*return visitors can jump direct to the latest blog update:: [Jump]

2nd Dec 2007 - Introduction::

Hi!! Welcome to the g7nbp blog area on the softrockradio.org site.

My initial purpose for blogging here is to describe the build process of my recently acquired Version 6.2 RXTX kit for 20/30m - but who knows - these things have a way of finding their own direction!

First a bit of background:-

Blogging::

Ive been an intermittant blogger for quite a number of years now, in fact since before blogging was called blogging - but thats another story. Lets just say Ive been around on the internet for a loooong time! Hopefully visitors will find my blog useful and occasionally entertaining. (I try my best anyway!)

SDR::

Out of curiosity Ive previously built a couple of minimal SDR projects previously (Zeta SDR - http://www.qrz.lt/ly1gp/SDR/index.html - and another hybrid minimal SDR project based on several published designs.) At the other end of the SDR spectrum (figuratively and literally) Ive also been fortunate enough to have had some input on a current research project utilising ultrafast ADC units sampling DC to 1GHz with direct number crunching using SciLab http://www.scilab.org/ (Opensource replacement of MatLab). Ive also had some recent experience with the GNU radio project - http://gnuradio.org/trac - again on some fairly high end kit for specialist research. Im not however an expert (plenty of real experts on this site - ask them not me!) - and certainly when it comes to the maths side Im very much a the bottom end of the learning curve.

Somewhere between these two extremes lies the softrock series of designs and this blog.

Im building the v6.2 RXTX kit for 20/30 - partly as a personal first look at the TX side of SDR, but also because I will shortly be co-presenting a talk at our local radio club (Salop Amateur Radio Society - http://www.salop-ars.org.uk ) on simple SDR and DSP technology. The softrock v6.2 RXTX and Lite kits look to be a good starting points for quite a few of the club members so thats where my current experimentation is going to be. Im going to try as far as possible to pitch the build notes and experiences with the relative newcomer in mind to transfer to my talk later in the year.

Hopefully the info will be of use to others following the same path too. Im adding some photos as I go along - these have been reduced to around 640 pixel width and compressed quite significantly both for fast download for those not on ADSL, and also as a courtesy to the site owners - bandwidth costs money! If anyone needs a hi-res version of anything posted, drop me a message and I will mail you one.

73's - Chris W - G7NBP

Ok, time to come clean - the kit arrived a week back now, and this section of the blog is pretty much setting the scene and playing catch up with where I am now. This section may well be a bit more "wordy" than ones later on in the build. Apologies if you are dozing off at this point!

The kit:

The kit arrived in a somewhat smaller than expected package - ie it actually fitted through my letterbox! It was well padded and despite the best efforts of our wonderful UK postal service, arrived intact. The day it arrived I was tied up with other work on the bench and apart from a cursory look into the bags I didnt have chance to take much notice of the project until the following morning.

When I did get round to finally having a proper look I noted that the boards were well produced, components sorted by type into the usual gripfast poly bags etc. The next thing that was obvious was that there wasnt the usual wad of folded A4 instructions with the kit. There was however a CD. Having quickly spun up the CD it was clear that it contained construction notes in pdf format and some software.

I noted however, that the constructional notes and schematics etc did not include the 20/30m version of the kit that I was building. There was however a link to the yahoo group site ( http://groups.yahoo.com/group/softrock40/ - the main files section being at http://groups.yahoo.com/group/softrock40/files/RXTXv6.2%20files/ - subscription required ) where the latest version of the files live. I must admit I had already joined this group in anticipation of the build.

A quick browse to the files area highlighted a second issue - did I have the August or the November version of the kit??? - there was nothing apparent on the kit to tell me that. Common sense suggested that I probably had the August one - as my kit was purchased via a UK retailer (Waters and Stanton) who would probably have had the kits in stock for a little while. A quick look at the two versions of the BOM (bill of materials) showed that an easy way of telling was to count the number of 4.99K resistors - I had 6 confirming it was the pre-october 30th version. (Im sure there are other differences, but was a quick easy way of telling).

