Repair of the Mystery Toaster

This Instructable deals - more or less as a pretext - with a repair process.

But no matter what you're repairing, understanding the internal workings of the unit to be repaired in advance is beneficial. Sometimes you learn it the hard way by repairing more than once, but you also might learn 'on the job', while performing the actual repair. In this instructable I tell you about how I managed to understand the entrails of a defective toaster while repairing it.

As in medicine, you shouldn't just fight the symptoms or even suppress them, but rather find - and cure - the cause for the disease producing the symptoms. If you're lucky the symptoms clearly show you what the cause is, but sometimes you have to dig longer and deeper, in order to understand the reason of the problem first. And once this happens, finding the solution will be rocket science no more.

I like repairing (almost) all things electrical and electronic. In addition to the gadgets I build for myself - that often need a repair right after being finished - I work in the repair workshop of a community center once a week, where we do repairs free of charge, have fun at the same time, and manage to make customers happy in about four of five cases.

A few weeks ago a lady brought her toaster that, because of its colourful design, has a high emotional value for her. I admit that for myself it rather was some cheap crap, but she wished to have it repaired nevertheless.

She said that it once had produced a loud, crackling noise. She still could toast bread with it since then, but she had to manually hold down the lever until her toast was done. This lever should, of course, be automatically held down and be released after a preset time.

Since I never before had had business with defective toasters, I had not the foggiest notion about their internal workings. This resulted in a rather shallow and long diagnostics/learning curve for me. But once the principle was finally understood, it was very easy to repair, and, after some additional thoughts, to beef it up by an upgrade that hopefully will keep the toaster functioning for many years. (I won't be here anymore for the next repair anyway :-)

I am confident that many different toaster brands and models work according to the same or a similar principle, so what you read here might very well apply to your particular toaster model, too.

After a first test that confirmed the lady's statement, the toaster's enclosure was opened. Of course I had to use a special screwdriver matching the abominable, tamper-proof Tri-Wing screws.

Apart from the heater and the timer electronics PCB, the main components of this unit are a solenoid and a mains switch. This switch is operated by the same lever that lowers the bread slices into the unit. Once this lever is pressed down, it is automatically kept down by the solenoid controlled by the electronic timer circuit. As soon as the timer circuit cuts the power to the solenoid, the lever is released, and the toasted bread jumps up. In the same moment the mains switch disengages, and the whole toaster is powered off. The pictures above show the mains switch open and closed.

At a first glance, on the electronics PCB a blown-up electrolytic capacitor and a cracked CD4541 CMOS universal timer chip were easily visible - which suggested that a heavy over-voltage condition must have occured, resulting in a high current flow. To my surprise, all the other components on the PCB were ok.

However, I was rather unsure about the reason for this overload. Fortunately I managed to find it - some time later.

After having located spares for the defective parts, I unsoldered the blown-up electrolytic capacitor (C1) and removed the timer chip (U1) from the PCB by cutting its pins along the package, turning it into a little 'coffin' (yes, it's much easier to unsolder the individual pins when the package is removed already). As long as you know that a chip cannot be saved anyway, you can do that with a clear conscience.

After having installed the spares, I tested the PCB outside of the toaster, using an external 12 V DC power supply. It worked nicely: after powering up, the output transistor (Q1) driving the solenoid was activated for the time set with the potentiometer (RV1).

So, everything ok?

Not yet.

In order to understand the circuit's function, I had drawn (an old-fashioned term for 're-engineered') the circuit diagram from the PCB, as shown above. This being a rather simple, straightforward circuit, I quickly succeeded. I also found (i.e. downloaded) the timer chip's data sheet for more information, which was not difficult as well.

Once I had learned the function of the timer circuit, I still wondered how the rest of this unit works. As you can see from the pictures in step 1, the lever not only moves the bread down towards the heater, but also closes two contacts that supply mains voltage to the heater.

There was neither a mains transformer nor any other supply unit contained, and I asked myself where the low supply voltage for the timer electronics might come from. While the timer PCB still was removed from the unit, I pushed the lever down (supplying the heater with the mains voltage) and measured the voltage between the two wires that went to the timer PCB before. Amazingly enough, this voltage was sometimes about 11 V AC (which was rather plausible to me), but sometimes I found the full mains voltage of 230 V AC between these same wires. This suggested to me that something was badly wrong, and also was an explanation for the blown-up capacitor and the cracked timer chip.

This toaster model has two slots for toasting two slices of bread simultaneously (as most toasters have). In order to do so there are four heaters provided, one for each face of each slice. These are simple resistor wires, wrapped around some heat resistant and electrically insulating material (mica sheets, in this case). The four wires are connected in series (see diagram above). Total resistance of all the heaters is around 80 Ohms, resulting in a current of somewhat below 3 A (Ampère) and a power of about 660 W (Watt) at our regular mains voltage of 230 V (Volts).

Fasten your seat belts, now it gets interesting: On the outer side of the left bread slot there was an additional tap in the heater wire, serving as a plain, old-fashioned potential divider (aka voltage divider). Its partial resistances were about 18 Ω (Ohms) and 3.6 Ω. The voltage drop across the 3.6 Ω part, then, was used to supply the timer electronics PCB.

This is rather important: What I tell you above means that, during operation, not only the heater wires but also the whole timer PCB is electrically connected to the mains voltage, and thus this PCB must NEVER be touched while powered up!

