shunt needed for victron bmv 501

[Pykebird]

I need to buy a shunt for the battery monitor I have had fitted. How do I work out which one I need and where can i get one please. I understand I need this to ensure the reading is acurate?(I will not be fitting it)

I have a 3kw pure sine wave, sterling invertor.

Thanks in advance.

[frangar]

You need a shunt rated for more amps than you will draw and a 50mV rating so I would suggest a 500A 50mV shunt

 

HTH

 

Gareth

Edited November 6, 2012 by frangar

[jenlyn]

I need to buy a shunt for the battery monitor I have had fitted. How do I work out which one I need and where can i get one please. I understand I need this to ensure the reading is acurate?(I will not be fitting it)

I have a 3kw pure sine wave, sterling invertor.

Thanks in advance.

What frangar said, that's what I have on mine. Though it came with the BMV?

[Chris Pink]

You need a shunt rated for more amps than you will draw and a 50mV rating so I would suggest a 500A 50mV shunt

 

HTH

 

Gareth

 

 

Not a suggestion, a requirement. Any other current rating will give wrong readings.

 

500A 50mV is supplied as standard.

[smileypete]

Not a suggestion, a requirement. Any other current rating will give wrong readings.

 

500A 50mV is supplied as standard.

Try ebay looks like they start at £30ish. If you're on a tite budget there may be cheaper ways (but more hassle.)

[by'eck]

Not a suggestion, a requirement. Any other current rating will give wrong readings.

 

500A 50mV is supplied as standard.

 

Theoretically a 250A 25mV shunt would allow accurate readings as well, accepting the current limit, though this is rather academic.

Edited November 6, 2012 by by'eck

[smileypete]

Theoretically a 250A 25mV shunt would allow accurate readings as well, accepting the current limit, though this is rather academic.

A 500A 75mV shunt would read 50% too high, and so the voltage across it would need dropping by a third, how best to do that?

 

One way would be 3 identical small value resistors in series across the shunt, taking the meter input across two of them.

 

Or even a 90cm length of 1mm² wire with a crimp spliced 30cm in from the end, connect across the shunt and take the meter input across from the larger 60cm portion.

 

Maybe worth it for a twenty odd quid saving.

 

cheers, Pete.

~smpt~

[frangar]

Surely the only requirement for a accurate reading is the 50mV part....the current rating is only relevant if the current flowing through it is more than the rating as essentially the meter is measuring the voltage drop over the shunt? e.g. if 100A is going thro the shunt as long as the shunt is rated at 50mV it doesnt matter to the meter if its 250A shunt or a 500A shunt?

 

Cheers

 

Gareth

[Gibbo]

One way would be 3 identical small value resistors in series across the shunt, taking the meter input across two of them.

 

No. Just no.

 

OP. As others have said, the correct shunt is a 50mV/500 amp. But it comes supplied with the meter. Why didn't you get one?

 

Surely the only requirement for a accurate reading is the 50mV part....the current rating is only relevant if the current flowing through it is more than the rating as essentially the meter is measuring the voltage drop over the shunt? e.g. if 100A is going thro the shunt as long as the shunt is rated at 50mV it doesnt matter to the meter if its 250A shunt or a 500A shunt?

 

A 50mV/250A shunt will drop 50mV with 250 amps going through it.

 

A 50mV/500A shunt will drop 25mV with 250 amps going through it.

 

They can't both be right because the meter will display 250 amps with the correct shunt and 500 amps with the wrong one.

[frangar]

 

A 50mV/250A shunt will drop 50mV with 250 amps going through it.

 

A 50mV/500A shunt will drop 25mV with 250 amps going through it.

 

They can't both be right because the meter will display 250 amps with the correct shunt and 500 amps with the wrong one.

 

 

Ok....but....

 

With the Adverc DCM I have fitted to my boat...before I really understood such things...... I have different Amp values of shunts on the different "loads" eg the circuit measuring the inverter has a 300A shunt where the ones measuring the domestic supply and alternator are 100A... I thought they were all 50mV I suppose they might not be...they were the ones supplied by Adverc so I guessed they matched the voltage to the amp rating to ensure the meter was accurate...which according to my calibrated clamp meter it is!

