Saturday, 21 February 2015

Dad! What are those Pixies on our dining room table?

Quite correct Helen! But not the type you were thinking about, that live with the Elves down the bottom of the garden?

I got involved in this one with G3XBM recently, when he had ordered a kit from a different dealer at a slightly higher price, I was determined to get the price down by using various search method tricks I have come to learn over the years.   
The kit(s) arrived from China, within 10 days of ordering. The PCB quality for the kit is excellent! Included is all the components, One Crystal (7.023MHz), and even the schematic and component list were supplied from this ebay dealer. All for a penny pinching £4.70p at the time I ordered!  * Note the price has increased slightly to £5.20p ($8 US) as I write this. This does ask the question though, is this the worlds cheapest transceiver kit including delivery?

I intend to do a couple of mods to my Pixie as I build it, I will replace the 47K preset with a 47K Variable Pot so I can have a couple of KHz Frequency swing or pull to hand when I mount it inside a case. I will also switch in a couple more Crystals 7.030 & 7.040MHz so I can have a bigger spread of the 7MHz CW Band to work with..

The Pixie transceiver has a lot of history too its name, and various mods and developments have been done over the years. If you want the full facts, grab a copy of The_Sprat_Pixie_File from the G-QRP Club Website.

Think about becoming a member of the G-QRP club while you are there, as this club is well worthy of joining! Without them the Chinese would not be producing these kits today!

For those who are after full circuit details etc for the Chinese version of this kit try this link here.

Saturday, 14 February 2015

Building the M328 component tester (2)

Calibration of the M328

I have previously explained how to go about building the M328 in Part (1)

Now this is where it all gets very interesting!

When I watched the VK3YE  video (A number of times may I add), I had always wondered about the 35pF offset and why Peter couldn't eliminate it? I had read the manual written by Karl-Heinz Kubbeler, upside down and inside out around the area of calibration (Section 3.2). I had felt maybe Peter had missed a step, or maybe two? But it would only be when I finally built the project I would be able to put my theory into action! After all I couldn't just pop over to Melbourne and explain, it is not a couple miles down the road from the UK is it, and doubtful some nut with a theory would be welcome asking for a phone number anyway??

Before you start calibration you will need to make yourself a three pronged shorting link, and a get hold of  a 100nF (0.1uF) Capacitor. The value can be greater, but not less than 100nF, which will be required for the final part of the Cal procedure:

You then insert the link into all three terminals (1,2,3) on the bottom terminal block, make sure the connection is tight. Press the On button this will put the unit into Self Test Mode:

When it gets to "T4 Isolate Probe" which I assume means Test 4? Remove the Link smartly! Do NOT touch any parts of the unit as it resumes its test and calibration cycle:

Let it move on, until it gets to the last screen which at this point you will require the 100nF Capacitor and Insert it between Terminals 1 & 3:

Leave the capacitor connected, you will see the value pulsate on the bottom line of the display as it is calibrating (Mine read 91nF). It will keep pulsating the value until the test concludes (it can take about 2 mins?). When the test finishes it will exit the calibration mode as indicated on the display, reverting to testing the Capacitor and indicating the value being tested as below:

(Sorry about the breadcrumbs on the bench, this is how one works when into resolving problems)

The unit then times out after about 30 secs and switches off.

Now if you have gone about this procedure correctly, when you next turn on the unit it will no longer display the 35pF offset as shown at the start of this blog,  the display will show the following after the initial Volts check and sign on etc:


You are now ready to start testing your components. Which we will come to in Part 3.


Shortly after I proved my theory about the elimination of the 35pF offset, I contacted VK3YE via email. Peter came straight back with a BIG thank you, after he too had tried my method out, and it had worked on his unit first time! The only problem he had found, he couldn't then test small value capacitors? Eg: 47 pF measured fine (48pF) but 22pF was not recognised? Hmmm!

I spent my teatime reading the manual again, and found it!:

Part of the introduction in the German manual states:
Chapter 1
10: One capacitor can be detected and measured. It is shown with symbol and value. The
value can be from 35pF (8MHz clock, 70pF @1MHz clock) to 40mF with a resolution of up to
1 pF (@8MHz clock].
So nothing below 35pF will work. Or will it?

Peter mentioned in the last part of his email:

Still you can get around this by making a small jig with (say) 100pF for use with small capacitors and just deduct 100pF. 


Friday, 13 February 2015

When certain stages are not viable to construct any longer!

