Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

High side current measurement

Status
Not open for further replies.
Hi Max0412,

I was referring to Humber's circuit. It does not feel like the current mirror from Zetex, or maybe I am wrong.

Cheers
 

Hello Humber,

I have a query regarding the circuit. Is the o/p voltage (viz current ) independant of the common mode voltage across the sense resistor ?
Can you suggest any OpAMP which can be used upto an common mode voltage of 45V.
TIA.

Added after 2 minutes:

The circuit works fine,but the transformer is actually giving me an output of about 42V,so at low currents the circuit does not work.
As the current goes higher and the voltage drops to <38V its working fine.
Hence the queries.
 

Hello Phaedrus,
I don't know that it has a name. It was in use long before Zetex packaged it.
As for the op-amp, 45V is probably not available outside of specialised devices.
The sensed voltage is 'impressed' across the collector resistor, and virtually independent of the common-mode voltage.
The higher voltage, and to some degree the large change between the loaded and unloaded states, makes a big difference.
Over what range of current do you need the 1% accuracy ?
 

Hi Humber,

I need the accuracy over 0~300 mA i.e. the entire range of measurement. I guess I would have to get the transformer rewound to get the lower voltage to 30V.
The current being measured is actually drawn from a diode bridge and capacitor. The source is not regulated as such.
Do you think there will be variations due to this which would not allow me to get the accuracy ?
I had asked the name , because i was not able to figure out its working and wanted to google.

Cheers.
 

I think that 1% over such a range will be difficult. Let's say that you supply the op-amp with +30V, obtained with a zener diode, and use resistors to reduce the common-mode voltage.
Even if the sensing resistor is shorted out, there will be an output due to common-mode errors, that changes as the supply rises or falls with the load. In practice, the output will be the sum of the wanted output and this error, meaning at the very least, that the output will not be linear.

Rather than me explaining all of the pitfalls, I suggest that you read Zetex Application Note AN39.
They describe how to use their devices at higher voltages. You could apply these methods to an op-amp version.

Added after 1 hours 35 minutes:

Phaedrus,
I think the required level of accuracy almost demands low-side monitoring. Given the difficulties of the high-side approach, perhaps it may be easier to overcome the problems that caused you to reject the low-side method?

Cheers
 

Phaedrus,
I think the required level of accuracy almost demands low-side monitoring. Given the difficulties of the high-side approach, perhaps it may be easier to overcome the problems that caused you to reject the low-side method?

Cheers[/QUOTE]
Reviving this thread again:
Well I finally did go for the high side voltage sensing.I was looking for some help on making the circuit have a programmable gain.I could add a relay and switch the value of R3 so that I get different amplifications.But is there any other more elegant way, using possibly, analog switches such that the analog switch resistance does not affect the output value ?

 

phaedrus said:
Phaedrus,
I think the required level of accuracy almost demands low-side monitoring. Given the difficulties of the high-side approach, perhaps it may be easier to overcome the problems that caused you to reject the low-side method?

Cheers
Click to expand...
Reviving this thread again:
Well I finally did go for the high side voltage sensing.I was looking for some help on making the circuit have a programmable gain.I could add a relay and switch the value of R3 so that I get different amplifications.But is there any other more elegant way, using possibly, analog switches such that the analog switch resistance does not affect the output value ?

[/QUOTE]

Sorry for the delay Phaedrus. Changing the sensitivity is not easy, and would be better done by changing the sensing resistor, R1, but I suspect that is not what you want. You can change R3 as you suggest. If the analog switches are placed between the emitter of Q1 and the highside of the resistor, then that will not affect accuracy, but complicates the switching. That is, always measure directly across the new R3.
Another approach would be to divide the voltage to the + input of the op amp.
So, a divider between the collector of Q1 and the - input, with the +input connected to the divider. The circuit is simple, so not much range will be available, before offsets start to erode the accuracy. If you have any questions, then I will try to help you.
 