 

I've not really thought about it before but I see what you say makes sense.....as an aside on the Victron VE Bus system you can now say what the mV of the shunt is...its still a lot more "bitty" than the Mastervolt unit tho!

 

I therefore was right in my first post but stand corrected in my second!...You really never stop learning however much you think you know!

 

Cheers

 

Gareth

[Gibbo]

I thought they were all 50mV I suppose they might not be...they were the ones supplied by Adverc so I guessed they matched the voltage to the amp rating to ensure the meter was accurate...which according to my calibrated clamp meter it is!

 

The Adverc uses 100mV/100A shunts. Or 200mV/200A shunts. etc. The losses are too high for most high power systems.

[smileypete]

No. Just no.

 

:lol:

 

Truthfully, I don't mind what the OP uses, or whether you explain why you said that.

 

cheers, Pete.

~smpt~

[Gibbo]

Truthfully, I don't mind what the OP uses, or whether you explain why you said that.

 

Well do you want an argument about it (which you'll lose) or do you actually want to learn something?

[nicknorman]

Well do you want an argument about it (which you'll lose) or do you actually want to learn something?

I'd be interested to know your justification because Pete's suggestion seemed reasonable to me assuming the sensing device has high input resistance / insignificant input bias current, as a decent DVM would. Maybe that is the issue (ie it doesn't)?

Edited November 7, 2012 by nicknorman

[Gibbo]

I'd be interested to know your justification because Pete's suggestion seemed reasonable to me assuming the sensing device has high input resistance / insignificant input bias current, as a decent DVM would. Maybe that is the issue (ie it doesn't)?

 

The easiest way to explain it is to ask you to design a circuit that will do the required job, then once you've got that cracked (if you can even do it - that's not meant as an insult - just the knowledge that it's nowhere near as easy as most people believe). Then do what Pete suggested and take a look at the mess it makes.

 

Spice is perfectly acceptable. You don't have to actually build one.

 

You'll be in for a shock

[frangar]

The Adverc uses 100mV/100A shunts. Or 200mV/200A shunts. etc. The losses are too high for most high power systems.

 

Can you explain what you mean by losses being to high? Surely this would only become a problem on a 500A & above shunt and even then only on the load side. The charging side would be compensated by the sense wire if a controller such as an Adverc was fitted?

 

I'm not trying to be clever it's just something I haven't given much thought to before and might help explain a few readings I get from the system.

 

Cheers

 

Gareth

[Pykebird]

Cheers guys, I will look on ebay, now I know what to look for! do not mind paying for the right piece of kit, the monitor did have a shunt, but was "lost" under the ton of other stuff that was stored! So in the great scheme of things, £30, is ok.

[Gibbo]

Can you explain what you mean by losses being to high? Surely this would only become a problem on a 500A & above shunt and even then only on the load side. The charging side would be compensated by the sense wire if a controller such as an Adverc was fitted?

 

I'm not trying to be clever it's just something I haven't given much thought to before and might help explain a few readings I get from the system.

 

Ok. Take the shunt used on the BMV501 (so we're staying on topic ). It's a 50mV/500 amp shunt. Switch a 200 amp load on. The shunt drops 20mV. The load is now receiving 0.02 volts less than it would have been without the shunt. The shunt is dissipating 4 watts in heat as wasted power. Pretty insignificant in the whole scheme of things.

 

Now take the 200mV/200 amp shunt used with the Adverc. It's a 200mV/200 amp shunt. Switch the same 200 amp load on. The shunt drops 200mV. The load is now seeing 0.2 volts less than it would without the shunt. That's a fifth of a volt. A substantial amount of loss. The shunt is dissipating 40 watts as wasted heat. That's a lot. It's like having an additional 3 amp load on the 12 volt system. It's just being wasted.

[by'eck]

Surely the only requirement for a accurate reading is the 50mV part....the current rating is only relevant if the current flowing through it is more than the rating as essentially the meter is measuring the voltage drop over the shunt? e.g. if 100A is going thro the shunt as long as the shunt is rated at 50mV it doesnt matter to the meter if its 250A shunt or a 500A shunt?

