Outdated Computers - Spare Parts?

Over time I have collected a bunch of old laptop computers, all too old to run just about anything remotely modern, even decent Linux distributions.

So what to do with this? 

One of the older laptops has 2GB memory and a Pentium 4 (2.4GHz), I could probably run a Windows XP with some "Windows only" programs that I still want to use, e.g. the control program for the 35 - 4400MHz signal generator, some spectrum analyzer software etc, possibly WSJT or QRSS software.

The processors of those old laptops simply don't run fast enough, and the RAM is too low to run any modern applications, so that is really useless. 

What else can be used of such old laptops? 

Batteries can always be disassembled, and the cells used for rechargeable battery packs, providing that the cells have not died. Some testing of their capacity will be needed.

Old hard disks should be checked for usable information on them, and backed up. Really old ones, i.e. used PATA disks should probably be discarded and disassembled for parts. (Stepper motors and magnets come to mind, maybe a few more items.)

From the main boards a few components could be of interest as spare parts. Switch transistors, possibly some capacitors or inductors for the switch mode circuits.

A few modules, such as WiFi modules could possibly be used elsewhere

I would save some small speakers from laptops.

One thing seems quite useful: The laptop screens. With an interface they should be useful for some Raspberry Pi projects. I have a few from old netbook PCs with only 1024x600, a 800x480, and one 1366x768 screen from a recently disassembled defective machine.

With all the propagation monitoring I want to do those screens should prove useful in the shack, and in the lab.

Right now I have two dead Asus machines (one with dual core i5 and one with quad core i7). absolutely nothing happens when the (19V) power supply is connected, and the PS is working. From the description of other faulty laptops etc I suspect one or two defective switch transistors, as I have have had some power cuts with possible transient voltages. Here the disassembled old laptop main boards might provide spare parts. If not I should try to get some replacements.

Fault finding and repair is something I have not done a lot of, but learning something new is always of interest, especially if I can get some useful equipment up and running again. Yes, I have had the tendency to just buy new (or good used) stuff, but the stock is quite high now, so I should get to use more of what I have.

Additionally, I have some very old - and very large/heavy - test equipment. I have got some new and smaller equipment, so the really old/heavy stuff, like an ancient 1.7 - 4GHz signal generator with a **klystron** is not too useful. Very heavy and bulky, and unstable. The new ADF4351 based generator is much better and takes up a fraction of the space. Any useful parts from the generator, like the attenuator(s), will, of course, be salvaged, as they are useful in the lab. 

So this is an effort combining tidying the lab, making space and improving my test equipment, all while I am building some ham radio projects. Absolutely no time to be bored (what is "bored?)

Especially the lab is in an attic with essentially no useful vertical walls, so space is at a premium. Also, while the newer/smaller test equipment is not of professional standard, it will work nicely for my amateur use.

Slow Progress on the 10HGz WFM Project.

 The past few weeks I have been working a bit on the modulator for the 10GHz HB100 module. The modulator worked, but I needed to build it up on some prototype PCBs and that is almost there. A CW (audio) generator is under way, and then it needs to go into a box, so it can be operated portable.

I may also make myself a test TX ("beacon") with the HB100 and just a CW audio generator. I think I have found some directions where I can test the TX on top of a 12m telescopic fibreglass mast and then walk on road and path to test the path for (near) line-of-sight propagation. First tests will be in the garden, of course, with a max range of about 35m ;)

Then I will have to take some walks, I suspect that I have found a possible distance of about 600m, and maybe one of about 1.6km. 

This will all be using the HB100 with the built-in dual patch antenna on the PCB, and the LNB just with the integrated feed horn. Not a lot of gain, but some people have claimed the range of such a simple arrangement to be 8, maybe 15km. The distance to a fellow amateur who is interested in participating is less than 10km, and we suspect that we could possibly make a line-of-sight contact from garden to garden.

Initially I just build a single TX and RX, but having made one set it is not too difficult to make another.

For longer distances some improved antennas or higher power is needed for WFM contacts.

I hope to be able to make the first one-way tests in my garden this week end, and hope for some decent weather to make the tests on longer distances.

