Why does the current keep increasing when using an N-mosfet ?

[rob77]

1) I understand now when my IRF540N starts opening VGS(TH) = from 2V to 4V. (Gate to Source Threshold Voltage) correct ?

it's not defined, the only clue is the Rdson

the threshold voltage and Rdson is only for nfet working in saturation mode. i'm not an expert of nfet in linear mode, but i'm suspecting you are running the nfet in linear mode based on your circuit setup similar to bjt high side emitter follower. so Vg (linear) doesnt have to take full Vgth to be working in lieanr mode and Rds (linear) will be much much greater than Rdson.

if you look at the diagram below. Vgth is only to overcome the "capacitance charge resistance" (or somesort of term) of the gate (horizontal line), once you get passed that threshold, nfet will be in full saturation mode, or very close to it... since you havent passed that threshold, i suspect you are working in fet linear mode. you'll be blessed if mosfet datasheet provides greater details about linear mode working condition, just as how bjt's datasheets have. but as usual, IRF540's nmosfet datasheet doesnt.

edit: its ok with your emitter follower circuit, except you are using n-mosfet for it instead of a npn power bjt. your circuit is intended to be working in linear mode anyway, so yeah, as far as the nmosfet linear mode and the datasheet is concerned? the specification of your circuit is basically.... "undefined".

and as Hero999 suggested, 5532 may go into unpredictable output when given not the right input (beyond common mode range of the working condition) such as phase reversal. i'm not sure 5532 if it has phase reversal problem i forgot, but opamp with phase reversal is something that i want to get rid of in anything, or else great care must be taken on its inputs.
http://www.analog.com/media/en/training-seminars/tutorials/MT-036.pdf

that diagram is a "Typical Gate Charge vs. Gate-to-Source Voltage" tells very little about Vgs relation to Rdson, you can see where it enters saturation mode but you can't tell anything about the lowest Rds from it (fully ON state). the OPs question was at what voltage the mosfet if "fully on". to answer that question you would need a Vgs vs. Rdson diagram and that's not in the datasheet.

and could you please elaborate on your statement ?

the threshold voltage and Rdson is only for nfet working in saturation mode.

what did you mean by it's "ONLY for nfet in saturation mode" ? pfets have Rdson and thershold voltage too ! do not discriminate pfets

basically we could describe the mosfet's Vgs to Rds this way: as the Vgs is increased from 0, at the threshold voltage the Rds start to drop (mosftet start "opening") and in fact the mosfet is in the linear region, Rds drops in somewhat linear(-ish) fashion at some rate. as you increase the Vgs further at some point the Rds drop "flattens out" but is still decreasing, the mosfet enters it's saturation mode. further increase of Vgs will give you a small drop in Rdson till the point of lowest Rds and further increase of Vgs toward max allowed Vgs will do nothing. so the answer to "at what Vgs the mosfet is fully ON" is exactly the point when minimal Rdson is reached - but that parameter is not defined in the datasheet, it indicates only the Vgs where the lowest RDson is guaranteed, not a voltage when it's reached.

[Mechatrommer]

because OP is using nfet, so i specifically explain to n-type without going to complicate things explaining pfet, which is just no more special than nfet symetry (except poorer spec). and why would the OP want saturation mode? which he wont be able to get to from his circuit. if there is saturation mode, there will be near 5V across the 0.2ohm (+0.077ohm Rdson) = 18A current flow, give or take.

[DuncanSteel]

So now is the time to choose the op-amp.

I have couple questions regarding calculation.

As an example lets use the NE5532

My power supply will be +5V & since I dont have a negative power supply I will use the same op-amp in inverting op-amp config to obtain negative voltage on the negative supply pin.

1) First how do I know how much voltage can I apply to the positive & negative input pins?

[bktemp]

Input: Look at the common mode range. Here it says typ. +/-13V for +/-15V and guaranteed +/-12V for +/-15V. So it will work down to typical 2V (worst case 3V) above negative rail up to typical 2V (worst case 3V) below positive rail.
At 0V/5V supply voltage it will therefore probably operate from 2 to 3V input voltage or worst case it won't work at all.
For the output voltage the given Output voltage swing is the same as the input voltage for the NE5532. But this voltage also depends on the load current (in this case >600ohms to ground (=0V =(positive+negative supply voltage)/2). Since a mosfet is a high impedance load, it may be higher, but it is not specified.

Rail2Rail opamps have a common mode voltage of slightly below negative rail to slightly above positive rail and an output voltage range of a couple of mV near the supply rails for low output currents.

[DuncanSteel]

Input: Look at the common mode range. Here it says typ. +/-13V for +/-15V and guaranteed +/-12V for +/-15V. So it will work down to typical 2V (worst case 3V) above negative rail up to typical 2V (worst case 3V) below positive rail.
At 0V/5V supply voltage it will therefore probably operate from 2 to 3V input voltage or worst case it won't work at all.

How do you calculate this for +5V/0V supply voltage?

At +15V/-15V supply voltage it will work if input is between +/-13V & guaranteed to work if input is between +/-12V. correct?

I don`t understand how did you get 2V offset from 5V.  tried to do the ratio calculation, but the numbers don`t match up.

15V-100 %
12V-80 %
5V-100 %
2V-40 %

[rob77]

Input: Look at the common mode range. Here it says typ. +/-13V for +/-15V and guaranteed +/-12V for +/-15V. So it will work down to typical 2V (worst case 3V) above negative rail up to typical 2V (worst case 3V) below positive rail.
At 0V/5V supply voltage it will therefore probably operate from 2 to 3V input voltage or worst case it won't work at all.

How do you calculate this for +5V/0V supply voltage?

At +15V/-15V supply voltage it will work if input is between +/-13V & guaranteed to work if input is between +/-12V. correct?

I don`t understand how did you get 2V offset from 5V.

2V above 0V is 2V and 2V below 5V is 3V probably ?

[DuncanSteel]

2V above 0V is 2V and 2V below 5V is 3V probably ?

I understand that, but how did you get those 2V to begin with?

simply 15V - 13V ?

[rob77]

2V above 0V is 2V and 2V below 5V is 3V probably ?

I understand that, but how did you get those 2V to begin with?

simply 15V - 13V ?

it's a property of the NE5532 opamp - it can't go closer than 2V to supply rails (both input and output).

[bktemp]

2V above 0V is 2V and 2V below 5V is 3V probably ?

I understand that, but how did you get those 2V to begin with?

simply 15V - 13V ?

Yes.
But this is only a very rough estimation, because the opamp is only specified at +/-15V. Therefore for any supply voltage other than +/-15V all the specifications are not guaranteed. Most of them won't change much, but you have to measure them yourself to be sure.
Opamps designed for low voltage operation often have different specifications for different supply voltage ranges. TLC271 datasheet for example has specifications for 5V and 10V.

[Mechatrommer]

My power supply will be +5V & since I dont have a negative power supply I will use the same op-amp in inverting op-amp config to obtain negative voltage on the negative supply pin.

its not gonna work.for opamp to output negative voltage, it must have -ve power supply at Vcc- to begin with. you cannot create food from yourself, overunity doesnt work here. the device for that purpose is inverting buck or boost converter. with only 5V power supply, you have to work within 1Vpp common mode range that is 2V-3V, even in inverting config, opamp will not be able to output anything beyond that.

[DuncanSteel]

I see, there is in this case 1 more thing I don`t fully understand. How big will be the output voltage if the supply voltage is +5V/0V and input voltage at pin (+) is say 2,5V.?

Input Offset Voltage is 4mV, so 5V-0.004V= 4.996, but I measured the real time output & it was 3,6V. How comes?

[Kalvin]

You can also think about your problem like this:

- Let's say you want to have a maximum of 0.1V at the sense resistor at the full current.

For a sake of example, let's say that you are designing the constant current load for a maximum of 2.5A load.

- At the maximum current of 2.5A and 0.1V across the sense resistor, the sense resistor needs to be R = U / I = 0.1V / 2.5A = 0.04 ohms.

- At the maximum current the power dissipated in the resistor is P = U * I = 0.1V * 2.5A = 0.25W, so you should select a 1W or 3W resistor.

Let's say you will use the constant current load to the a 5V or 12V power supply or car battery.

- At 5V the MOSFET will need to dissipate (5V - 0.1V) * 2.5A = 12.25W.
- At 12V the MOSFET will need to dissipate (12V - 0.1V) * 2.5A = 29.75W.

In order to the MOSFET pass the 2.5A current it is quite safe to assume that the Vgs has to be something like 4V more that the source voltage. In this particular example, the op amp is required to be able to drive its output ie. the MOSFET gate voltage to be at least 4V + 0.1V = 4.1V in order the MOSFET to pass the 2.5A current. If the Vgs needs to be more that 4.5V at 2.5A it is quite probable that you need to choose another MOSFET. You need to check the MOSFET datasheet for actual numbers.

In order to make the MOSFET close completely in practical terms, the Vgs must be below the threshold voltage Vth. Let's assume the Vth is 2V for the MOSFET.

What we have now found out is as follows:

- The sense voltage across the sense resistor will be 0.1V at the maximum current.
- The sense voltage across the sense resistor will be 0.01V at 250mA.
- The sense voltage across the sense resistor will be 0.001V at 25mA.
- The op amp driving the MOSFET gate needs to be able to drive its output up to 4.1V when the maximum 2.5A current is required.
- The op amp driving the MSOFET gate needs to be able to drive its output below 2V in order to close the MOSFET completely.

From the following facts we can say that we need an op amp that needs to be able to operate on input signals very close to the negative rail, its input offset voltage needs to be less than 1mV, and provide output signals that are quite close to the positive rail, and the minimum output voltage needs to be 2V of the negative power supply rail or less when the op amp is operated from 5V power supply.

When you take a look at the datasheet of the NE5532 you may find out that some of the requirements are not satisfied**. So, you may want to check some other op amps that are targeted for a single supply operation, have low offset voltage and which have RRIO specification. You may also tweak the sense resistor value somewhat higher if the op amp's input voltage range cannot go close enough to the negative rail and you want to be able to adjust the current in milliampere range. If you want to have multiple ranges, you may want to consider something like this: http://electronicdesign.com/displays/resistive-dummy-load-draws-constant-current-12-50-v Notice also how the op amp is compensated as it is driving a MOSFET gate.

To answer your original question: The probable cause is that the op amp cannot maintain the feedback loop so that the voltage across the sense resistor is constant. Either the op amp cannot drive the MOSFET gate high enough or the input of the op amp cannot measure the very small voltage close to the negative rail across the sense resistor. Or both.

** From the NE5532 datasheet one can see that at +/- 15V power supply the Common-mode input-voltage range is typically +/- 13V. That means that the input signal needs to be 2V less than the positive operating voltage and 2V more than the negative operating voltage. In the +5V single supply case this means that NE5532's input signals needs to be in range (0V + 2V) ... (5V - 2V) ie. between 2V ... 3V. In a similar fashion one can see from the datasheet that the Maximum peak-to-peak output-voltage swing is 26V which means that the output will swing 2V less the positive operating voltage and output will stay 2V above the negative operating voltage. In a 5V single supply case that the NE5532's output will be able to swing typically only in range 2V ... 3V.

[Mechatrommer]

I see, there is in this case 1 more thing I don`t fully understand. How big will be the output voltage if the supply voltage is +5V/0V and input voltage at pin (+) is say 2,5V.?
Input Offset Voltage is 4mV, so 5V-0.004V= 4.996, but I measured the real time output & it was 3,6V. How comes?

you dont provide enough information. please measure whats the reading on -ve input pin, ie the 0.2ohm/0.01ohm top side. are you refering to 0.2ohm? or 0.01ohm?

[DuncanSteel]

Well i kind of understand now what I need, the problem is to understand the datasheet.

Let me try. I found this Op-amp TS1005IJ5T, If I understood this correctly,

1) The output of this Op-Amp can range from Max0.8V+30mV = (0,83V) to 5,5V-60mV = (5.44V), Correct?

2) And input can range from (max 5mV?) 0.8V+5mV=(0.805V)  to 5,5V-5mV=(5.495V), correct?

[Mechatrommer]

1) The output of this Op-Amp can range from Max0.8V+30mV = (0,83V) to 5,5V-60mV = (5.44V), Correct?

if you are suplying the opamp with 0.8V at Vdd, max output is 0.8V-60mV = (0,74V)
if you are suplying the opamp with 0.8V at Vss, min output is 0.8V+30mV = (0,83V)
if you are suplying the opamp with 5.5V at Vdd, max output is 5,5V-60mV = (5.44V)
if you are suplying the opamp with 5.5V at Vss, min output is 5,5V+30mV = (5.53V)

2) And input can range from (max 5mV?) 0.8V+5mV=(0.805V)  to 5,5V-5mV=(5.495V), correct?

https://en.wikipedia.org/wiki/Input_offset_voltage