Idea Box: Putting the 10GHz Amplifiers to Good Use.

Now that I have some decent amplifiers, it is time to make a project to make good use of them. since they have a high gain it is essential to avoid over-driving them. 

Here are some ideas:

The easiest modules to employ are the 2-400mW modules, as they only need a single, +15V power supply. The modules are rather small, and heat up quickly, so some kind of heat sink is necessary for those. Not unexpected when I saw the size, and they do have holes for mounting on a heat sink.

The 1-2W module needs a -12V and a +12V supply, as does the amplifiers in the up-converter module I have tested earlier. Those units *need* the -12V to be supplied before the +12V, so some kind of sequencing/safety circuit is necessary.  

For this reason the first construction tests will be done with the low power modules. Here are some options for experiments. 

1) a WBFM modulated DRO, such as a modified HB-100 module driving the amplifier. The HB100 module needs to be modified for use as an exciter for the amplifier, and a modulator is needed, too. The modification would consist in disconnecting the RX mixer circuit and the transmit antenna on the PCB, adding a semi-rigid cable at the "antenna output" and building the HB100 into a shielded box. One more thing is needed, an attenuator in order to avoid over-driving the amplifier.

Adding the attenuator has one more advantage, it provides some isolation of the DRO, giving less drift and external influences.

Such a transmitter can be used together with a satellite LNB and a WBFM capable scanner running in the 600-700MHz range. Separate antennas can be used for simplicity.

It could also be used as aWBFM test TX ("Beacon") when used with an audio frequency keyed system, e.g. with a microprocessor, such as the Arduino with a suitable program. Of course, this could also include a keyer input for manually keying the audio tone.

The range with such a transmitter will be significantly improved when compared to a simple system with the HB100 "stand-alone".

2) A set-up using a cheap 70cm transceiver as the source for the TX signal. This is a bit more involved, but some of the circuitry can be used for other experiments later.

The circuit needed is a frequency multiplier chain, for example 432x2 -> 864x2 ->1728x2 ->3456x3 ->10368MHz (x24 multiplication, so other orders of the multipliers could be useful). A bit of offset from 432 would be good, as I want to avoid running WBFM in the narrow band segment. The advantage of this will be the improved frequency stability, as the reference frequency is usually synthesized. Yes, an attenuator is needed after the 70cm TX, but that is not too difficult, since such a TX can run a power as low as 0.5-1W, and the load impedance is not overly critical for the TX.

Otherwise the multiplied and filtered signal can be used as in example 1 above.

This was all wideband, and this does have its limitations with respect to range, so I would like to use the amplifiers for narrow band transmission as well.

3) Narrow band CW (FM) transmitter. Using the multiplier from example 2 it is then a matter of generating a 432MHz signal with sufficient stability. For a simple, single frequency CW TX, e.g. a "beacon", I have some TXCOs running on 14.4MHz. As those oscillators are likely to generate a square wave, taking out the 5th harmonic on 72MHz and amplifying it should not be tricky. After that a frequency doubler, followed by a tripler will generate a signal on 432MHz. As I have done a bit of VHF construction before I would not expect this to be too difficult, especially as this comes after trying making the higher frequency multipliers. Good shielding and decoupling is a known construction technique for me. Keying can be done in one of the multiplier stages, and this can also be used as marker signals for 2m, 70cm, and 23cm.

200mW CW should be quite good from a hill top.

Because I do not know any people near me running WBFM, this might be a better option to make a first QSO on 10GHz, if I do not build a second simple WBFM set, or find someone who has one or wants to build his or her own.

Now, although I know some, not all people doing microwaves can do CW, so what comes next?

4) Up-converter from a VHF or HF station, so all modes are possible. The receive side can still be a satellite LNB (modified for better reference frequency), so here goes:

Among the stuff I ebay'd lately, there is also a 10-11GHz mixer circuit. The IF connector is not a standard SMA, so it looks like I am going to make a modification so that all three connectors will be SMA. Otherwise I can use the one from the up-converter I already have (and tested), modified, so that the DRO is removed and a multiplier chain (as above) is used in stead. I will probably need a new multiplier chain, but that should be doable, if I can get the first version (above) working. In the up-converter there is (what looks like) a decent band pass filter centered on about 10330MHz. This has to be re-tuned as 10368 is outside the band pass.The amplifier chain in the up-converter as it originally was, is quite bulky, and could be replaced by the 2-400mW amplifier. I suspect there will be sufficient gain.

