PDA

View Full Version : Generator waveforms



.RC.
26th Apr 2019, 09:41 PM
I purchased one of those cheap portable digital oscilloscopes to have a look at the various waveforms of mains and generator power.

Safety being a factor I looked up how to do it properly and the waveforms were interesting.

I never expected it to give a completely accurate waveform so here is the mains waveform I get. The scope has a problem showing the bottom of the waveform properly. You can modify it to fix the problem, but I have not bothered.

379198

Here is the unloaded generator I am making. I never checked it loaded up. I was happy to see no voltage spikes, it has an AVR with brushes.

379199

Here is my Miller welder generator that is about thirteen years old. Few spikes here. It is capacitor regulated with brushes.

379201

Here is the Miller generator under load with an inverter welder drawing some decent current. It knocked the peaks off.

379202

.RC.
27th Apr 2019, 12:15 PM
I think what the last image shows is where the inverter welder draws it's current. Only when the voltage is at the peak. Not through the entire cycle. The welder being a cheap $130 one probably has very poor power factor.

I should put the plasma cutter on it as it has power factor correction and see what it looks like.

.RC.
27th Apr 2019, 05:39 PM
Here is a comparison between a El Cheapo Cut40 plasma and a Unimig razorcut 45.

Cut 40 you get a waveform like this

379206

Unimig like this

379207

Karl Robbers
29th Apr 2019, 03:21 PM
Out of curiosity, where did thee get your oscilloscope and how many shekels did thee need to grease the sellers palm?

Michael G
29th Apr 2019, 08:19 PM
Looks like one of these - https://www.ebay.com.au/itm/JYE-Tech-DSO-Shell-DSO150-2-4-TFT-Handheld-Pocket-size-Digital-X0N6/132636097124?hash=item1ee1ba3a64:g:J58AAOSwl0VbB-~i

Around $50 with free post.

Michael

.RC.
29th Apr 2019, 08:41 PM
It is a DSO Shell made by JYEtech. There are original ones and counterfeit ones.

It is a kit or pre-assembled, you can buy the parts with the surface mount devices already soldered in, you just have to solder in the through hole parts.

The original comes with a X1 and X10 probe. You have to supply power as well, a 9V battery clip soldered in works well and it is then an isolated scope, so a bit safer depending on how you want to measure your mains voltage waveforms.

You can use a low voltage step down transformer which may or may not distort the waveform due to harmonics in the transformer.

I just measured directly with the probe on X10. The DSO Shell can only handle 50V input, so on X10 mains is 24V (although peak voltage is higher)

I would suggest anyone thinking about this to look at youtube videos first and understand what you are doing and the risks involved. https://www.youtube.com/watch?v=xaELqAo4kkQ and https://www.youtube.com/watch?v=oMqhrRUjxOg

I have put the plasma cutters on my big generator now, mostly to see what the cheap cut40 would do to the waveform.

Here is the cut 40 waveform at full power on the 11kVa generator

379220

Here is the Unimig razorcut45 at full power

379221

Now here is my 250 amp MIG flat out.

379222

Ok so what is happening with these cheap inverter machines.

From my internet travels they are essentially switch mode power supplies without any sort of power factor correction. And they only draw current at specific times during the wave cycle. So in the first photo of this post, for each sine wave cycle which takes 0.02 seconds, the plasma cutter is only drawing power for 0.012s at the tops of the peaks. On the image each vertical background line represents 0.002 seconds. So the cheap ones are not very efficient at drawing electricity.

Ropetangler
29th Apr 2019, 10:21 PM
I might be right off beam here, (it wouldn't be the first time, and likely not the last either), but I would have thought that any switch mode power supply would basically have an electronic switch controlling power into a bridge rectifier circuit, the output of which would then go into a capacitor bank. From here the power out would also be controlled by another switching circuit, depending on the needs of the primary device. The initial switching would I think be done as the voltage waveform was crossing the zero volt line, in order to eliminate the generation of harmonics which would likely happen if the current was switched at any point other than the zero crossing point. This should minimise any waveform distortion being imposed on the input supply, so I am not sure what is going on in your plasma cutters RC, unless the capacitor bank is not up to the mark. Others with more smarts and knowledge hopefully will put us out of our misery with a succinct explanation of what is the likely situation.

.RC.
29th Apr 2019, 11:13 PM
I do not know either, all I have is what I have read

Power Topics for Power Supply Users: What is PFC and why do I need it? (http://power-topics.blogspot.com/2008/02/what-is-pfc-and-why-do-i-need-it.html)

Looking at the first image the dotted line is current draw showing the short bursts of current consumed. I extrapolated that on the generators consuming that current in short bursts pulls down the voltage and hence the waveform I am seeing.

morrisman
30th Apr 2019, 09:39 AM
If the voltage peaks are being cut off , maybe that means generator isn't up to handling the work load ? Or poor generator design maybe ? Have you compared the cro waveform between the gen and the mains when using the same welder . Just some ideas ! Some years ago I was given a huge 100kva gene , with a 6 litre Cummins turbo diesel. I eventually passed it onto my neighbour who has a 200 acre farm. It was a standby gene for ABC radio Longford . I never ran it. Thr fuel bill put me off !

RustyArc
30th Apr 2019, 01:34 PM
For a switching/inverter power supply with no PFC, there's no switching on the "primary" side - just mains into a rectifier and then a bank of capacitors. On the other side you have the switching and a transformer to produce the desired voltage and waveform. With no load, the capacitors will sit at around 320V. As the load increases, the capacitors' voltage will drop as energy is used, and get recharged 100 times a second by the incoming mains.

However, the capacitors will only take a charge when the incoming voltage exceeds their voltage level, so the load might be bringing them down to, say, 280V, which means they'll only start to charge when the incoming rectified AC reaches this voltage, and will stop once it drops, hence they only draw current around the crests of the rectified AC wave, leading to the flattening off of the troughs and peaks of the supply. This equates to a low power factor, although not in the same way as power factor is associated with inductive loads like motors, where the current lags the voltage , but both are sinusoidal. In the case of an inverter, the voltage waveform may be distorted, but more significantly, the current waveform will be discontinuous.

.RC.
30th Apr 2019, 09:46 PM
Here is the Cut40 running off mains.


379224

RustyArc
1st May 2019, 10:11 AM
The lesser sagging of the peaks makes sense, due to the lower impedance of the mains compared to a (comparatively) small generator.

.RC.
1st May 2019, 02:47 PM
Yes, the mains has more oomph factor. But interestingly the mains curve has a similar profile to the 11kVa generator.

The voltage rise stops abruptly, then slowly rises again, before free falling.

RustyArc
1st May 2019, 10:08 PM
Yes, the mains has more oomph factor. But interestingly the mains curve has a similar profile to the 11kVa generator.

Yep, while the mains is lower impedance, there's still the street cabling and house cabling between you and the local transformer, which adds some impedance, so you're going to get a similar sag during the period where the mains voltage is high enough to charge the capacitors.