Hall effect sensors read exactly 10% of real value

Post Reply
maniago
Posts: 68
Joined: Mon Mar 12, 2018 10:34 pm

Hall effect sensors read exactly 10% of real value

Post by maniago »

My two 100amp Hall effect sensors read exactly 10% of the actual amps running thru them - ie when they should read 5amps they read .5 amps. I know what they should read as I have an amp meter on the high side of my 120AC/12DC power supply to measure total load. Clearly its a setting, but I havent found what bit to flip for fix this.

I have only one wire running thru each toroid (measuring A&B Buss) and their set using the 100amp SF164 and Offset309 (to zero out when no current). Aux1&2 are being used for the amp readings; EIS SW Version91.

What am I doing wrong?
Bobturner
Posts: 444
Joined: Mon Mar 11, 2013 6:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by Bobturner »

Is the current variable listed as an integer or a decimal? Try the other one.
maniago
Posts: 68
Joined: Mon Mar 12, 2018 10:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by maniago »

Bobturner wrote:Is the current variable listed as an integer or a decimal? Try the other one.
Well since I can read .5 amp, that would be decimal. Why would that make any difference anyway? Besides, I want to be able to read decimal values for the amps.
maniago
Posts: 68
Joined: Mon Mar 12, 2018 10:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by maniago »

From the sheet on the Hall effect....
"For sensing current in the -100 to 100 Amp range, or -50 to +50 Amp range, an auxiliary input which does not include a decimal point in the display is the best choice, as the displayed value will have a resolution of 1 Amp."

....I infer that I can get 10ths on the amp reading if I want, because its says "best" choice, not "only" choice. But I changed the display to integer and it reads right so I see that indeed 10ths resolution is not an option.

Jeff - this is misleading, so recommend you guys change this line in the documentation and republish it to state that resolution of 1 amp is all thats available on the 100 and 50 amp scale and integer must be used to read correctly. Obviously 10 turns for the 10 amp scale (as you already note) then validates this.
Bobturner
Posts: 444
Joined: Mon Mar 11, 2013 6:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by Bobturner »

maniago wrote:
Bobturner wrote:Is the current variable listed as an integer or a decimal? Try the other one.
Well since I can read .5 amp, that would be decimal. Why would that make any difference anyway? Besides, I want to be able to read decimal values for the amps.
The EIS uses old technology. It has very limited memory, you need to program it in something which is almost machine language. I think you can display tenths of a volt, but you will need to read the manual to get the new settings to enter. Understand that the Hall sensor itself is probably not accurate to 0.1% of full scale (0.1 amp out of 100). Also, that last tenths digit will probably be jumping around a lot. It’s distracting if nothing else. The plus side of the older technology is that you can leave the EIS on during engine start. Even if it reboots it does so in a couple of seconds, so you can see the oil pressure coming up (or not!).
maniago
Posts: 68
Joined: Mon Mar 12, 2018 10:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by maniago »

The sensors also seem to be progressively inaccruate the more amps run thru the wires. For example, its off about half an amp best I can tell at around 3 amp actual draw, but when I get to pushing 25 amps or so, the EIS is only reading about 18 amps.....its kinda useful in that the numbers do increase as the amps increase, but def not what one would want to call accurate.
GRT_Jeff
Posts: 802
Joined: Tue Dec 11, 2012 12:11 am

Re: Hall effect sensors read exactly 10% of real value

Post by GRT_Jeff »

I don't know what typical installation accuracy is, but the sensor data sheet says it's supposed to be accurate to 2% full scale, so that would be 2 amps. Is it possible the current is affecting your ground voltage level at the sensor or EIS? At the low end it could be the zero offset, which you can adjust.
maniago
Posts: 68
Joined: Mon Mar 12, 2018 10:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by maniago »

I've messed with the zero offset, but that doesnt help. I will say tho that its pretty clear that the sensors are affected by any other magnetic field impingements, for example, the two toriods near each other have affects, and I'm guessing that having the other wires running next to the outside of the toriods - grounds or power wires - also have an effect. I'd also suspect that the magnetic fields generated by the wires inside the toriod arent super happy if the wires are dead centered. I'm going to have to play with their location some more and see if that helps, so for now I've moved on and will take this back up when I'm doing engine/alternator live runs. Right now I'm still in the power supply phase.
Bobturner
Posts: 444
Joined: Mon Mar 11, 2013 6:34 pm

Re: Hall effect sensors read exactly 10% of real value

Post by Bobturner »

I have not made any measurements on mine, other than adding up the published current requirements of everything that’s turned on and comparing to the efis number. But I’m sure it’s not off by 7 amps when I’m pulling about 25 amps. My torroid is on the cabin side of the firewall, not too close to any other high current wires. And I run the alternator output thru it just once. Do you have multiple turns thru it?
BTW, in principle the location of the wire (dead center or near the edge) shouldn’t matter.
One thing to check: the actual Hall effect detector is on a small circuit board, inserted into a small slot in the torroid and glued. Make sure it hasn’t backed out.
Post Reply