After some more disassembling - and several cuts in my fingers due to the razor-sharp edges of the internal sheet metal enclosure - I found the cause for the problem: One end of the 3.6 Ω heater wire was broken and made contact only from time to time - which explained the varying voltage across this resistance.

Of course I also thought about why such a wire might break, and I have two different explanations for this fact (the first of which seems to be somewhat more plausible to me):

1. When heating up, the wires expand thermally, i.e. they get longer, while the riveted contact points on the mica sheets remain fixed. So every time the toaster is powered up, the wire is bent at the riveted contact point, and after a lot of use the weakest point next to one of the rivets breaks. Both expansion (when heating) and contraction (when cooling) are clearly visible in the video clip attached below.
2. A different reason might be that the contact between the heater wire and one of the rivets could be less than optimum, and so there is some contact resistance. As soon as current flows through the heater (around 3 A) this results in excess heat at the contact point, which might melt the heater wire at this point.

Resistance wires normally consist of a metal alloy that refuses being soldered. With the wire becoming red hot in operation, any soldered connection would automatically unsolder itself anyway. That's the reason why all the connection points to the resistance wires in this toaster are riveted. Soldering was out. But riveting was out as well due to restricted access, and because I don't have any matching rivets in my toolbox.

I fortunately found a second, hollow rivet next to the one from which the wire was broken away, allowing me to feed the heater wire through the rivet and clamp it using an M2 screw and nut. In this way the wire is only a few Millimeters shorter, which has very little effect on its resistance. The first picture above shows the wire (white arrow) that was broken from the rivet circled in red and is re-connected to the second rivet circled in green. As you can see in the second picture taken from the opposite side of the heater unit, the two rivets in question are interconnected. The excess wire standing up from below the nut was, of course, cut before reassembling.

When carefully testing the repaired heater - still without the timer PCB - it heated up normally and reliably.

Everything ok now?

Of course, the unit might already be assembled and used now. But there is still more to it.

I still tried to understand why the supply voltage was much too high when the heater was broken, therefor I followed the current path in the circuit diagram with and without the broken wire; the two versions are given above.

As you can see from the bold current path highlighted in black, during normal operation the current flows from the Line (L) mains contact, through all the heating resistance wires connected in series, and back to the Neutral (N) mains contact. Across the smallest resistor (R6) there is a voltage drop of about 11 V that is used, via the half-wave rectifier D1, to supply the timer circuit.

Under faulty conditions - R6 interrupted - the current tries to find a different path, highlighted in red in the second diagram. The current now flows from the Live mains contact via resistors R7 to R10, but the path through R6 is blocked. The alternative path goes through D1, and then through C1 in parallel with the timer chip U1, back to the mains Neutral. Since the combined resistance of R7 to R10 is rather low, a high current flows through C1 and/or U1. These two are low-voltage components and can sustain this grossly elevated voltage and current only for a very short time. Then they blow up quickly with some light and sound effects, followed by a rather unpleasant smell - similar to fireworks.

BTW, that's the reason why I sometimes say 'Electronics operates with smoke. As soon as it breaks the smoke comes out'.

I not only wanted to repair this toaster. The ultimate goal was to avoid this problem in the future. The way to go is by completely separating the heater circuit from the timer PCB.

First of all, the two glass-silk insulated wires going from both ends of R6 to the timer PCB are made redundant. The timer PCB will be fed from a different source later. I didn't want to crop them (in case they might be required at a later date) but rather removed the large insulating sleeves from the wires, folded them back on themselves, and then slid the insulating sleeves back over the wires, insulating their ends without cropping them. Fortunately the insulating tubes are somewhat elastic; nevertheless some patience was required doing this. A picture of the wires treated in this way is shown above.

I had a small 9 V/200 mA transformer in my junk box that I could find room for inside the toaster's enclosure. I added a (full-wave) bridge rectifier for supplying the timer PCB (see the picture and the circuit diagram of the upgraded unit above). This allowed me to reduce the value of C1 from 470 uF to 220 uF (I didn't have a 470 uF cap in my junk box anyway); apart from that, transformers in general don't suffer half-wave rectification gladly since they are then forced to deliver DC instead of AC, and small transformers can become very hot when abused in this manner. So full-wave rectification is the way to go anyway.

Due to this, D1 on the timer PCB was no more used and replaced by a wire bridge.

As you can see, this addition was made in a rather amateurish, quick & dirty way because I was running out of patience, and I admit that during this process I used several words not fit to be reproduced in print.

Please note that, in operation, the interior of a toaster can become rather hot. It is important, therefor, that any additional wiring you install within a toaster has to be heat-proof. There exist several different heat-proof materials used for wire insulation, such as glass silk, PTFE or silicone. I used the silicone-insulated wires I had at hand (two brown and blue wires each in the picture). A green/yellow wire, not shown in the picture, connects the protective earth (PE) contact of the mains plug with the transformer's mounting bracket.

After reassembling the whole unit (using Phillips screws from my junk box instead of the much-hated Tri-Wing screws) it tested ok - and could be given back to a delighted customer :-)

Total cost for this repair and mod was only about CHF/EUR/US$ 15 - in addition to quite a few hours, bleeding fingers and a lot of patience from my part.

The best of it all, however, was having saved some materials from going to the recyclers or the landfill, together with the feeling of success!