Dual Rail Op Amp Shutdown design

[lk.dgironi]

Hello,

I'm working on a board where I need to shutdown an op amp.
It's used as unipolar to bipolar converter, 0 5 V to -10 0 10 V. The op amp is driven by a DAC, I've to shutdown it waiting for the microcontroller to set the DAC to 2.5V output, which is the output that lets the op amp sit to 0 V.
Op amp supply is -12 0 12 V. I'm using a common LM358.

I came up with three ideas:
1) to cut the +12 and -12 supply line to the op amp
2) shutdown the output using a N-Channel to GND
3) use an op amp with shutdown pin, but I've I prefer a jelly bean op amp (like the LM358)

About the point 1) you can find attached the schematic, I think it could work.

I'm more on the software side, so I ask your help here.

Thanks!

[Kleinstein]

What is the purpose of the shut down ? 
There are some OP with an shut down function. This can be used for 2 purposes: one is reducing the power usage and the other is to get a higher impedance at the output. This can still be tricky because of the feedback path.
For the 2nd case it may be easier to keep the OP powered and have a Photomos switch at the output.

[lk.dgironi]

Thanks @Kleinstein

When 0 V is feeded in the non inverting input, the output of the op amp reachs -10 V.
When 2.5 V is feeded, 0 V is the output.
The output is connected to an industrial device (let's say a motor), peak uncontrolled voltage (like the startup -10 V) may damage the machine.
At the startup stage, the micro is not driving the DAC for a few cycles, but I need output of the system to be 0 V, so input of the op amp needs to be 2.5 V.
I hope I explained well.

I've search for op amp with shutdown, but I prefer to use common op amp cause in production this is simpler. Of course, if it will be possible.

[Kleinstein]

So the requirement is to have a safe startup, not saving power or getting a high impedance output.  In this case I would prefer to have the extra switching between the DAC and amplifier. E.g. have the signal clamped down by default. Some DACs may have a shutdon function that could do this job.

It is unusual to a machine controlled with -10 V, the usual way is 0..+10 V.

[lk.dgironi]

Yes, a safe startup! Power saving or high output impedance is not a problem.
The problem is that if I shutdown the DAC, or clamp the signal down, I've a 0 V output, which means -10 V output from the op amp.
Find attached the voltage conversion diagram.
Reference design is TI paper SLAA869 https://www.ti.com/lit/an/slaa869/slaa869.pdf

That said, the op amp supply switch may seems the only option, right?

I know 0..10V or 4..20mA is the standard, but the -10 0 10 V is in some way common in the testing machines which I'm working on 

[Nominal Animal]

I know nothing, but:

Why not use something like a pair of AD5761Rs instead of the AD5663?  Then the opamps would simply be buffers, without any offset needed, as the AD5761R can only output a few mA.  The AD5761R can be daisy-chained, and use the same single one ADUM1401 isolator for the SPI communications to both.  Each device takes 24 bits on the SPI bus, so if two are daisy-chained, 48 bits must be transferred.  When the /SYNC line is pulled high after those 48 bits, both AD5761R's will switch to the new output voltage.

At startup and reset, AD5761R has the output clamped to AGND (via an internal 1kOhm resistor) and output amplifier unpowered.  Power up state (and if you want to use it, /CLEAR state) can be chosen from minimum, midscale, and maximum.

[TaylorD93]

Does the Opamp output ever need to go negative during normal operation? I.e. you only use 0-10V output, but have +/-12V to get near zero volts out?

If so, fit a series diode on the output of the LM358D, then take feedback resistor from R64 to the Cathode of the Diode instead of the Anod (see attached sketch) then replicate on the other channel with R61

That way the opamp can still be driven from a +/- supply and will get 0V to 10V out as the feedback will compensate for the Vf


If you do need +/-10V output, then ignore this 

edit: re-read OP, adjusted post.

[TaylorD93]

Or maybe another option, use a DG411 Analog switch.

On power up or when you want the opamp "disabled" connect its input via the DG411 to a Vref to give you 0V out (instead of -10)

Then switch over to the DAC output after a trigger pulse is received. But have it so the enable of one is the inverse of the other, that way you never short Vref to DAC Out via the DG411

[lk.dgironi]

@Nominal Animal

Thank could be an option, we have some AD5663 around and I've worked on that to speed up the development.
If I'm not able to shutdown I could think about shifting the DAC.

[lk.dgironi]

@TaylorD93

Thanks for suggestion. The output must work between -10 to 10 V during operation, so the diode could not be an option.
About the DG411, wouldn't be simpler the mosfet solution?

[Nominal Animal]

How about using a pair of AD5663s instead, each driving the differential inputs of a single buffer opamp?

With ±5V output on each, the difference would swing ±10V, and since the reset state is the same for both outputs, you'd get default zero output too.
(Sure, there would be a lot of code pairs to create the same output, but at least it would behave very nicely at startup, without any need for the MOSFETs.)

Since AD5663s cannot be daisy-chained, you would need separate /SYNC signals, though; so ADUM130 or Si8640 or similar 4:0 isolator needed.

[lk.dgironi]

@Nominal Animal

That's nice, also I will have one more bit of accuracy.
But the problem is that the MUST work in symphony, cause a + - signal may happens in the same time. And i doubt that's a bit too dangerous. I've to think about this.

