The Best Value Wattmeter

[Deleted member 4366]

The problems with cheap wattmeters is that they require fairly heavy wiring, and they're not very waterproof. Luckily there's a solution that costs virtually nothing to solve those problems. You need one of this type:

http://www.ebay.co.uk/itm/Free-Ship-GT-Power-LCD-RC-130A-Battery-Watt-Meter-Power-Analyzer-Ver-2-0-/230933611688?pt=UK_ToysGames_RadioControlled_JN&hash=item35c4b744a8

It's already fairly waterproof because it has no buttons. Everything is on the one back-lit screen; however, it has a grille underneath to let the heat from the shunt out. Wattmeters work by measuring the voltage drop over a shunt resistor. The greater the current, the greater the voltage drop. There's a CPU that takes this information and uses it to compute all the info shown in the display. We can solve both problems if we remove the shunt and place it near the battery. We then only need to send only the voltage drop information to the wattmeter. The wattmeter needs a power supply, so it needs two wires from the battery to power it, so that makes three wires if we also include the voltage drop info. the wires only need to be thin.

Here's what it looks like inside. The shunt is that coppery coloured thing at the front. All the current goes through it on the negative side. There's two PCBs. You have to bend them apart to get at the shunt:



To remove the shunt, you need a thin-bladed knife to lever it up while you use a soldering iron to melt the solder - one side at a time. It helps to transfer heat if you put a blob of solder on the iron's tip.

Here it is with the shunt removed. You also need to remove the thick red wires from the back side:



You could simply connect the battery wires and three thin wires to the shunt you removed, but I prefer to make my own out of 14g wire. 10cms = 0.010 Ohms, or 5cms = 0.005 Ohms. here's a close up of a 0,010 Ohms shunt that I made for my other wattmeter. The red and black power wires for the wattmeter go on the battery side and the white signal (voltage drop) wire on the other end of the negative wire:



The wattmeter above requires a .005 Ohm shunt, so only 5cms long.

Here's a similar wattmeter that I did that shows the three thin wires soldered to the PCB where the shunrt was removed from, and the red positive on the back side where the red battery wire was before:



So after making the 5cm shunt and soldering the wires to the PCB, it's ready for testing. I'm lucky to have a battery tester that puts a variable programmable load on the battery, but you can check yours before and after modification by connecting it to your bike, and giving it full throttle with the brake on to slow the motor right down. this will be the max current from your controller, which is constant.

I had to adjust the length of the black wire slightly to get it exact. At 10 amps:



So here it is finished. Next stage is to seal it all round with silicone sealant and make a bracket to fix it to the stem before screwing it back together:



Here's the schematic:

 

[melb-ourne]

hi d8veh

http://www.ebay.com.au/itm/251354753090?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

i brought one the cheap watts meter and did the shunt mod..and worked perfectly for the BBS01 350w36v 18A... getting Peak Amp of 28 and the rest of reading ..aH,voltage looks normal as expected.

Two months later, the Peak Amp jumped to 55 Amp, i got a bit concerned... but assumed the watts wass wrong. is there a way to calibrate or check the peak amp.

I order a same replacement watts meter last week but i might order this one as well.

thanks for tutorial

 

[Deleted member 4366]

Inside your watt-meter are two piggy-backed solid state 0.003 ohm shunts like this:




One of the resisistors must have blown or become detached, so the resistance has doubled and therefore so has the voltage drop, so the CPU will calculate double the current. The other resisistor will be working above its rating, so is likely to get very hot.

My photo shows one of the resisistors as 3 miliohms. If the two are the same, the total resistance will be 1.5 miliohms, which is very low - equivalent to 15mm of 14g wire. I guess you've got nothing to loose by removing the shunts and making your own remote one. If you do that, be very careful not to bridge solder joints on the PCB close to where the shunts are. IIRC there's something close to the shunts. Once you've established the values of both shunts (probably both R003), You can either make your own or buy a remote one.