All in all, it will be interesting to see what I can do with the surplus equipment/modules I have got now. A bit more is on the way, more on that later. There is enough work for quite some time with the speed I normally build stuff. For now, the HB100 modulator is back in the front again. I received some (SMD) 4001s, so the morse tone generator can be made, even if I need to use an adapter from SMD to "normal" (old fashioned) PCB soldering.

Solar Panel Re-Location.

For a while I have had a 50W solar panel running with a sealed (gel) lead-acid battery indoors, running with a charge controller.

Last year I realized that the panel was part in the shade during the deep of winter,  and the battery had been discharged, just by supplying the controller. but at the time it was too late already, so nothing was done.

Looking outside today around noon I saw that the panel was only  about 2/3 illuminated, so I got to change it. 

The panel is now about 1m above ground level, and placed a bit further from the building shading it. Yes, the sun is very low above the horizon at this time of the year here, somewhere around 10-15 deg. at winter solstice, so it does not take much to put the panels in shade.

There will only be a few hours of charging, but it is dfinitely better than at the previous position.

This brings me to another project that should be worked on this winter (apart from the 10GHz equipment). I have been working on re-conditioning some gel-batteries, but they are not yet ready for deployment. I have now decided that as a start, I will get some new batteries, so I can increase the storage capacity,.

Apart from this, I do have a small store of a few 30W (small, flexible) panels, and some very small solar panels. Some of the smaller panels I should use with some LiIon cell chargers (yes they are with BMS) to supply some low power consumption stuff, like a HB100 TX or the like.

The aim is to have some simple low power equipment for monitoring purposes powered entirely by solar, and with the panels and batteries I already have that should be possible.

To be sure, this location of the panel is temporary, and a more permanent location (more mechanically stable) is needed.

Oh, well. I am probably never running out of things to do or try. No time to be bored here. Was there ever?

Successful Test of Amplifiers for 10GHz.

Recently I received some amplifiers capable og working on 10GHz:

Having finished the test set-up I could finally make a rough test. Here are the results.

One claimed to deliver 2W (33dBm) with a gain of 47 dB.

My test resulted in an output of close to 1W (30dBm), with a gain beteween 40 and 45dB. 

A couple claimed to deliver 400mW (26dBm) with a gain of 47dB.

My test result showed about 200mW (23dBm) with a similar gain as the "2W" amplifier above.

Admittedly, the calibration of my test equipment may be about +/- 3dB off when the accuracy of attenuators, and most likely some loss in a SMA to N transition. Given that both tests showed less output I am willing to accept that the result may easily be up to the specification. Until I gan get it measured on a better calibrated set of instruments I will accept that the amplifiers are probably within 1-2dB of the specification.

Now I need to create a good signal, so I can put the amplifiers to good use. 

More 6GHz Attenuators Tested on 10GHz.

 More attenuator testing of nominal 6GHz attenuators measured on 10GHz with the up-converter and the HP432 power meter :

- 10dB attenuators: 2 of them very close to the 10dB mark, the 3rd 9.5dB

- 20dB attenuators: All 3 within 0.5dB of the 20dB mark

- 30dB attenuators: All within 1.5dB of the 30dB mark

All this is good enough for my amateur radio lab. Part of the purpose is avoiding overdrive and/or destruction of input devices in sensitive microwave amplifiers, as well as getting estimated gain figures.

Not bad for relatively low cost attenuators, tested out-of-spec. Now some rough gain testing of amplifiers can be done. Those 6GHz attenuators are now all in the box with 10GHz test equipment.

Error #40 ;) .

 We all do it from time to time. Error #40, also known as "User Error".

