Hi, I have been working on a duracell 3000 w inverter model drinv3000, long story short, my dad used it with power tools and now it does not work properly, he measured the voltage, and it delivers 104V AC instead of 115.

I already checked almost every component on the board (desoldering them and testing), and I haven’t found any problems, some of the components just were a little off standard value but I replaced them.

So far, I have an idea of where the problem is. The inverter consist of three main parts: a circuit that uses 18 mosfets and several transformers to increase the voltage of the battery (12V), then, that increased voltage is rectified with 3: FMS MUR1620CA and 3: FMS MUR1620CT ultra-fast diodes, these two parts deliver at the end 160V DC, and lastly (the part that I think is going wrong) is where those 160V are converted to 115VAC,I think that because according to the formula of VRMS, VRMS= Vpk/ √ 2 in this case Vpk = 160V so VRMS=113.13V (but I’m having 104VAC). This last part consists of 8 N-Channel mosfets that simulate a 60Hz sine wave.

I purchased an inexpensive oscilloscope to check the output signal and I noticed something is off, the output signal instead of being perfectly square, has a lot of peaks and look a bit unstable/erratic, I compared it with another Inverter and the other one look just fine.

I also checked the two signals from the microcontroller that trigger the gates on the mosfets and they look fine, I reverse engineered the diagram of this part of the circuit so you can see it (I already checked every component here too) What do you think it’s causing the peaks in the output signal?, Do you think even though the components I checked work fine, they misbehave at fast switching?

(attached circuit) two of this exact same circuits generate the AC output, they just have an out of phase signal input from the microcontroller.

Heres a video where you could see the output signal and the spikes I talk about

 

 

Your scope us showing what is called “modified sine wave” and looks perfectly OK. What are you using to measure the voltage? If it is not a true RMS meter, the reading will not be accurate.

From specifications on the Duracell web site:

Output waveform

Modified sine wave

 

Your scope us showing what is called “modified sine wave” and looks perfectly OK. What are you using to measure the voltage? If it is not a true RMS meter, the reading will not be accurate.

From specifications on the Duracell web site:

if you look closely, the square signal has overshooting, while the signal from another inverter ( which is working properly) , does not show this.

, if you compare this with the other video , this one looks steady and does not have overshooting.

I used a fluke 15b+ to measure the voltage, im not sure if it measures true rms but i think the ifference is clear

 

I used a fluke 15b+ to measure the voltage, im not sure if it measures true rms

According to what I read, it does not.
You need a True RMS meter to accurately measure those output waveforms.
Most meters that measure True RMS say so on their front panel.

Your scope should be measuring True RMS but, for some reason it is not, since Vrms and Vave show as the same voltage.

 

Note that the duty-cycle of the quasi-sinewave is not arbitrary but is selected so that the peak volage and the RMS voltage are close to that of a sinewave, so it will usually properly operate devices that either work from the RMS value (motors, heaters, etc.) or the peak value (electronics with peak rectification for the internal DC power).

 

inverter model drinv3000, long story short, my dad used it with power tools and now it does not work properly

How it was noticed that it was not working properly?

the output signal instead of being perfectly square, has a lot of peaks and look a bit unstable/erratic

Those peaks are typically called ringing which is normal. That should not affect to power tools operation. Are the "fully functional inverter" and "not working inverter" the same model?" If they are the same model, then there can be some issue for example in the gate driver circuit.

I compared it with another Inverter and the other one look just fine.

Is the another inverter exact same model than the problematic inverter? And works with the tools using otherwise the same setup like same battery etc.

The scope pictures show that DC voltage peak-peak level (ringing ignored) is about 34-40V lower in the functional unit than in the non-working unit. Are the supply/battery the same, load and the device model the same? If they are, it would be good to find out what is causing so much different DC voltage level. DC voltage level affects to duty cycle and it could affect to your meter reading.

The scope shows RMS level but there seems to be a bug. It shows the same value as average in two images and a low value in one image, perhaps samples are not squared in the calculation.

 

Note that the duty-cycle of the quasi-sinewave is not arbitrary but is selected so that the peak volage and the RMS voltage are close to that of a sinewave, so it will usually properly operate devices that either work from the RMS value (motors, heaters, etc.) or the peak value (electronics with peak rectification for the internal DC power).

