Regulating Inverter output

Thread Starter

jj_alukkas

Joined Jan 8, 2009
753
yesterday I built an inverter for using mobile and mp3 player chargers in my car using a cd4047 ic and a pair of IRF540 Mosfets.. The circuit is capable of driving a 100w transformer but I'm using only a 1A transformer in inverted config.. I guess it gives abt 20W All my chargers are working well on the setup. The problem is that the direct output from the transformer is about 280vAC with a 400v small value capacitor.. Though SMPS chargers will take it easy, I would like to regulate the output a bit more to the safe zone using 2 or 3 components.. What should I do??
 

SgtWookie

Joined Jul 17, 2007
22,230
A MOV is a Metal Oxide Varistor; it's designed to break down at a certain voltage - in this case, 240V. They're used to protect devices from overvoltage.

As your circuit exists, there are no provisions for output voltage regulation; the 4047 simply has complementary outputs. Also, there is no "dead time", which may create a problem with a phenomenon known as "flux walking" if the duty cycle isn't precisely 50/50. Google "Flux Walking" to find out more about it.

Off the top of my head, one way to fix the problem would be to use a comparator that turns off the appropriate MOSFET when the output reaches the correct voltage, or the 4047 outputs change states (whichever occurs first) - but that's going to be more complex than you have in mind.
 

Thread Starter

jj_alukkas

Joined Jan 8, 2009
753
Thanks a lot for the reply. Learning new things I didn't know. Well I think I'll leave the design as is as its only gonna drive a small charger for now. And what would that VR1 of 100K do in that circuit? Seems like it controls the oscillation of 4047 say frequency?? When I put the 390k resistor, the tolerance resulted in a final value of around 420k.. so i omitted the 100k variable to test the circuit.. And would adding a MOV be safe for the time being on the output?
 

SgtWookie

Joined Jul 17, 2007
22,230
What would that VR1 of 100K do in that circuit? Seems like it controls the oscillation of 4047 say frequency?? When I put the 390k resistor, the tolerance resulted in a final value of around 420k.. so i omitted the 100k variable to test the circuit.
R2 and C2 set the minimum time period, VR1 allows increasing the time period.

According to the datasheet, the time period is calculated by:
4.4RC
Since R2=420k and C2=0.01uF, the time period is 0.01848, and the frequency would then be 1/0.01848 = 54.11255 Hz. This is approximate, of course; capacitors rarely measure exactly what they're marked. If C2 was exactly 0.01uF and R2 measured 455k, the nominal output frequency would be 50Hz.

As the original schematic was drawn, the adjustment range would be roughly 46.4Hz to 58.3Hz with ideal components.

And would adding a MOV be safe for the time being on the output?
It would certainly "square off" the output! Any voltage over the MOV's rating would result in it conducting, dissipating the excess power as heat. I suspect that it would at least get very hot, very quickly - and likely burn to a crisp.
 
Last edited:

Thread Starter

jj_alukkas

Joined Jan 8, 2009
753
Ok then frequency is fine as long as smps chargers dont start crying about it.. About the output voltage, I feel I need to drop it by a bit.. Maybe a small resistor on the primary side...
 

Audioguru

Joined Dec 20, 2007
11,248
The CD4047 has a digital divider that produces outputs that are exactly 50-50 duty cycle.
The outputs switch at exactly the same time so one Mosfet will turn on while the other is turning off and they will create a huge shoot-through current spike.
The inverter has a simple square-wave output. Maybe your meter cannot measure the p-p voltage of the square-waves accurately.

The low voltage winding of the transformer is only 9V so at low current the inverter's output is close to 320V p-p. The voltage will drop as the battery voltage runs down.
 

SgtWookie

Joined Jul 17, 2007
22,230
The outputs switch at exactly the same time so one Mosfet will turn on while the other is turning off and they will create a huge shoot-through current spike.
Nice to know that it's exactly 50% duty cycle; I hadn't read that far in the datasheet.

Were it an H-bridge configuration, I'd agree with you about the shoot-through.

However, since the drains are at the opposite end of a center-tapped transformer, with Vcc supplied via the center tap, I'll suggest that shoot-through considerations are moot in this case. Both MOSFETs being on concurrently will likely reduce the efficiency, but I'd be surprised if there was a noticeable current spike.
 

DickCappels

Joined Aug 21, 2008
10,152
One thing you might consider is putting a 10.5 volt pre-regulator in series with with the connection to the car's electrical system, and if necessary, add a few turns to the secondary windings to compensate. This way, when the car's electrical system voltage rises (such as when the engine is revved up) the output of your power supply will not rise.

Consider a switching regulator to minimize heatsinking requirements.

Once you feel confident in getting switching regulators to work, throw out the CD4047 circuit and the pre-regulator and use an inexpensive switching regulator chip to drive your MOSFETs.

Seriously.
 

Thread Starter

jj_alukkas

Joined Jan 8, 2009
753
Ya, there might be a possibility that meter readings can vary, Ill try a higher value cap and check again.. If I had an Oscillo, I would have checked the timings of both the MOSFETs, but sadly I dont.. Voltage dropping of battery is not possible cos I tested this circuit from my car battery with engine off.. when its on, it'll go up again I think..
 

Thread Starter

jj_alukkas

Joined Jan 8, 2009
753
I thought of adding a regulator but an inverter draws massive current from the battery but I guess in my case it'll work fine as m usin only a 1A transformer.. I also searched for switching supplies before I soldered this thing out but they seemed to be complex for such a simple 15W application.. If it outta be something heavy then switching mode is sure the way to go.. So ill go for input regulation for now..
 

italo

Joined Nov 20, 2005
205
:mad:
The CD4047 has a digital divider that produces outputs that are exactly 50-50 duty cycle.
The outputs switch at exactly the same time so one Mosfet will turn on while the other is turning off and they will create a huge shoot-through current spike.
The inverter has a simple square-wave output. Maybe your meter cannot measure the p-p voltage of the square-waves accurately.

The low voltage winding of the transformer is only 9V so at low current the inverter's output is close to 320V p-p. The voltage will drop as the battery voltage runs down.
not quite right manufacture claim that 50% duty cannot be quarantee:D
 

Audioguru

Joined Dec 20, 2007
11,248
The datasheet for the CD4047 says that the oscillator does not guarantee a 50:50 duty-cycle but the digitally divided opposed outputs (which are used in the inverter) have a 50:50 duty cycle.
 
Top