Since I have quite a bit of data to show I will upload it using several messages over time so please bare with me as I'm not glued to my computer!

So firstly, some pics and a general description.

Behringer inuke NU6000:

I'll be amazed if I find anyone who thinks this is attractive? Maybe they let their new graduate designer sculpt these fine lines!

The case is made from folded aluminium (except the front moulding which is plastic fastened to the punched aluminium inside). Despite it being quite thin it doesn't feel too flimsy and given its very light it should be just fine.... The chunky front panel adds a lot of depth to the front of the amp and it ends up with the knobs 42mm from the rack ears so make sure your flight case has enough space!

The non-Neutrik connector worked OK for me but I wasn't keen on the XLRs which have no locking pins!

There is no ducting for the two 80mm fans on this model. All the power devices are clamped to the base plate, which is acting as the heat sink. The fans just push air through the box to take away the heat build-up of the case and the components. The base plate does get warm when delivering a lot of average power. The fans are speed controlled but even at the lowest setting they are quite noisy, when you work it hard it starts to sound like a hair dryer!

The mains input filtering looks chunky enough and the NTC soft-start resistor is bypassed by the relay after start-up. There is a 12A thermal breaker on the back panel which may be there to protect the IEC inlet and power cable as this thing can drink some juice!

The circuit boards all look nicely made with SMD parts used in preference to through-hole. The amp sections are based around the IRS20957S and there are four channels used in permanent bridge mode to give the two outputs.

The Behringer doesn't have any PFC (Power Factor Correction) and would appear to be a fairly standard forward converter. It uses a fairly substantial ferrite transformer with a resonant switching choke (well glued to the PCB). Power rails measure at +/-84V and there is a reasonable amount of energy storage around which is formed of 4 x 2200uF/200V on the primary side and 4 x 3300uF/100V plus 8 x 1000uF/100V on the secondary side. The effective primary capacitance is therefore 2200uF/400V and secondary side is 10,600uF/100V per rail. Total energy storage is therefore approx. 200J at nominal working voltages.

Amp output inductors used a decent gauge of wire and are securely glued to the PCB. Each channel has its own local reservoir caps but they are common to the same power rails.

KAM KXD7200:

The KAM is 1U and thus it has been a lot harder to fit everything in. There are multiple instances of stacked PCBs which appear to be held in place by a posh type of twist tie! Four variable-speed 40mm fans pull air from front to back over the heat sinks and guts. Casework is steel and feels substantial enough.

The KAM has PFC which uses the modest size toroidal choke on the right hand side.

All controls are recessed (to the rack ears anyway) and it looks a lot nicer than the Behringer!

The KXD7200 is a single-ended output type and thus can be bridged if you really want to get some voltage swing! The IEC inlet is fused with an F15A fuse which means it would probably fail before the 13A fuse in the plug, as it should.

The amplifier sections use two boards stacked on top of the main PCB and the top board has been coated in black paint or similar for reasons unknown. Probably to prevent easy copying I guess?

The output inductors here have a much smaller gauge of wire than the Behringer and, as we'll see later, that leads to a lot of heat build-up in these parts.

The power supply in the KAM seems smaller that the Behringer despite its extra complexity. The Ferrite transformer is smaller at just 5cm x 5cm x 3.5cm. Primary side capacitors are 2 x 330uF 400V and secondary side consists of 10 x 1200uF 160V, so 6000uF per rail. DC rails run at +/-145V and seem very stiff even under heavy loading. They could be regulated or perhaps just benefiting from the stable output of the PFC stage. Assuming 380V for the PFC DC bus output (I didn't check it) then total energy storage would be approx. 173J

The secret amp boards complete with what looks like paint brush hairs :)

PFC inductor and part of the input filter. Driver circuit is again painted over to disguise it's secrets!

Certainly an impressive power to weight ratio if it lives up to it's specs....

So in the next section we'll look at power output.....

Stay tuned as I'm off for my dinner :)

Thanks for sharing..

+1

NU6000:

The instruction manual for the Behringer actually has rms figures in it as well as their advertised inflated figures. I'm still not sure where they get their multiplier from to convert one to the other but it seems to be x1.36.....

