Author Topic: Can anything be done about AC load "ripple"?  (Read 2441 times)

0 Members and 1 Guest are viewing this topic.

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Can anything be done about AC load "ripple"?
« on: July 06, 2023, 09:11:42 am »
My off grid inverter powers my office.  It's usually between 200W and 450W.  The issue is, it's noisy as hell.  The load graph looks like a random sawtooth like pattern.  Lots of noise, 5-10W of it.

It's on the ACOut and makes it through to the DC side and is seen on the battery current and voltages too.  So it's real.

My understanding is there is very little that can be done about it.  You can't just store AC power in a capacitor to smooth it.

An example this morning.  Its just "load noise", the power factor isnt that bad either, given it's 90% SMPSUs.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline m k

  • Super Contributor
  • ***
  • Posts: 2651
  • Country: fi
Re: Can anything be done about AC load "ripple"?
« Reply #1 on: July 06, 2023, 10:20:25 am »
What is X-scale, minutes?

Low pass, 10th order Butterworth, passband 75Hz, stopband 119Hz.
https://tools.analog.com/en/filterwizard/
Advance-Aneng-Appa-AVO-Beckman-Danbridge-Data Tech-Fluke-General Radio-H. W. Sullivan-Heathkit-HP-Kaise-Kyoritsu-Leeds & Northrup-Mastech-OR-X-REO-Simpson-Sinclair-Tektronix-Tokyo Rikosha-Topward-Triplett-Tritron-YFE
(plus lesser brands from the work shop of the world)
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #2 on: July 06, 2023, 10:49:09 am »
The X axis is "clock time".

Y axis is Watts and VA
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #3 on: July 06, 2023, 10:54:26 am »
Interestingly, if I look at the voltage chart, it does not have that noise.  it's 230V, 50Hz stable.  it droops and spikes with high load changes, but it doesn't have the noise.

This makes sense as it's load driven noise, the power requirements are altering, probably constantly, the graph resolution is only 0.2Hz tops.

So if voltage is stable and the VA is bouncing it means the current is bouncing.

I need an AC capacitor and I know they don't exist.  How do you store alternating current?  With an alternator, aka a flywheel.  I don't think a ton block axle and some AC induction motors/generators is worth it though LOL
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline m k

  • Super Contributor
  • ***
  • Posts: 2651
  • Country: fi
Re: Can anything be done about AC load "ripple"?
« Reply #4 on: July 06, 2023, 11:16:21 am »
Battery is a flywheel, so is grid.
Advance-Aneng-Appa-AVO-Beckman-Danbridge-Data Tech-Fluke-General Radio-H. W. Sullivan-Heathkit-HP-Kaise-Kyoritsu-Leeds & Northrup-Mastech-OR-X-REO-Simpson-Sinclair-Tektronix-Tokyo Rikosha-Topward-Triplett-Tritron-YFE
(plus lesser brands from the work shop of the world)
 
The following users thanked this post: Slh

Offline gf

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: de
Re: Can anything be done about AC load "ripple"?
« Reply #5 on: July 06, 2023, 11:17:23 am »
As long as the inverter can keep the voltage approximately constant, what's the problem? The battery is your energy storage device. The inverter likely also has a capacitor bank on the DC side to respond to load transients faster than the battery can do, and to handle reactive power that flows back and forth between the inverter and the load.
« Last Edit: July 06, 2023, 11:24:15 am by gf »
 
The following users thanked this post: tom66, paulca

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Can anything be done about AC load "ripple"?
« Reply #6 on: July 06, 2023, 11:30:14 am »
Isn't that variation in your power use? If voltage and frequency are stable, it indicated the inverter is correctly responding to load changes, hence doing it's job.

Quote
I need an AC capacitor and I know they don't exist.

The battery+inverter is exactly that! It works like a capacitor, responding to load changes, keeping voltage stable. You are just looking at the wrong metric, you should be looking at voltage.
 
The following users thanked this post: newbrain, paulca

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #7 on: July 06, 2023, 11:40:04 am »

The battery+inverter is exactly that!

I suppose.  I mean in flat areas with no pumped hydro they use batteries for grid transient suppression.  Previously they used more heavy turbines, but in the days of renewables those are getting lighter and lighter.

You are right.  If I don't like the noise I can low pass filter it out by adjusting the graph interval :)  Then I wont see it and it wont distrurb me.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #8 on: July 06, 2023, 11:44:04 am »
Battery is a flywheel, so is grid.

