X10 probe mystery; distorted full-wave output!

[fezder]

OK, so I have really simple measurement; feeding 12vac from transformer, and using db103 as bridge. I get just what I expect when using x1 attenuation, but x10....what is this? :O
Frequency is low, so shouldn't be bandwidth issue (50hz and 100Mhz probe) since it works in x1 which should suck in bandwidth. Even 60Mhz gave this result. Using transformer, so no ground issue either. I measure directly from output of bridge. And just in case tested also with analog 'scope; SAME thing!
Any ideas?I can post photo of circuit, but since on x1 it measures just fine, I doubt It's circuit but something else I can't think of.
x1:

x10:

edit: changed pictures from .BMP to .JPEG

 

[fezder]

Odd! With 1k load problem is gone. But why is this happening?

 

[Nigel Goodwin]

No idea, but post the pictures as something sensible, not huge BMP's, use PNG instead (or even JPG).

 

[JimB]

I suggest that there is something a bit odd about the rectifier.

Look at the X1 trace, alternate "dips" do not go down to 0v.
Maybe a leaky diode?

Combine the higher impedance of the X10 probe with the leaky diode, and the alternate dips can just float at a higher voltage.

When you put your 1k load, the small leakage of the diode is completely swamped by the current drain of the load.
Also, with the 1k load, I guess that the alternate dip discrepancy disappears.

JimB

 

[fezder]

Nigel, changed pictures now, Didn't realise such size difference!.
Jim, yeah, that dip on x1 was improved with 1k load. Leaky diode doesn't sound good. This happened also with bridge made of 1n4007 diodes. This has never occured before.