Signal chain overload effects

Recently had a clients venue change Formula Sound PM100 to a Pioneer DJM800. This was due to pressure on riders from 'DJ's. Pioneers sound crap! Personal AB listen again last week. Bleedin' awful. Ruined a great sound system, however all the 'DJ's are 'happy' 

Nice one Mick.

And a very pretty mixer that!

OK then, looks like it's high time to see how the "Rider Friendly" DJ Mixers perform with the same tests!

So is anyone with one of these expensive boxes able to spare about an hour to run and verify the same tests as performed on the other Mixers so we can see properly any differences? The Test Bench is located in South West Bath.

Can assure that the tests are non-destructive and that covers etc don't need removal - that is unless we find a fault which needs attention perhaps

Please PM if you are able to contribute.... evenings for tests are OK!

thanks in advance
Mik

Good interpretation!

Can be fun though with one of my Signal Generators to write suitable words into the waveforms which can then be displayed on the 'scope screen........ something for another thread perhaps

Mik

I'm sure there must be someone near here who could pop over with a DJM for an hour or so perhaps one evening; can include me showing the owner of it how to test it as well!

Now to put you off with tekky things used to do the tests, here is a pic of the PHAT with some of the instruments I use to do the tests


However whilst this might look really complicated, amongst the stack of instruments are those used to generate the three tone test signal, an oscilloscope for monitoring the input and output signals for the mixer and the Spectrum Analyser to look at the contents of the mixer output for each input level tested.

Now a quick explanation of what the Spectrum Analyser shows so that those reading this Thread can see differences between the Mixers being tested.
The Spectrum Analyser displays the signal level (vertical) and frequency (horizontal) as a graph with 500 sample lines width between 10Hz and 50kHz, much like the response graphs provided by 'speaker manufacturers for example as it displays the level in dBm of each frequency sampled, not quite in real-time but with much greater resolution than PC based equivalents used for 'speaker tests etc.

Firstly, one of the 'clean' signal Spectrum:

The top line on the display is +20dBm and each horizontal line is 10dBm lower with the bottom line of the display at -80dBm - so showing a 100dB range.
The Vertical lines are at each decade of the Frequency, starting at 10Hz on the left and increasing to 100kHz although the maximum measurement frequency is actually 50kHz.

Here we can see the three test signals combined to the input of the Mixer appearing at it's output as the three large 'peaks' in the display. Two of the signal components are quite close together at 960Hz and 1120Hz and appear as the 'twin peak' on the 1kHz vertical line in the center of the display.
Lower down, the various small peaks are the distortion products of these input signals, noise and mains Hum; we could call these the 'unwanted' bits of the Mixer output since a perfect Mixer would have none of these visible.
Comparing this display then between different Mixers gives a good idea of their differences in performance using this single measurement.

Next is the 'overload' signal Spectrum:


Now this one shows all of the added harmonics and other distortion parts of the Mixer Output Signal with the input overloaded as well as the noise and Hum which is still somewhere amongst it all!
Try spotting the higher frequency input signal peaks as these are becoming 'lost' amongst all of the distortion products caused by the overloaded Mixer input!

Comparing these displays between the different Mixers tested then gives a good indication of what each Mixer will do when an overloading signal is applied to it's input.

Now for all of the Mixer Tests (for the Cloud I used a different test method for this as it was to show what happens to the outputs of the LMS's) I have not adjusted any measurement settings so as to be able to directly see like-for-like results.
The Spectrum Analyser is also set to read RMS values for the levels of each frequency component; something I explained in an earlier post in this Thread so we can see the real power being sent to the Amplifiers and Speakers at each frequency.

Hope this assists in explaining things further!
Mik

PS: tis a bit late at night whilst posting this; so please reply with any comments about this test method and/or if I've missed something or not explained things clearly


Edited by audiomik - 20 February 2012 at 4:12am

Mik-not forgot you, but have landed another week down the road here..