Stereo 8 Ohms per channel: 1500W or 1100Wrms

Stereo 4 Ohms per channel: 3000W or 2200Wrms

KXD7200:

Now if these amps were class-AB/H/G etc. with 50Hz transformer based supplies then measuring power output is normally easy, just use a continuous tone and measure what you get. As we know amps with switching power supplies sometimes restrict the output power under sinusoidal tone conditions as they do not have the thermal capacity or overload capacity to suck it up for a minute or so whilst the engineer measures it. In anticipation of this happening here I choose to make two measurement types. Continuous sinusoidal output for 1 minute (i.e. the traditional way) and output under 33% duty cycle conditions, which is a kind of burst rating but where the burst is repeated so as to deliver 33% maximum output. This is similar to saying that an amplifier must deliver 33% of its power indefinitely so that music signals with a low crest factor can still be reliably produced. 1/8th power is typical used for domestic equipment but 1/3rd seems safer for professional use where those dub bass lines can get quite demanding.

Speaking of bass, we also know that amplifiers cannot always deliver the same power at low frequencies as they do at 1KHz so I also measured both the conditions above at 31Hz and 1KHz. In order for the burst rating not to take a unfair advantage of the energy stored in the capacitors I chose the repetition rates for the burst to be relatively long so that the power output is what the power supply can sustain, not what it can do for 1ms!

So the 33% duty test signal used was as follows:

For the 1KHz measurement the signal is on for 70ms and off for 140ms.

For the 31Hz measurement the signal is on for 450ms and off for 900ms.

Sorry for boring those non-techies of you but there will probably be some debate about what tests are appropriate so I thought I needed to explain where I came from.

So, on to the results, first up, the continuous tone tests.

The Behringer was happy to drive 8 Ohm loads at full power (limiter active) for 1 minute without any change in power output. With both channels driven it produced 1.17kW at 1kHz and 1.08kW at 31Hz

The Kam was more interesting, it was happy for 30 seconds at 1kHz but had reduced output slightly by 1 minute. The reduction seemed to steady out after that. With both channels driven it produced 1.05kW at 1kHz at the 30 second mark and then fell to 950W at 1 minute.

Strangely, at 31Hz, the reduction did not happen, possibly because it started at a lower value in the first place. At 31Hz the output was stable at 980W.

The Behringer was happy to drive 1 channel at 4 Ohms with no reduction in output at 1 minute. Power output was 2.27kW at 1kHz and 2.12kW at 31Hz. This demonstrates that the amp section should drive high power into 4 Ohms without any short-term limitation. However, when driving two channels simultaneously the amp would trip its protection and shut down, needing a power on/off cycle to reset. Since the power amp part was happy to deliver the power then I conclude there is a total power limit on the supply part and rather than engaging the limiters to reduce output the amp switches off. This seems a strange decision on the part of the designer, better the show go on at reduced level if the amp finds itself producing sine-wave like power outputs. It is only fair to remember however that this is a harsh test and the amplifier was pulling 24Arms from the mains at the point it cut-out! (Note: the 12A breaker on the rear did not operate as it has a time delay, this was an internal electronic limitation).

Whether this ever occurs in a music situation is debatable and the burst tests below will demonstrate its actual ability with more realistic signals. That said it does mean it is possible to trip the amp if you abuse it into 4 Ohms, not ideal.

If the output power was kept below 1.7kW per channel then the amp would not trip. To leave a margin I left the amp running at 1.5kW per channel to see if I could get a 1 minute figure. Unfortunately after 15 seconds the circuit breaker opened, not surprising at it was drawing 22A from the mains at the time. So in the end I didn't get a 1 minute figure for 4 Ohms running but I expect it will be around the 1.2kW per channel mark as that produces a current draw of 15A which would eventually trip the breaker.

The Kam takes a different and far more cautious approach. Regardless of whether you use one or both channels the output power into 4 Ohms is limited on a time delay basis. The limitation here does not appear to be the supply but the amp sections themselves.

Power output at 1 minute was restricted to 650W at 1kHz and 540W at 31Hz. The limitation comes in after around 4 seconds:

Again, this may not affect the operation with music, see the results from the burst testing below.

At least it doesn't trip out leaving you with no sound however it does seem a large restriction when it should do 1.8kW!

So now it's onto the more music like signals to hopefully get an idea of what these amps will achieve in practice. First up, the Behringer get the 1kHz test.

