Can I use 5W wire wound resistor as a shunt?

Thread Starter

markzz

Joined May 25, 2016
16
I have a BOM that requires a wire shunt, 0.005ohms and 5W, I can buy a wire wound power resistor that is 5W and 0.1 ohms. I could place some in parallel to achieve the 0.005ohms. Will this work or not. The circuit is a current sense circuit that is placed on the output of a power supply like a meanwell. Very simple circuit.
 

ErnieM

Joined Apr 24, 2011
8,196
Yes, a very simple circuit but very dependent on some subtle details.

First, to get .005 ohms from 0.1 ohm resistors you need to place 20 resistors in parallel. and you need to get them exactly balanced so they don't build up and resistance so the total is still in spec, and you have points that accurately represent exactly where this resistance is.

Simply put, you will never get there even with exotic instruments such as mili-ohmmeters.

What to do? Buy a current shunt.

Current shunts are large resistors made to pass large currents with little loss. They have terminals for the current and terminals for the voltage sense points.

Here is one from EBay:

https://www.ebay.com/itm/Shunt-Resi...h=item4acacf0b5b:g:6AgAAOSwzppaRO1Q:rk:7:pf:0
 

Reloadron

Joined Jan 15, 2015
6,206
What you really want, is as mentioned, a current shunt. Here is an example below:
50 Amp Shunt.png

Since it is a 50 Amp = 50 mV we have a .001 Ohm resistance between the sense marked terminals. New current shunts can be had pretty inexpensive as they flow in from China.

Ron
 

oz93666

Joined Sep 7, 2010
737
I have a BOM that requires a wire shunt, 0.005ohms and 5W, I can buy a wire wound power resistor that is 5W and 0.1 ohms. I could place some in parallel to achieve the 0.005ohms. Will this work or not. The circuit is a current sense circuit that is placed on the output of a power supply like a meanwell. Very simple circuit.
Why go to all that trouble ??? what wire do you have in your scrap box ??? for example if you have some 4mm square copper wire , that has a resistance of 4.61 Ohms /Km ... so 1 meter will be .00461 Ohms ... you need 1.08 meters .... wind it into a compact coil.

If the wire you have is thinner , put multiple coils in parallel , since you will need a fair sized surface area on the wire to dissipate the heat.

The resistance of copper does increase with temperature (only 1% every 3C rise) .. if you calculate the resistance at 35C that should allow for the slight heating from the current.

This will be much more accurate than using standard resistors , which typically have an accuracy (tolerance) of 5% !!!
 
Last edited:

oz93666

Joined Sep 7, 2010
737
What you really want, is as mentioned, a current shunt. Here is an example below:
Since it is a 50 Amp = 50 mV we have a .001 Ohm resistance between the sense marked terminals. New current shunts can be had pretty inexpensive as they flow in from China.
Ron
Look at the printing on the shunt itself ...50A ...50 MV Mega volts ... the times I see M when it should be m ... I doubt the owners of the company who made the thing even understand the difference !

I'd be interested to know what material that shunt is made from (the black part) .. if copper as I suspect , the resistance will change substantially under different currents since it's surface area is small so it's temperature will get hot.
 
Last edited:

Thread Starter

markzz

Joined May 25, 2016
16
Thanks for the replies.

I will use the wire method for the shunt.
3ft of 12awg from the resistance calculator. Should I use solid or stranded?


For wire I have everything from 10awg, 12, 16 and very thin, I can always buy the wire I need.
The BOM is from mouser, but I want to buy local from B&E Electronics.

Here is the shunt, from mouser.ca
https://www.mouser.ca/ProductDetail/66-OAR5R005FLF

I can find everything at B&E except the shunt and the the exact 512-LM78L12ACZX
however I found a suitable replacement 7812UC from ST MICRO IC REG VOLT POS 1A 12V TO220-3 7812UC
it is a different form factor, but all the specs look close.

A small hurdle is the pcb normally bolts onto the output terminals of the psu. They made their own pcb, but I am just buying through hole bread board. It has a 2 pair hole, will fit the dim



 

oz93666

Joined Sep 7, 2010
737
Thanks for the replies.

