Op-amp reference at zero volts is wrong?

[grizedale]

Hello,

The schematic of figure 2, page 4 of the following shows current regulation using a low-side current sense resistor, being read by an op-amp with a reference of ground.


https://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00003244.pdf

Surely , opamps cannot read things near their ground potential , and this circuit is wrong?

 

[Prototyp_V1.0]

Surely it can.

 

[grizedale]

i am very sorry to say it but those are just absolute max and mins...theyve nothing to do with what the opamp can do in performance terms

 

[FvM]

Surely , opamps cannot read things near their ground potential , and this circuit is wrong?

Many OPs can, mainly single supply or rail-to-rail input types. The TSM1011 specification clarifies, that the CC opamp has a common mode voltage range of 0 to Vcc -1.5 V.

i am very sorry to say it but those are just absolute max and mins...theyve nothing to do with what the opamp can do in performance terms

There are more specifications than maximum ratings. Paragraph 4 ELECTRICAL CHARACTERISTICS has the answer.
P.S.:

It depends on the input structure of the op amp.

Yes, in this case it's a bipolar PNP input stage, most likely similar to LM358/324.

 

[mtwieg]

It depends on the input structure of the op amp. If the inputs use p channel devices or complementary paired inputs than they should be able to work with inputs at 0V. Most devices can. But the datasheet will say for sure.

 

[grizedale]

i understand that it can work with input at zero volts....but surely the opamp will not be performing in a quality way with an input at zero volts?...


surely there must be degradation in performance.?.....after all, why would we use so many split supply opamps if operation with inputs at ground was ok?

 

[FvM]

but surely the opamp will not be performing in a quality way with an input at zero volts?

You can analyze the transistor level circuit of the LM324 input stage if you are interested in details. In a short, the performance will be almost independent of common mode voltage over the specified range. In so far, your assumption doesn't approve. Of course, these amplifiers aren't highest performance, but serving their purpose.

P.S.: For LM324 transistor level circuit, see https://www.edaboard.com/threads/221016/#post941668

 

[Prototyp_V1.0]

Now you ask a different question than in your first post and then you claim i'm wrong.
I answered you on a direct question that wans't to misunderstood. I call 0.3V to be near ground level. You didn't used the words "at" or "below".

If you uncertain about performance then just use a breadboard and make run a test on it. There should exist some datasheet with any neccesary curves that also can answer that.

 

[grizedale]

sorry prototyp_v1.0

my confusion was making me respond with out thinking, as i was in a state of confusion.

-thanks for the tips on the op-amp issue though

I am actually looking for an op-amp like the following, but with the current source and reference inside a single chip, but i cant find it.....

https://i56.tinypic.com/153uqfs.jpg

this circuit is "invisible " during normal circuit operation, but would step in and provide average current mode control in the event of an output overload.

the output is a full bridge secondary.

This circuit is needed because even though i have a primary side current sense transformer to handle output overloads.......that, in conjunction with the overcurrent input on the MC34025, would give problems due to the capacitor in series with the primary....this is due to the fact that the full-bridge becomes unstable when controlled by cycle-by-cycle current limiting, as occurs in overload........overload is a normal operation of this full-bridge smps, becuase it powers a guitar amplifier.

.....so the above average current control circuit would step in and take over, and provide smooth operation of the full bridge smps in times of overload....but i cant find a chip with this stuff integrated...and its too many components to do discrete.

 

[Orson Cart]

The LM358 / LM324 series of op-amps operate correctly for inputs down to -0.3V below their negative supply. This is why they are great for current sensing!

 

[grizedale]

The thing is, as Prototyp_v1.0 kindly pointed out, the pin has a limit of -0.3V.

...so what if a huge current surge suddenly happens.......the pin will go below -0.3V ...and bang!

....this is only average current mode control, and not peak current mode which would suddenly stop the current.......so i certainly cannot use the circuit as in my first thread.

I must use a circuit like i just showed, but its loads of components.

i am surprised nobody in the world makes anything like this

 

[FvM]

...so what if a huge current surge suddenly happens.......the pin will go below -0.3V ...and bang!

Your post reminds me to emphasize the difference between maximum ratings and specified operation conditions once more.
- you should keep the operation conditions, e.g. common mode voltage range to guarantee it's working correctly
- you have to strictly keep the maximum ratings to avoid permanent part damage

For the said LM324, the datasheet does not guarantee correct operation with input voltages below 0. It specifies 0 as minimum common mode voltage. But I can confirm, that the part will typically work downto -0.3 input voltage. It's no problem however, to design a current sense circuit, that keeps the 0 V as well.

Maximum ratings are a different point. The -0.3 V limit is a standard specification, that can be translated as follows:
- the pin has a substrate diode that is forward biased when applying a negative voltage
- exceeding the 0.3 V specification won't damage the chip as long as the current is limited
- the tolerated substrate diode current is ruled by different mechanisms
- diode current capability
- risk of chip latchup by forward biasing parasitic transistor structures (can also destroy the chip)
- below possible damage, forward biasing of parasitic structures can affect the operation of other amplifiers on the same chip

Modern OPs or other analog parts are often designed to utilize the substrate diodes (of both polarities) as protection means, they come with a maximum current rating for the inputs, e.g. 10 mA. It's your part to place suitable current limiting resistors. In your above linked circuit, the 1 k resistor can act this way, so there's little risk to damage the input amplifier, even in case of overload.

 

[Orson Cart]

so what if a huge current surge suddenly happens

perhaps you forget the current is always limited by the o/p inductor - so it cannot jump to huge values very quickly, a well dsigned average current mode control circuit will protect the mosfets against an o/p short quite well - we have designed and built a 600W unit for a client (half bridge) that works very well - it has been tested into an automated repeated short circuit (large relay) for 1000 times. It has very good transient response also and uses the large electrolytics to split the rectified 230 volt line.

Also regarding the input to the opamp at -0.3 volt, if there is a large over current it is the resistors into the opamp that limit the worst case current into the opamp pins, so say- 5V is worst case (-0.3 is normal maximum + 20%) then a 5k1 resistor will limit the pin current to 1mA which the opamp will survive quite happily during the short term overload,
Regards, Orson Cart.