question

ceriw avatar image
ceriw asked

Quattro buzz/rattle

I have a pair of Quattro 8k in parallel operation. They have started making a loud buzz/rattle sound. I've isolated the mains inputs and outputs but it still does it so can't be any bad load/poor Mains etc.

It's really loud and distracting when trying to sleep.

I've attached a sound clip of it.

Anyone got any ideas?

ThanksVID_20190722_210857.mp4

Ceri

offgrid
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

4 Answers
Kristof G. avatar image
Kristof G. answered ·

Hi,


I had that too on my 48v 8000 quattro , one time! But the mains whent down and than he started that noise... (also i have solar on my output with feed-in = think that was the problem for me) didnt hear it since.

But you have disconnected mains and at same time no loads connected or solar on output?, so then.. i dont have a solution.


Will follow this up...maybe i have it again in the future.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

ceriw avatar image
ceriw answered ·

I've been working on this for a few weeks now. I've striped my battery pack down and changed its configuration from 16s6p to 14s8p so am very confident there are no loose connections there.

As an experiment, I've enabled "has DC system" on the Venus GX. The only thing connected to DC is the REC BMS and the Venus GX.

When the Quattro units make the loud buzz, the system reports a spike on the DC system. Please see attached screenshot.

I'm beginning to wonder if one of the Quattros is faulty. Any opinions or things to try to rule this out?


1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

ceriw avatar image ceriw commented ·

Screen shot of dashboard showing phantom DC Power whilst units are buzzing

0 Likes 0 ·
1566033545006.png (114.1 KiB)
Daniël Boekel (Victron Energy Staff) avatar image
Daniël Boekel (Victron Energy Staff) answered ·

Hi @Ceri.w

Do you have pictures of the installation? especially the wiring?

13 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

ceriw avatar image ceriw commented ·

Hi, thanks for the response.


Complete system. BMS and Venus mounted on reverse wall.


Top row of batteries and covers removed. Back right is 500A fuse connected to battery with 6 x

25mm² cables. It is then connected to contactor with 4 x 35mm² cables. Each Quattro then has 2 X 35mm² cables to it's input.

On the negitive side, each Quattro has 2 x 35mm² cables to the current shunt. The current shunt is connected to the negitive of the battery by 6 x 25mm² cables. Battery pack is linked in the centre by an Anderson connecter (in parallel) wired both ends with 6 x 25mm² cables.


An internal shot of the bottom battery pack before it was fitted showing the bus bar layout. 50mm x 6mm aluminium bar was used.


Are there any areas you want more details about?

Many thanks

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ ceriw commented ·

Hi @Ceri.w yes, the AC-wiring is just as important, see:

https://www.victronenergy.com/live/ve.bus:manual_parallel_and_three_phase_systems#dc_and_ac_wiring


I notice you use a contactor, what do you use to pre-charge?

0 Likes 0 ·
ceriw avatar image ceriw commented ·

Its a bit harder to get a photo of the AC. The grid comes into the house downstairs and goes through an RCD and then 2 mechanically coupled 40A MCBs. It then goes from there to the Quattros upstairs using 2 lengths of 16mm2 twin and earth (6m long) and connects to AC input 2. Output 1 of each Quattro is connected by a 16mm2 twin and earth cable to a single RCD on the consumer box downstairs (6m long). the 4 16mm2 cables are all within 10cm length of each other.

I also have AC 1 inputs connected to the generator shed via 2 lengths of 2.5mm2 twin and earth. These are fed from 2 mechanically coupled 16A MCBs and fed from a single 4mm2 feed from the generator. The generator has not been commissioned yet though.


For the pre-charge on the contactor, I use a REC pre chage resistor. https://www.rec-bms.com/datasheet/UserManualPrecharge.pdf



0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ ceriw commented ·

the point I'm trying to make: these sort of things usually have to do with AC (and DC) wiring, did you read the document I linked to?

0 Likes 0 ·
ceriw avatar image ceriw commented ·

Hi @Daniël Boekel (Victron Energy Staff), yes i have read the documents you linked to. I believe my system meets the guidelines. If i fit some current clamps to the AC in and AC out of the 2 Quattros, i'll be able to see how well they are load balancing. I could also run them one at a time in stand alone mode and see if that still shows the noise and phantom DC charging. Do you know of any other diagnostics I could do?

The big clue seems to be the phantom DC power- every time this happens, you can hear the transformer flux field density change in the Quattros..