Unfortunately I've lost the picture that goes with this measurement but the result was 1.37kW into 8 Ohms with either one or both channels driven. At 4 Ohms the output was 2.05kW both channels driven and 2.45kW one channel driven. For the single channel case the waveform is shown below:

The upper traces are the + and - speaker terminals, measured separately as this was a balanced (bridged) output. The lower trace is the AC mains current draw with a scale of 20A/division. The lack of PFC means the current peaks are already 36A for one channel driven.

Now changing to 31Hz, the NU6000 now manages 1.16kW into 8 Ohms with one or both channels driven. The traces below are for 2 channels driven and the current is now at 50A/division!

At 4 Ohms with 1 channel driven we get 2.05kW:

And at 4 Ohms both channels driven we get 1.80kW:

Peak mains current is now at 80A!

The Kam results are similar but lower. Starting with 1kHz and 8 Ohms it will produce 1.12kW with either 1 or 2 channels driven. Trace shows 2 channels driven:

Notice the mains current is more sinusoidal though not perfectly so (more on that later) and a scale of 20A/division is adequate :)

At 4 Ohms the Kam produces 2.08kW with one channel driven and 1.77kW with both driven. Trace shows both channels driven:

Peak mains current is around 30A.

Changing to 31Hz, at 8 Ohms the KXD7200 produced 1.11kW with one channel driven and 1.02kW with both driven:

At 4 Ohms the output was 1.95kW with one channel and 1.62kW with both channels driven. I appear to have lost the trace for 2 channels so here it is for one channel, 4 Ohms, 31Hz:

Neither of the amplifiers showed any signs of progressively restricting the output during these tests so I'm inclined believe that they should be good to do this indefinitely. As a caveat though I did not test them in this mode for long as my load box is not capable of these power levels for more than a couple of minutes at a time.

To summarise, for music signals, both channels driven, you can expect an output of:

Behringer:

1.37kW into 8 Ohms, 2.05kW into 4 Ohms for full-range or mid-top duty

1.16kW into 8 Ohms, 1.80kW into 4 Ohms for Bass duty

Kam:

In the next instalment we will look at the frequency response......

I didn't check the power supply sections (though I can if needs be) but the output stages were working at 348kHz for the Behringer (both channels virtually the same ) and the Kam has one channel at 310kHz and the other at 330kHz.

I believe in both cases that the frequency varies with instantaneous output voltage.

I will be including some traces for the output noise with no signal soon.

 

S

The frequency response of the Behringer is load dependant because it takes it's feedback before the output filter. You can see the result here:

The flattest response is for 8 Ohm loads where the response is just 0.27dB up at 20kHz and otherwise flat to below 10Hz, -3dB is at 42kHz. 4 Ohm loads over-damp the output filters causing a loss of 2.2dB at 20kHz, -3dB is now at 22.7kHz. Fortunately there are virtually no speakers out there with a 4 Ohm impedance at 20kHz as the voice coil inductance, even for compression drivers and tweeters, has started to push up the impedance.

For open circuit loads you can see the output filter is very under-damped, with only the Zobel network for damping. Probably no importance in practice however as long as you don't drive it with a 38kHz sine wave with no load 

The Kam amplifier uses a different technology and obviously takes it's feedback from after the filter as evidenced by the well-behaved response:

The response is pretty much load independent with a -3dB point at 39kHz. Since a high-pass filter option was included I measured it also and you can see it produces a second-order roll-off with -3dV at 33Hz, handy to protect speakers if you're not using an LMS.

The response variations can also be illustrated with a square-wave measurement and the Behringer shows the expected results:

The low frequency square wave (110Hz) shows gentle sloping due to the high-pass filter at sub 10Hz. You can see a slight transient at the start because this is for an 8 Ohm load. To see more detail we go to a 5kHz square wave:

So the slight under-damping is evident. Changing to a 4 Ohm load:

The over-damping is also apparent as well as the bandwidth limitation giving a nice slope to the edges. With an open circuit:

The excessive ringing can be seen, this is due to the energy content of the square wave exciting the 38kHz resonance. In practice, despite its appearance, this is not really an issue as there is not likely to be any frequency content in the signal to set off this resonance. Remember also that this is with no load......

