My application is to use the LTC 2946 in a serial connected string of batteries to measure total string current and voltage appliyng it in two distinct design:One for -48V as commoly used in telecom industry, and another one to be used to read from +4 to +500V.
The wide range voltage operation is what I'm looking for, considering that with only a few manual circuit board configuration changes, will enable me to design a common circuit for a wide range of voltages measurements in distinct strings.
The voltage and current measurement is all I need, The Coloumb counter is a plus feature. I will try to desgin two distinct applications, one for telecom voltages i.e. -48v operation strings, and another one for 4 to 500V operation, this last one I will need a help too.
My doubts about my the IC use are:
As already answered before in another post, LTC 2946 cannot process negative numbers when Rsense is submitted to a backwards current, i.e. its is not bidirectional:In this situation, having a backwards current;
Can we still reading the current converted by A/D and sampled as a positive number?
The LTC2946 doens't know the Rsense resistor (shunt resistor) value and only process the voltage potencial across its pins, considering this:
Can I use an analog open loop hall effect current transducers instead of a shunt resistor, directly or indirectly to have a full scale range of 102.4mV?
Its possible to derivate the datasheet example of -4 to -500V voltage meter to accoplish my needs to +4 to +500 using an external power supply for the IC?
Thank you for attention!
FernandezAnahia said:Can we still reading the current converted by A/D and sampled as a positive number?
If Sense- is above Sense+, the current ADC will give a reading of zero and no energy accumulation will occur.
FernandezAnahia said:Can I use an analog open loop hall effect current transducers instead of a shunt resistor, directly or indirectly to have a full scale range of 102.4mV?
This can potentially be done, however we have not attempted it here. If you attempt to do this, pay special attention to the currents going into the sense pins as well as their imbalance. Do you have a part in mind that you plan on using for this?
FernandezAnahia said:Its possible to derivate the datasheet example of -4 to -500V voltage meter to accoplish my needs to +4 to +500 using an external power supply for the IC?
From the LTC2946 datasheet:
Use a shunt resistor in the GND path. Rshunt must be able to stand-off 500V (use size SMD 1210), and dissipate a boat-load of power: > 20mA x 500V = 10W. The circuit in figure on p38 can be adapted for +500V, actually, since the SENSE+ and SENSE- inputs are allowed to go up above the VDD input, so you could drop all of the power in M1 (careful, it will get HOT). In that case the circuit will be referenced to GND, so won't need the opto-isolators.
Note that in all of these ground-shifting schemes the I2C bus and GPIOs will need to be isolated and level-shifted down from 500V to 3.3V ground referenced. The opto isolators will be slow because of the isolated ground (figure 17).
Also be aware that if the GND is not at 0V (because it is floated up) then your ADC readings of voltage will need to be VDD referenced. This is what register bit CA does.
Another method to achieve this would be to use a circuit similar to the one found on page 41 of the LTC2947 datasheet.
This would require swapping out devices Q1 and R13 to appropriately bias the circuit and simulating in LTSPICE.
Let me know if I can be of any further assistance or if you have any other questions!
Thank you and best regards,
Thank you for the answers.
About my first question about the use of a current transducer instead of a shunt resistor, you asked if I have some part to use in mind, yes I have.I'm planning to use the These LEM HO Series Current Transducers. About my answer to get measurements in the rang of 4 to 500V I will understand correctly what you are changing to get this to work.Thank You very much!