I picked up the April 2015 QST and found an article written by Veikko, N7VK about a mini curve tracer and decided to build it. It took some learning as I haven't touched the 8051 in a while but with a little help from Veikko, I got it running. First, I am amazed that more people haven't built this curve tracer. Veikko wrote almost 6k lines of code on the PC application side and overall it works very well. I've been using it to explore various active components and plan to use it to check for leaky FETs in HP volt meters I am repairing.
In playing around with it I found something that I've never noticed about zener diodes. First, they aren't that accurate, around 5%. Second, as you can see in the attached trace, the knee varied between zener voltages with the lower values having a smoother knee vs the higher values. I checked several packs of the cheap eBay specials and they track very closely.
The trace on the left is for a 2.2V and the right, 6.8V. The little circle on the 6.8V trace shows the zener voltage at 7.004V and the 2.2V trace is actually 2.275. Both of these voltages compare closely to a test circuit I made.
Is this typical?
Thanks.