On a good one, perhaps.
The ones I have looked at have an unregulated step-up circuit, which gives a DC bus that varies with the battery voltage. Then they vary the duty cycle to keep the RMS voltage correct. It becomes a 230V peak square-wave at low battery voltages.
I suppose the argument (if anyone even thought about it that deeply) goes that switched-modes can usually cope with a wide range of voltages so that the actual peak voltage doesn't matter much.

 

How it was noticed that it was not working properly?

To be honest I did not test it under load myself, I trusted my dad's word and I believed it because of the voltage, my dad says that after being 3 days working with it, being used to feed a hammer drill of a similar model to the GBH 4-32 of the bosh brand suddenly stopped delivering power (it turned off), and stopped using it, then, being at home he measured the voltage and saw that it was lower than normal, according to him he had measured the voltage before that and gave "127V", this inverter has been "broken" for months so I do not know if I trust much in that information.

Those peaks are typically called ringing which is normal. That should not affect to power tools operation. Are the "fully functional inverter" and "not working inverter" the same model?" If they are the same model, then there can be some issue for example in the gate driver circuit.

No they are not, They are completely different models, the "fully functional inverter" is built in a toyota truck I just checked its output to have a reference to compare and i noticed the ringing so i thought it was abnormal

Is the another inverter exact same model than the problematic inverter? And works with the tools using otherwise the same setup like same battery etc.

The scope pictures show that DC voltage peak-peak level (ringing ignored) is about 34-40V lower in the functional unit than in the non-working unit. Are the supply/battery the same, load and the device model the same? If they are, it would be good to find out what is causing so much different DC voltage level. DC voltage level affects to duty cycle and it could affect to your meter reading.

The scope shows RMS level but there seems to be a bug. It shows the same value as average in two images and a low value in one image, perhaps samples are not squared in the calculation.

That "another inverter" is the same as the "fully functional inverter", because is less powerfull (1000w) I didnt test it, I just measured its output to compare.
Its a different setup because the "fully functional inverter" is using the trucks battery and the "not working inverter" its being tested with a lab power supply (im not putting it under load, just measure some signals and output so its not demanding more amps than the lab power supply can deliver)

Yeah because it is a cheap scope maybe it does not read true RMS


My main question here was the thing about the ringing, I dont know to much about it, I though it was completely abnormal.
Thank you for the input!

 

Yeah because it is a cheap scope maybe it does not read true RMS

The display says Vrms so that's a significant error in its operation, since it obviously is not measuring that.

 

So, are we all in agreement that the scope show what is expected for that type of inverter?

The next step is to test it with increasingly greater loads.

 

So, are we all in agreement that the scope show what is expected for that type of inverter?

I think it looks ok based on calculating RMS value from the pulse amplitude and duty cycle. The equation for single pulse was given on few sites as Vp*sqrt(duty cycle) where Vp is amplitude of the voltage. But this is a bipolar signal and I calculated using only positive pulse (imagining that negative pulse would also be positive). The scope shows duty cycle which I doubled. This was based on looking the image and without doubling RMS values are only about 80V on both units. I read peak-peak value (ignoring ringing) from the screen image and used half of that. This removes effect of the offset which is a few volts.

Non-working: 150V x sqrt(0.55) = 111 VRMS (image in post #1)
Functional: 134V x sqrt(0.718) = 114 VRMS (duty cycle varies a lot and I just took one, but seems to be reasonable)

 

Non-working unit is supplied with a lab supply so it's possible to change voltage for example from 12V to 14V and see if the
DC voltage level and duty cycle change according to input voltage. I don't know if it has much value but it's easy to test and gives more understanding of it's operation. It's also possible to compare multimeter readings. Perhaps input voltage affects to meter reading so when measuring output voltage with a multimeter one should measure the input voltage too.

The unit turned off which means overloading according to manual. If we assume that the problem is caused by drop in output voltage it could be increased. The meter reading and also my coarse RMS calculation show that there is room to increase output voltage. I think +10% is ok and even more than 10% should be ok to power tools. It would be good to find if there are some failed components affecting to output voltage/duty cycle before adjusting the output voltage level.