[TaylorD93]

Depends, you still have a discrete enabling your Opto on the schmatic, so you could use the same signal but control the inputs.

That way with the PSU's on all the time, your less likely to get any "odd" goings on.

Slightly different, but I have a Cambridge Audio amp which when you power it down, the opamp input stage oscilates and you get a "zip" noise like a jacket zipper, through the speakers unless the volume is at 0.

So IMO switching PSU rails is less desirable in any control setup as the output of the opamp becomes unknown possibly unstable.

Hence changing the inputs only you IMO have less risk of a noisy or uncontrollable output.

[langwadt]

switch the DAC to the version that powers up midscale instead of zero

[lk.dgironi]

Depends, you still have a discrete enabling your Opto on the schmatic, so you could use the same signal but control the inputs.

That is right  I've not think this
I think the solution is to switch to a dual rail output DAC like @Nominal Animal and @langwadt suggested!

Monday I'll take a look on this... and when done, I'll let you know.

[Nominal Animal]

That's nice, also I will have one more bit of accuracy.

Well, not really.  (Basically, it is lost in the common mode values, which comprise half of all valid value pairs.)

Hey, I think I have a better option.

Use a single MOSFET to switch the 5V line based on the optoisolator output; configured so that at startup (and when the microcontroller side is not powered but this side is), we have a 0V line, and 5V only when we get the enable signal from the microcontroller.  Let's say there is a ferrite bead and a capacitor to filter this MOSFET-controlled 5V line, and call it the reference voltage.
(One could use another LM358D opamp pair to buffer the 5V reference line, but I don't know if that is necessary; I too am just a hobbyist on the electronics side.)

This reference voltage line is used for the VREF, and for the noninverting inputs of the two opamps.

The AD5663 inputs are connected to the inverting inputs of the opamps, with the opamps in differential amplifier configuration.  Each output can handle a load of 2kOhm (or larger), so let's make the resistor between the AD5663 output and the inverting input of the opamp at least that.  To get amplification by factor 4, the feedback resistor for the opamp needs to be four times that.  Let's say they are 2.5kOhm and 10kOhm (0.1%).  On the noninverting side, we need a voltage divider where the resistor between MOSFET-controlled 5V line and the noninverting input is exactly one and a half times the value of the resistor between the noninverting input and ground; for example, 5.0kOhm and 7.5kOhm.  This way it will be doubled.

With this, when the reference voltage is low (0V), both outputs are 0V too.

When the reference voltage is 5V, the output swings from +10V to -10V.  DAC code 0x0000=+10V, 0x4000=+5V, 0x8000=0V, 0xC000=-5V, 0xFFFF=-10V.

You could even use only +12V, GND, and -12V, by using a voltage reference between +12V and GND for the 5V reference voltage, perhaps buffered by an opamp per noninverting input to a difference amplifier (and the VREF to AD5663 directly from the shut); but with the MOSFET as a switch between the reference and +12V (possibly with a transistor to invert the gate logic if using a P-channel MOSFET) basically turning off the reference unless ENABLE signal is present.  This would give better precision, I think.

[David Hess]

If the AD5663 is replaced with the AD5663BRMZ-1, then power up reset will result in a midscale output of 2.5 volts instead of 0.0 volts.

[lk.dgironi]

@Nominal Animal
I think I've to setup a breadboard, and make a few tries, although AD5761R seems a better option in this case. Attached a AD5761R in daisy chain version, I think a unity gain buffer like this should work.

@David Hess
AD5663BRMZ-1 seems a reasonable option! I've not found this info in the datasheet until you point it out. I'll try this.

[David Hess]

@David Hess
AD5663BRMZ-1 seems a reasonable option! I've not found this info in the datasheet until you point it out. I'll try this.

I had to look it up but a lot of modern DACs include versions which reset to zero or half scale.  On some it is configured by strapping a pin.

[lk.dgironi]

I've decided to build a board with two design.
Then I'll decide which one to maintain for bipolar.

Find attached both the board.
Do you find any issue?

Thanks!

[Ian.M]

Does anyone apart from David *EVER* read the datasheet before designing in a part or commenting on a design?   
This one line mic-drop was beautiful to see, though I do wish I'd seen it before reading page 15 of said datasheet then posting the original of this comment! 

If the AD5663 is replaced with the AD5663BRMZ-1, then power up reset will result in a midscale output of 2.5 volts instead of 0.0 volts.

Octopart claims Rochester Electronics and a few non-authorized distributors still have AD5663BRMZ-1 stock. 

[langwadt]

Does anyone apart from David *EVER* read the datasheet before designing in a part or commenting on a design?   
This one line mic-drop was beautiful to see, though I do wish I'd seen it before reading page 15 of said datasheet then posting the original of this comment! 

it is on the front page of the datasheet: "Power-on reset to zero scale or midscale"

[lk.dgironi]

Guilty, I'ven't read the datasheet well. "Funny" part is that as @langwadt says, it's on the from page ("Power-on reset to zero scale or midscale"), but I've noticed it only after David comment 

Anyway I'll use the AD5663BRMZ-1 of course on the first design. And on the second a dual rail DAC, the AD5761R.

[lk.dgironi]

Hi all, just for info.
I've finally used AD5663BRMZ-1. AD5761R has too long delivery times
It works, but not yet tested on production.