Voltage Reference with Zener Diode

Go To Last Post
9 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello,

I'm on a university vehicle competition team, and we inherited a set of hardware that we didn't design. I've been trying to figure out what's wrong with a certain piece of the circuit, and I was hoping you guys could help me out.

We have an ATMega48 with some sensors attached, which use the analog-digital converter to measure voltage and current (via Hall Effect sensors). The problem is, the circuitry to provide the reference voltage isn't outputting nearly the correct voltage. It uses resistors and a Zener diode to step 3.3 volts down to 2.5; but the pin is only seeing about 0.7 volts. The schematic is attached. Sorry that it's a little messy, but the diode in question is U9.

I doubt the actual design of the circuitry is at fault, but I don't know what the most likely places it would fail are. I tried replacing the diode and checking the resistors; both seem to be okay. How likely is it that the diode was damaged when it was soldered?

The software, by the way, seems to confirm the fact that the voltage reference is too low -- the ADC maxes out at a much lower voltage than it should, and the current sensors are reading 0x03FF when they're only outputting 1.4 volts.

Any thoughts as to where the failure might be?

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Firstly, the LM385 is probably quite a bit better than the average zener diode. I can think of two possibilities in regarss to your problem:
1. the LM385 reference is in backwards.
2. You've enabled the internal reference in the AVR and that's being output on the AREF pin.

Isolate the LM385 from the AVR and see what you measure - that will identify who is at fault. Divide and conquer!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would guess that the series resistor is too large. With 3.3V in and 2.5V out, and 5.1K., your maximum load it 0.8V/5.1K = 160uA.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The 385 is spec'd from 20uA to 20mA so if it is reading too low and it is the right way around, something else is stealing the current. AREF needs 70uA so there is still 87uA left for the diode. What else is connected to the "ref" node?

Klave

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Actually, that 0.7V is a bit suspicious, being the approximate forward voltage drop of a regular diode. Is it installed with the correct polarity?

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the advice everybody. I'm not actually going to be able to test anything until later this week, but I will post results once I do.

"ref" doesn't actually lead anywhere else on the schematic, strangely enough. I was puzzled about that. My best guess is that it's just a marker to signal where to measure the reference voltage -- but I'll check out the hardware once I have access.

I should also add that I haven't actually taken any circuits courses yet, so there are some gaps in my knowledge about this stuff. The diode has only one correct orientation on the circuitboard, but it's possible that the board was designed incorrectly. What is the best way to test the actual orientation of this type of diode? Just put a voltage less than 2.5V across the terminals and see which way allows current to flow?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Don't be surprised if its broken. I've had rotten luck with some precision references in the past.

oddbudman

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Which package is it. There is TO92 which looks like a plastic transistor with one flat side. There is the SOT23 which is rectangular surface mount with 3 legs, two on corners on one side and one in the middle of the opposite side. The third is the SO package which looks like a small 8-pin surface mount IC. Uggh, there is also a small metal can version (TO46) and a 16 "pin" surface mount version.

It is most likely you have the TO92 version and the center pin is the cathode (the positive end of the "zener" diode). If you have the metal can one, the cathode is closest to the little orientation tab on the rim of the metal case.

You can find the spec sheet with all of the pinout information here:

http://www.national.com/ds/LM/LM...

The spec sheet actually snows that "forward" drop at 200uA of current is just under 0.6V, typically.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I got a chance to test the system out a couple of days ago, and it turns out that the board was actually designed incorrectly -- the diode (in an SOT23 package) was on backwards. Luckily it was easy to fix since the diode doesn't use the side of the package with only one pin.

Thanks again for all the help.