The grid is a very, very, very expensive flywheel.

kWh import:  0.41 GBP.
kWh export:  0.05 GBP.

However, my system is not "notified" and I have export completely disabled.  As I only have a basic mechanical meter running it backwards is considerd a criminal offense of "utility tampering".  The electric operator is aware of what Im doing and have said it does not need approval.  However, no export and no "grid tied" operation.  it treats the grid exactly like a caravan would treat shore power and even then, only when absolutely necessary.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Can anything be done about AC load "ripple"?
« Reply #9 on: July 07, 2023, 10:06:44 am »
Current noise and power noise is because the consumption is noisy. Producer (inverter) has to accurately generate this "noise" in current and power, otherwise the voltage would get noisy and that's something we don't want.

I suppose you are running computers. Sudden changes of up to tens of W are to be expected, CPU workloads are like that. Electrolytic capacitors on motherboards and PSUs only store energy for tens to hundreds of milliseconds.
 
The following users thanked this post: tom66, newbrain, paulca

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #10 on: July 07, 2023, 11:23:54 am »
... and any attempt to store even more (on the DC side) will come with it's own worse problems.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #11 on: July 07, 2023, 11:33:05 am »
On load response...  it's rather amusing what happens when it's running the 250-300W office and I hit it with the 300W garden strimmer.

Basically, I hit that button and the inverter goes "Nope, no f'ing way dude" and drops snaps the transfer switch closed pass-through'ing the grid.

However, very quickly after that it comes back up, transfers out the grid again.  It's just that huge VA spike from the cold static motor start up that freaks it out.  Thanksfully it does "accept AC" at that point instead of tripping out.  Through out the on/off, on/off cycle of the strimmer the inverter dropped in and out of grid a few dozen times, just using the grid to take the load spikes when the strimmer went from stop to run.  Impressively it did not need to accept grid to dump the -VA coming back on spin down.  That appears to have gone to the battery.

The only "miss performance" comes when you monitor the AC voltage live and introduce a 250W load suddenly.  That load is probably more like 600W for a quarter of a second.  The neon indicators almost blink out and a multimeter is even fast enough to show the voltage ramping from 180V to 230V over about 2-3 seconds.

I suppose that is expected.  Unlike the grid it doesn't have that "near ideal voltage source" property, it will take time to respond.  The fact the voltage bounces very rapidly and nothing crashes or reboots is a testament to it.  Also the slow ramping from 180 up is most likely deliberate to prevent overshoot, especially if it's a short transient spike from an inductive load that will disappear just as quickly.  Undervoltage drops are usually far less damaging than transient over-voltages.

The only times I have plunged myself into darkness has been my own fault.

1.  Testing how low the battery would go.  Couldn't get the remote config re-uploaded fast enough to stop a low voltage disconnect while "Ignore AC-In" was active.  BANG- blackness.... oops.  That one I had to get dressed and go out in the rain to the garage to turn the battery back on and bring the system back up.  The cons of being your own micro-grid operator...  you get the call at 3am to fix it in the rain.

2.  Top balancing a battery using the inverters build in charger (reversing the inverter).  The unit is apparently not able to cope with the battery terminating charge input sudden while it's feeding it with 16 Amps AND running a 350W inverted load.  I imagine those 16 amps current in the transformer coil has suddently nowhere to go and instead of letting it spike the ACOut well over voltage it has no choice but to trip out completely.  BANG darkness.  This one did self recover though, the power came back on 1 minute later automatically.
« Last Edit: July 07, 2023, 11:43:50 am by paulca »
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Can anything be done about AC load "ripple"?
« Reply #12 on: July 07, 2023, 01:22:34 pm »
... and any attempt to store even more (on the DC side) will come with it's own worse problems.

Well, I think you could quite safely increase the capacitance of the 400V (PFC) electrolytic capacitor in PC PSU, to make that tens of milliseconds into... maybe just a hundred. (This is the best place because it's the only one where voltage ripple is acceptable and normal, and capacitor energy capacity is directly proportional to delta V, so can't store a lot if you just parallel capacitors on some 12V bus which is always regulated to 12V.)

But the big question is, why? Everything's working just as expected. Power changing is natural and causes no issues whatsoever. If the inverter is too small to handle those say 500ms peaks, then by far the easiest is to get a bigger inverter. Or if the battery can't handle them, get a bigger battery.