The Kam, as the frequency response tells us, shows a much less variable output. Firstly the 110Hz square wave:

This does not have the slight overshoot at the start because of the controlled damping but otherwise is the same as the Behringer. The 5kHz response at 8 Ohms:

and 4 Ohms:

and open circuit:

Spot the difference  OK, so there isn't any change..

So to summarise, the Behringer has a somewhat load dependant response at the highest frequencies which may mean a little tweaking of the EQ is required for different speaker systems, no big deal. The Kam however has an outstanding performance being essentially tolerant to any load condition with little effect.

Of course there is a secondary result of these two approaches, the output impedance of the amplifier itself (Damping factor). The Kam, with its feedback after the filter, should have a very low output impedance across the band but especially at low frequencies (where open-loop gain is at its highest). The Behringer cannot correct for the impedance of it's output filter so will have a higher output impedance across the board but will get especially bad at the highest frequencies, hence the variability with load. The real significance of this is probably minor for most users but it's useful to know. Incidentally, the damping factor specified by the manufacturers is 120 for the Behringer and 800 for the Kam which ties into the results here.

If I find time i'll measure and plot the output impedance against frequency for interest.

Next up will be THD....

In the absence of an AES17 filter option I used an external 30kHz second-order low-pass filter to reduce the switching frequency content and allow the AP to measure the distortion. Analyser bandwidth was left at 80kHz (In hindsight the 30kHz setting would have been more appropriate but hey-ho, it doesn't make a big difference).

I did some scope traces of the residual so the nature of the distortion could be viewed at specific conditions. Then two measurements were undertaken, THD+N (Total Harmonic Distortion + Noise) versus frequency at 1/3rd power and THD+N vs output power for 1kHz and 31Hz.

Finally for the Behringer is the plot of THD+N versus output power:

Because the limiter comes in as clipping starts the THD doesn't actually rise too much, we see the level just stops increasing and the curve fluctuates around a constant power point. The limiter is not defeat-able so I couldn't generate the classic types of curves you normally gets from this measurement with a sharp increase in THD above clipping.

Onto the Kam:

THD was consistent with the Kam pretty much regardless of level, so there was no sweet spot to be obtained at low power outputs.

At 1kHz, 8 ohms, 340W, the THD+N was 0.11%:

Quite a bit of harmonic content and fuzz here so not as impressive as the Behringer.

At 1kHz, 4 Ohms, 640W the figure was 0.126%:

Similar to the above and perhaps looks indicative of some induced distortion from power supply tracks since its asymmetrical and has a large component at the zero cross points.

At 31Hz, 8 Ohms, 340W the THD was 0.115%:

This makes it very clear that the positive half cycle has a much larger interference effect than the negative half despite there similar shapes, re-enforcing the thought of power trace related distortion. Pure speculation based on personal experience but it may be down to circuit layout allowing a node to be disturbed by current in another part of the circuit.

31Hz, 4 Ohms, 640W:

No real change.

Plotting the THD+N vs Frequency we get:

This shows the independence of THD on frequency further suggesting a well defined coupling mechanism of distortion into the signal.

Moving on to the THD+N vs Output power. This caused a headache since, as we have already seen, the amplifier limits power into 4 Ohm loads after 4 seconds. Unfortunately the measurement of THD vs Power output takes longer than that to do so the results are messed up. Still, for the fun of it, here they are:

Both 8 Ohms traces are OK and you can see the power limiting on 1kHz but not on the 31Hz as already noted in the power testing. The 4 Ohm traces get messed up by the power limiting kicking in during the test. At 1kHz it tops out at 650W but it maintains the 31Hz output for long enough to get up to reasonable power but it wasn't stable.

The traces are, in the main, flat again at around 0.13%. It does increase at high frequencies but not until 10kHz so the top end is not bad at least.

You can draw your own conclusions but I think the Behringer has the better chance of sounding nice.

As expected.

Good on behringer, keep up the good work.

As already noted the Kam has PFC which really helps keep the AC input current under control. However, it's not a perfect looking PFC performance so I recorded a couple of shots of AC voltage and current with a 1.05kW output power (single channel).

So, the Kam with output of 1kHz, 8 Ohms, 1 channel driven. The Ac voltage and current waveforms are shown:

Now I don't have any fancy EMC test equipment but I can connect a scope to the speaker outputs (with no input signal) and see what's leaking out! We expect to see some carrier signal remaining after the output filter but it should close to sinusoidal as any high frequencies are likely to cause common-mode or differential-mode radiation from the speaker leads.

