Two kits? Why not make it three. I’m enjoying this trip down memory road – but this one is rather interesting as well, as I’ve seen some of these units around but never had one of my own. This kit is almost as old as I am!
This kit is yet another “pink” Test Gear kit, the K7220 Transistor Tester kit based on the circuit published in the now defunct Electronics Australia, December 1990 issue. In the 90s, through-hole transistors were still rather common in electronics and considered rather precious components which could easily be damaged by electrostatic discharge (ESD) or fail in-circuit, so having a tester could well pay for itself. I can remember seeing one at my school’s electronics lab at one stage, but it was never used. It did attract my curiosity along with the ESR meter kit (which I don’t have).
This kit claims to be capable of checking most discrete semiconductors including BJT and FETs. It claims to be easy to build with a battery test facility and error-check function. As usual with boxed kits, it comes with all of the components and hardware, but this one also has a pre-punched silk-screened front panel.
Inside is the bag of components, as usual, accompanied by a retail package for the low-cost signal meter, the pre-punched and silkscreened metal front panel, PCB, full-retail jiffy/zippy box package (with a surplus lid and aluminium plate) and the text material.
The unit comes with a metal front panel which comes with the required cut-outs and printed labels for a very professional touch. It also makes the kit easier to assemble, but is more likely to have been included as most people would not have had the ability to cut out a rectangular slot for the panel meter. Often times when a kit calls for some “difficult” step like this, they would have it “pre-done” in some way.
Everything is made on a single-sided PCB with very much straight-tracks as if to have been drawn out by hand somehow. The donuts are very thin around the multi-position selector switches, and some traces run very close to the edges of the PCB. This seems to be helpful in reducing the cost of the PCB by reducing waste material, but at the risk of having the occasional reject because manufacturing errors near edges can sometimes occur.
The PCB is the standard fibreglass-based board, single sided without silkscreening or solder resist, thus some care in construction is necessary. Constant reference to the supplied assembly manual is necessary as well, as the component values are not marked on the schematic, thus cross-referencing is required.
Not much to see from the top-side of the single-sided PCB, as expected, but this particular PCB is different to many others for having two sizes of drill-holes rather than one. The design allows for the integration of the rotary switches directly onto the PCB, allowing the PCB to mount directly behind the front panel. I suppose this is one reason why it is considered “easy to build” as this reduces the necessary wiring.
Despite the age of the kit, this assembly manual claims to be the second edition. The project is from Electronics Australia magazine, now defunct, but it’s nice to see their version as illustrated on the front page of the manual. Their front panel is a lot brighter and their signal meter looks to be a bit taller as well – so it seems Dick Smith didn’t just “borrow” the design without changing it to meet their needs regarding what sort of components were available.
At the end of the manual, there is a list of stores and phone numbers – very nostalgic to see (008) and 888 free-call numbers. At this point, it seems they really were across all of Australia, but may not have begun their expansion into New Zealand or USA just yet. A snapshot in time, most definitely.
By now, I am getting more used to the wider spacing of components on these older kits and the occasional issues with oxidation, so that’s no great issue. It’s just a matter of populating the components and soldering them down – this kit did not have PCB pins for connecting wires to, which was a nice change.
Only a handful of components needed to be mounted aside from the two large multi-position rotary switches. The instructions contain information about clipping off the lugs from the switches to allow them to directly fit into the board, but the versions supplied were PCB-mount so did not have lugs to begin with, so that’s one step saved.
The switches actually can be configured for the number of selections – the nut has to be undone and a selector ring inserted into the right position to get the four position + three position set-up necessary.
The underside of the PCB shows my soldering handiwork. I used my own “fluxy” Multicore solder which leaves some brown residue. However, soldering the multi-position switch did require a fair volume of solder, and some of the thinner pads required some care to actually get the solder to wet them. Ideally the pads could have been larger diameter, but I think the risk there would be solder bridging.
From there, the front panel had the banana jacks populated and soldered to wires connected to the PCB. The switch was then populated and wired in, along with the panel meter and battery snap/barrel jack input. The assembly manual shows the barrel jack being wired for centre negative input – I didn’t like this so I changed the wiring to centre positive. There was no supplied method of mounting the panel meter to the front panel, so hot melt adhesive glue was used. The LED was placed into position to protrude through the front panel, while the rotary switches were fed through the front and secured to the front panel to complete most of the kit. The final steps are to add the knobs and secure them to the shafts by using a jeweller’s flat-blade screwdriver.
The supplied front panel replaces the lid of the kit entirely, thus leaving the lid and an aluminium plate as spare – good to have in case of making a mistake with another kit. Unfortunately, the banana jack holes were a little too small, so I had to ream them out slightly to allow the insulating bushings to fit properly. The kit looks rather processional from the front and it’s definitely nostalgic to have that EA logo showing. Unlike some later kits, Dick Smith didn’t slap their name on the front of this one which is a nice change.
On the topic of “mistakes”, in my desperation to complete the kit, I drilled in holes for the external power socket a little too high in the case. At the original position, the socket interfered with the PCB which was mounted directly behind on the front panel. I drilled a second set of holes further down, which were just fine – covering the first set of holes with some black electrical tape. But often times, you do almost complete a kit and make a stupid mistake like this … then you kick yourself … and get tempted to buy a new case. Not that they’re around anymore …
The completed meter with the test leads inserted looks as follows.
The use of those darned banana sockets was a bit difficult to solder to (again) but after developing a method and some patience, I was able to create the test leads above with the remaining wire in the kit.
Unlike the previously constructed thermocouple adapter, this transistor tester requires no calibration at all. Unfortunately, combined with the low-cost signal meter supplied, this meter is not highly accurate in measuring transistor gain but instead, is more useful for relative checks between several units to verify if they have broadly similar gain characteristics.
That being said, in my brief testing, it seemed to work as expected, as did the battery test. The latter feature is quite important, as the readings from the unit and the accuracy will depend on the battery voltage. The error check LED is useful to check for shorted leads which would otherwise result in an off-scale reading on the meter which could damage it by slamming the needle into the end-stop. Considering most modern multimeters don’t have transistor gain testing features anymore, this could be useful, although there are other kits which use microprocessor-based algorithms which could perform even more tests for an inexpensive price.
This is truly one of the kits which can be constructed in such a way to look almost professional, as if store-bought. While the tester isn’t exactly calibrated in any way, it’s more useful as a troubleshooting tool for some qualitative testing of transistors in a rather “manual” fashion. For that purpose, it seems to work acceptably and the inclusion of an external power socket feature makes the issue of battery replacement less problematic.
Unfortunately, I made a mistake while drilling out the power socket, so there is another mark of an amateur. However, as I wasn’t going to buy another jiffy box (Dick Smith, the “electronics” shop is long gone), I settled for some black electrical tape to cover it up. I made a variation to the wiring to have the polarity as center-positive (as this seems more popular nowadays) rather than center-negative as originally specified. While it did come with a pre-punched front-panel, fitting the meter required some glue (I used hot-melt adhesive) and reaming of the banana socket holes because they were a bit too tight to fit. The knobs also showed their age, being a little dirty in appearance as the plastic appears to be breaking down.
This is, however, superseded by modern microcontroller based testers which can automatically identify component types regardless of connection order and can measure threshold voltages, gain values and more. The newer meters are even more inexpensive and would be a dream to have back in 1990.