Blog di Mauro Munzi

 

With some real measures…

I recently bought the EPX4000 to drive my two bass unit built on the CIARE HW380 woofer (yes, it is a “home” component), which specifications are 300W RMS and 500W max; some month before I bought an EPQ2000 to drive an old CIARE PW322, a 12” unit.

These amplifiers belong to a recent line of “traditional” amplifies, with the only difference the the EPQ does not provide 2Ohm stereo and 4Ohm bridge working condition, and they represent the highest model of their series; basically they are the last two lines of “analog” amplifier while the latest is the iNuke one, based on digital amplifiers (which I don’t like).

Before going to the datails of the two amplifiers, particularly to the EPX4000, I would like to comment a little the way Behringer uses to declare the power ratings since some time, which is the result of my experience on less recent models (EP2000) and on discussions with their technical support; first of all I was told that they mean “ideal” test conditions, from which I though they refer to stabilized 230V (in Europe). Second, some models show declarations like “XOhm per channel, stereo” and in this case, confirmed by the tech support, it means max power in stereo configuration with 1 channel driven; to be honest I found myself a little bit disappointed for the EPQ2000.

That being said here are the specifications of the two amplifiers:

As we can see the EPX4000 is declared for 530W @8Ohm with both channels driven; within a little we will see that this is too optimistic. Going back in time we can see that for the old EP1500 the specifications were in line with the real measures, as explainer in this test. Same thing for the EP2500 tested on the same pages. With later model instead, like the EP2000 I had, Behringer began to drift with the RMS power ratings, and today we can consider a 20% less for model like the EPQ (“XOhm per channel, stereo”) and around 15% les for models like the EPX (“Stereo, both channels driven”). So, when I bought the EPX4000, mindful of the previous experience with the EPQ, I was thinking to an amplifier of about 450W @Ohm compared to the 530W @8Ohm of the specifications.

That said, I can confirm that both models are well made, even if they use electronic components of poor quality for the capacitors (DECON or XUNDA) and fair for the ICs (JRC4580, also used in the Crest CPX); on the other hand the QSCs, which the Behringer seem copy of, use ELNA capacitors and the famous and reliable NE5532. with a minimum expense and a little work with a soldering iron the supply capacitors and the ICs could be replaced, with a consistent quality gain; but this is another story … or another blog 🙂

Here are some images of the internals of the EPX4000

This confirms that the EPX4000 is simply a “renamed” EPX3000, and also the specifications are the same; You can see also an EPX2000 mark (now EPX2800).

This is the detail of the supply voltages, 65V/100V, which are values with no load; they are almost the same values as the QSC CX-702 and DCA-2422, which are declared for 425W @8Ohm and 700W @4Ohm (keep in mind this values). The EPQ2000 values are 55V/85V ( @idle we have 55V/95V, probably under load the higher one lose about 10V), but I didn’t find an equivalent schema on the QSC site; considering the dimensions and the specifications it could compare to the GX series.

These are the capacitors on the high voltage supply, 4x2200uF in series two by two, like on many QSC, while the EPQ2000 has only 2x1500uF.

This is the detail of the supply capacitors of the 65V/100V lines: a total of 16x470uF capacitors while the QSC CX and DCA have 20, but considering the working frequency of the supply the “drive” capacity should be minimum. I imagine what could happen replacing them with some very low impedance PANASONIC … 🙂

Last, the detail of the cooling tunnel, a little smaller on the exit probably to compensate the lower air velocity at that position

After this inspection, I did the same day I bought it, I connected the EPX4000 to my scope and each exit to an 8Ohm load built with 4 2Ohm/50W aluminum  resistors, fixed to a heatsink with mounting grease; they count for 200W (each channel), but based on their specifications for a “duty cycle” of 5s (on/of) each package can handle till 1000W. During my tests with sin waves I use a 5s On and 10s Off cycle, so the real limits are a bit higher.