So the behringer was scoped up (one channel on + and one on - speaker terminals due to bridged output arrangement):

The waveforms look nicely smooth and show a 348kHz carrier is used for the power amp stages (at least at idle). Since there are no strong high-frequency elements present I would surmise the Behringer is not going to create much interference, a good result!

Now the Kam:

Hmmmm, not so good. There's a good breakthrough of switching noise and ringing from the output stage suggesting a less than ideal layout and probably a lot of contamination of the internal ground reference. It's possible the Kam could cause issues with some equipment and in fact I did experience this when I did listening tests using this along with a Void QX5 amplifier. The Void amplifier was emitting a quiet but audible tone through the speakers when the Kam was running and this tone varied as the unit warmed up. The effect was just like an intermodulation tone or beat frequency as it started high, reduced right down to DC and then started climbing again, classic symptoms...  The noise ceased if I disconnected the Kam from the LMS and so just a ground connection into the system was enough to cause an issue (the amps were not fed from the same LMS output).

The DC offsets were noted:

Behringer Ch1: +11mV, Ch2: +56mV

Kam Ch1: 6.5mV, Ch2: 2.2mV

Nothing too nasty there then.

The Kam limiter was speedy, cutting down a 200Hz signal which was 6dB too hot in two half cycles (~5ms):

Apologies that it was 200Hz, it was supposed to be 1kHz but I messed up!

I then tested the Behringer at the originally planned 1kHz (D'oh!) and it was also fast, taking 2 half cycles to drop the gain:

This speedy response, without overshooting on gain reduction, should help keep signals clean when overdriven.

I took some thermal images of both amps whilst running a fairly high output level for a minute or so to see if anything was getting worryingly hot!

The Kam had the hottest component, the output inductors were wound with quite thin wire and when driving 4 Ohm loads the wire was very hot and the core not far behind:

118 degrees for the wire itself and the core was getting up to 100 degrees as well. A short time later I stop the test as I was getting worried:

138 degrees!

The power supply heatsink also warmed up nicely but the transformer itself was also hotting up:

Transformer:

Overall view from top:

Otherwise it was OK but those output inductors certainly do get warm! Cramming so much into 1U certainly doesn't help get the heat out but under music conditions I assume it works OK otherwise we would have heard about it wouldn't we?

The Behringer fairs better, no doubt due to the colossal airflow that takes place when you pump the signal up a bit, those fans sure are noisy!!

Transformer:

Bridge rectifier:

Fan control circuit with dropper resistors:

Power amp section top view:

Output inductors:

Close up of FET driver chip:

Nothing too alarming here at least 

Well thats about it for what I've done already. I'll re-read the other posts and see if theres any test I can do to answer question. I did do some listening tests but since these things are subjective I haven't decided whether to write about them or not. Better you make your own minds up with these things 

Great writeup, but the current comparison in the PFC section isn't so clear (although I can tell, doesn't make clear which is Berry, and which is Kam)

Feel free to give your opinion on sound quality (perhaps compared to the QX?), complete with disclaimer 

 

Totally agree with the overating IEC connectors and heat on the reservoir caps. Not to mention the temperatures on poor quality inductors.

 

I agree with U Viktor. A crash is inevitable.

 

With such high current variation I would not like such a device in any rack with any brand name. Also miffed at the linearity of the amps. Very poor.

 

The Behringer's storage is somwhat less at operating voltage. 10.6F x 84V = 890.3 Coulombs x 2 (rails) = 1780.8 Coulombs = 74.8 Joules. This is 17.9 calories... They are deffo LOW CALORIE 

totally agree with this, not seen a post as usfull and as filled with actual facts for a long time on here.... well done 

As far as the current draw goes its not going to be any worse than any other amplifier that doesn't have PFC built in, and that includes virtually all of the heavyweight units and a lot of the switch mode supply ones too. With no PFC they all use rectification into a set of capacitors which will always pull peaky currents. Hence I wouldn't be worried, it's not a case that the Behringer is poor, more that the Kam is Good! For domestic appliances the rules are changing such that anything above a certain power needs PFC but for "professional applications" it's not mandatory, hence multi-kW amplifiers with no PFC.

I have amps from QSC, Void, Matrix and Ohm and none of them have PFC so will all draw peaky current from the mains.......

