GSSL THD measurements. Weird behavior.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Ilya

Well-known member
Joined
Feb 25, 2005
Messages
750
Location
Moscow
I'm stuck. I hope somebody of the gurus here could chime in and give some kind of explanation.
What happened was this. I decided to measure my clone for frequency response, thd, imd etc. The thing measured really good but...
When in bypass, if I turn the threshold knob all the way clockwise, thd is around 0,004% which is pretty cool. Now, when I turn it fully counter-clockwise, thd grows up to 0,37% (92 times!). I'm using Spectralab for measuring, the fft graph shows only odd harmonics in cw position, and odd (but greater amplitude) and even harmonics in ccw position.
I'm usung 2181LAs all the way through. And I've done all the mods and calibration suggested by martie-08 (I can't find the link right now).
Oh, and the unit is trimmed for minimum thd (which is actually 2nd harmonic only).
Actually that's not all. I've measured the second gssl which is identical to the first one but doesn't have the mods and hasn't been calibrated yet. In cw position it measures slightly worse, 0.007% thd. And the threshold behaves the same except it has max thd in the middle position (I think that's because I don't have 47k resistor soldered to the threshold pot - look at marties site for more info). And it acts kinda like thd "blender": in cw it has only odd harmonics, in ccw - mostly even, and in the middle it has both.
I have no idea if this behavior is normal or if I did something wrong (though I'm pretty sure I haven't screwed anything). Looking at schematic I can't tell if this is the right thing or wrong.
Maybe somebody can measure their clones as well to see how it reacts.

P. S. I'm using RME FireFace 800 as a reference audio interface.
 
EDIT: Distortion figures are as follows: from 0.02% (threshold fully clockwise) to 0,16% (threshold fully counterclockwise)
 
UPDATE.
I found 2 more interesting facts regarding this fenomena:
1. Threshold pot vs. THD increase behavior is highly dependant on ratio. At 2:1 THD changes from 0.02 to 0.17; at 4:1 from 0.003 to 0.30 to 0.17; at 10:1 from 0.004 to 0.56(!) to 0.25. Very confusing.
2. I've traced the distortion to sidechain. When I pull out the sidechain VCA the THD measures about 0.004% and doesn't depend on any controls settings.

I've remeasured both clones in every possible combination (girst/second motherboard/daughterboard combinations), and the results they are showing are pretty the same.

As I said before I'm using 2181LAs throughout. So I guess I'll try to trim the sidechain 2181 and see what I get.

Any ideas?
 
That's absolutely correct.
But this thing I'm describing happens in BYPASS mode.
I can't understand that...
 
I realize this is a bit of an old thread, but I have the same problem. When the unit is bypassed there is a lot of noticeable odd harmonics added, when the threshold is turned all the way to the left (CCW). When the threshold is turned CW to the right, sub harmonics are generated. Again, this is only in bypass.
 
Back
Top