Best place to store energy is that battery DC side of the inverter; batteries are like 1000 times more energy dense than capacitors.
« Last Edit: July 07, 2023, 01:25:24 pm by Siwastaja »
 
The following users thanked this post: nctnico, paulca

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #13 on: July 07, 2023, 06:43:17 pm »
Bigger inverter.  Indeed.  I downsized due to inavailable stock on a 1600VA inverter from another brand.  800VA from Victron will do 800VA until it overheats and then it will drop the voltage ... if the load responds by increasing current, it trips.  The cheap 1600VA, reviews see trips outs frequently for half of that.  Victron 800VA is thermal, but "realistic temps" and smooth derate.  cheap brands, a 1600VA will probably not even run 800W for longer than 5 minutes before it overheats and trips.  (Small/multi transformers, dozens of mosfets and parallel inverter sections).

So I got the right brand, or rather I spent the right amount.  Half the VA, same price, thus better quality.

i 'should' have put the purchase off a month and bought the bigger victron 1600VA  Only £250 more..

the most annoying thing ... I cannot run the coffee machine on it!  In fact, outside of the office and the multitude of small consumers ... all my loads are momentary, short term and 1.5-3kW.

I am weighing between, going self "project funded/managed" and source my own parts and labour to extend this system onto the main roof for solar panels ~3kW and into 10kWh+ of DIY lithium + 4kW hybrid inverter.... then try to get that "notified" by an expensive spark (Like andy's off grid garage)... or...  put a normal "package deal with finnace" grid tie system up there with 0% finance (hopefully) and ...  keep the two separate... although ican run the battery charger on the "off grid" system when there is "house solar" available.  An anti export cross feed if you will.
« Last Edit: July 07, 2023, 06:45:50 pm by paulca »
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Can anything be done about AC load "ripple"?
« Reply #14 on: July 08, 2023, 05:57:19 am »
Wire two identical coffeemakers in series, so that coffeemaking power is reduced to half (quarter per coffeemaker), which your inverter could handle, making double the amount of coffee in 4x the time. Then drink all that coffee and come up with more interesting threads here.
 

Offline gf

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: de
Re: Can anything be done about AC load "ripple"?
« Reply #15 on: July 08, 2023, 06:50:36 am »
Top balancing a battery using the inverters build in charger (reversing the inverter).  The unit is apparently not able to cope with the battery terminating charge input sudden while it's feeding it with 16 Amps AND running a 350W inverted load.  I imagine those 16 amps current in the transformer coil has suddently nowhere to go and instead of letting it spike the ACOut well over voltage it has no choice but to trip out completely.  BANG darkness.  This one did self recover though, the power came back on 1 minute later automatically.

Strange :-//, because

* Battery charging usually does not terminate abruptly, but current decreases gradually when the battery becomes full. At the end of the absobtion stage, the battery current should be already pretty low. [ If the BMS happened to turn off unexpectedly, you should check whether your charger settings possibly violate the battery/BMS specification. ]

* While the battery is AC-charged from mains, the transfer switch relay is closed and the load is powered from mains, too. The load cannot be powered from inverted battery power at the same time. The Victron inverter/chargers are not double conversion (AC -> DC -> AC) devices, like many UPS, but the bidirectional inverter either acts as inverter, or as charger, but not both at the same time.
« Last Edit: July 08, 2023, 06:56:03 am by gf »
 

Offline paulcaTopic starter

  • Super Contributor
  • ***
  • Posts: 4364
  • Country: gb
Re: Can anything be done about AC load "ripple"?
« Reply #16 on: July 08, 2023, 09:57:30 am »
A single cell reached 3.650V before the pack was at the absorb voltage.  The BMS ended charge.

You are right, in charger mode the AC is passthrough.  It still doesn't help it with the current coming through the transformer.

However, it is capable of inverting while running on ACIn.  It has a weak AC boost feature.  This is used by boats when they come up to very limited shore power, like some might only be 10 Amps.  While charging the battery from that and you switch on a 20 Amp stove, the current will switch to coming out of the battery to reinforce the AC.  As this involves grid-tie operation, even though it shouldn't export, I have avoided it... I don't need that feature and I haven't put "grid codes" into it, the tick box is there is config even with no grid codes, so maybe it is able to operate in some way that it guarantees no export.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline gf

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: de
Re: Can anything be done about AC load "ripple"?
« Reply #17 on: July 08, 2023, 10:42:34 am »
However, it is capable of inverting while running on ACIn.  It has a weak AC boost feature.