I will use the wire method for the shunt.
3ft of 12awg from the resistance calculator. Should I use solid or stranded?4.764
That sounds about perfect , comes out at 4.764 Ohms (at 25C) ... should be 5 Ohms at around 31C .... maybe make it an inch shorter, on reconsideration operational temp might be closer to 40C , assuming room temp is 25C

I would use solid , you don't want any plastic insulation on it inhibiting heat transfer , either bare copper or enameled . if wound around a one inch pipe former it should make a nice self supporting spring like coil ... remove the former , important to have good air circulation .
 
Last edited:

ebp

Joined Feb 8, 2018
2,332
Copper wire makes a very poor shunt if you care about stability. The temperature coefficient of copper is large (0.393%/°C) and self heating can easily produce an error of several percent. This may be acceptable.

What magnitude of current are you actually trying to measure? What voltage do you require from the shunt at full scale?
 

Reloadron

Joined Jan 15, 2015
6,206
Look at the printing on the shunt itself ...50A ...50 MV Mega volts ... the times I see M when it should be m ... I doubt the owners of the company who made the thing even understand the difference !

I'd be interested to know what material that shunt is made from (the black part) .. if copper as I suspect , the resistance will change substantially under different currents since it's surface area is small so it's temperature will get hot.
While when it comes to current shunts I never gave it much thought but I have several laying here. The shunt pictured above was made by GE (General Electric) as seen by the GE logo on the shunt. I also have a S.E.C shunt (Simpson Electric Company) and a Weston Electric Company shunts and all of them use the upper case M rather than the lower case m to denote milli.

As to the actual conductive material there are several in use and popular for their thermal characteristics. This Vishay paper is a good read on the subject.

MISC Shunts 1.png

MISC Shunts 2.png

MISC Shunts 3.png

Both Weston and Simpson Electric as well as GE used the upper case M so go figure?

I doubt the owners of the company who made the thing even understand the difference !
I doubt I would take it that far considering the manufacturers in question.Again, the link I posted to a Vishay paper covers things fairly well.

Ron
 

Thread Starter

markzz

Joined May 25, 2016
16
OK full explanation.

What this device will do, is control the output current of the Mean Well psu. Basically you install one resistor onto the pcb which has a certain value to attain the current you want. So the Mean Well psu has a fixed current and a variable voltage. Replace the one resistor for the current you want and instead of 48V and 20A it can now be 48V and 10A or whatever the corresponding resistor value is. Some Mean Well psu's go full current at first until it reaches the voltage set on the pot, then the current tapers off and the voltage is constant ( CC / CV ) which is what is used in charging lithium batteries.. I have been using MW's for the last few years at 16A and they work like a charm, yes lots of people use this method. Its hard to find reliable well built yet affordable chargers that are not junk and will hold up in a backpack.

The circuit diagram has 2 or 3 choices of selectable parts. Choose a zener diode for the voltage range, choose a resistor for the current range, double up the shunt to double your current (optional).

In my case, I build my battery to suit the charger. If I have a 36V 10Ah battery, I do not want to charge beyond 10A of charge current, to achieve a 1C charge rate. My batteries have changed over the years but I have always been at either 36V or 48V, and the Ah capacity has changed from 20Ah, 15Ah, 10Ah so accordingly I change the chargers. Of course I keep in mind the charge rate for the cells. The Mean Wells I bring along with me so I can charge up anywhere with 120Vac outlet (gas stations are common, starbucks, malls, parking lot block heater outlets [in colder climates like Montana/Alberta]) and the MW's are very stout, of course not meant for vibration or what I do but they have worked very well. Others have used LED supplies, I havent yet. My current setup was putting 2 MW's in series, which can be done right out of the box, no isolation required. I used two 24V MW's to achieve 36V, the range was 20-30V and I needed 42V. Nominal of course.

Not much else to explain other then that.
Issue 1 was the Shunt which I will use wire to achieve the 5W 5mOhm which I know I should get pretty close to for the measurement of current.
Issue 2 was the 12V voltage regulator chip which I found a suitable swap for. I will post it below, everything is close except a few things.
Issue 3 is just the connection method of the board onto the psu which I will just use suitable wire for the amps used/ampacity.