I downloaded and printed off all the associated pdfs and jpegs for the correct version of the kit. I also had a quick scan through the messages section of the yahoo group to see if I could dodge any potential pitfalls. There are quite a few thousand posts on the group as of this moment in time, so its a little difficult to see the wood for the trees. Sensible search clues based on the band of choice and words like problem, issue, fail etc brought back a smaller subset of messages which I read with interest.

The construction notes, BOM and schematics are well produced, though those constructors used to the heathkit / elecraft style of notes are going to find them perhaps a little minimal. This is almost certainly an unfair criticism however, as I doubt most constructors will have too many problems following the instructions given. If I was being super critical I would say that a printed component overlay both total and in smaller closeup sections matching the particular part of the board being populated would help loads. The silk screening on the boards is excellent, but as i will come to in a minute - on the daughter board it is perhaps a little difficult to see the correct orientation of the resistors in hairpin layout. builders need to look closely at the pcb and I advise looking closely at the images of the populated boards on the groups site, both in the files area and in the general pictures section - it helps a lot - especially later on when placing some of the SMDs on the main board.

Its perhaps a good time to mention SMDs. Yes this project uses SMDs, and SMDs are small. If you have never worked with SMDs before, they can be a bit offputting until you know how to wrangle them. If anyone reading this blog is still unsure if they want to tackle the project - because of the SMDs my advice is - YES - give it a go. They are not as difficult to work with as they look. There are plenty of good tutorials on the net for first time SMD constructors - google finds loads.

I consider myself lucky in that despite having passed the "Big Four Oh" birthday and spending most of my working life "close working" with either electronics or on a PC screen, I still have better than 20:20 eye sight. The components in this project are inescapeably small. 1/8th Watt resistors are not toooooo bad in good light, but I tend to sort those with a meter (more on that in a min). Some of the caps have markings on the limit of my vision. The SMDs are either marked with very small (and often incomprehensible) digits - or have no markings at all. Despite good lighting here, and eyes youthful beyond what I deserve, I would suggest some form of optical enhancement is needed. Im lucky in that I have a low power binocular microscope, which is an amazing tool for these jobs. However, for quickly looking at the components at checking time, and also as an aid while soldering the best tool Ive found for the job is actually a pair of those cheap reading glasses you find at supermarkets etc. A pair of those with a +5 dioptre make an ideal lightweight close working magnifier comparable to those head worn units, but without the weight and the price tag (I paid a pound for mine!).

Anyway - back to the kit. As is usual with pretty much all projects of anything past minimal complexity, it is broken down into stages. The first of which in the 20/30 project is building the daughter board oscillator which runs the XTALs in overtone mode to get the required frequencies. This is a nice introduction to the project as it is self-contained, easy to test features some minimal SMD work as a warm up for the main board population.

A useful tip I learned a long time back is to pre-sort the components you need for the forthcoming section of a build into compartment trays. I have a multi-compartment case with a close fitting lid I bought some time back now which is ideal. This allows you to check that you have everything you need, limits the chance of confusing component values and actually speeds up construction. It also means that if I have to stop unexpectedly I can just close the lid and nothing will get lost or disturbed.

I must admit that while sorting the components for the daughter board I was a little lazy and quickly sorted the Resistors into piles by final colour band, then used my multi-meter to read off the values rather than reading each band. Having spent a few mins working through them I started seeing values that were not on the BOM. In fact I spotted three values of R that were on the face of it wrong. In point of fact they were all one order of magnitude out. My first instinct was a simple packing error - or that perhaps the values might ave been changed as an update/correction to the BOM. I dashed off a quick post to the yahoo group to this effect right away. However, it was at this point I popped on the +5 glasses and had a closer look at the bands. All showed correct color code for the values on the BOM. Strange.... I then noted that the low batt indicator was flashing on my multi-meter... A quick change of batts and a re-test suddenly shows all values correct. Seemingly the decimal point doesnt get moved correctly when the batts are low on my auto ranging meter. Doh!! The moral of this is - double check before you post. (I did apologise to Tony!)