Today it was mu turn (again) (Nothing was destroyed).

When testing attenuators for 10GHz I was setting all up (I thought). setting output power of the signal generator, sending the signal through the up converter, got the DC power connected - sending the output to the power meter, and - no power out... Did I break the up-converter? 

No! I had forgotten to set the frequency to 70MHz. The generator starts up with a default 300MHz. the resulting mixed frequency on 10GHz is far outside the 10GHz band pass filter, so no output could be detected. (This does show that the filter is quite good, so there is that).

Set the frequency to the "ordained" 70MHz, and all was good.

I chuckled and got started with the testing of the attenuators (see previous post)

Rough Calibration of Attenuators When Used on 10GHz.

 

A bit of activity in the lab today:

While I have two SMA attenuators capable of working at 10W with 10dB nominal on 10GHz, I wanted to test some lower cost attenuators specified up to 6GHz. How to test that without a true, calibrated 10GHz signal generator. Some thinking was needed. Here is what I did:

Using my calibrated (well enough for me) RF generator on 70MHz entering the signal into the up-converter described a few days ago, except omitting the output (2-stage) amplifier, just using the DRO/Mixer and the first (buffer) stage module.

This got me an output on 10GHz close to the output on 70MHz, e.g. 0dBm on 70MHz -> 0dBm +/- 1dB on 10GHz. It looks like the buffer on 10GHz essentially compensated for the loss in the (passive) mixer. OK, now I have a reference signal, showing -1.5dBm on the HP432 power meter.

Here are the values measured with the HP432:

10W/10GHz/10dB:-10.5dBm   ->9dB

2W/6GHz/10dB:    -11dBm      ->9.5dB

2W/6GHz/6dB:      - 7.5dBm    ->6dB

Non-brand

6Ghz/6dB:             - 9.5dBm    ->8dB

Return loss has not been tested, so the impedances are not well known, but given  the values I got from the brand attenuators I used, those are probably close enough for my purposes.

So now I have a good set of SMA attenuators capable of handling up to 10W, and with sufficient attenuation to avoid destroying the thermistor mount of my HP432 power meter.

Before anyone aks: Yes, I did terminate the input and output of the 200mW output stage, to avoid those stages oscillating. They might be unconditionally stable, but I do not know. With components/modules like this, better safe than sorry.

All this took some time because I removed +12V DC power from the unit every time I changed attenuators etc.

I do have 2 more attenuators capable of operating on 10GHz, those have N-connectors. They will be tested together later, as one of them is mounted in my 2.4GHz up-link transmitter system. Now, at least I know that I can safely measure power with sufficient accuracy on 10GHz.

Why do all this? Today I received a 2W amplifier from another seller in Italy, and I do want to test that one.

New 10GHz Modules Are In. One Tested.

A little while ago I found some 10GHz modules at an Italian seller on Ebay:

- high gain amplifier, 47dB gain, claimed 26dBm out, covering 4 - 10GHz

- a 10GHz passive mixer module

- a tuneable filter, w/cavities

- a TX up-converter module claimed power output 22dBm

All are announced to be tested, but I should test them all.

Some of those need work, some not:

The high gain amplifier is complete with a single supply and SMA connector, so no mod needed

The IF port connector of the mixer needs to be changed into the SMA type, so I can make decent measurements on it with my test equipment.

The tuneable filter should be re-tuned to 10368MHz

The TX up-converter needs a b it more, but I got started with testing it. It is built with modules mounted on a "sandwich" of 2 PCBs.

First of all. The converter uses a dielectric resonator oscillator, so unmodified it is only useful for wideband applications Specifications:

-70MHz IF

- 10240MHZ output

- DRO tuneable from 10.2 - 10.8GHz, so frequencies above 10270MHz all the way up to the 10500MHz band edge should be possible to generate.