See EDN magazine "design ideas" section 12/15/2010 edition,
Pg43 for a decently simple high side monitor. It uses a local
regulator (zener shunt and resistor) to make the high side
negative supply. You might substitute a small MOSFET for
the Darlington and eliminate the residual base current errors.
 

Hi Humber,

Thanks much for answering.I did some simulation and trials with a circuit as attached and it served the purpose.The R4 resistor has a trim pot in series to help calibrate.


Thanks again.

---------- Post added at 17:42 ---------- Previous post was at 17:30 ----------

Hi dick_freebird .

Any links if can download it ? I googled but could not locate it.
Thanks.
dick_freebird said:
See EDN magazine "design ideas" section 12/15/2010 edition,
Pg43 for a decently simple high side monitor. It uses a local
regulator (zener shunt and resistor) to make the high side
negative supply. You might substitute a small MOSFET for
the Darlington and eliminate the residual base current errors.
Click to expand...
 

Please notice, that according to the LF411 datasheet, the OP is not guaranteed to have the positive supply included in the common mode range.
 

FvM said:
Please notice, that according to the LF411 datasheet, the OP is not guaranteed to have the positive supply included in the common mode range.
Click to expand...

Thanks for replying FvM.
I was going by the common mode graph of LF411 attached.At +/-9V the common mode is +10V/-7V. I am measuring current in only one direction.

Is my assumption correct ?Pl. let me have your opinion.
Is it that the common mode supply should be different from the OpAMP power rails ?
 

The EDN circuit is here
**broken link removed**

This will work well, I expect.
It will not be any easier to change the range or sensitivity of this circuit than the other, and still needs an op-amp with inputs that can be at the rail voltage.
The LF411 is not guaranteed to have this capacity, but the internal circuit indicates that it will for the likely range of this application. The initial request was for cheap and commonly available components.
ETA:
The EDN is a neat solution to the problem of the base current. The LF411 should work too.
 
Last edited:

Hi Humber ,
Is the graph I posted the correct one to check this ?Does this mean that the graph is characterized and not actually tested ?
Thanks !!


humber said:
The EDN circuit is here
guaranteed[/I] to have this capacity, but the internal circuit indicates that it will for the likely range of this application.
Click to expand...
 

phaedrus said:
Hi Humber ,
Is the graph I posted the correct one to check this ?Does this mean that the graph is characterized and not actually tested ?
Thanks !!


humber said:
The EDN circuit is here
guaranteed[/I] to have this capacity, but the internal circuit indicates that it will for the likely range of this application.
Click to expand...


Yes. Phaedrus, that is the correct chart. The margin is small, and perhaps cannot be guaranteed to some particular specification, because there will be some change in the offset voltage as the rail is approached, rather than failure of the common-mode capability. Again, it depends on the accuracy you seek.
Click to expand...
 

I agree, that common voltage equal to V+ will most likely work, but it's not guaranteed. This is no problem for a DIY design. But you won't make a production design based on typical parameters.

The graphs are clearly said to give the typical behaviour. Curiously, the graph specifies a typical common mode voltage limit 0.5 V above V+ while the table in the datasheet specifies 0.5 V below. You'll also notice that a number of JFET input OP's has a similar common mode specification, e.g. TL084 or LF356. The TL084 e.g. has a datasheet headline "Common-Mode Input Voltage Range Includes VCC+", but the detailed data table says +/-11 min (= Vcc-4V guaranteed).

The margin is small, and perhaps cannot be guaranteed to some particular specification, because there will be some change in the offset voltage as the rail is approached, rather than failure of the common-mode capability. Again, it depends on the accuracy you seek.
Click to expand...
How do you know that it can't fail completely?

My understanding is, that the manufacturer will keep the option to ship devices with a lower FET threshold voltage to cut costs. If you get a production lot of this kind, you can't complain referring to typical datasheet parameters.
 
Last edited:
There is an inconsistency between the text and the chart. That is not uncommon.

The text says:
"The amplifier will operate with a common-mode input voltage equal to the positive supply; however, the gain bandwidth and slew rate may be decreased in this condition. When the negative common-mode voltage swings to within 3V of the negative supply, an increase in input offset voltage may occur."

I don't think you will find data on input offset against common mode voltage, so that would limit the device to operation within 3V of the rail, if guaranteed performance is a demand. That depends on the required accuracy and speed.
As far as I know, there are no problems using the device in this manner, but Phaedrus should be aware the the simulator model will not include such effects.

In this design, it is not necessary for any input to exceed the rail, but to equal it, and that will only occur when there is no load. If the supply is taken from the input side, then both inputs will be below supply when there is a load.

It is not necessarily true that typical specifications may not be used in mass design, because all other components must be considered if the failure is based on pcb or other unit, and according to cost of rework, versus the cost of and availability of the higher grade components.
TL082's have been used in huge quantities in mixing desks even though the specifications suggest they may not be. Of course, a more expensive device is an option, and there are many single chip solutions.
 
I didn't read the comment on degraded performance near the positive rail. I agree, that it sounds like like a guarantee for non-failure. But getting no exact number about the behaviour, I personally won't rely on it. According to my experience, most trouble in rework respectively redesign is with circuits, that should work according to typical data and other optimistic assumptions. And some is in addition with properties, that should never show, if the datasheet would be true. The bad thing with problems related to process variations is, that they show suddenly and unexpected, but are often affecting a full production lot, because all chips are from the same reel.
 

FvM said:
I didn't read the comment on degraded performance near the positive rail. I agree, that it sounds like like a guarantee for non-failure. But getting no exact number about the behaviour, I personally won't rely on it. According to my experience, most trouble in rework respectively redesign is with circuits, that should work according to typical data and other optimistic assumptions. And some is in addition with properties, that should never show, if the datasheet would be true. The bad thing with problems related to process variations is, that they show suddenly and unexpected, but are often affecting a full production lot, because all chips are from the same reel.
Click to expand...

I don't disagree, but the decision can be based upon labour costs alone.
Anyway, the LF411 has a long pedigree of working as described, but accuracy with a small value sensing resistor may not be adequate.
One obvious solution is to use a charge-pump or some other means to increase the supply, but if that cost or complexity can't be tolerated, then any other means of improvement will face the same problem.

The EDN circuit is good, but the matter then moves to the output limit. In the LF411 circuit, the opamp is clamped to within Vbe of the output, but the opamp in the EDN circuit can only swing to within (some) 30mV of supply.
 

humber said:
There is an inconsistency between the text and the chart. That is not uncommon.

The text says:
"The amplifier will operate with a common-mode input voltage equal to the positive supply; however, the gain bandwidth and slew rate may be decreased in this condition. When the negative common-mode voltage swings to within 3V of the negative supply, an increase in input offset voltage may occur."

Hi Humber,FvM ,
The discussion has become a tad esoteric, atleast for me I think :).
1)"When the negative common-mode voltage swings to within 3V of the negative supply, an increase in input offset voltage may occur."
I could not understand this , what does the "negative common-mode voltage" mean ? Is it the voltage at the load side of the sense resistor(for high side sensing) w.r.t. V- supply of the OpAmp? So if this voltage is always greater than 3V of V-,then there would be no offset voltage increase ?

2) I don't think you will find data on input offset against common mode voltage, so that would limit the device to operation within 3V of the rail, if guaranteed performance is a demand.
So if I am using an op-amp powered by +9V/-9V , then the common mode voltage should be less than 6V to get the specified offset ?
Click to expand...
 

Negative common mode voltage of course is of no meaning for this circuit. The discussion has been about the simple question, if the datasheet guarantees the cicruit to work with input at positive supply, I don't think this is esoteric. Although he datasheet doesn't guarantee, strictly spoken, there are good reasons to assume, that the circuit will work unconditionally anyway. But you are able to read the datasheet and can decide on your own.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top