Once I had the required components sorted the daughter board build itself was pretty uneventful. SMDs on the back of the board first. The construction guide suggests to tack these down at one side first, then to correctly solder both sides and tidy up with some solder braid. Personally I use a few different techniques when it comes to SMD work - depending on the design. one of my favourite ways of getting a neat job is to use the end of a toothpick to put a tiny dab of superglue (crazy glue for those of you on the other side of the pond) between the pads and to offer up the SMD into the correct spot. Only a tiny amount is needed. This holds the SMD in place while you solder it. I usually lay out thw whole board this way before soldering. If a component is not in the correct place, a gentle sideways tap with a flat bladed screwdriver is usually enough to dislodge without incident. Its much easier to correct a component placement before soldering than after soldering.

This technique offers a number of soldering options. for small numbers of components I tend to use a 20 watt iron with needle tip and fine LMP solder, but for greater numbers of joints I tend to use solder paste, applied via its syringe to the joints and then use a small gas torch waved over the board to make the joints. I have seen people use the glue technique, but rather than a gas torch lay the PCB out on a hot plate or in a mini oven to "cook" the joints.

The remaining components are conventional thru hole devices placed on the other side of the board. These are easy enough to place and solder - though some care needs to be taken to correctly identify the orientation of the resistors - refer to the overlay, schematic and the photos to ensure these are correct. The circle denotes the body of the R, with a little tail pointing towards the hole to the other leg.

Solder pads are small and some care is needed to avoid solder bridges. Be careful with the plated through holes. component removal may damage a plate thru, requiring careful top and bottom soldering to fix. A good quality pair of side cutters is a must for kits like this too.

Some care is needed when winding the RFC onto the binocular core. As described in the construction guide, if you are not careful the enamel gets scraped from the wire during the winding process - giving rise to potential short circuited turns. The other coil on the board L100 is a simple wind. The only bit which is a tad tricky is the XTAL placement as care needs to be exercised to stop the bottom of the case shorting onto one of the thru-hole vias. The xtals also need an earthing wire soldering to the case which is then fed through the board and soldered to ground. This is a bit tricky and use of minimal heat is important.

Total build time on the daughter board was a shade over an hour and a half, though this was at a gentle pace with breaks for coffee/phone/the cat. The completed board looks something like this

Testing:: After a quick visual inspection, and quickly checking there was not a low resistance DC path between the +v line and ground I was happy to connect up the daughter board to a current limited 12V supply and to check for an output. Initially results were very promising. A nice steady oscillation at 40.5 MHz (actually 40.497 by my counter)

The output on my scope was solid

and a quick peek on my home made spectrum analyser looked, as expected very clean

So far so good....

I then switched the jumper over to the other XTAL, fully expecting to see something very similar at 56.19MHz.... Nothing...

I had already seen some comments in the yahoo group regarding some people finding problems with the 18.73MHz XTAL not wanting to oscillate, so after a couple of quick checks to see that nothing had come adrift I carefully lifted the XTAL off the board and popped it into my XTAL tester. At its fundamental frequency the rock was showing pretty much normal activity, but in overtone mode I couldnt get it to budge, either on the board or in my tester. As per other people comments on the group I did quickly try a few minor mods to the number of turns on L100, but nothing seemed to bring it to life. Reluctantly I concluded that my XTAL was also one of those found to be problematic. Tony has kindly offered a replacement for these for as long as his stocks last.

For the time being my tests are limited to 40.5Mhz until I am able to swap the XTAL.

I also note my output is somewhat lower than expected. Testing into the 4.99K load Im seeing slightly under 1V peak to peak (1.5-3.5v expected). My shack scope is only a 40MHz one, so this might be a little bit of roll off at its upper freq, though Im sceptical of this. A double check of the board and circuit tracing doesnt show anything amiss. I will test further when I can get to my more usual 100MHz scope on monday.

Update: Scope rated to 40Mhz... what A shame Im using my old probe which isn't! Just checked again with my newish scope probe (100Mhz rated) Now if anything Im seeing a little overvolts at about 4v P to P. Much more like it. :)

Thats about all I have to report at this stage. As they say - watch this space!