I have never own these devices but some possibilities to adjust voltage which might be usable in this case:
-there is a pot to adjust output voltage/duty cycle
-if DC voltage level affects to duty cycle, change voltage divider resistor values so that the duty cycle control sees lower voltage
-if there is an understandable duty cycle generation circuit, adjust duty cycle on that

 

Non-working unit is supplied with a lab supply so it's possible to change voltage for example from 12V to 14V and see if the
DC voltage level and duty cycle change according to input voltage. I don't know if it has much value but it's easy to test and gives more understanding of it's operation. It's also possible to compare multimeter readings. Perhaps input voltage affects to meter reading so when measuring output voltage with a multimeter one should measure the input voltage too.

You were right!, I didnt know about this feature here is a video where I try it

The unit turned off which means overloading according to manual. If we assume that the problem is caused by drop in output voltage it could be increased. The meter reading and also my coarse RMS calculation show that there is room to increase output voltage. I think +10% is ok and even more than 10% should be ok to power tools. It would be good to find if there are some failed components affecting to output voltage/duty cycle before adjusting the output voltage level.

I have never own these devices but some possibilities to adjust voltage which might be usable in this case:
-there is a pot to adjust output voltage/duty cycle
-if DC voltage level affects to duty cycle, change voltage divider resistor values so that the duty cycle control sees lower voltage
-if there is an understandable duty cycle generation circuit, adjust duty cycle on that

I will look for the circuit that reads the battery voltage.

 

I will look for the circuit that reads the battery voltage.

I thought about high voltage DC but battery voltage measurement is there anyways to monitor input voltage so it makes sense that duty cycle is adjusted according to that.

The unit seems to be working with no-load and was working with the tools earlier so possible problems in the unit are:
-output voltage level has decreased from the initial condition. Induction motor torque is to the square of voltage. 10% voltage drop means 0.9*0.9 = 0.81 -> torque is only 81% of the torque that would be get with 100% supply voltage. This makes starting time longer...more likely to have an overload shutdown. Modified sine wave inverter has low fundamental (60 Hz) amplitude by design so there is not so much room to have additional lower voltage anymore like with sinusoidal supply.
-overload protection has become too sensitive from the initial condition

It's possible that the change in performance/circuitry has developed slowly and finally there was an overload and the unit stopped. What if the unit was and is working and the problem is external to the inverter? It would be good to check that the running condition like battery voltage, connections and especially that the tool was known to be working earlier. It would be even better to test again if possible. You can test with some other loads like more resistive loads to see if overload supervision is too sensitive, already suggested by BobTPH, if you have a good battery.

 

I thought about high voltage DC but battery voltage measurement is there anyways to monitor input voltage so it makes sense that duty cycle is adjusted according to that.

The unit seems to be working with no-load and was working with the tools earlier so possible problems in the unit are:
-output voltage level has decreased from the initial condition. Induction motor torque is to the square of voltage. 10% voltage drop means 0.9*0.9 = 0.81 -> torque is only 81% of the torque that would be get with 100% supply voltage. This makes starting time longer...more likely to have an overload shutdown. Modified sine wave inverter has low fundamental (60 Hz) amplitude by design so there is not so much room to have additional lower voltage anymore like with sinusoidal supply.
-overload protection has become too sensitive from the initial condition

It's possible that the change in performance/circuitry has developed slowly and finally there was an overload and the unit stopped. What if the unit was and is working and the problem is external to the inverter? It would be good to check that the running condition like battery voltage, connections and especially that the tool was known to be working earlier. It would be even better to test again if possible. You can test with some other loads like more resistive loads to see if overload supervision is too sensitive, already suggested by BobTPH, if you have a good battery.

I found the circuit responsible for voltaje dividing, it turned out that it doesnt read the battery voltaje directly, it reads the already stepped-up voltage (160 VDC), here is the little schematic and the modification I did.


Here is a video of my testing

In conclusion, maybe all I did wasnt necessary after all, as you can see in the video, the voltage rises as soon as theres a load.

This whole situation that my dad thought it had a fault was because it shut down which could even be because he had been working for a while with the tool and maybe discharged the battery enough for the inverter to protect itself, as he measured the voltage afterwards and saw that it was lower than what was supposed to be then he assumed that the inverter was malfunctioning.

For my part I just accepted that as a fact without having done any test run under load.

Anyway, I learned a lot about these devices during the diagnosis.

Thank you all very much for your answers, they helped me to understand more about inverters.