Thanks for your help

Ceri

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ ceriw commented ·

The fantom DC power is usually a result of a delay in the power / current reports, and that's why you don't want to enable this setting if there are no DC users.


DC current is a calculated value, so not to be used as an exact source.

measuring in a parallel system is not difficult, just measure the differences between the AC-in connections with a multi-meter on all clamps. while a big load is active.

0 Likes 0 ·
ceriw avatar image ceriw commented ·

Scope across the 500A 50mV current shunt

0 Likes 0 ·
ceriw avatar image ceriw commented ·

Hi @Daniël Boekel (Victron Energy Staff),

AC input and AC output are current sharing to within approx 3% so I don't think it is that giving me any issues.

I connected a battery powered scope across the 50mV / 500A current shunt and attached the screenshots.

During this time, the Victron Console was reporting charging the battery at 5A. As you can see, the AC noise is well over 100A even at 5A charging current and causing the batteries to make the buzz/rattle. I've isolated the batteries from the system as now I know what the noise is - I can see it destroying the batteries pretty quickly.


Do you have any suggestions on what to do next?

Thanks

Ceri



0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ ceriw commented ·

Hi @Ceri.w

I'd first swap master / slave Quattro's (this is only a software change)

if that doesn't change anything: isolate the Quattro's, test one at the time as stand-alone

this way you can systematically find the point where it goes wrong.

0 Likes 0 ·
ceriw avatar image ceriw Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi,

I swapped master slave around and it made no difference. I have now removed the AC outputs from both units and powered them individually in stand alone mode.

When the battery is charging at 1.5A, there is still 6A of 50hz AC flowing through the battery. This is identical on both units. I have confirmed my scope measurements of the current shunt with an AC clamp multimeter connected to the battery cable.

What, if any, AC current would you expect to see flowing through the battery when charging at 1.5A?


Thanks Ceri


0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ ceriw commented ·

Hi @Ceri.w

There is always a 100 Hz (or 120 Hz on 60 Hz systems) ripple on the DC side.

I'd be more interested in the numbers on a higher load.

And the DC ripple (also readable when you connect the inverter to a laptop with VE.Configure) is an interesting value, saying something about the battery resistance and wiring.

I've seen that your system made some big 'feed-in' events from battery to grid in the past, this is usually due to wiring of configuration errors, but can of course also be due to a defective unit.

I'm interested in seeing if this stops happening with the inverters not parallel.

On noise coming from the batteries: this I haven't seen (heard) before, can you try to find out where exactly this noise is coming from? (housing moving due to magnetic effect?)

0 Likes 0 ·
ceriw avatar image ceriw Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Hi, thanks for the quick reply.

The DC ripple when the unit is idle is being reported as 0.2v according to the laptop. This seems quite high given the very low impedance of the battery pack. What range would you expect to see on a 3milliOhm battery/cable impedance?

The noise is coming from the steel lids of the battery cases moving with the change of flux comming from the AC. If you touch it, you can feel it vibrating and the more I load the system, the louder it gets.


I'll monitor the ripple on higher loads in stand alone mode and send you my findings.


Thanks Ceri

0 Likes 0 ·
ceriw avatar image ceriw ceriw commented ·

Udc ripple does not change from 0.2V on either unit in stand alone mode - even loading it up to 5kW. One thing that is interesting is with a kettle and a florescent light on it, the florescent light keeps going on and off. The UOut is set at 235V, but when the units make their extra loud buzz, goes to 238V. This happens independently on both units...

0 Likes 0 ·
kai avatar image
kai answered ·

Hi @Ceri.w

This is a bit of a hunch. If you have the configuration software, are you able to temporarily install the silent fan assistant and see what the result is?

(btw if this works I'm not suggesting its a permanent solution, the assistant is just to localise the potential issue)


2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

ceriw avatar image ceriw commented ·

Hi @kai, It doesn't seem to be related tot eh fans at all - they work ok with load and are not noisy etc. Does the fan assistant cause any other "side effects" that are useful to help diagnose?

Many thanks

0 Likes 0 ·
kai avatar image kai ♦ ceriw commented ·

I experienced a (possibly) similar scenario. My hypothesis is that the fan PWM was causing some interesting accoustic side effects. The sound I heard is similar and doesn't appear to be emanating from the fan (I was having trouble localising it at all, seems to be coming through the whole enclosure), but when I turned the fan off it disappeared.

0 Likes 0 ·

Related Resources