Jump to content

Rectify 3 phase before or after cable run to battery?


Antfarm
 Share

Recommended Posts

Greetings,  I have a 3-phase y-connected 48VAC PMG that I will connect to a 48V battery bank, located about 40m from the turbine.  Must I rectify the current at the generator and then run a 2 core cable, or should I rectify at the battery bank and run a 3-core cable?  I already have the cable, which is surplus from a previous project, being 2 x 40m lengths of 3core+earth. If I were to rectify at the battery bank, I would configure the cable to run 3 circuits in parallel.

Efficiency is important here as it is small turbine and I need to get max power to the battery. The cable I have is aluminium with a diameter of about 4.5mm

Thanks

Link to comment
Share on other sites

Hi 87, I have a water turbine connected to a permanent magnet generator (http://www.motenergy.com/mepmge3.html), which generates 3-phase y-connected 48VAC, and should produce about 600W at the peak water flow available here.  I need to connect it to an existing off-grid solar installation (48V BYD-BBOX batteries, Victron Multiplus).  I plan to connect it directly to the battery bank, and use a tristar to dump excess power.

So I need to run cable from the turbine to the battery banks, which is about 40m away.

I have 2 choices for the cabling:

1) Rectify the current at the generator side and transmit +-54VDC single-phase via 2-core cable to the site of the batteries

2) Transmit the 3-phase power on 3-core cable at +-48VAC to the site, then rectify before connecting to the battery

My question is which would be more efficient, given the cabling I specify above

Link to comment
Share on other sites

 

I would run 3 phase to the battery banks. It will allow you to access the AC to test/measure/monitor the rectifiers. 

I assume that you will be adding some smoothing capacitors after the rectifiers ? Also you would be using a charge controller to prevent overcharging of the batteries ?

Use the thickest cables that you can afford to mimimize the energy lost in the cables. Even if it is a small amount, it will eventually be sizeable assuming the system is in place for a number of years...

Post some pics when you get a chance.

Link to comment
Share on other sites

4 hours ago, Antfarm said:

Thanks Greenman.  That's a good point you make about being able to monitor the AC.  I'll run 2 sets of cable in parallel, with each core being about 6AWG. I still need to get the rectifiers. Any recommendations?

For your 600W generator (well actually an alternator 😀)  you are looking at 12.5A AC max so a 35A 200V bridge would work nicely like this FB3502 - Communica [Part No: FB3502]

I would mount each bridge rectifier on a decently sized heat sink. Or you could mount all 3 onto a big one for ease of mounting. 

Link to comment
Share on other sites

  • 5 months later...
On 2021/07/22 at 10:06 PM, Antfarm said:

Greetings,  I have a 3-phase y-connected 48VAC PMG that I will connect to a 48V battery bank, located about 40m from the turbine.  Must I rectify the current at the generator and then run a 2 core cable, or should I rectify at the battery bank and run a 3-core cable?  I already have the cable, which is surplus from a previous project, being 2 x 40m lengths of 3core+earth. If I were to rectify at the battery bank, I would configure the cable to run 3 circuits in parallel.

Efficiency is important here as it is small turbine and I need to get max power to the battery. The cable I have is aluminium with a diameter of about 4.5mm

I would definitely make the long 40m cable run in AC and then rectify to DC right by the battery using a 3 phase rectifier like this…. 
 

https://www.mantech.co.za/ProductInfo.aspx?Item=72M3140

Link to comment
Share on other sites

Out of interest, for those that might know, any reason you can't just use 3x full bridge rectifiers for something like this?  (ie. this linked one is 1000v @ 50 amps, so you get a lot more bang for buck).   The 3 phase rectifiers look expensive... (probably because they are pretty niche)

Edited by Gnome
Link to comment
Share on other sites

15 hours ago, Gnome said:

Out of interest, for those that might know, any reason you can't just use 3x full bridge rectifiers for something like this?  (ie. this linked one is 1000v @ 50 amps, so you get a lot more bang for buck).   The 3 phase rectifiers look expensive... (probably because they are pretty niche)

Strange, 3 phase rectification requires 6 diodes, a full bridge requires 4, so 3x full bridges have double the number of diodes and should be more expensive. That said, you can do 3 phase rectification with 2x full bridges.