Sure, but then the inverter does not charge, but discharge the battery.
[ Only an additional charger (e.g.a solar DC charger) could then charge the battery simultaneously. ]

Quote
You are right, in charger mode the AC is passthrough.  It still doesn't help it with the current coming through the transformer.

The passthrough current does not flow through the transformer. In bypass mode, AC input and AC output are directly connected via relay. Only the AC current flowing into or out of the bidirectional inverter passes through the transformer.
« Last Edit: July 09, 2023, 07:10:03 am by gf »
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Can anything be done about AC load "ripple"?
« Reply #18 on: July 08, 2023, 11:26:59 am »
Interestingly, if I look at the voltage chart, it does not have that noise.  it's 230V, 50Hz stable.  it droops and spikes with high load changes, but it doesn't have the noise.

This makes sense as it's load driven noise, the power requirements are altering, probably constantly, the graph resolution is only 0.2Hz tops.

So if voltage is stable and the VA is bouncing it means the current is bouncing.
Did you ever consider the load isn't constant? Just look at your process monitor on your computer. You'll see it jumping around which translates to more or less power consumption. There is nothing wrong with your setup. The problem is you not really understanding how electricity works  :box:
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2852
  • Country: us
Re: Can anything be done about AC load "ripple"?
« Reply #19 on: July 10, 2023, 01:45:21 am »
I need an AC capacitor and I know they don't exist.  How do you store alternating current?  With an alternator, aka a flywheel.  I don't think a ton block axle and some AC induction motors/generators is worth it though LOL
Well, a great big inductor in series with a big capacitor will store energy at its resonant frequency.
Jon
 
The following users thanked this post: EPAIII

Offline EPAIII

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Can anything be done about AC load "ripple"?
« Reply #20 on: July 10, 2023, 10:50:46 am »
"Clock Time" What are you, a politician who answers a question without giving any real facts?

So, the X axis numbers are HOURS and MINUTES?????

That would seem to say the "spikes" we see are changes over time periods of somewhere between 5 and 30 seconds.

OK, another question: At what rate, IN SECONDS, is this data being sampled?

If the sample rate is also in the order of a few seconds to perhaps 30 seconds, then those jumps are only changes in the load over that time period and they mean ABSOLUTELY NOTHING in terms of noise.

It is important to understand what we are actually looking at.



The X axis is "clock time".

Y axis is Watts and VA
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline EPAIII

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Can anything be done about AC load "ripple"?
« Reply #21 on: July 10, 2023, 11:04:19 am »
That's what I am expecting. Except I would say you don't understand what your graph actually means.



Interestingly, if I look at the voltage chart, it does not have that noise.  it's 230V, 50Hz stable.  it droops and spikes with high load changes, but it doesn't have the noise.

This makes sense as it's load driven noise, the power requirements are altering, probably constantly, the graph resolution is only 0.2Hz tops.

So if voltage is stable and the VA is bouncing it means the current is bouncing.
Did you ever consider the load isn't constant? Just look at your process monitor on your computer. You'll see it jumping around which translates to more or less power consumption. There is nothing wrong with your setup. The problem is you not really understanding how electricity works  :box:
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline EPAIII

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: Can anything be done about AC load "ripple"?
« Reply #22 on: July 10, 2023, 11:06:27 am »
Resonant circuit stores AC. YES! Beat me to it.

But that is not what is needed here. What is needed is an actual understanding of what his graph means.



I need an AC capacitor and I know they don't exist.  How do you store alternating current?  With an alternator, aka a flywheel.  I don't think a ton block axle and some AC induction motors/generators is worth it though LOL
Well, a great big inductor in series with a big capacitor will store energy at its resonant frequency.
Jon
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Can anything be done about AC load "ripple"?
« Reply #23 on: July 10, 2023, 04:21:36 pm »
I need an AC capacitor and I know they don't exist.  How do you store alternating current?  With an alternator, aka a flywheel.  I don't think a ton block axle and some AC induction motors/generators is worth it though LOL
Well, a great big inductor in series with a big capacitor will store energy at its resonant frequency.
Jon

Indeed, this is how ferro-resonant transformers work.
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf