Maker Pro
Maker Pro

Voltage Reference Supply Help

Good Morning,

I have inherited a board design that will take Ethernet packets and convert the commands to analog voltages. My task is to make it work (as it doesn't currently work). I'm have an extremely hard time getting a specific part of the board to operate as I'd expect it to (and I assume as the original designer would expect). BLUF is that the 2.5 voltage reference device is spitting out 10V instead of 2.5V!!!! I presume it's a bad device, but I've changed it three times now with no success, I think it's frying as soon as I power up. See the details below.

I have attached a schematic of the portion that is causing me to pull out my hair. You should know that the "2.5V_REF" is coming from an AD1582 voltage reference, the "DAC_OUT#" is the output of a DAC, feeding the OpAmp channels. The OpAmp is an LT1053 device and is powered (+V) by 12V and (V-) not connected (R-to-R is 0V to 12V). "An#_VO" simply fo to output pins that will drive an analog device off-board. Basically, the "client" expects a voltage between 2.5V and 7V (hence the 2.5V bias). Like I said up front, the issue is that the 2.5V_REF (used to bias) is putting out 10V instead of 2.5V. Thinking the OpAmp matrix was causing feedback, I used the extra OpAmp on the LT1053 as a buffer for the 2.5V reference (as noted in the attachment), but this did not fix the issue. I know there is a design issue, but nevertheless, I cannot figure out why this is happening.

I appreciate your help.

R/
Nick
 

Attachments

  • OpAmp_Matrix.png
    OpAmp_Matrix.png
    59.2 KB · Views: 106

davenn

Moderator
HI Nicholas
welcome to EP :)

agree with Bob ....

you have told us that the 2.5V reference is your problem
but you give circuit info for everything except the 2.5 V reference

cheers
Dave
 
Ha, my apologies! I figured out the issue about 5 minutes ago, but I appreciate your interest. For completeness, I will answer your questions and explain the huge mistake the original designer made.

As mentioned, the device is an AD1582 and it is driven by a 12V supply. Basically, 12V is fed to the input, the output (2.5V) biases the DAC output and feeds the OpAmp (as shown in the schematic, "2.5V_REF"). The GND is connected to GND... well, it is now! The original designer was "decoupling" the GND pin of the voltage reference source to GND, haha (see attachment). I think he meant to decouple the 12V supply to GND, so I made that fix. Everything seems to be working good now, not sure how I missed that when I first inherited this project.

Thanks, and I will be sure to think harder about what information to include!

R/
Nick
 

Attachments

  • VoltageRef_Mistake.png
    VoltageRef_Mistake.png
    10.7 KB · Views: 128
Top