I will need the weather to be dry and not too windy for these experiments, as the HB100 module(s) need to be in the open air, or at least with very low loss material in front of the patch antennas.

Now we start the idea box:

Improved antennas can be the following:

- mounting the HB100 and the LNB (or just the LNB) on a dish. Not easily portable if I need to up a steep road to a hilltop, or up a narrow staircase to a watch tower or the like. Also, for the dish a tripod is needed, and it should not be too flimsy.

- The second option involves more work: Horn antennas. In this case both the HB100 and the LNB need to be "connected" to horn antennas. For this some coax-to-waveguide transitions are needed. I have one.

It may be possible to "glue" a horn antenna in front of the LNB. I will have to test how efficient that will be.

The HB100 *could* be mounted inside a "horn". I do not know how efficient that will be, so modifying a HB100 with a connector/coax to the coax/WG transition.

- The other option for keeping the size/bulk of the equipment low also involves more work: Going narrow band. This could be NBFM as we use it on 2m and 70cm, or it could be a CW transmitter, to keep it as simple as possible. Both would likely involve generating a much lower frequency, and using frequency multipliers to get to 10GHz. Much more complex electronically, but still relatively simple mechanical construction.

We shall see what we will do, but first some short range tests will be done, likely this autumn.

10 Meters Is Opening, Slowly, But Surely.

 For the last few weeks I have been monitoring 10m FT8, and lately the beacons on 28.200.

A few weeks ago the band started opening to South America from here. There has not been a day within the last 3 -4 weeks without receiving SA stations, mostly from Brazil, and often stations were also coming in from Southern Africa. I suspect that some of this propagation was trans-equatorial propagation (TEP). This fits well into the pattern of TEP being strongest around the Equinox times.

Then stations started coming in from the Far East and Oceania, and the South American stations started coming in during day time, and several stations from Australia. 

The past week the propagation has moved closer to the North American continent, Northern part of South America, and Caribbean stations coming in, and the past few days South Eastern and also North Eastern United states have come through, as well as regular propagation to the Middle East and Atlantic islands like the Canaries and the Azores. These are clearly single hop F2 propagation  

The past week or so I have re-started monitoring 28.200, and  daytime propagation has been detected on this frequency. Almost daily signals from CS3B and ZS6DN, and occasionally very strong signals from the 4X6TU beacon.

To me this means that the F2 propagation on 10m has returned after the deep solar minimum, and I am happy that my favourite HF band is coming to life again.

I suspect that we will see more North Atlantic propagation in the next few months, unless the sun goes very quiet again. That may happen in short periods, but I suspect that we will have good, maybe excellent 10m propagation the next 3 - 5 years or so. Who knows, with the new weak signal digital modes we may even see weak F2 propagation on 6m. That would be exciting. 

The most general predictions for the next solar maximum (cycle 25) is much like the previous cycle, but we can always hope that it will be larger. As a Danish poet said about 100 years ago: "Prediction is difficult, especially about the future" ;)

Spectrum Analyzer Extender Using the HB100 Module as Down Converter.

Tonight I made a test of using the HB100 module as a down converter, so I should be able to "see" 10GHz band signals on my (low-cost) Tiny SA spectrum analyzer.

I have a few of the modules, so here goes:

With a previous test I had used a LNB with a more expensive spectrum analyzer I had access to. I tuned some HB100 modules to different frequencies, so I had some simple signal sources on frequencies from 10300 up to 10500MHz in 25MHz steps. I also made one for 10368MHz (narrow band) and for 10489MHz (QO100 segment).

Today I wanted to test how the HB100 module did as a simple dowm converter. Because it has been used with (broadcast) FM band receivers I am aware that it can be used as a general down converter, and sure enough, using a 10450MHz HB100 and a 10375MHz module, connecting the spectrum analyzer to the IF output of the module produced a 75MHz IF output. It does work as I expected.

However, I do not know the RF or IF) bandwidth of the module, so I have not checked that.

The base bandwidth of the Tiny SA is 350MHz, so I tested with a 10300 module and a 10600 module, and got a nice 300MHz signal. Using the modules is a bit tricky because the apparent output varies a lot when tuning.