As for the sustained output, I may re-visit that and see if it does eventually limit the power or not but as I've shown, for music signals it should be just fine.

U.Viktor wrote: i have seen the original manufacturer of those KAM internals somewhere in the china backstage of prolight&sound in frankfurt. It still pulls the hot air to its capacitors quickly dry them out and let you buy another one of this scrap. About the iNUKE: the whole thing has built around the IRS20957S which barely hits the entry level, it has made for cheapest car radios and does not even have real output feedback. The sample frequency is not 348Khz (or only in low signal or steady-state) because it is free running variable freq. controller with the lack of sync. capability. As I saw it would make a voltage doubler circuitry by a relay and center tapped primary caps for low line operation. good luck with that! And deal with 200A+ current spikes and heavy AC line distortion :-D

U.Viktor I've noticed that you have a very black or white view of things! I agree that pulling hot air across the capacitors is a bad idea and they are only 85 degree rated types too. The generally accepted rule of thumb is that if the rated life of the capacitor is, for example, 2000 hours at 85 degrees then the life will double for every 10 degree reduction in temperature. Ultimately it's going to depend on what temperature they run at but thanks to the efficiency of the amp and taking account of the nature of music signals I doubt they will run that hot. I'll take a guess (for the sake of calculating a number) at 65 degrees in which case life might be 8000 hours which is the equivalent of 333 days of continuous running or 1000 days at 8 hours a day. Lets remember this amp is targeted at musicians, DJ's etc. and not touring companies! I suspect their usage might be a few hours a week in which case I don't think they have much to worry about, do you?

With regard to the iNuke, the IRS20957S is hardly entry level, it is used in some reference designs which have extremely good audio performance, I know because I've tested them! Yes, the typical application circuit is a self-oscillating second order delta-sigma type modulator but it happens to work quite well. You can of course use it differently if you wish, it's up to the designer. It can also be synchronised to an external clock if needed but not over it's entire PWM modulation range.

If you wanted you could turn it into a UcD type circuit with feedback from after the output filter, maybe that would appeal more to you?

The voltage doubler arrangement for 115V supplies is, I agree, a bit (no, a LOT) of a tall order! I can only assume they don't use a 10A IEC socket for the US market, otherwise it's going to be very limited in power before that 12A breaker goes! Stick to 230V I think...

U.Viktor wrote: Does the 2x 2000W@4R with 33% burst ratings mean 2x 700W@4R CONTINUOUS real power?

During those particular test yes it is but I'm not saying that's the best the amplifier can do, that's just the result of the 33% test signal I used to represent worse case audio abuse. As I stated I think the Behringer will probably do about 1~1.2kW per channel continuous sine wave, limited by its 12A circuit breaker. For music it will happily produce 2kW into 4 Ohms per channel.  The Kam can only do 650W into 4 Ohms continuous sine wave but its happy to deliver 1.6kW under music conditions. Should be absolutely fine for most people...

audiomik wrote: David looking at your results for the iNuke: "At 4 Ohms both channels driven we get 1.80kW: Peak mains current is now at 80A!" Then checking the Behringer website info, and your images earlier in this thread, for this Amplifier it appears to be using a single IEC type 10A (rms) power connector. Now your measurement is peak current, albeit with a reduced duty cycle, but the peak current rating of a 10A IEC connector is still only 14.14 amps..... When I last looked up CE Compliance Regs, this would be a 'Fail'! The KAM seems to be substantially better performing in this respect with only a 2:1 overload compared with over 5:1 for the NU6000. OK, like me you have limited loads in terms of continuous power dissipation measurements, but how long will it take for the IEC connector to heat up to an unacceptable level and fail do you think? Anyway, an interesting set of test results, but would like to see the sinewave output just into clipping and 33% voltage squarewave responses of both Amplifiers to be able to compare their performance with other Amplifier tests. Don't know which type of Current probe you have, but the Amplifier output current waveform into a Resistive Load with no signal input can be interesting  Then again, output current measurements into a Load with an Inductive component, just like the Le of a 'Speaker, give some 'spectacular' results with Class D.......  A couple of MilliHenrys and 4 Ohms in series with 50% output Voltage at 1kHz? Wouldn't want to overload the output stages with too much back EMF now would we Just seen UVictor's posts whilst writing this one - max continuous output without 'tripping' for 5 minutes would be interesting - and the test described above with the Inductive component added can show marked differences in performance where the various clocks aren't properly locked or synchronised respectively in different Class D/SMPSU Designs. Keep up the good work Mik

