Cheap current sensing shunt resistor

[Gregte]

The wiring harness from my battery to the controller is 16 gage copper wire. According to wire tables and also to an online calculator 3 inches of 16 ga stranded copper wire would have 0.001 ohms of resistance.

Therefore it would seem that all you would need to do to measure the current, in amps, going thru the wire would be to attach the two leads of a voltmeter to the wire 3 inches apart. The reading in mili volts would be the number of amps.

The reason for detecting current this way as opposed to using a regular mili Ohm resistor is that this method requires no new parts to mount on the bike plus it is free.

Can anyone here see any reason that this would be unsatisfactory? There are minor issues such as temperature variations of resistance of the wire but that must also be true of standard mili ohm resistors I would think.

EDIT: I have since learned that the 3 motor wires are 14 gage, not 16 gage.

 

[fechter]

That works reasonalby well, and I've used it in the past. The connections for the meter need to be soldered onto the main wire so there is no variation in contact resistance there. You can calibrate it by moving the taps. The accuracy is temperature dependent, but most of the time you just want a general indication of current, so accuracy is not a huge issue. The measurment section of wire should be located where it will not be subjected to much temperature change.

 

[Drunkskunk]

By the AWG table, I get 0.0041 Ohms per foot. I don't think you can calibrate the resistance off a table. you'll need to actualy measure the resistance. One batch of drawn wire is going to have slightly diffrent resistance from another's.

I think it could be done. Copper isn't all that thermaly stable, but its close enough if you aren't going to use the measurment for anything critical.

The only real problem I see is.. 16 gauge? on the battery? WTF?
Thats rated at 3.69 amps (AWG rated) to 10 amps (Other sources). 16 guage on the battery side is tiny.

 

[Gregte]

The only real problem I see is.. 16 gauge? on the battery? WTF?
Thats rated at 3.69 amps (AWG rated) to 10 amps (Other sources). 16 guage on the battery side is tiny.

This is my first ebike kit. I just used the wiring that they sent.

Would I see much benefit to change the 3 power sires from the front hub motor to the controller from 16 ga to 14 or 12 gage? What do you recommend? (500 watt controller, 750 watt motor, 16 Ah battery)

 

[onemintyulep]

If you have a fuse (which you should) you can use this. It has a slightly higher resistance.
I don't have the requirement for metering amps anymore, sinced I created an expanded scale voltmeter. It tells me all I need to know about the state of mind of my battery - c/o the voltage sag at varying loads. It really gives you 1:1 feedback. Just no hard numbers. Works best with lead or LiMn though.

 

[Gregte]

I mainly want to be able to observe the relative difference in current draw at the various speeds and load conditions so I can choose more wisely the speed I ought to travel with regard to how much battery power I have remaining.

I also want to know how much relative power is left in the battery. A volt meter is the tool here of course unless you have recorded amp hours used.

I like the idea of the expanded scale voltmeter. Why is it now desirable for the LiFePO4... is it's discharge curve to flat?

Another idea I had concerning using a section of the battery cable as the resistor, if you tap into a place that you know is excessively long (will give a high voltage drop) you could then devise a dimple voltage divider (use a trim pot) to feed your mili volt meter. Then if you have a way to verify your actual ampere reading you can just adjust the trim pot and be set. At least it sounds correct in my mind.

 

[onemintyulep]

However we do it, using part of the harness as a shunt means it's all based on an imperfection of the setup. I mean, if there is noteworthy resistance or voltage drop I would rather feel pressed to improve it

Yes, with LiIon or LiFe the voltage drop is not really telltale of the remaining capacity, rather they drop dead when they're empty. NiXX only show their real voltage under load, as I understand. It would still show the load on the battery though, just not be explicit about the remaining capacity.
With the SLAs that I currently use I really get to see the impact of cranking up the throttle, and it's not at all intuitive. There is always a thin line of throttle that makes for a sudden drop from 100 to 80%, for example. So I can avoid "crossing the line"
I have the meter zoomed in on 20V-26V for 24V of lead.

 

[Drunkskunk]

This is my first ebike kit. I just used the wiring that they sent.

Would I see much benefit to change the 3 power sires from the front hub motor to the controller from 16 ga to 14 or 12 gage? What do you recommend? (500 watt controller, 750 watt motor, 16 Ah battery)

Gratz on your first Ebike. 16 Gauge wire for a 250watt motor would be small, with a 750watt..

What brand of motor is this?

Yeah, many have found good improvements switching to a higher guage motor wire, but that won't have much effect if the battery wire is choking what can get to the controller.


back to your original question, you can get one of the round guage automotoive amp meter gauges for $10-15 bucks at Autozone that can be wired in line.
A better alternitive would be a watts up meter which does measure amps, voltage, and amps used, giving you a better idea of how much capacity you have left.
Certianly, your idea can be made to work as well, though it might cost more to build your own than it would be to buy a watts up meter.