Using a 10600MHz module should make "seeing" 10300 - 10900MHz with a small gap around 10600 (zero IF). Not too bad for a $5 module. For the putpose I also made a 10500MHz module (10200 - 10800MHz coverage) a 10300MHz module can be added for coverage all the way down to the 10000MHz band edge (10000 - 10600MHz)

I am well aware that I do not have any image rejection, but as long as I can calculate expected frequencies I should be able to estimate the spectrum of an oscillator or a transmitter.

The Tiny SA has a base bandwidth of 350MHz (LOW input), but also has an extended coverage up to 960MHz, so it should be possible to see more than 1800MHz with a single module. The question is how much IF bandwidth the module can provide. 

SM6WHY estimates on his blog that the IF bandwidth of the HB100 might be up to 2GHz. I have my doubts, but if that is the case I might try using my Chinese 35 - 4400MHz (primitive) spectrum analyzer which uses direct conversion to sweep the band. If I recall correctly, only 350MHz can be displayed at a time, but the 10500MHz version then might be able to cover 8500 - 12500MHz with a small gap around 10500.

Even if this is far from calibrated, it is much better than not being able to see anything.

For the initial test I just used the modules lying on the desk, with the antennas there. I do think that removing/disconnecting the antenna(s) and adding connectors, and put the module in a stable casing will improve the set-up considerably, both with respect to stability and reliability.

What do you think? a spectrum analyzer extender to the 10GHz band for about $5? Not bad. 

Mini-Lab Downstairs.

While I do have a mostly complete lab upstairs, it will be nice to do some home construction at the operating position downstairs, and have at least some test equipment there. This mostly for simple constructions and kit building.

Here is an overview of what I have for work downstairs, with a few ideas for improvements.

- Soldering iron and antistatic mats.

- 2 Multimeters (DMM).

- single channel portable oscilloscope (100MHz?, more like 30MHz, but OK)

- LCR-multimeter (not for very low L and C values, but a start, a better one is upstairs)

- Nano VNA (900MHz Fundamental wave generator up to 300MHz)

- Nano VNA (SAA2) (3Ghz)

- (Both Nano VNAs need to be tested with a simple diode multiplier for signal generation above 3GHz, e.g. 3.4, 5.7 and 10GHz)

- Tiny SA, low cost spectrum analyzer up to 960MHz, base scan about 300MHz. (The bigger/better one is upstairs, anyway).

- frequency counter, up to 1.3GHz. Under warm-up-test. Looks like it needs a bit of burn-in to age the OCXO crystal a bit more. It has moved 5Hz @ 10MHz for now, and seems fairly temperature stable at normally changing room temperatures, < +/- 1Hz. If better stability is needed, a GPSDO can be adapted (10MHz out to 1MHz in as counter reference frequency).

- a GPSDO, 10MHz and a distributor/extra OCXO. More power supply needed.

- (I should set up some more power supply downstairs for testing. I have an older one (up to 1A w/fold-back at over-current/short circuit) that needs to be tested. Output voltage should be 10 - 15V. Claimed max current 2.5A, but looks like insufficient cooling for that.)

- (For more test equipment a 12V (not 13.8) supply with several Ampere output should be available. Must be linear to avoid noise/ripple on the output.) (Also usable for the microwave converters/transverters.) Some equipment does not like the 13.8V, and should have just a 12V supply voltage.

- for now, a few battery cases with 3x 18650 Li-Ion cells (9.5 - 12.5V), and a few Ni-MH rechargeable sets are available. (4.8V, 7.2V, 9.6V available, a casing for 10 cells (12V) can be made.)

- Charging for Ni-MH cells set up.

- (Charging for Li-Ion cells needs to be set up.)

- I think I will need a more advanced linear power supply for the downstairs lab. (There is one upstairs) I should go see if I can find something suitable. Something with both voltage and current limiting. A simple SMPS or DC/DC converter with those functions may be sufficient for quick experiments. I think I can find one in my pile of modules - 12V input buck/boost converter with 3 - 24 (30) V out, with digital meters and settings ;).

- A small set of components for experimentation.

If you can think of further improvements please consider leaving a comment.

I do have more test equipment upstairs, and that will come into use for more elaborate testing and building. It does need some (a lot of) tidying there, though.

10GHz HB 100 Test With Modulation.

 Today was an active day with the soldering iron.

As the very first test I just connected the HB100 modules, preset to 10375MHz and 10450MHz on the desk, just to see if I could hear the carrier with my LNB mounted outside the house, and pointing away from the house. This is the LNB I have used for beacon monitoring, so I know that it works.

A primitive modulator using the LM386 PCB module with gain control (from China) connected to the 5V power supply (yes, 7805) was tested, making sure that the peak voltage would not be too high for the transistor in the HB100 module. This initial test was done with a 150 ohm "DC-dummy-load" in place of the HB100, and the LM386 input was connected to the earphone/headphone connector of a transistor radio. The voltage swing was about 50mVpp, so well within the limit.

Time for an on-air test. I connected my QO-100 base radio for TX, without RF output, and used the CW side tone to modulate the system on 10450MHz. When the HB100 was in an optimal position the wideband FM modulated CW signal was loud and clear in the receiver. Not very strong, but with significant quieting and a clear CW. My callsign has been sent out on 10GHz.

What I need to do now is getting the system into a box with switches, a tone generator the modulator, likely an electret microphone and a bias tee for the LNB, plus connectors for getting signals and DC out to the LNB and HB100, and the IF signal in from the LNB, and I should have a working system.

Right now I do not have any stations to test with, but a local amateur has a HB100 module and LNB somewhere, so he might be available. Otherwise I may have to build another system, so I can get some tests done.

The transmission distance tested right now is about 8m, so there is plenty of room for improvements.

First HB100 Tests on 10GHz. Simple Low Cost Field Strength Meters.

 The first experiment with very simple equipment, using the low cost HB100 Doppler radar module has been completed: A simple field strength meter (FSM). The idea comes from F6HCC's website http://f6hcc.free.fr/10ghz.htm .

A non-functioning HB100 (with no output when supplied with 5V supplied) was modified according to F6HCC. Another as yet untested HB100 module was used as a source.

But the initial test was done by using a 2m/70cm hand held radio near the module's receive antenna. The "IF" output of the module uses a 10nF capacitor (non-critical, F6HCC uses 22nF) as decoupling for the RF to the meter. The meter used is a low cost digital multimeter (DMM) from the local DIY (home improvement) store. 

The reading with 10cm distance from the source (TX antenna) to the sensor (FSM RX antenna), and the reading is a modest 5mV, clearly seen when changing the distance.

Because most hams do have a DMM this provides for a low cost method of testing the functioning of a 10GHz transmitter and/or antenna. The HB100 module can still be had for less than $5, and the modification is simple and well described (with images) by F6HCC. Admittedly, the sensitivity of this meter is low, but it does work.

On the same page F6HCC also describes a more elaborate FSM for 10GHz, involving a diode inside a wave guide, and an amplifier for an analog meter.

However, for a much more sensitive FSM I would likely use a low cost (surplus) low noise block converter LNB, usually used with a satellite TV receiver dish. This has lots of gain, and you can probably salvage one from a discarded satellite dish. The signal on 10GHz is amplified and converted to a much lower frequency, usually around 250 - 750MHz, where a diode detector is much easier to make. Also a simple multimeter (analog or digital) can easily be used in this application. 

For measuring close to the lower band edge (about 10.0GHz) it may be a good idea to modify the LNB with an external DC supplied, not via the IF cable, but directly through a hole in the casing, and disconnect the DC "RF-choke" on the PCB from the IF connector.

Further, if there is a filter on the 11-12GHz side, it may be an advantage to by-pass this for greater bandwidth.

The gain in such an arrangement may be rather high - too high - but reducing gain in a system like this is easy. Just put some RF (microwave) lossy material between the LNB and the signal source...

There is, of course, the possibility of bypassing the RF (10GHZ) amplifier and the RF filter in the LNB, and connect the sensor antenna directly to the mixer via a piece of coax (semi-rigid coax will be the best choice), thereby reducing the gain.

If you do this, you might salvage some useful GaAsFETs, if you can avoid destroying them with static electricity.