 

Cheers

 

Gareth

 

Since both current and millivolts dropped across the shunt (which equate to a specific resistance) are relevant and implicit to the correct display of amps flowing through it, your answer is wrong. This is basic Ohms law stuff I=E/R

 

The only variables are if you choose one with the same resistance but a different current rating i.e. 500A 50mV same readings as 250A 25mV same readings as 100A 10mV within the current limits, although this is probably academic.

[nicknorman]

The easiest way to explain it is to ask you to design a circuit that will do the required job, then once you've got that cracked (if you can even do it - that's not meant as an insult - just the knowledge that it's nowhere near as easy as most people believe). Then do what Pete suggested and take a look at the mess it makes.

 

Spice is perfectly acceptable. You don't have to actually build one.

 

You'll be in for a shock

Easiest for you! Not having ever used Spice in anger (gave up professional electronics in 1987) and as you rightly say not having the knowledge to design such a circuit and get it right first time, and certainly not the spare time, it would be easiest for me if you explained it!

Edited November 7, 2012 by nicknorman

[Gibbo]

Easiest for you! Not having ever used Spice in anger (gave up professional electronics in 1987) and as you rightly say not having the knowledge to design such a circuit and get it right first time, and certainly not the spare time, it would be easiest for me if you explained it!

 

Right, take the 50mV/500A shunt used with the BMV501. That meter resolves and displays down to 0.1 amp. In order to stand a chance of doing so it needs to aim to measure down to 10mA so that what it's trying to measure is actually more than its own noise and offsets.

 

10mA on a 50mV/500A shunt is 1uV. It has to reliably measure 1uV on a shunt that is probably wobbling round with a few hundred millivolts on it, in both directions. It can only stand a chance of doing this with a differential input. It has to get this 1uV up to something it can actually measure. Say 0 to 3 volts. Assume a 12 bit A/D and range switching on the amplifier. To measure down to 1uV the gain therefore needs to be around 800 with a maximum input offset of 900nV and an offset drift substantially less than that. That in itself is a tall order but just about doable with modern zero drift op amps.

 

The differential input resistors will have to be 0.01% tolerance at worst, anything more than that and it doesn't stand a chance of working.

 

Now add Pete's resistors on the shunt and see what happens. The common mode rejection will plummet from the required 120dB or so down to about 60dB. 1000 times not good enough. And that's assuming he intended to use extremely low value resistors: about 1 Ohm each. As the shunt voltage, with respect to DC negative, wobbles around, the common mode voltage will get amplified so much that it totally swamps the wanted differential voltage.

 

It absolutely will not work.

[smileypete]

Now add Pete's resistors on the shunt and see what happens. The common mode rejection will plummet from the required 120dB or so down to about 60dB. 1000 times not good enough. And that's assuming he intended to use extremely low value resistors: about 1 Ohm each. As the shunt voltage, with respect to DC negative, wobbles around, the common mode voltage will get amplified so much that it totally swamps the wanted differential voltage.

Just put the meter input I- (analogue ground) directly to the shunt load side where the meter - (digital ground) is connected. Then put the meter input I+ across two (away) of the resistors, easy really.

 

cheers, Pete.

~smpt~

[Chris Pink]

Just put the meter input I- (analogue ground) directly to the shunt load side where the meter - (digital ground) is connected. Then put the meter input I+ across two (away) of the resistors, easy really.

 

cheers, Pete.

~smpt~

 

why low value rather than high value resistors Pete? would you not want to take less of the current?

[Gibbo]

Just put the meter input I- (analogue ground) directly to the shunt load side where the meter - (digital ground) is connected. Then put the meter input I+ across two (away) of the resistors, easy really.

 

Everything is easy... until you learn something about it. Read what I actually wrote, then go and spend a few days studying until you understand it.

 

Better still, I know you have LTSpice, get that out and design a circuit for me to do this job. You'll eventually realise where you're going wrong.

[Pykebird]

Thank God I didn't ask a difficult question!