So I've decided to try to "calibrate" a little my Rigol DS1052E o-scope. I'm using stock probes, both connected to the 1khz square wave generator (calibrator?) on the front panel.

I've really zoomed in the voltage, and noticed things that obviously should not be.

1. The voltage readings are different. There are resistors inside the probe, right? Maybe they are just not matched super closely?

2. The bottom of the square wave is not flat. This is a problem with capacitance adjustment in the probe? There is a small adjusting hole on the probe that I haven't touched since it came from the factory.

Any ideas welcome.

 

Get a small screwdriver and turn the trimmer capacitor and see what happens.
There is no harm done in adjusting the trimmer as far left or right or completely 360°.

When scope probes are shipped from the factory they are not adjusted to your scope. This is something the user is required to do.

 

Since the test signal is a square wave, be sure the channels are set for DC coupling as well.

 

Good point, ifixit. I forgot about that one.

 

ifixit, I'm most definitely in DC mode. Even though the voltage clearly goes into the negative territory. -110mv for one probe, and -48mv for the other probe.

Could be that the square wave generator is just "cheap"?

MrChips, I saw the page in the manual for calibrating the probe, but if I follow their procedure, my probes look just fine as they came from the factory. The manual tells to look at the full wave on the screen, with vertical set at 1v, and horizontal at 200us. So the full wave is visible. At that resolution it looks perfectly square.

In my test I set vertical to 50mv per division, so I could see the bottom of the wave much better. And it's certainly not very flat. I'll try tweaking the adjuster still though.

 

Are these cheap probes from China?

 

Are these cheap probes from China?

Given that Rigol is a Chinese company, I'd say there is a good chance that you are correct. When China lets the currency loose, they won't be cheap, just Chinese.

 

Some things to try...

1. Measure cal signal by directly connecting it to the scope input with a wire.
2. Using the probe, measure a simple battery of known voltage (Both polarities).
3. Post the results.

 

Some things to try...

1. Measure cal signal by directly connecting it to the scope input with a wire.
2. Using the probe, measure a simple battery of known voltage (Both polarities).
3. Post the results.

+1 This is standard procedure with me.

 

ifixit, I'm most definitely in DC mode. Even though the voltage clearly goes into the negative territory. -110mv for one probe, and -48mv for the other probe.

Could be that the square wave generator is just "cheap"?

MrChips, I saw the page in the manual for calibrating the probe, but if I follow their procedure, my probes look just fine as they came from the factory. The manual tells to look at the full wave on the screen, with vertical set at 1v, and horizontal at 200us. So the full wave is visible. At that resolution it looks perfectly square.

In my test I set vertical to 50mv per division, so I could see the bottom of the wave much better. And it's certainly not very flat. I'll try tweaking the adjuster still though.

When you adjust the vertical scale to 50mV/div, you're causing the input circuit to be overdriven when the voltage rises to the top of the pulse. Older analog scopes used to be very good at recovering from this overdrive condition. Many (most) modern digital scope do NOT recover well from overdrive. That is likely part of what you're seeing. In other words, don't overdrive the inputs.

 

When you adjust the vertical scale to 50mV/div, you're causing the input circuit to be overdriven when the voltage rises to the top of the pulse. Older analog scopes used to be very good at recovering from this overdrive condition. Many (most) modern digital scope do NOT recover well from overdrive. That is likely part of what you're seeing. In other words, don't overdrive the inputs.

That's a very interesting theory! Is this a side effect of being a low end o-scope? As I'd imagine there are many situations where one would want accurate readings while over-driving the inputs.

 

Swap the probes around (Ch 1 probe to Chan2, Chan 2 probe to Chan 1) and see if the condition follows the probe or stays on the same channel.