It's worth making a remote shunt out of 10 cms of 14g wire. You can connect a voltmeter to each end of the wire, and the voltmeter millivolts equals the number of amps, so your 25 amps should give 25 mv.

 

[trex]

it would be nice if someone makes a bluetooth or wifi watt-meter then it can be tucked away in the controller box or the battery case. BTW, what you you use to sketch your circuits?

 

[Deleted member 4366]

it would be nice if someone makes a bluetooth or wifi watt-meter then it can be tucked away in the controller box or the battery case. BTW, what you you use to sketch your circuits?

I use Open Office Draw to do schematics. The whole Open Office Suite is a free download (open source) as is compatible with MS Office.

They do exactly what you want. It's the Speedict. The battery info, speed and distance etc all work well, and so does the data logging. Some of the other functions, like speed limiting and file handling are not so stable. Still not bad for £60:

http://www.speedict.com/

 

[jackhandy]

My speedict's playing up again: 3 out of 4 rides it doesn't log the data & I have to remember to write down everything before I switch off or synchronise, or it's all gone down a black 'ole.

Totally unreliable for any sensible data gathering.

 

[Deleted member 4366]

Totally unreliable for any sensible data gathering.

But it still shows live data ok. It's just the transfer of logged data that gives problems, isn't it? The original Speedict only had the basic battery and speed functions, and everybody was happy at £60. Then they added a load of advanced features that don't work for everybody, although I never had any problems. I think it depends a lot on which phone and Android version you have.

 

[jackhandy]

Quite right, but that's the strange thing - I use a Galaxy Note & android 4:1 (or 4.2, can't remember); a very common combination - nothing unusual & I've tried several different phones.

Seemed significant that they wouldn't publish a list of phones & o/s that it definitely worked with, despite being asked to do so.

 

[melb-ourne]

Inside your watt-meter are two piggy-backed solid state 0.003 ohm shunts like this:

One of the resisistors must have blown or become detached, so the resistance has doubled and therefore so has the voltage drop, so the CPU will calculate double the current. The other resisistor will be working above its rating, so is likely to get very hot.

My photo shows one of the resisistors as 3 miliohms. If the two are the same, the total resistance will be 1.5 miliohms, which is very low - equivalent to 15mm of 14g wire. I guess you've got nothing to loose by removing the shunts and making your own remote one. If you do that, be very careful not to bridge solder joints on the PCB close to where the shunts are. IIRC there's something close to the shunts. Once you've established the values of both shunts (probably both R003), You can either make your own or buy a remote one.

It's worth making a remote shunt out of 10 cms of 14g wire. You can connect a voltmeter to each end of the wire, and the voltmeter millivolts equals the number of amps, so your 25 amps should give 25 mv.

Thanx d8veh for your help, like most people i am lost with electronic stuff and my soldering skills are crap, so mostly like my joint had pop/destory something.

i give this a try and check the resistor/wire gauge so i don't screwup the next meter.

 

[danielrlee]

Dave, great guide as usual. I've been meaning for ages to do this on a Turnigy watt meter, so last night I did. Kind of.

Reusing the original shunt, I managed to get the amp reading within about 15% under what it should read, but after this point, any additional fine tuning just resulted in me adding more solder to the shunt. This, of course, only leads to lowering the shunt resistance and causes the power meter to further under-read the amps.

I'm out of solder now, so will have to await delivery of more before I try this again. I think I'll take your suggestion of using 14 gauge wire instead of the shunt, as I imagine it'll make fine-tuning the ground sensing wires much easier due to the higher tolerance in regards to its length.

On a side note, do you know of a source of flat multi core wire, similar to servo wire, but black in colour? I remember seeing something of this description with maybe a dozen or so cores, that can be split to requirement, but I can't for the life of me remember where, or what it is called.

 

[Deleted member 4366]

As you can see, I'm using servo wire. I'll spray it black afterwards.

I made a bracket out of 1mm aluminium. It's finished now, and looks a lot neater than it did before with big wires sticking out of it, plus it's waterproof.

