Duratool hotair/soldering iron - Display dim non functional

[tom66]

I've had this unit for only about 6 months. I bought it used off TopLoser so there was no warranty of course and the price wasn't bad.

I didn't really use the soldering iron that much, but I did want to use it recently, and came to find the display backlight was very dim and the unit did not heat the iron.

I left it running for about 10 minutes and it fired back up, and worked fine, although the display was very dim and the temperature seemed incorrect, for example it had to be set to about 400C to melt solder.

After leaving it to cool down, it again would not power up, so I decided to do some diagnosis.

The 5V regulator 78L05 (marked only with the refdes of "VOL") takes ~14VDC from the transformer and regulates an internal 5V rail for the logic and backlight. When the unit did not work this regulator had an output of 2.8V, clearly too low. When it worked, with a dim display and inaccurate temperature readings, the output of the regulator was 3.5V.  By heating the regulator with a hot air gun (the same one the unit has) I was able to get the output voltage to rise to 4.5V, where the unit worked properly, aside from the backlight being a little dimmer.

The regulator has no heatsink. The backlight is a single or multi-parallel blue LED with a 22 ohm resistor, and a forward voltage of ~3.1V. So the backlight current is about 85mA. That alone would produce a power dissipation of approx 800mW in the small TO-92 package (far too much, at about 160degC/W for a TO-92, it makes it run about >120°C above ambient...), not including the other current from the microcontroller(s) (there look to be two, but it's possible one is just an LCD controller... they are both Holtek devices.)

I will probably replace both regulators on both boards, which are identical for the hot air and soldering iron, with TO-220 devices, which should be able to dissipate the power using the tab only.

Overall build quality: 5/10 - The soldering is surprisingly good for a cheap product. However, there are definitely some corners cut. In particular the pump is secured only by rubber spacers - which work great to reduce vibration and noise, until it falls off them! There needs to be some kind of retention mechanism. There appear to be no glaring safety violations. Component choices are mediocre, cheap caps and part numbers on ICs you'll never find.

[KL27x]

At 85mW plus, say at least 30mA for the micro, plus say 30 or so milliamps for the display, itself, that's 150mA.

14V-5V = 9V of overhead. at 150mA, that's 1.35W.

The datasheet of a 7805 T220 says its good for up to 1.5A. But that's at minimal overhead of say 2V and while heatsinked, IMO. So 3 Watts maximum while heatsinked. I would put a heatsink on those regulators, to boot. An alternative to a bulky heatsink is to use a series resistor on the input. All you have to do is figure out your maximum load... in this case you can safely overestimate it to be 250mA, let's say. Then take your 14V input - 7V = 7V you can safely remove from the equation and still have the regulator to function properly (5V +2V overhead). So you can use a series resistor that will drop no more than 7V at 250mA = 28ohms. It will need to dissipate 1.75W maximum (7V x .25A). If the circuit tries to draw more than .25A, the regulator input voltage will drop below 7V, and the output will sag. But anything under .25A, and the regulator output will be a true 5V and the regulator will need dissipate a maximum of around half a watt. (At maximum 250mA draw, it will be dropping only 2V. At lower circuit draw, it will have to drop more voltage.... so the heat dissipation of the regulator will not be a linear function of current draw, but the max will nonetheless be around half a watt.)

Because the datasheet states the output capacity in amps, it's easy to look at amps and say "I'm good." What you need to look at is heat dissipation. And that's where Vi-Vo comes into play. Whoever designed that circuit didn't do this. If you only look at amperage, then the TO-92 package should work fine.

I have tried to use a TO-92 package 7805 (unenclosed, open to air) to drop 15V, only to run a 12-15mA load 99.9% of the time, and to occasionally switch on a power FET. The regulator died within days.

[tom66]

I've ordered a TO-220 replacement, but I figure it will be OK without a heatsink. The thermal resistance is about 63C/W for a TO-220, and the ambient environment is cool.

It sounds like from the design, they originally intended for a 9V input to the regulator, as parts of the board appear to be marked 9V. The dissipation would be still too high but at least it wouldn't die as quickly.

I wonder what the actual failure is - why the regulator works after warming up. Some kind of on-die failure which is temperature sensitive...