Voltage translation and AD623

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi Folks,
Need some advice on voltage translation. I have a pressure transducer, output 1-6V. What I need is a circuit to simply translate this to 0V-5V ( no gain initially ) for input into an ADC with possibly some over volt protection. I was under the impression I could use the AD623 to do the job simply, but after looking at the datasheet I'm not so sure. With a 5V supply to the AD623, if I apply 1V to one of the input pins and connect the other pin to the transducer that outputs 1V at 0psi then I should get 0V out of the amplifier, fine. But according to the datasheet ( graph of voltage output VS common mode input ) I can only expect a maximum of 2.5V output. Am I reading this correctly? If both inputs start off at 1V and the other increases, am I right in assuming that the common mode voltage in this case 1V? Another unkown is that will the AD623 allow an input of 6V when the supply is only 5V?

Thanks.
 

Attachments

t_n_k

Joined Mar 6, 2009
5,455
You have a single +5V supply?

The input common mode range is then from -0.15V to 3.5V. See page 4.

The output swing depends on the supply value and the loading.

With a 100kΩ load the output can swing to within 150mV of the +ve supply rail. If the load is 10kΩ the output can swing to within 0.5V of the +ve supply rail. So the differential input range (5V) would be beyond the device capabilities with a +5V supply. With unity gain the input differential limit would be 4.85V at 100k and 4.5V at 10k load.
 
Last edited:

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi, Apologies for the late reply, I hadn't subscribed to the thread... :rolleyes: As for the transducer, unfortunately the product datasheets I have don't give the output impedance, I've attached the .pdf to give you some idea. I have access to a 12V auto battery supply which I could use with the AD623 but as it is unregulated ( could go to +13.5 ) it may end up destroying the IC as it's only rated for 12V max. I was thinking of using a 7V or 8V regulator off the 12V battery ( allowing a couple of volts drop but with enough head at 10V to still drive the regulators ) as the supply to the AD623 which I believe would help towards achieving the input/output ranges I require, but I'm not sure.:confused: This would also mean that the potential +5V from the AD623 would need to be 'clipped' as it will be fed into an ADC pin on a microcontroller. I'm assuming I could use some form of zener diode to achieve this but as far as I understand, they take a relatively large amount of current to drive and am not sure if the AD623 can supply enough current and/or how this will affect the input voltage signal to the microcontroller...

Thanks.
 

Attachments

Thread Starter

saiello

Joined Jul 1, 2007
24
Yes, that would be the most obvious solution.... unfortunately I have a number of these pressure transducers that I obtained on the cheap and am not about to throw away a few hundred pounds worth of equipment save for want of a simple circuit to make use of them....
 

Ron H

Joined Apr 14, 2005
7,063
We can design a level shifter for you. How accurate does it need to be?
What do you have in the way of test equipment?
Is it OK if it has a pot or two that you have to set for initial calibration?
 

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi again,
I've attached a schematic of the sort of thing I'm after. It is more of an idealised schematic in that it indicates ultimately what I would like the circuit to do rather than electrical correctness...( I'm more of a software guy than electronics...;)). The battery supply is from an auto battery, so I would like the circuit to be able to operate using a potentially wider voltage window than the ideal 12V. I would also like to be able to use 0V-5V transducers with this circuit, hence the jumper. Also, there is a jumper to include a pot for potential gain control. The output from the AD623 will need to be clipped for input to the M/C hence the zener diode, but I'm not sure if this is a valid way of going about things. I would like the circuit output to extend the full span of 0V-5V ( or as close as possible without compromising linearity too much, if that is an issue ) to match the input voltage range of the M/C's ADC. In essence, I am trying to 'future proof' the circuit so that I have something robust, reliable and flexible enough so that I may use it again with other types/ranges of sensor without having to make any modifications.
In terms of accuracy, linearity is the most important, the lower voltage reading corresponding to 0.0psi doesn't really have to be exactly 0V as there will be the opportunity to calibrate the sensor with a program running on the M/C.
This circuit is going to be part of a larger 'in-house' microcontroller/graphic LCD based unit that I am designing for a group I belong to which will be used to monitor engine operating parameters such as speed, temperatures, pressures, etc ( see attached pic ). The circuit needs to be relatively tight too, as it has to squeeze onto a single proto board ( below M/C in pic ) amongst all the other electronic components which will use wires to interconnect.


Thanks.

P.S. the only electrical testing equipment I have is a cheapo multimeter...:)
 

Attachments

Last edited:

Kermit2

Joined Feb 5, 2010
4,162
Doesn't the micro controller you are using have an A to D conversion ability?? Maybe you should go that route and let the micro do the decoding in software.
 

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi,
Yes, it does ( 10-bit ADC ), but I want to be able to input the full range of a transducer. As well as this, I'd like to have the facility to increase it's sensitivity by using gain so that I can get better resolution for a lower portion of it's range. If I connect the transducer directly to the M/C then I'll be chopping off the top 20% of it's range as it'll only be able to input 1V-5V and I will still need to address the issue of clipping the transducer output to protect the M/C ADC inputs. I thought it best to have a single 'package' that will do all the jobs needed for a safe signal input to the M/C and be flexible enough that will allow it to be useful for other sensor connectivity in the future..:)

Thanks.
 
Last edited:

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi,
No, I just thought an instrumentation amplifier like the AD623 is designed for this type of job and would be best suited. I'm open to other solutions though, as long as the component count can be kept to a minimum as it's going to be a tight squeeze what with all the other sensor electronics...
 

Ron H

Joined Apr 14, 2005
7,063
What is the required bandwidth (frequency response)? Or, how fast can the output of your pressure transducer change?
 

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi,
According to the limited information in the leaflet that accompanies the transducer, the response time is <1ms but this would be far and above my requirements. The types of pressures I will be reading will be relatively slow changing, I'd be happy with a response time of no greater than 50ms.
 

Ron H

Joined Apr 14, 2005
7,063
For what you want, I would use a single supply, 5V inamp with rail-to-rail output. Then you don't need overvoltage protection. My choice would be INA326, but it's a little pricey.
You could probably use a dual, single-supply, rail-to-rail op amp running on +5V, configured as a differential amp. It won't be as precise as the inamp, but it's cheaper. See attached.
A couple of pots will allow you to vary gain and offset. Varying gain will vary offset, but varying offset will not affect gain.
Digikey has MCP6242 for US 40 cents in quantity of one.
I've included the .ASC file, if anyone wants to run simulations in LTspice.
 

Attachments

Thread Starter

saiello

Joined Jul 1, 2007
24
Hi,
The INA326 would be better as you say, but doesn't appear to come in a DIP package and is a tad expensive for what I want. For your circuit, would the use of high precision resistors be of any benefit? For pots, I am assuming these can be used to replace R5/R6 for offset and R3/R4 for gain? If I use a gain pot how would I go about re-calculating the needed offset?

Thanks again.
 

Ron H

Joined Apr 14, 2005
7,063
High precision resistors can give you better temperature stability, but 1% resistors are typically 100 ppm/°C, which ain't bad. I think you would need a pot for offset in any case. I just put fixed resistors in the original schematic to illustrate the technique.
As I mentioned, changing the gain does not change the offset, but I think I see what you mean. Changing the gain changes the required offset.

The equation for the output voltage is Vin*(R2/(R1+R2))*(R3+R4)/R3 - Vos*(R4/R3).
 

Attachments

Top