Super Cheap Mini Inverter
I was sent this tiny cheap inverter to test. It costs around US$4 retail.
It is specified to be modified sine wave 12V input and 220V 70W nominal output.
Nothing much else to say, so, on we go with the pictures.
This is the inverter. It is quite small and light.
Sharpie for size comparison.
Specifications are embossed on the plastic case.
No stickers here to save costs!
Unit is marked 50Hz but as per testing below, it is a 60Hz unit.
Case simply snaps close.
No screws to open it, just pry it apart - it takes a bit of effort to open though.
Interestingly, it uses a polyswitch resettable fuse for overcurrent protection on the DC input side.
This appears to be a 6A rated part.
Solder side is reasonably clean.
Solder mask is removed on the high current tracks to improve current capacity by adding solder to the traces.
Output H-Bridge devices are IRF730.
The two 16pin devices are TL494.
One is for the input DC-DC converter. The other for the AC oscillator.
There is a bunch of rosin flux gunk on one of the heatsinks.
This must have been from the wave soldering process.
Flux is quite corrosive so the heatsink will look even uglier eventually. IPA / laquer thinner or flux remover should remove it easily.
DC power switching MOSFETs are 2SK4145 devices.
These are rated 60V 84A so it is pretty overkill for the job.
Idle current draw with no load at 12.5V input is 270mA, so a little over 3W.
It does not change a lot with battery voltage.
At 10.5V input, it only puts out 188Vrms.
At 12.5V it is quite acceptable.
At 13.8V, it reaches 240Vrms!
At 14.8V (fully charged lead acid battery) It is right at the top end of acceptable.
But this is at no load so it should drop a bit when loaded.
At 15V (lead acid battery at equalize) it gets pretty high.
For heating devices, it would get warmer than normal so watch out.
Although most universal plug packs/adaptors should handle this fine as it is MSW and peak voltage is much lower than a pure sine waveform.
This is the primary side step up converter waveform.
Frequency is rather high at 164kHz to achieve high power with a small main power transformer. Waveforms are quite clean with no ringing and spikes.
At 10.5V, peak voltage is 200V.
Duty is stuck at 45.9% so the TL494 is maxed out.
At 12.5V, peak voltage has increased to 240V but duty is still at max.
Frequency has also increased slightly.
At 13.8V, duty cycle has reduced to attempt to drop the RMS voltage but as the measured value above, it is not as effective.
Further increase of input to 14.4V, Duty cycle further reduces.
Frequency further increases.
At 15V, peak voltage is still at 304V.
Rectified 240V is at about 340V so it is still very safe for switching adaptors and plug packs (Cellphone and laptop chargers)
Frequency is still increasing! But for AC chargers, this is not a problem as the output is simply rectified to DC again. Not good for AC driven clocks but the waveform might damage them anyway so it won't matter.
Here is the waveform using the math function of my DSO.
This is at 13.8V and no load.
This is with a laptop charger plugged in and with a 20W load.
Just for kicks, here is with a 15W AC fan load which uses a shaded pole induction motor.
With the 20W charger load, it remains warm except for a pair of 0806 SMD resistors.
One outer leg of the main transformer also gets warmer than the other side which means the ferrite on that side is saturating. It means the ferrite core faces are not machined to mate perfectly flat or the colder side has dirt stuck between it causing the other leg to take most of the magnetic flux.
It gets really toasty so I believe this will be the weak point for this inverter. Being placed beside an electrolytic capacitor does not help reliablity at all.
It should have used a 1206 resistor or larger for the high dissipation rating.
It can also eventually arc over as the 0806 resistor are used well above their voltage rating.
These are the resistors in question.
They are R1 and R2 which are 120K 0806 size.
A pair of normal 120K 1/4W through hole resistor would work, a pair of 1/2W resistors are better but fitting them on the solder pads could be tricky.
These are basically a resistor from each pin of the output jack to battery - so they see up to >300V peaks across each of them or about 180Vrms, yep, 0806 will eventually arc over and blow up. But the inverter could still possibly function without them? Not 100% sure though.
This inverter is very cheap. It runs with a very low idle power consumption as with typical MSW inverters which is great for making low power portable charger/generators. Output voltage and frequency fluctuates a lot with varying battery voltage but is still within reasonable limits for switching plugpacks and wall warts. Not sure how linear adaptors would deal with it but for most gadgets these days, it should be of no issue. It does have soft start at turn on and the power LED lights when output is at the end of the soft start stage.
Page created and copyright R.Quan ©22 Nov 2016.