LEM SENSOR NOTES

where to get parts for building your own transmitter.

LEM SENSOR NOTES

Postby N9NEO » Sun Mar 21, 2004 8:14 am

Seems there's some interest in LEM sensors using Hall Effect devices. They are a good choice as they are fast enough to protect a mosfet, and provide a nice isolation barrier. The Hall Effect was discovered not long after electricity itself. (I know, I was there.) There are various choices when it comes to the Lem sensor.

The Lem sensor uses a gapped magnetic core material to concentrate the flux. The Hall device is then placed in the gap. Cores of both ferrite and propriatary metallic mixes in the tape wound fashion are both used. Their are two main contenders - open and closed loop. The open loop devices just convert the magnetic flux into a voltage directly. The closed loop devices have a secondary winding wound on the core. The principal here is to use a current source to drive the flux to zero in the core. The current is then displayed as a voltage signal over a resistor - which in most cases the designer selects.

Either type of part should work well for protecting mosfets. In general the more expensive closed loop part is used in applications where precise current measurements are needed in currents which are rapidly changing. A specific application would be a servo motor drive where the motor currents need to be controlled to a very high degree.

Beyond the open and closed loop arenas there are still many flavors. Many parts require both +&- supplies to run, while newer designs may only need +5v. Some parts can be mounted on pc boards, while others for measuring really big currents are panel mounted with large windows to pass the wires through.

There are other options beyond the packaged devices which deserve a looking at.

A small sense resistor is a good way to go, but some attention should be paid to the parts inductance as well. A wirewound part would be a bad idea if the currents have any edges. A few 1/4w metal film resistors in parallel would be a good choce if the necessary values were difficult to obtain. The problem with the resistor approach is there is no voltage isolation. It's difficult to sense a current and respond to a fault condition when the sensor is 100v away from the control circuit. In most cases a bottom of the barrel opto-coupler would be more than adaquate to get the signal down to the controller in order to protect the mosfet. Agilent produces a HCPL-7840 part which is a very fast serial link sent over an optocoupler. This could be used to send the fault signal down and could also be used to send audio up to a controller sitting at positive bus potential if that exctes you.

Surface mount devices are available which mount next to or over current carrying traces. I don't think there are many surface mounters here so I will stop now.

73
Bob
User avatar
N9NEO
Class E Enthusiast
 
Posts: 322
Joined: Tue Apr 08, 2003 11:52 am
Location: Uxbridge MA

Return to Parts and Sources

Who is online

Users browsing this forum: No registered users and 1 guest

cron