- The mixer output is filtered through a 4 or 5-cavity filter. I suspect it will be tuneable in the whole specified frequency range After the filter there are 2 amplifier blocks:

a) what looks like a single stage amplifier in aluminium casing, probably milled. Likely gain about 10dB.

b) the next amplifier module was opened (Taking the lid off, and contains 2 stages, likely gain: 15-20dB

c)  the specified output of the unit is 22dBm (about 150mW)

- the unit needs both +12V and -12V

Knowing that GaAsFETs need to have the negative gate bias before applying the drain voltage, for "proper use" I will need to make a DC voltage sequencer in order to protect the amplifier(s)

Test equipment:

Recently I purchased two attenuators capable of working on 10GHz, with 10dB attenuation and a 10W rating each. With this I can test/measure up to 1W output without destroying the thermistor mount for my old HP432 bolometer. Since the claimed output is 22dBm/about 150mW this should be no problem for testing the up-converter.

For the initial test the 70MHz exciter is my signal generator, capable of an output level of +5dBm, just what the mixer needs, so the test set-up is:

1) signal generator

2) the up converter module supplied with +12V and -12V, using two 3-cell battery cases with each 3x 18650 cells

3) the two 10dB attenuators daisy-chained

4) the HP432 bolometer with a probe capable of measuring on 10GHz

Here are the initial test results:

Entering +5dBm (70MHz) at the mixer gives 200mW (23dBm) output, so the power specification holds, within the estimated accuracy of my test equipment. changing the IF frequency from the signal generator shows about 1dB higher output at 60 and 80MHz, and about 3dB less around 55 and 85MHz, indicating a filter bandwidth of about 30MHz. Not bad if a higher IF is used.

When changing the IF power level the linearity below 150mW looks good, so with a different local oscillator (synthesizer) there should be no problems running SSB/CW through this up-converter. This is definitely encouraging. A relatively simple 200mW WBFM transmitter using this module with a WBFM modulated oscillator on 70 - 80MHz and using a satellite LNB as down converter should provide a decent basis for wide band experiments with a bit more power than the low cost HB100 module.

This first lab test is a success, More experimentation with this module is in order, but that is foe a bit later. The first of those new tests would be a different LO, and a re-tuned filter, so the module can be used at 10360 or nearby. More to come later.

The 10m Band, Update.

 Last night I forgot to switch off my receiver for the International Beacon Project frequency on 28.200MHz. As a result, about 0815 local time, 0715 UTC the 4X6TU beacon popped up with a good signal.

With the FT8 monitoring (24/7), every day for the last month or more, the band has shown spots at my receiver from several continents, many days for 6 continents. North America is the most tricky, because it requires a higher solar flux than the others at my Northern latitude.

South America has not failed on any day, with PY being there daily, and often LU and CX. Some days also CE. and a few times YV, PZ and HC were spotted.

On 28.200 the 4X6Tu has essentially been heard every day, often peaking at S9. Other beacons heard fairly often are rhe ZS6DN and CS3B. A few times the RR9 and the VK6 have been heard. 

Sometimes I hear some signals with a long dash followed by a 2 or 3 letter ID in morse code. As far as I know these are buoys used by fishermen in the Mediterranean Sea. Not quite according to the rules, but a good propagation monitoring tool. One of those was spot on 28.200 one day.

The OTHR (over the horizon) radar systems are now active on 10m, and sometimes obliterating all other signals in a 30kHz or more wide band.

A week or so ago I worked a few North American stations with CW.

Oh, yes, 10m has woken up, and I will expect regular openings for the next few years during the solar maximum.

A Bit of Progress With the HB100 Modulator(s).

Just to remind possible readers of this blog, much of what I write is also a store of ideas and information that I want to be able to find again.

Lately I have not been so active with the soldering. However, since the first test of a modulator I have built another one on a new PCB. This one is a version shown on the website of F6HCC .

My version is simplified, as the idea was to use it with one of my MP3 players for modulating a test signal. This could be voice or (WBFM) modulated morse code.

As far as I know there is a kit out there for the F6HCC modulator, but I built my version on an experimental PCB, as modifications are a lot easier than on such a board.

Compared to the original F6HCC I omitted the (microphone) audio amplifier stage, assuming that the MP3 player should have sufficient output for modulating the HB100. 

The initial test, like the test of the first modulator, shows that the output of the IC-821 transceiver when using the CW side tone, works nicely, and can be adjusted to a voltage deviation of +/-100mV at 4.9VDC.  This looks like a good start. Now, the MP3 players seem to have a too low output for the purpose, at least for the modified F6HCC modulator. Now I have two options:

1) Add the second audio amplifier stage, or

2) Build the "digital" morse generator. This could be:

- a keyed CMOS oscillator, like the F6HCC one, but with a simple keying circuit. I have ordered some CMOS 4001 (NOR gates) IC. Amazing that I did not have any in my stock of components, but they are on the way.

- a simple beep-beep generator as described in the F6HCC modulator, or

- the CMOS oscillator with a microprocessor generating a repeated CW keying signal. Somewhere I should have a program for that in a PICAXE. The other option is using the Arduino. I think I found a program for a "beacon keyer" with dual output, a simple CW key output and a (side) tone output. With a (hardware or software) modification a simple input for a CW keyer, even a straight key, is possible. The microprocessor road will require me to learn to set up programming of the processors, and I have to learn some programming, too. The PICAXE with the BASIC programming language that I have a decent grasp of, is probably the simplest way, and the smallest of the PICAXE processors is an 8-pin DIL package, so there is plenty of space on the PCB to have that as well as the keyed oscillator.

It looks like I will be building a third modulator PCB on a slightly larger PCB, because this time I want the full circuit on the PCB, with space to spare. The very first modulator used an external AF amplifier for modulation, but I want to avoid that for mechanical simplicity - and also avoid lots of wires between PCBs, switches etc.

The end result will have the microphone amplification, so it is likely to be usable for the MP3 player, too.

Later, I expect to make myself a "WBFM modulated CW beacon", that I can set up in the garden and test the range of the setup. That could be a simplified version using the microprocessor keyer/audio generator and the F6FCC modulator without the extra audio amplification.

The microprocessor should also be usable as a true CW beacon keyer.

Oh, yes, I do need to (re)learn some programming and get the things set up for PICAXE and/or Arduino programming. Everything takes time.

Idea Box: Single Band SDR Receiver With Old SDR Kits.

 

I have some old "single XO" SDR kits from Germany. I think I have some Softrock kits somewhere, too, they use discrete crystal oscillators.

As indicated they have a canned (DIP/DIL) crystal oscillator running on 4x the RX operating frequency, and the quadrature signals are generated using a 74AC74 dual flip-flop as divider/phasing, providing quadrature local oscillator signals on the receiver boards. 

The receivers cover only one small frequency range around the LO frequency, so with a 48kHz sound card interface the max bandwidth covered is a bit less than 48kHz. Sound cards with higher sampling frequencies can, of course provide more coverage, but the 48kHz sampling frequency is now ubuquitous in computer sound systems.

Here is the idea. Not particularly complicated or original, but this could make better use of those old kits.

My intention is to replace the fixed XO with a Si5351 module controlled by Arduino. This could be programmed in, say 10kHz steps for the LO, or indeed for a single frequency. There are several Arduino programs for making a VFO with the Si5351, so I should be able to modify the program to suit this purpose. 

I have one finished kit of this kind, and a few (2 or 3) not yet assembled, so I should be able to make simple SDR system for rarely changed frequencies, such as beacon frequencies or digital mode frequencies. All for monitoring purposes.

In addition, if all the wanted frequencies are within 45kHz, the system could be used as-is for a few frequencies, branching out the "stereo" I/Q outputs to more than one sound card/computer for dual monitoring, or alternatively, use a local webSDR server. 

I think that the splitting of I/Q signals is the simplest solution to this, though.

The single frequency can be generated by a suitable XO, if that cannot be found, the Si5351/Arduino can be used to generate a single frequency.

I do have a finished kit for 80m, and I will check if I have a suitable XO for covering the FT8, JS8, WSPR/QRSS frequencies. It does look like a 14318kHz XO is usable for covering these frequencies with the 48kHz sampling frequency, although JS8 could be a bit tricky.