 

[danielrlee]

I think that looks quite tidy mounted to the stem. I might have to do something similar.

Last night I tried making a shunt from 14AWG wire, but the result was the same as when I tried reusing the original shunt - I'm down by around 15-20% on metered power. The original shunt was 0.005 ohms, so I used 6cm of wire to give myself a little leeway if needed, but even using the full length still wasn't enough. Maybe the 14AWG wire I'm using isn't quite on spec.

Is there a limit to the gauge of wire used for the shunt sensing wires? I'm using a repurposed USB cable and the wires are very fine, maybe only several strands each.

Also, should the position of the shunt sensing wires on the watt meter pads affect the readings?

I had a feeling this wasn't going to be as straightforward as you make it look ;-)

 

[Deleted member 4366]

The thickness of the sense wire shouldn't make too much difference because virtually no current flows down it. The same applies to the position on the pads I think my shunt was just over 6cm, which I put down to mine being slightly non standard, so try a longer piece. Also, when you solder it, the solder gets sucked in to the strands, so you could lose 1/2 cm each end.

To complete the story, here's some photos of the three button type wattmeter, which has long wire shunts. The principle is the same: prise up the display module, unsolder the Deans connectors and two shunts ans solder on a remote one. These wattmeters are hard to waterproof because of the three buttons. The shunts are 0.010 Ihms, so you need about 10 cm of 14g wire.

 

[jackhandy]

Evening D8veh,
My Speedict has finally failed completely, following weeks of becoming less & less reliable at recording & downloading ride data.

I intend modifying a r/c wattmeter to remote shunt operation, following your tutorial, but the main purpose for mine will be to act as a "fuel guage" for the bike battery; for which I'd like to be able to keep it powered when the bike battery is switched off if I leave the bike parked-up for an hour or 2.

Is it possible to use a couple of 18650 cells to keep it ticking over & will it retain the ah used figure reasonably accurately & carry on recording when the main power is restored?

I assume, if this is a practical proposition, that I'd need blocking diodes of some kind to prevent the 37v from the main battery blowing-up the littl'un: Would they need to be high-power, or would they just need to handle the small current from the small battery?

 

[Deleted member 4366]

That's a good question, and I''m not sure of the answer.

These wattmeters keep counting as long as they remain switched on, so if you put the shunt between the battery and main switch it'll stay on. You can then put another switch on the red power wire to the meter, which would switch it on/off and reset it. You can add a small switch to the meter itself, to make a convenient way to switch it. No need for a separate power supply as long as you don't switch the battery off with a key-switch or other switch upstream of the meter. I don't believe that the meter draws a lot of power, so it would be OK for a few days. The cheapest ones with no back-light draw about 7mA (45 days to flatten main battery). The Turnigy type have a backlight, so consumption will be a bit more.

 

[NZgeek]

I wired a 9V plug up to the pins for the external power supply on the cheap blue ebay unit. Worked brilliant, and when my BMS wa shutting off, it saved my total consumption etc.

 

[jackhandy]

Thanks - that would seem to be a goer then

Ah - except that mine doesn't seem to have an external power facility

 

[Deleted member 4366]

Oh yes it does. You only have to connect a battery to the source side. Anything 5v or over should work.You have to tie the two 0v wires together for it to display correctly. you could probably get one of those 1/2 AAA size 12v batteries inside with a switch on the outside. It's the same battery as they use in a car remote key.

 

[jackhandy]

Right - But won't the 37 volts from the bike battery blow-up the littl'un?

 

[Deleted member 4366]

No, because it isn't connected. You don't connect your battery red wire, so on the source side you have 12v from the little battery plus the two zero volt connects. On the load side, you only have the sense wire (either remote shunt or internal shunt), which is a few milivolts above zero volts. The meter only checks the difference in voltage between the two sides of the shunt, which is a fewmilivolts, but it needs a power supply of at least 4v to power the CPU and LCD. You can use either the red from your battery or any other voltage source.