d input. I use a precision 2.5v reference source.

I use software calibration; cable the RF out to the RF in, push the CAL
button, and
the ATmega328 s/w does a software calibration to "level" the input readings.

A pleasant side-effect of that approach is the software calibration
process can
compensate for an external network's transfer characteristics.

The instrument has two modes of operation: dBc, and dBr. The dBc mode is
very
close to dBm. When you enter the dBr mode, the dBc reading of the moment is
used to established the reference point. Thereafter, measurements are
displayed
in dBr, which directly shows loss or gain, eliminating the need to do
mental math.

Used with an external return loss bridge, the instrument is very
flexible and useful.

I need to reiterate: for real-world antenna measurements, and many bench
measurements, it is not necessary to use lab-quality gear. An example is the
compromise 16 MHz LPF serves well for the most of the HF spectrum. The
AD9850 is
clean enough in the 3 to10 MHz range, and the compromise LPF cleans up
the spurs
nicely up to about 30 MHz.

In addition to the 16x2 LCD display, the instrument uses a serial port
to output
frequency (with 1Hz resolution) and amplitude (with .1 dB resolution)
for logging
to a PC.

I wrote data collection software for the PC that puts the output into a
file format
useful for a very powerful (and open-source) data plotting program
called KST.

I built the instrument because I needed to characterize h/b crystal
filters and preserve
the results on a printed page. My HP 140/8553 specan is a nice
instrument, but it
does not provide printed output with digital frequency output.

Attached is a file showing multiple high resolution crystal filter
sweeps, to give
you and idea of how versatile this configuration is.

I am open to sharing the software under a GNU license.

Let me know if this is useful to you.

Regards,

Rod KM6SN
attached: crystal filter scans


11587 2015-09-05 10:17:13 Ashhar Farhan Re: Source Coupled JFET - Oscillator
see the sweeperino http://hfsignals.blogspot.in/p/sweeperino.html, it is a fairly robust instrument, you have to be aware that the bandpass sweeps will show the odd harmonics. as long as you keep that in mind, you are safe.

- f

11588 2015-09-05 10:28:08 Ashhar Farhan Re: Source Coupled JFET - Oscillator
ravi, 

there are many ways to get there. 
1. Use a VFO that goes from 15 Mhz to 30 MHz and divide this down by 2 for 7-14, by 4 for 7-3.5 and by 8 for 3.5 to 1.6 Mhz. For each octave, you will also have to use at least a three section low-pass filter to keep the harmonics down. The HP8640jr by W7ZOI does precisely this.
2. Hetrodyning a 45 to 75 Mhz VFO with a 75 Mhz oscillator. But you must be aware that all is not well with this scheme. For instance, if your VFO is tuned to 70 MHz and the beat oscillator is at 75 Mhz, you must be prepared for spurs at 10 MHz, 15 Mhz, etc due to harmonic mixing. The VFo stability is not going to be great either.
- f