Oh, I see you linked a banggood part. I would be wary of using them for electronic components that are easily faked (or likely does not meet spec).

Here is a 3 phase that is cheaper than the banggood part (but you have to buy 5)

https://za.rs-online.com/web/p/bridge-rectifiers/9179023

Link to comment
Share on other sites

1 hour ago, P1000 said:

Oh, I see you linked a banggood part. I would be wary of using them for electronic components that are easily faked (or likely does not meet spec).

I haven't had issues buying parts from Banggood.  On eBay you can get fakes but what is more likely with a fake is they scratch off the markings.  If it still has the markings on it, there is a point where "faking" it and it working intersect so that it is more expensive to fake it.  For parts that are unbranded, you aren't very likely to get a fake, it isn't worth it.  The stuff that does get faked are usually high volume parts and rarely discrete components like this one.

That said, go ahead and buy from RS (I've bought plenty from both RS and Banggood, never had an issue with parts from either)

1 hour ago, P1000 said:

Strange, 3 phase rectification requires 6 diodes, a full bridge requires 4, so 3x full bridges have double the number of diodes and should be more expensive. That said, you can do 3 phase rectification with 2x full bridges.

Each phase must be on its own full bridge.  If you connect more than one on a full bridge you will have a phase to phase short.  My question was more related to how the dynamo responds to that VS. a 3 phase rectifier that is typically only half bridge per phase.

Link to comment
Share on other sites

4 minutes ago, Gnome said:

Each phase must be on its own full bridge.  If you connect more than one on a full bridge you will have a phase to phase short.  My question was more related to how the dynamo responds to that VS. a 3 phase rectifier that is typically only half bridge per phase.

Look at the schematic for 3 phase bridge and full bridge. You need 1.5 full bridge rectifiers to create a 3 phase rectifier.

Link to comment
Share on other sites

14 hours ago, P1000 said:

Look at the schematic for 3 phase bridge and full bridge. You need 1.5 full bridge rectifiers to create a 3 phase rectifier.

Indeed, but since a full bridge rectifier has 2 legs at the AC side and 2 legs at the rectified side (single chips and not discrete diodes).  How would you connect 3 phases to 2 full bridge rectifiers?

Link to comment
Share on other sites

6 hours ago, Gnome said:

Indeed, but since a full bridge rectifier has 2 legs at the AC side and 2 legs at the rectified side (single chips and not discrete diodes).  How would you connect 3 phases to 2 full bridge rectifiers?

BR1: AC1 to gen PH1, AC2 to PH2

BR2: AC1 to PH3

Tie the DC+ of BR1 and BR2 together and do the same with DC-.

Now you have a 3 phase rectifier with one spare input. If you draw the diagram for the bridges in the correct way, it becomes obvious:

image.png.6dfca2c8012a7f1f12c5dd368872d89f.png

Pin 1,4,3 and 5 form a full bridge. Pin 1,2,5 half of the second.

Link to comment
Share on other sites

Technically DC is more efficient to transmit because you don't get affected by the reactance of your cable and you will only have resistive losses. At this relatively short distance and low voltage the difference between the two options is pretty much negligible in my opinion. Where you will probably see better gains in efficiency is in your rectification semiconductor choice. Look for the lowest Vf you can afford that matches your needs, good thermal dissipation also helps.

Link to comment
Share on other sites

3 hours ago, P1000 said:

BR1: AC1 to gen PH1, AC2 to PH2

BR2: AC1 to PH3

Tie the DC+ of BR1 and BR2 together and do the same with DC-.

Now you have a 3 phase rectifier with one spare input. If you draw the diagram for the bridges in the correct way, it becomes obvious:

image.png.6dfca2c8012a7f1f12c5dd368872d89f.png

Pin 1,4,3 and 5 form a full bridge. Pin 1,2,5 half of the second.