Also causing me grief at the moment is getting my soundcards working correctly. Ive decided to build a new PC for this project (Im actually planning on building the TRX into a diecast box inside the PC itself for demo purposes).

Currently I have a reasonable 1.8GHz P4 I picked up for the job. Onboard it has an intel based soundcard which Ive turned off in the bios. (its not bad but I believe is only a 12bit card which isnt really good enough for SDR work). Ive added two internal Soundblaster Live 24 cards, which while cheap and cheerful do seem to be a workable option.

The problem Im having is getting the two of them to work in Windows 2000. A single card works fine, and in fact the two cards work fine under Linux (I have it set up for dual boot as I intend to do some development work in linux for SDR shortly). However two cards in win2k is a nono... They are both spotted OK in safe mode, and there are no conflicts etc. There are no other PCI devices in the PC. Booting in standard mode gets me a bluescreen every time.

Ive posted a question about this on the soundblaster forum, but so far not seen any replies. I realise that having two identical cards in a machine is probably a little confusing, but I know that this option does work as sat on the other side of the room from me at the moment is a "non-linear editing suite" which has two live24 audio cards and runs on win2k. Its a bit hard to get to the windows options on it as it has an OEM setup, but as far as I can tell has default drivers.

I plan to do a bit more searching in a bit, as this must be a common problem. I will report back if I find anything. Any suggestions are of course welcome!

Well... Just time for a quick post while Im sat at the PC for something else. Replacement 18.730 XTAL arrived friday morning - Thanks Tony - thats brilliant service!!. Today (sunday) however has been first chance to get the iron warm and swap out the other one.

The change over today was very straight forward - just a case of dropping the XTAL in and soldering up the leads. Getting it out previously for test however is perhaps worthy of a bit of comment. Both the XTAL cases are earthed to 0V via a short wire from the side of the can to a single hole on the pcb. It right in the middle of the board and surrounded by components topside. When you come to desolder it, work on the assumption you will have to unsolder both xtal grounding leads before you can desolder and lift the XTAL to be replaced. It makes life much easier!

As mentioned previously the pcb makes generous use of plated through holes. Removing the XTAL carefully is vital as if you tear out a plate thru via when removing the xtal it will become VERY difficult to make the required connection as the tracks for the rocks are topside.

Obviously no one designs a kit with the expectation of having it re-worked too often, and frequently the fault finding and replacement process ends up making a real mess of a project A less dense pcb layout makes such repairs easier - especially for those not used to re-working round faults, but then you dont end up with a nice compact project. In this case however all was well - which speaks volumes for the quality of pcb fabrication in this kit. I now see the expected 56.19MHz signal. Output is about 0.5v lower than 40.5MHz one, but well within tolerance.

Ive not had chance to start on the main board yet as things have been a bit hectic here over the last week. Possibly later tonight I may make a start if time permits.

As for my quest with the using two live 24 soundcards.... Ive had an email back from creative labs with the stock answer saying not to use two cards at all. I mailed back the helpdesk guy and pointed out that I have already seen it done with multiple cards and need more tech info to work out if this is a driver or config problem. In the reply to that he conceded that yes it can be done, but that the tech support desk only provides typical level support, not detailed driver level support. He did however give me the address for the creative forums (which I already have a post on, and more usefully the developers area as used by software designers etc. I will be dropping a post to the developers forum a bit later, so hopefully we will see some progress on this soon.

Not had much spare time this afternoon, but have managed to add the SMD caps to the PCB. As previously mentioned I have a few ways of working with SMDs depending on the board layout etc. Today I tacked and hand soldered as per the instructions. Result is not as neat as working with solder paste and hot air, but is more than acceptable and for the novice constructor should not cause too many problems. This stage build time: 45 mins again - no rush with plenty of slurps of coffee. Cant do any more now, my turn to cook evening meal now so off to make a mess in the kitchen... :)

10th Dec 2007 - [remaining caps added]

Managed to get back to the bench for a couple of hours just before midnight, so 01:14 GMT as I add this update - hence the less than brilliant photo! Wire links, the two small 1/8th watt resistors, ceramics and electrolytics now added. Only issues encountered have been getting the two resistors to sit tidy on the pcb when the hole spacing is too small for them. The temptation is to mount them upright, as they would have been had the daughter board not been there, but having checked the clearance it really is too small - the instructions say lay flat for a good reason! Newcomers usually find some capacitor markings a bit confusing - the same is true of us old hands! A digital capacitance meter make life much easier. Most of the ceramics are on card strips still - easy thing to do is carefully pull one off the strip - measure it, return it to the strip and write the value onto the card. If no capacitance meter available it is easy enough to work out which are which by a process of elimination. All the electrolytics are the same value which makes life easy, as does the fact that the PCB has the plus side of the cap marked rather than the negative - why is this easier... well, although the caps have the usual negative marker stripe, the plus wire is the longest, so you just pop the longer leg into the hole where the plus sign is :) Too late now to even think about adding the resistors... more to follow :)

13th Dec 2007 - [SMD soldering...]

No more construction to report at the moment, however - Ive just had quite a lengthy chat on the phone with someone else who is just about to embark on a v6.2 RXTX kit. They were asking me about the differing ways of SMD soldering and which is the best to try first.

I have to admit that Im in two minds about this, I generally use solder paste and hot air and feel that even for a newcomer this is actually a technique that produces very tidy results. However, the tack and tidy technique as listed in the construction notes works well for most people, though you do need a steady hand and a bit of patience, especially if your previous soldering exploits have largely consisted of heating up an iron poker in a furnace and melting roofing lead ;)

I quickly rounded up a few vids off youtube which illustrate the various ways of soldering the SMDs for them, and have added them to the blog below too. Hopefully they will help those who are trying to decide which method they want to try.

acknowledgements to the respective owners of these videos - if linking not considered appropriate please contact me to have them removed

Results are very good - especially when combined with a warming plate to avoid there being too much thermal shock. This is my preferred way of working with SMDs.

A good source on info about the hotplate and hot air method can be found here http://www.zianet.com/erg/SMT_Soldering.html

(I have no connection with them by the way - just a useful resource I found while googling)

Cecil also has a page up showing his tests using a simple oven for SMD work - http://www.hpsdr.com/Public/Projects/SMT/SMT.html - which shows very good results.

The main advice I would give is to have a stab at each of these methods with some scrap SMDs (melt a few off an old board) and practice soldering them between the tracks of some stripboard before tackling the softrock project.

The bottom line is - Its really not that difficult :)

One word of advice specific to the softrock project though... the construction notes have you add the SMD caps first, then other components and finally the ICs and other active devices. The reason for this is to avoid problems with static electricity - a few of the devices are prone to damage if you dont have reasonable ESD precautions in place. If they are soldered on last, they are protected from a lot of potential damage during handling of the board when not all the susceptable inputs may be grounded or protected.

However, if you plan to use the solder paste methods, especially with a plate warmer and hot air you will want all the SMDs to go on at once. If you do - be sure to observe good ESD precautions throughout the build - or risk finding at the end you have a brick :(

Hope thats some help in deciding exactly how to cover the SMD aspects of the build.

16th Dec 2007 - [Resistors]

Just a quick blog update - I had a free hour this afternoon so fitted the resistors to the main board.

Little of critical importance to comment on here, other than re-stating my previous comment about checking resistor orientation on the board. Circle represents the body of the resistor - the little tail represents the remaining wire leg and points to the hole needed. Most of the resisitors are placed vertical hairpin style, some however are placed horizontally to fit under heatsinks etc.

It is important that the hairpin layout resistors do go in as marked as the tops of some of the resistors are used as test points and if the resistor is inserted with the body and wire leg reveresed although the circuit will work fine, test voltages will show as wrong.

Its good layout practice to standardise on orientation of the resistors - if vertical so that you read the colour code from the uppermost band downwards, and if horizontal so that you always read value markings from the top left. This makes checking for incorrect values etc much easier.

Of course if you are careful and have pre-sorted the components into trays first there should never be a need to check components for wrong placement - but its a good habit to get into.

