Surface Mount Soldering – Tools

This set of tools is (in my opinion) some great tools for SMT soldering without going the reflow route (which requires an oven).This shopping list costs approx. £180+VAT and is complete.

Tools required
A temperature controlled soldering iron (approx 50W), e.g. Antex TCS approx. £45+VAT
A small tip for the iron £2.65+VAT
Thin (approx. 26SWG) solder,and preferably _not_ unleaded. £28.65+VAT
An extremely good pair of tweezers, e.g. VOMM PSF SA ESD or (not tried) these ones £12
Rosin flux dispensing pen e.g. Chemtronics CW8200 £8
Desoldering braid, aka Solder Wick 1.5mm e.g. Chemtronics 60-2-5 size#2. Buy several, each spool is 1.5m £2.87
A cheap lens preferably with lamp, e.g. “UltraOptix 7x Aspheric LED Lighted” (search eBay) £9
Isopropyl Alcohol (IPA) Just need a few hundred ml to last a long time. 1 liter is £12.56
Antistatic brush £17.74+VAT
Scalpel handle Swann Morton No. 3 £2.50
Scalpel blades Swann Morton 0203 No. 11 (pack of 100 will last forever) £8+VAT
For non-surface mount, add these:
A very good pair of full flush cutters, e.g. CK 3780HEF or (not tried) Lindstrom 7191 £36+VAT
A very cheap pair of cutters (for when you don’t want to damage the expensive ones) e.g. from Maplin, or this  (not worth spending so much though – maybe just £5). £9.24+VAT

Total approx £180 + VAT

Posted in general electronics | Leave a comment

Long-tailed-pair oscillator

I needed a quick alarm tone generator for a piezo speaker, and with no 555 at hand, this turned out to work quite well, and it’s just as small. It operates with just a few mA current. It is a 1950’s Baxendall circuit from this paper.

A toroid from Amidon (FT-37-43) was used and 40swg enemelled (about 36AWG). Anything thicker won’t allow the turns to fit. There are 100 turns on the collector coil (ends painted green in the photo below).
18 turns of feedback coil (connected to transistor base) was wound on top, spaced out. The circuit consumed about 6.5mA for me (change R2 for a louder sound if needed, but it was loud enough). R1 is not a critical adjustment, it could be replaced with a couple of resistors if needed. It can run from a lower voltage with no modifications. At 1.5V, it was still emitting sound (lower volume), with 0.7mA current consumption.

C1 is to some extent dependent on the piezo capacitance, but you’re virtually guaranteed to hear a tone at some frequency in the correct ballpark, with the capacitor mentioned and the same toroid and windings.
The coil was partially wound, then the winding was held in place with some impact adhesive, then the rest of the coil was wound and secured with adhesive again. The ends were marked for this coil (just used some acrylic modelers paint). The 18-turn feedback coil was just wound on top, and was left unmarked. The values of 100 turns and 18 turns were just guesses – to have enough inductance on the collector winding (it will be around 3.5mH with this coil incidentally), and to have enough coupling and feedback with the second winding. With the inductance set, C1 was then picked for oscillations at a few kHz. The trace across the piezo is shown here.


Posted in analog, audio | Leave a comment

EAGLE CAD Toroid drawing tool

This small bit of software can auto-generate a toroid outline for EAGLE. You can select horizontal or vertical mounting, and specify the toroid dimensions and the number of wire turns. It will generate an EAGLE script file (.scr) that can be imported into EAGLE. Incidentally, another use for the program is to be able to visually see what the toroid and winding will look like, before spending money on the wire. You can visually see the effects of different numbers of turns and different wire thicknesses.

Here are some examples that were auto-generated:


The first example above is a T50 sized core with 7 turns spaced 1mm from each other at their closest point. Since there was still lots of empty toroid space left over, the second example shows the same toroid but with the 7 turns spaced 1.8mm from each other at their closest point (the closest point is at the inner diameter of the toroid). The last example shows the same toroid with 7 turns (spaced 1.8mm apart) mounted vertically.

As a possible future improvement, the program should create a few holes for allowing to tie the toroid to the PCB for some mechanical strength. It also is only designed for one winding (i.e. inductors), so it would need to be adapted for transformer use. These should be easy to do (the source code is below).

Windows executable:

Source code:

The values that were entered for the second example shown above: toroid_example_run1.txt

Posted in eagle, mixer, radio | Leave a comment

EAGLE CAD layers reference

This document eagle-cad-layers-reference3.pdf explains what all the layers in EAGLE mean when it comes to PCB component and board design and how the PCB manufacturer will interpret it, with real examples. The EAGLE documentation is awful for this. Personally I think my explanation is a lot better.


Posted in eagle | Leave a comment


After getting fed up manually creating device packages for EAGLE, the following tool was created. It allows for the automatic creation of many SMD packages. Here are some examples:

p1.jpg    p2.jpg   p3.jpg

The tool just needs some values entered, which can be obtained from component datasheets or this very good website.