A single chip full bridge looks like this

image.png.d1bfeda4a040ef1012e83d08d588af32.png

So how does that factor into your diagram?  My point before was, if you connect AC1 to 1 of the phases and AC2 to another phase you have an output voltage of roughly (phase voltage x2 / 1.14).  In OPs case ~84 volts (instead of ~48v).  The phases are typically separated by 120°.  If you connect AC1 to 1 of the phases and then parallel another phase to one of the legs you have a phase to phase short (at 84 volts potential).  There isn't any way to connect it without discrete diodes that I can see.

Edited by Gnome
Link to comment
Share on other sites

55 minutes ago, Shadders said:

Technically DC is more efficient to transmit because you don't get affected by the reactance of your cable and you will only have resistive losses. At this relatively short distance and low voltage the difference between the two options is pretty much negligible in my opinion. Where you will probably see better gains in efficiency is in your rectification semiconductor choice. Look for the lowest Vf you can afford that matches your needs, good thermal dissipation also helps.

The voltage and current determine the loss in the cable.  For DC the voltage is equivalent to roughly the RMS AC voltage.  Not only is the rectified voltage going to be lower than the AC peak to peak voltage but the current draw will be roughly 3 times what the 3 phase current would be.  There are good reasons why AC is used for transmission lines (other than being easy to convert).  And yes I'm aware of HVDC but those benefits aren't relevant for a small scale installation using (uncontrolled) rectified AC.

Link to comment
Share on other sites

42 minutes ago, Gnome said:

A single chip full bridge looks like this

image.png.d1bfeda4a040ef1012e83d08d588af32.png

So how does that factor into your diagram?  My point was before that if you connect AC1 to 1 of the phases and AC2 to another phase you have an output voltage of roughly (phase voltage x2 / 1.14).  In OPs case ~84 volts.  The phases are typically separated by 120°.  If you connect AC1 to 1 of the phases and then parallel another phase to one of the legs you have a phase to phase short (at 84 volts potential).  There isn't any way to connect it without discrete diodes that I can see.

Rotate your diagram 90° left, then compare it with mine:

rectif.png.1d02d3fda10d8dceacc51f1503047a5e.png

Not the best schematic, but it should get the idea accross.

Edited by P1000
Link to comment
Share on other sites

10 hours ago, Gnome said:

For DC the voltage is equivalent to roughly the RMS AC voltage.  Not only is the rectified voltage going to be lower than the AC peak to peak voltage

I beg to differ. Your DC voltage is equal (or rather much closer) to the peak voltage not the rms voltage... as long you are using full bridge rectification and of course smoothing it, which it would not make sense to not do and Vdc = Vpeak = Vrms *sqrt(2) so your DC voltage is higher than Vrms (even if we factor in rectifier drops). The way Vrms is defined leads to this confusion as it is defined to be equivalent to the same voltage DC as required to generate the same power when applied to a resistive load.

Your 3 phase current statement is misleading too. AC current draw is less per phase/conductor than the DC yes but it's not a simple factor of 3. Consider the data given of power generated 600W, Vline of 48Vrms as Antfarm is connected in Y - the 3 phase power =sqrt(3)*Vline*Iline and Iline =I phase so doing the maths we find the current per phase = 7A flowing in each of 3 conductors. Consider if we rectify and smooth the AC. DC power = Vpeak*I and using the same 600W and Vpeak=67V, I=9A flowing in 2 conductors. Losses have been ignored to simplify matters but this shows it's not a simple factor of 3.

For transmission lines AC systems require 1 more conductor than the DC and all components must withstand the peak voltage, however as AC power is calculated using RMS this means that effectively because DC lines can run at peak voltage it can supply more power for a given transmission line design. If you factor in reactance and skin effect and not just the resistance DC lines have lower losses for a given current and are hence more efficient.

All that being said at these low voltages I stand by my statement that it doesn't really matter.

Link to comment
Share on other sites

On 2022/01/05 at 10:39 AM, P1000 said:

Rotate your diagram 90° left, then compare it with mine:

rectif.png.1d02d3fda10d8dceacc51f1503047a5e.png

Not the best schematic, but it should get the idea accross.

Aah, they aren't using a neutral line at all, I forgot about that.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...