And of course reading the instruction through before starting is the key to avoiding making mistakes :)

Total build time to add all the resistors was a shade under an hour - as ever with intrerruptions by the cat and a break for coffee.

I may get time to add some more components later, but its a busy evening.

No photos at this stage, not much new to see.

23rd Dec 2007 - [Semis #1]

Well... I managed to find a few minutes away from the pre-festive family madness and furious bouts of present wrapping, so Ive added the diodes and the first batch of transistors - up as far as the voltage reg.

Instructions are clear and easy to follow, no problems as far as component ID and placement are concerned. Silk screening makes both diode and transistor orientation very simple - but it pays to double check!

One thing I would comment on here is that the pads for the transistors are very small. If you have not had much experience with this density of work you may find this bit of the project challenging - more so than the SMDs. If you can find an old board (PC or similar) with pads of a similar size a few soldering and desoldering sessions might make you feel more confident. Its probably a good idea to check what tips your iron can take and downsize to suit.

Another thing to note is that you probably want to use a higher wattage iron than youd perhaps expect. Im using a temp controlled iron which holds the tip temp at 410 degrees C. Its actually about a 25watt iron, but is designed to keep the tip temp at the required point by varying the actual wattage over a 12-25 watt range.

Even with this Im finding that the large earth plane needed for RF circs of this type tends to suck heat away from the joints. You can always tell which pad is the 0v one while soldering!

So far no dry joints and no cooked devices, but now that we are moving on to the more delicate active components Im probably going to use my other higher wattage variable temp iron, as despite the higher wattage - localised component heating is actually lower due to the faster heat transfer to the joints.

Once you get as far as adding the 78L05 reg you can do a couple of DC resistance tests to see if there have been any constructional errors or soldering issues up to this point. Checks are simply a case of using a DMM to see what the resistance between 0v and the +12v and +5v points are. I found around 4Meg ohms between the +12v and ground (substantially greater than the 100K minimum) and around about 820 ohms on the +5v rail, slightly above the expected 800 ohms.

The remaining FETs, and other transistors have yet to be added, and these do not go on all at once as some of the placements will make adding the inductors difficult.

Im hoping to get some more time late tonight and over the rest of the holidays - so check back for updates fequently.

26th Dec 2007 - [Semis #2 + Coils #1]

Once again I managed to find an hour or so at the bench. As the title suggests, todays activities have centred on adding the remaining semis and making a start on the coils.

About half of the remaing FETs are added conventionally - ie upright. However the remaining three are added a little unconventionally, lying on their backs. Read the instructions closely before fitting them.

The last three fets are fitted to the board bent over horizontally, so that TO220 heatsink can be screwed down on top of them with their flat sides against the bottom of the heatsink to help dissapate heat from the power stages Its not very clear in this image, but if you look at the board layout all will make sense. The important bit is to make sure that the transistors are as close to the bolt hole as is possible - to ensure they sit completely under the heatsink The instructions suggest tacking the FETs in place first, then clamping down the heatsink, before finally soldering the leads and triming. However, my tip would be, push the bolt through the bolt hole first, then butt the FETs up against the bolt and tack a leg topside. Then offer up the heatsink, clamp down and then solder the reverse side of the board.

Also in the above image is T1 - (the toroid). T1 and T2 are added before the last of the transistors, (Q2 - the 2n2222a) as once this is fitted with its heatsink it would be difficult to insert T2.

Again, I would suggest that those who have cut their kit building teeth on offerings from Heathkit or Elecraft are going to find the instructions a bit minimal when it comes to the coil winding. Experienced constructors will have no problem following the details, but the newcomer will probably scratch their head for a while over this section. If I had to be honest I would say that this section is probably not covered as well in the instructions as it could or perhaps should be. Ive no doubt that my phone is going to ring in the next couple of days for a request to help the locals who are also working their way through this project a few steps behind me.

Specific comments would be that the diagrams in the construction notes are clear enough as far as which winding goes in which hole on the PCB, but there are no diagrams for the newcomer explaining how the transformers should be wound. Relative newcomers will probably not recognise the word bifilar and may struggle with exactly what it means in this context. Also that the required number of turns for each transformer is on the circuit diagram and not in the constrction notes. This means that you have to hunt around multiple sources for the info needed to wind each transformer.