Interlude time now, in between writing up the Calibration procedure for the M328.. We will push this little one in to keep you all happy.

I have a few various projects to build over the next couple of years, some that will require a regulated supply to feed various DC Volts  (low noise is a must!).  I had been running through my mind various ideas using regulators on some stripboard to mate up  the supply rails,  I had a few emails between my friend G0FTD  on the subject, he said take a look at these "5" seperate LM317 boards already made up on a PCB, complete with heatsinks for just over £5 ($8 US), which works out at just over £1 each! I couldn't even buy the heatsink for a £1 over here, never mind the LM317 and the PCB.  I ordered five they arrived within 10 days from HK, a couple of  the LM317's were not bolted up tight to the heatsinks, but a quick tighten with a screwdriver and all was ok, what a bargain!

Time to test one.
Most of what I will be requiring is between 3.3V and  5V with a 12V dryfit battery on the input end.

To prove I just rigged up a quick lash up with 15V feeding one module from my bench supply:

I was able to able to obtain a precision swing of between 1.25V & 13.5V as the the output is made variable with a multiturn pot attached to the PCB, ideal for what I am looking for. 

Conclusion: Another good purchase from China that is useful, money and time saving in project building!

Not only have I come across various types of regulator modules (some switching types so be careful!), but also LM386 Audio Amp modules, Timer modules and even a 3 Axis Magnetometer PCB built and ready to go!


Wednesday, 11 February 2015

Building the M328 component tester (1)

Having being inspired by one of Peter VK3YE recent video's, it was time to purchase a couple of kits and dust the soldering iron off the shelf and get down to building one:

This could prove to be a very useful piece of component test kit, for the constructor and repair bench, including the novice. It does, R, L, C  including ESR, Diodes and Transistors, giving the pin configuration detail of the Semiconductor device undertest as well as the useful gain figures. 

It is basically a copy of the Karl-Heinz Kubbeler design, centred around a programmable ATMEGA328 microcontroller which is very well documented. The kit took about 2 weeks to arrive from China, which contains a well made PCB, display module and all the components, including the blown main chip all for around £8.00 ($12 US). It comes with no instructions on how to put it together, but a Chinese manual is downloadable from the purchase site, which I will make a link available at the bottom of this Blog. Infact, really you don't need the manual for construction, as the component values are printed on the PCB, it just helps a little to get one or two things installed the right way around like the switch, and the circuit diagram can be useful for component reference and maybe fault finding later? Of course the Chinese manual is written in Chinglish, we are refered to welding not soldering! I don't think my old arc welder would prove very suitable for this project somehow? Hi!

I emptied all the components out of their anti-static packet into an empty biscuit tin, so I didn't lose any of them. I was away, it took around 2 hours of soldering, and sorting out the correct values, a DMM can help with the resistors values, as I found an orange band can look like a red, so its best to measure them to avoid confusion and getting one soldered in the wrong position. Transistors are marked to board values, and the marking of the outline makes sure you cannot put the devices in the wrong way around.

At this stage it is time to check the board over for shorts, man made solder links etc, and snip off component ends. All looked ok, time to connect the 9V battery, before inserting the main ATMEGA IC, at this point a DMM is required to check for regulated 5V at the IC socket pins 7 & 22, all confirmed correct and the regulator was doing its job!  
All in all it was quite a relaxing project to put together, I didn't find anything too difficult, although a bit of care is needed aligning up the pins of the main chip before pushing firmly home into its socket.


There is not much work to do with the display board as most of this is already constructed, just a strip of header pins that carefully require soldering in across the top of its PCB. This then mates up with the socket strip on the main board when it is pushed home and bolted together. 

All looked good time to switch on! If it fires up correctly one push of the On button should turn on the display. In my case it did, and didn't? When I released the Push To Make switch, it went out ? Some folk have had problems putting the switch in the wrong way around, I knew I hadn't done this and a quick check confirmed the switch had been inserted correctly, time to investigate further? The clue was the LED under the display board wasn't on, a quick check with the DMM around the circuit in this area confirmed my thoughts, I had put the LED in the wrong way, huh! Oh dear! I had to pull it all apart, split the two boards desolder and turn the LED around, and then put it all back together.

Great it then fired up correctly, and held in its On state after pushing and releasing the button, a quick adjustment of the contrast pot to get the display correct and all was looking well, time to calibrate..

To be continued in part 2.


This is where I purchased the kit from, although they are available from ebay too:  

Construction manual:

Design manual helps with Calibration and other stuff:

Newer version of manual: