What comes to DCX's other features, I noticed that only way to mess the sound is to use auto-align. Does your DCX also give different delay-settings every time with auto-align?...Where is the problem? Used mic is Behringer ECM8000 measurement mic.
I use the same mic. I think the auto-align stuff is just about useless. The corrections for my system were just plain wrong. I don't think its nearly sophisticated enough to deal with the real world acoustics of a small listening space or speakers with closely aligned drivers.
That said, the digital delays look pretty bullet proof. You just need to measure and input the distance you need and the DCX does the rest, including changing the delay based on room temperature (which is pretty cute, although I'm not going to measure the temp each day and adjust the delay to my sub-woofer
).
Handling and GUI is superior for the purpose of volume control from a dedicated plug-in - of course !
Yes, IIRC the BlueCat was one of those plugins that had compromised performance at attenuated levels.
Theoretically the problem starts already as soon as there is *any* attenuation – if i'm right on that its a calculation / interpolation problem.
I think I'm downloading again...
Michael
Last edited:
What you have measured seems to be pretty characteristic of the transformers I've measured in the past. I don't have data saved to present though. I've just made a quick trip to the lab to check an idea that's been rolling around in the back of my mind for a couple of weeks. The idea was to try using the Arta STEPS program to test transformers. Seems to work pretty well. Gives frequency response and harmonic distortion information on one graph. I'm willing to measure the transformers I have on hand to provide information on how the various transformers perform. I've got some Lundahl, Sowter, O-Netics, and a Slagle in a mix of core types.
Here is the measurement of a Lundahl LL1635 at 2.68Vrms voltage and 300 ohm source impedance.
An externally hosted image should be here but it was not working when we last tested it.
Things that would need to be finalized before running through the transformers would be:
Voltage level(s) of test
Impedance driving the transformers
Placement frequency for the cursor (determines what the set of data at the bottom of the graph presents)
Size of graph (This one is 800x600) Can be 1024x768
?
A baseline run of the measurement system would be included.
Do you find the information presented in the graph usable? The colors are hard to tell apart, but the way each line ends at the right side of the graph makes it easier to pick out which line is what harmonic, I.E. the blue line that extends the furthest is D2.
Gary
Hi Michael, do not know what to tell you. RMAA software doesn't provide calibration like some others do. There is no calibration process where you measure voltage at the output or input and than enter that value in the software prefs. How it does reading is that ask you to provide signal that is close to clipping and than it takes that as 0 dB. There is not much control in that software. And I do not exactly know what Fqs are used for "Intermodulation Distortion" Have to check that if it it shows. I still have to do measurement at 1:2 setting which is better than 1:4. Let me know what do you want me to measure and I will do it on HP network analyzer.
I'd be highly interested ! – I guess your offer would be gorgeous not only for me but also provide valid data in the context of that very thread.
Thanks a lot for all the effort - color scheme and information content is fine for me - also stimulus level and source impedance – all makes sense only for comparison anyway.
Bottom line of your plots show that 3rd order dominates by far the overall picture of that Lundahl transformer - almost not to distinguish from THD at frequencies < 1kHz (2nd order takes over past ~ 1kHz or so).
Second interesting fact I was not aware of is, that there seems to be a distinct frequency of lowest distortion at ~ 4kHz in this case.
What interests me most is, if substantially (> 10-20dB) different performance can be awaited from different makes - what you think?
Michael
Last edited:
Thanks for the test, Gary. I'd love to see some more. My thoughts on the test.
- 2V RMS max. That's the highest level of most DACs. Might even drop down 6dB for a more typical level. Not sure.
- Driving with 300 ohms seems good, unless someone can tell us otherwise. And I would say into a 20K load.
- Cursor at 1Khz
- 800x600 is quite legible.
An impedance plot might also be nice. I found a lot of difference in the transformers I measured. Different values, different kinks and wiggles. That should show up in the output, tho - right?
I did have a hard time interpreting this graph, as I am more used to seeing single or dual tones with the resulting harmonic fall out. Looking at your graph I think I see that H3 is prominent, and that the curves for the odd order harmonics are smoother than the even order harmonics. In fact I don't understand why the even order plots are so wiggly. Any ideas?
What would the plot look like driving a 20K resistor?
that's plain simple site ;
try with other browser , or behind some proxy server - maybe your ISP is allergic to Serbia situated servers
well - that really depends of your age - maybe you are old enough to attend (long time ago) oldest cinema , with alte Kinogerate ......
Yeah - dobro, knowing Todd-Ao (and some others) "to the last screw" I was not aware that ISKRA was many years ahead with 4 channel audio – what a shame, considering you being a such close neighborhood...
Was a nice reading about all that stuff from your perspective, thanks for putting up that page...
Michael
Last edited:
The min AC output load for the AKM DAC is 1K Ohm. Min DC is 1.5K Ohm. Unless you're adding a buffer stage, 300 Ohm should be overloading the DAC. How it reacts to overloading I don't know, but I doubt it continues to meet its published specs.
'I guess the min DC of 1.5 kOhm is meant not to create DC current overload from the 2.5V offset ?
We would not get into trouble with a transformer then..
I don't know either but would assume the usual resistive output decoupling of around 25-100 Ohm each side to apply to that voltage DAC's too - so anything from 100-1k Ohm seems fine for me for that measurement.
As the transformers in the context of this thread are meant to be rather output than input transformers I also agree, that it might be beneficial to simulate a load of - say - 10 kOhm plus a capacitor of - say - 1nF to count in for a standard amp impedance plus few meter of standard signal cable.
What you think?
Michael
Last edited:
DCR depends on how the transfo is wound, but the ones I've used tend to average about 150 ohms. Impedance rises rapidly above this, tho, because of the inductance.
But there is no DC load on the DAC, as the offset voltage is common to both output legs - thus no current flow.
I'm going to try to measure the output impedance of the AKM chip today. I'll let you know what I find.
But there is no DC load on the DAC, as the offset voltage is common to both output legs - thus no current flow.
I'm going to try to measure the output impedance of the AKM chip today. I'll let you know what I find.