Tony - Sorry if the above sounds a little negative, but I tend to say things the way I see them - and like to give what I hope is constructive feedback where I can. This is one of those occasions.

The image on the left comes from an elecraft manual, and as the saying goes is certainly worth a thousand words. Im going to go out on a limb here and say that this aspect of the instructions really is the only part thats letting down what is an otherwise brilliant project from the newcomer perspective. If anyone reading this has the required draughting skills and has some time to spare to run up some supplimental diagrams like this for this part of the kit Im sure they would be well received by all.

Although the transformers have been designed with a good degree of lattitude in construction (ie you wont notice too much degradation if you get an extra turn here and there) If you get the pairs in the transformers into the wrong holes on the PCB you are going to have at best a really deaf radio, and at worst a non-worker. Fault finding the tranformers once on the board is not a trivial task so you really do need to get them right. There has been quite a bit of traffic on the groups site about the coils and transformers so thats a good place to start. However I also note that quite a few of the recently discussed problems have resulted from incorrect coil winding / insertion so people are still getting this part wrong.

Read and re-read the instructions and the info on the group site before starting this phase if you have any doubts about it.

A couple of tips spring to mind here... Firstly if you can lay your hands on some extra enamelled copper wire of the correct diameter in different colours it makes identifying multiple windings on the same core much easier. The second tip if different coloured wire is not an option is to lable the windings with a scrap of paper and a bit of tape In this example the secondary has one amount of turns, and the two bifilar primaries have an equal number. Its important to make sure when you solder the transformer that you know which is the secondary - hence its markings. The two primaries are in essence the same so its safe to identify those just using a meter to ensure you have the correct pair of ends and that they are in the correct phase.

T2 is another transformer wound onto a binocular core. It is important that its wound onto the correct type of binocular core - there are two types! It needs to be one of the cores from the bag labled BN-61.

As with the binocular core on the daughter board, the construction notes suggest its a good idea to gently spin a small drill bit in each of the holes. This really is important as when I looked at my core there was a really sharp burr which would definitely have taken the enamel off and resulted in some short circuited turns.

The lengths quoted on the circuit daigram for each winding are generous - dont worry about coming up short while winding.

As I commented above - if you can find some alternate coloured wire it makes identifying the pairs so much easier.

My final comment would be to read and re-read the instructions well. The information there is more than sufficient to correctly orientate T2 on the board. Comments above re a few winding diags however do hold true. If you are an "newbie" and are struggling with this part of the build, my advice would be shout out on the yahoo group - someone will point you in the direction of some help and possibly someone local who can act as an "elmer" for you.

Incidentally - somewaht off topic - but valid :- we in the UK dont really have a word equiv of "elmer"... Im not sure of the origins of it stateside, perhaps someone will enlighten me... Anyway... the reason we dont seem to recognise the term "elmer" much in the UK is that we just dont seem to have many "elmers". Shame on us Brits - more people need to be there to help the new guys here in the UK. I owe a lot to the local amateurs who helped when I was starting out getting on for 20 years ago and I guess we all do. A gentle reminder to those who do have the skills but dont support the newbies - Guys, its time to give something back! (::End of soapbox rant!)

Ok, back to the build. Once you have T1 and T2 on the board its possible to add Q2 - the 2n2222a. It has a push on heatsink which needs to be carefully fitted before soldering to the board. Care is needed once this is on to correctly identify lead orientation - be careful of where the case tab goes. To act as a spacer there is a nylon washer - yes the one you coulnt figure out where it went when you put the mounting bolts onto the board.

Yes - thats why its a good idea to read right through the instructions a couple of times before you start. I did, and still missed out that little nugget of info on both occasions. I did however have to foresite to save what seemsed on the face of it an extra washer in the multi-compartment tray I use for the project and not to add it to the collection of useful hardware tubs.

Ok, not much more to add at this time. I hope to finish off the coils over the next day or so as time allows and then I can get on to adding the SMD ICs. I hope to grab some video footage of that. Watch this space as they say....