12V Voltage Regulator
Recommended - 100mA output current max, 5/6/8/12/15V with +/-5%, fixed voltage, monolithic IC
https://www.mouser.ca/ProductDetail/512-LM78L12ACZX
spec sheet - https://www.mouser.ca/datasheet/2/308/LM78L12A-1301396.pdf

Swap
https://www.be-electronics.com/ProductDetails.asp?ProductCode=7812UC&CartID=14
Spec sheet - https://www.st.com/content/st_com/e...gulators/standard-voltage-regulators/l78.html

looks close, I will snap pics
 

Attachments

ebp

Joined Feb 8, 2018
2,332
You might be able to find the 78L12 at MRO or at whatever the former Active calls itself now, if you don't mind an expedition to the northeast end of town.

Beware of a shunt that will have moderate inductance in a switcher. It really depends on where it is in the circuit, but it can mess with the frequency compensation. I'd recommend ordering a suitable shunt. Those bare wire types are inexpensive, The LVR03 and 05 series from Vishay are nice.

You could have parts from DigiKey tomorrow for shipping cost of less than a local courier.

EDIT - don't forget the pin sequence for the 7812 and L12 are opposite when viewed with the marked side toward you. I've seen people get burned (not quite literally) by that before.
 

shortbus

Joined Sep 30, 2009
8,959
Wouldn't both a coil of wire or a wire wound resistor act like an inductor, and not a shunt? It is my understanding that is why a meter shunt is carbon. And many of them have saw cuts in them to "tune" their out puts. Or at least the ones I've seen are done that way.
 

Reloadron

Joined Jan 15, 2015
6,206
They make and market 0.005 Ohm metal film resistors for use as current sensors. I have seen 4 watt versions and I would think you can find a 5 Watt version made to do what you want to do. I would even look for two 0.010 Ohm 4 watt and parallel them. I would not be too quick to use a wire wound resistor for the reasons already mentioned, it may or may not matter. I would look for metal film types.

Ron
 

MisterBill2

Joined Jan 23, 2018
9,813
Real shunts, that is, those manufactured to serve as accurate stable shunt resistors, are a lot more than a length of copper wire. The stability of the resistance as the temperature changes matters if the application requires any accuracy. AND the sense connections also matter if accuracy matters. if +/- 15% is OK, and a temperature coefficient of +12% is acceptable then ordinary copper wire, carefully selected and connected is fine. But if any higher accuracy or stability, and reproducability is needed then a decent commercially made shunt is a better choice. But if you just want to trip an overcurrent shutdown then the copper wire is OK, if you calibrate it.
 

ebp

Joined Feb 8, 2018
2,332
shortbus, Yes, wirewound resistors are inductive which makes them unsuitable for higher frequencies. They are generally OK if there is no frequency component above a few kHz. When you get into larger types, there are so-called non-inductive types where the winding reverses direction many times instead of always being the same way around the core. They aren't too bad, but they are more expensive. Lots of low value resistors that are quite good at high(ish) frequency are essentially just straight wires or metal strips. In surface mount low value resistors, some are available with the terminals on the long edges instead of the short ends to get lower inductance.

In a switch mode power supply, the location of the sense resistor matters. If it is in the switching path, it must be low inductance. If it is "outboard" of the filter capacitors, it can be quite inductive without causing problems in most cases.

I agree with MisterBill2 that in the particular application the temperature coefficient of resistance of copper is probably not a big problem. It has the merit that the current would drop with rising temperature.

Manganin is a popular alloy for making shunts that are quite stable. It is a pretty decent compromise between performance and cost.
 

ErnieM

Joined Apr 24, 2011
8,196
Wouldn't both a coil of wire or a wire wound resistor act like an inductor, and not a shunt? It is my understanding that is why a meter shunt is carbon. And many of them have saw cuts in them to "tune" their out puts. Or at least the ones I've seen are done that way.
Not if you specify a non-inductive wire wound resistor.

How would you make such a beast? Well, you start with a suitable length of resistance wire. Then fold it in half.

Wind the halved wire around your bobbin. Since every turn one way has a matching turn the other way the overall inductance is a small fraction of a conventional wind.

See here:
https://www.eetimes.com/document.asp?doc_id=1305529
 
Top