This document eagle-cad-smd-drawing-tool.pdf contains all instructions on how to use the tool. The Windows executable is available here: (no installer, just place the file at a convenient place and then launch from the Windows command prompt. The PDF document contains all details).

Posted in eagle | Leave a comment

Toroid (Amidon) calculations

The excellent website here and  here  allows you to figure out what core and number of turns for a given inductance, for ferrite and iron powder toroids respectively, and also to see the impact of adding or removing turns. It is really great.

I decided to create a table in Excel of the various ferrite and iron powder toroids, so that it is possible to type a desired inductance, and see the number of turns for all toroids at the same time, or to type in a number  of turns and see the inductance for all toroids. Actually not all toroids (I only picked a selection), but it is useful for quickly evaluating which toroid size, and then using the changpuak website for fine-tuning. Another thing – Amidon’s ordering pages specify AL values which are slightly  different to their PDF reference. I used the AL values from their ordering pages, whereas the changpuak site uses the PDF reference as far as I can tell. I don’t know which is more accurate.

The Excel spreadsheet can be downloaded here: Type a value in the green cell and results will be shown in the red cells alongside.


Posted in radio | Leave a comment

Quadrature detector coil

Here are photos of the elusive 235SU1 coils. One will be disassembled in the name of science!



It can be seen that the primary coil has the tuning slug pulled out a bit. This kind-of agrees with parts of the earlier analysis of the two coils, assuming that the two slugs are of the same material. I found this image from this website, of the Carver TX-2:


Here is an image of the inside of a NAD 4130 (detailed image). Also the balun for the antenna can be seen at the top-left of the image, made using some parallel pair of wire – nice!:


Posted in radio | Leave a comment

Tuned Front End for FM receiver

The Ryder paper (here) uses a tuned front end. The specified coils (Toko E526HN-100319) seem to be unobtainable, but surprisingly, E526HN-100119 are available (just about), and it looks like they are very similar (I don’t know what the difference is). Of course, any inductor could be used of the correct value, but these particular ones have a very high Q which is desirable for the front end. Coilcraft had some with a Q approaching similar values, so those could be used when the Toko ones are no longer available. The image below is of the E526HN-100119.


Posted in radio | Leave a comment

Modern take on Ryder’s FM Receiver

The link here shows an FM receiver with a high level of claimed performance. The diagram here shows a snippet from the paper.


The problem is, most parts are obsolete. It is an opportunity to try to improve areas of the design, such as to use a better PLL possibly. However the  core ICs are still available via eBay from a few sellers.

Quadrature Detector Coil

The most difficult part to obtain by far is the quadrature detector coil. A dual coil design is employed which just doesn’t seem to be available anywhere. However the diagram here goes a long way to explain how replica coils may be made. Looking at the second coil (marked S on the diagram), the resonant frequency of coil and capacitor across terminals 6-8 is 10.7MHz of course.  Since the capacitance is marked as 100pF, this means that the inductance between pins 6-8 must be 2.21uH.


The LA1235 datasheet suggests the external resistor should be 3.3k, but Ryder’s paper uses a value of 2.2k. With the replica coils, some experimentation may be required with this resistance (and the 4.7k resistance in the circuit too) to set the loaded Q to the optimum value.

The inductance between pins 1-4 and pins 5-9, in parallel with the capacitance between pins 1-10 forms another 10.7MHz LC circuit. The coupling is hard to guess unfortunately. However most small coils don’t have much space, so the chances are that the coil 5-9 was probably just wound on top of coil 6-8.

The coil between pins 1-3 is marked as 64 turns, and is marked as 26uH on the diagram.

It was decided to find two coil kits that, when wound for 26uH and 2.21uH would be approximately 64 and 18 turns respectively.

The  second coil (marked S on the diagram) can therefore be replicated with an adjustable inductor kit of approximately 7nH Al value. The Neosid “10 F 1”  kit seems ideal.

The first coil (marked P on the diagram) can be  replicated with an an adjustable inductor kit of about 6nH Al value. Again, the “10 F 1” would be ideal for this, with the tuning slug pulled out a bit more. The Neosid kit is shown here.


The diagram shows that the wire is 0.12mm for all coils except the 64 turn one, which is 0.1mm. These values correspond to  36AWG and 38AWG respectively (or 40 and 42 SWG respectively).

The tuning slug is 7mm in length, so for a coil to create 26uH and fit within 7mm of length, then 0.1mm wire is needed (0.12mm will be too thick).

Posted in radio | 2 Comments

Another Test Probe

This one used an SMA plug, so it is a bit more versatile. The SMA plug was from Maplin. The 1.02mm tips from Pomona fitted quite nicely into the center pin of the SMA plug as shown here, and it was soldered in place (a bit difficult to solder the steel, the solder found it hard to stick). Before it was soldered, the tip was shortened by 10mm with a hacksaw.


After the solder blob has been tidied up (cut or sand off the excess), it was a push-fit into the connector.


Finally,  a small amount of epoxy resin glue was used to keep it in place.


Posted in general electronics | Leave a comment