Here are some data I measured; with a 100Hz sin wave (it’s to easy to test @1Khz !) the EPX4000 shows the first clipping signs @59.5V with an 8Ohm load, which is 442W, and this come back to the 425W@8Ohm of the QSCs cited before; this is also the maximum voltage the limiter allow to deliver with the 8Ohm load. A little after I found that during the tests my wife was using a hairdryer … to further confirm the “real condition” operation 🙂

With a 4Ohm load clipping happens @54V, which translates into 730W; this is a further confirmation of the likeness with the specifications of the CX-702 and DCA-2422.

Considering the measured data and the above considerations the EPX4000 seems to ma a good “value for money”: I paid it 325€ here in Milan.

I write also the  8Ohm power which I measured sometime before on the EPQ2000: at the first sign of clipping it delivers 55.5V RMS which translates into 385W vs the 500W declared, but remember how the specifications are;by the way You can now take the EPQ2000 home with around 250€ and because I expect around 550W @4Ohm practically with around 500€, if you don’t have size problems, you can build a stereo amplifier of more than 1000W @8Ohm (using them in bridge mode), but remember to use it only with 8Ohm loads. The same is true for the EPX4000… around 700€ to have the equivalent of a stereo amplifier of more than 1400W @8Ohm (bridge mode), with the option of use it @4Ohm.

I’ll try to update this blog in the future with the 2Ohm test for the EPX4000 and the 4Ohm one for the EPQ2000, in order to complete also the figure of the bridged power.

Let’s see now a small negative note shared by the two amplifiers, which I found while they were under the scope; they are almost without polarization or with too few, and for sure also other model of the same series, the “old” EP and several amplifiers of other brands (except the Crest CPX, the Peavey PV and for sure other models of the same group). This comes to a consistent  crossover distortion as well as high levels of THD at high frequencies.

This is a 15Khz 2,83V RMS @8Ohm taken on the EPQ2000, and the one on the EPX4000 is almost identical; we can see a consistent crossover  distortion.

To confirm this, if we look at the following image, taken from the test of the old EP1500

we can see that at higher frequencies due to the poor polarization the distortion is very high, around 2%; the same happens for the EP2500, which delivers only 260W @8Ohm with 0.1% TDD compared to the 450W at 1Khz and 100Hz.

If we look at the Peavey PV2600 tested on another site (identical to the CREST CPX2600) we note how low is the THD at high frequencies, due to proper polarization.

Given this, once opened the EPQ2000 I found the bias trimmer of the output stage, and with a tester connected to one of the output stage resistors I set the trimmer in order to have a minimal polarization (few mA); the result is the following image where the crossover distortion has gone away.

I  currently made this change only on the EPQ2000, while the EPX4000 is still unchanged. Everyone with a minimal practice can try this operation and set a minimal bias current to better in a consistent way the THD figure at high frequencies of these amplifiers, few mV, between 3 and 5, are enough; remember that these are not the only PRO amplifiers suffering of this small “defect”.

And … sorry for my English.

Update on 29-Dec-2012

As stated before today I tested the EPQ2000 in bridge mode @8Ohm, using the same method as above, so 5s On and 10s Off; the RMS voltage reached by each side (so 4Ohm stereo equivalent) at the clipping has been 45,5V, a total of 91V RMS, which translates into a little more than 1030W RMS @8Ohm (around 515W/4Ohm stereo).

The strange thing i noticed during the test was the fact that the RMS voltage dropped to 87/86V (940W/920W) after a couple of seconds.

Update on  01-Jan-2013

After having done a re-cabling of the power string on the rack, as I wrote on the Crest CC4000 article, which allowed to me to gain some W, I tested again the EPQ2000 in bridge mode @ 8Ohm, and I got interesting results

As You can see clipping happens at 98.6V RMS, which translates into 1215W; with the limiter on the RMS voltage does not go over 96.5V RMS (around 1165W) practically around 10% more power than before and the most important thing was the the RMS voltage didn’t  drop as before after 2s.

The main voltage dropped from 238V to 232V