 

[Gregte]

I don't have the requirement for metering amps anymore, sinced I created an expanded scale voltmeter. It tells me all I need to know about the state of mind of my battery - c/o the voltage sag at varying loads. It really gives you 1:1 feedback. Just no hard numbers. Works best with lead or LiMn though.

After much consternation I finally decided to go the same route as you onemintyulep. I configured my meter to display from 32 to 43 volts as 41.8 is the full charge voltage of my lithium manganese battery and I believe 32 is the very bottom limit I think I want it to ever get to. I don't intend to ever go lower than 35 volts normally.

 

[liveforphysics]

I got a 500amp shunt with digital current meter on flea-bay for cheap. Something like $60 shipped if I remember right. They had 50amp and 100amp shunts for about $10-15 shipped if I remember right.

 

[onemintyulep]

I configured my meter to display from 32 to 43 volts as 41.8 is the full charge voltage of my lithium manganese battery and I believe 32 is the very bottom limit I think I want it to ever get to. I don't intend to ever go lower than 35 volts normally.

Nice scale you have
Does it show load with your battery? I'm going LiMn too. As soon as I receive the cells...

 

[Gregte]

Does it show load with your battery? I'm going LiMn too. As soon as I receive the cells...

It used to be a meter for a metal detector. It had a real odd bunch of scales on it. I spray painted it white them use rub-on letters for the new dial face.

I used two 16 volt zener diodes in series to establish the low end reference and then a pot to set the high end with a 43 volt signal.

It is connected directly across the battery (after the key switch) so it shows the battery voltage at all times when the key is on, while under full load, or being charged via regenerative braking and everything in between.

My next mod which I just finished but have not had the chance to try (raining) is I paralleled the 3 motor power wires (14 gage steel, about 5 ft. long each) with 14 gage copper wire. According to the wire tables the cross sectional area of my motor power wires is now the same as 11 gage wire thsu making each of the 3 wires 11 ga., half steel and half copper.

Do you think I will detect any difference due to this wire addition, going from 14 ga. steel wire to 11 ga. copper/steel mix?

 

[Grinhill]

Steel wire? What's wrong with good old copper?

EDIT: Is it tinned copper perhaps?

 

[Gregte]

Steel wire? What's wrong with good old copper?

EDIT: Is it tinned copper perhaps?

Nothing at all wrong with copper. It offers less resistance per foot which is of course good.

In radio antenna applications copper clad steel wire is often used for making antennas. The wire is made of steel for its superior strength but it is coated with copper for its better electrical conductance. But I don't know of any wire that is made of a copper core but plated with steel. Of course tinned copper wire does appear this way to the eye I guess.

The wire in used by my bike motor is bright silver color and feels stiff like it is steel. Maybe it is tinned copper but it did not feel like it. I did not want to cut it to find out since there is nothing I can do about it anyway.

I cut away its insulation for a half inch so I could wrap my new copper around its bare conductive surface, and solder it. Anyway, in scraping away the insulation, I did not gouge into the wire, like I would have expected to do if it were made of the softer copper, tinned or otherwise.

But, maybe you are right, maybe it is tinned copper which would be all the better.

 

[onemintyulep]

I spray painted it white them use rub-on letters for the new dial face.

Ah, that's it. It has a kind of solid look. Better than just inkjet on paper

I used two 16 volt zener diodes in series to establish the low end reference and then a pot to set the high end with a 43 volt signal.

The zeners should get 5mA at all times, to keep the flow going. If the meter has a high resistance/sensitivity, a resistor in series with the zeners and the meter parallel with that resistor is better.

I paralleled the 3 motor power wires (14 gage steel, about 5 ft. long each) with 14 gage copper wire.

Do you think I will detect any difference due to this wire addition, going from 14 ga. steel wire to 11 ga. copper/steel mix?

It's certainly tinned copper. Steel cables only in the piano :wink: I'm not going to say you will not notice a difference. Oops . Maybe in a HiFi sense

 

[Gregte]

]

The zeners should get 5mA at all times, to keep the flow going. If the meter has a high resistance/sensitivity, a resistor in series with the zeners and the meter parallel with that resistor is better.

I used a variable voltage bench supply and digital voltmeter to make a mark for each of the integer volt readings to make the dial face. Happily, the meter needle does not vibrate in an unusable way. It is quite stable, just rocks slightly with the handlebar movement.

Thanks for the clarification on the tinned copper vs steel. I should have known better.

Any idea on what 'unloaded' voltages to expect for the various amounts of remaining battery capacity? Battery is Lithium Manganese 36v 16Ah.

TIA

 

[Gregte]

It's certainly tinned copper. Steel cables only in the piano :wink: I'm not going to say you will not notice a difference. Oops . Maybe in a HiFi sense

I only now did the math and have concluded that adding the additional 14 ga wire will only gain me about 1.5% more power at full throttle. Not really worth the do.

But when I started I thought my motor wires were 16 ga and steel. It would have made a significant difference if that were true.