Hi Mik, I think the traces I put up for the limiter operation will give you some idea of the output under just clipping conditions since the limiter holds the output at the just clipping point. If you'd like to see some more detail i can use a phosphor scope and zoom in to see it properly 

The current probe I use is a Honeywell CSNR161,  DC-150kHz with 0.5% accuracy, it's nice little device 

I have a 0.65mH inductor here that will take 50A plus (an air cored beast from an old UPS) so I could use that but I don't have anything around 2mH. I don't suppose it would matter much, I could just change the test frequency to get the phase shift needed. I would expect the class-D to be very good with inductive loads (and capacitive ones) owing to its 4-quadrant operation. If I get the time i'll try something, what conditions would you suggest?

As already mentioned there are a lot of products out there just as bad, why would this one be any worse?

I doubt that a flatted-top sine wave caused by this kind of load would cause any harm to other equipment, why would it?

snafu wrote: so it doe's roughly what it says on the tin ?. As expected. Good on behringer, keep up the good work.

Yes, I think it does and for the "weekend warriors" like myself £270 gets you a powerful, efficient bass amp! Since it's not bad run full range either then it could get you out a tight spot if your mid-top amp went down too but I'd stick with class-AB for that myself I think.

U.Viktor wrote: Anything bellow 1% THD is OK.  THD is not the most important thing how well a professional amp would perform. However the lack of real output feedback in the iNUKE is a serious problem and renders this amp USELESS anywhere above driving sub sections. The undamped/uncontolled output LC filter causes unpredictable signal level variations with different speaker loads probably creates huge spikes in the mids/high freq. range. I have seen/heard first series PKN amps ~10years before with similar problems so Behringer should have learnt something :-) That KAM has at least two blatant design mistakes 1. see those electrolyte caps in the HOT airflow(!) 2. see the "PFC inductor" has made from single layer insulated magnet wire (light yellowish). Also the winding technology of that inductor deserves a possible failure due insulation weakness+bad construction. Those design mistakes will be paid by poor guys without understanding of electronics and believe that a sub ~200EUR amp could be reliably good for their applications. Of course NOT.

Whether 1% THD is OK or not depends very much on what the spectrum of it is. 1% of low-order (2nd, 3rd, 5th harmonic) is normally fine but if it was huge amounts of 9th, 11th, 13th, etc. then it would sound horrible, especially as a mid-top amp.

The iNuke does have "real" feedback, its whole topology depends upon it to function, but I guess you mean it's taken before the output L-C filter. I don't agree that's always a big problem, the DCR of the inductor is likely 20-30 mOhms and the output impedance of the closed-loop power stage will be very low, especially at low frequencies. The output impedance of the amp is therefore going to be very low until the highest audio frequencies and so it doesn't matter what the load impedance swings are like they will not significantly affect the response of the amp. To calculate a number let assume the output impedance of the amp is bad, say 100 mOhms. If the load impedance went from 2 Ohm to 1000 Ohms then the output level would vary by just 0.21dB, I wouldn't say that's a disaster! The resistance of the speaker cable is likely to be far more significant so the amp is pretty blame-free all of a sudden....

On the Kam PFC inductor I would entirely agree it looks a bit weedy and badly constructed, I would at least expect it to be dipped in varnish to stop the windings vibrating.

spongebob wrote: Teslaman - I can't confess to understanding all of the technical details that you have provided in these posts, but I would like to thank you for such an in-depth contribution to this forum. Do you think you will be reviewing other amplifiers in a manner similar to the above?

If there are any "interesting" amps out there that people want testing then I will consider it. The only problem is time as I don't have much and it takes quite a while to do all this stuff. The reason I was keen with the iNuke and Kam is because they are cheap and yet promise great power, if they lived up to their names then they are game changers for the weekend warrior types like myself. As it is I think they are good for bass, especially at 8 Ohms per channel, and not too bad for full-range or mid-top use though I would still prefer a class-AB in this role.

infrasound wrote: Thanks as well. Great writeup, but the current comparison in the PFC section isn't so clear (although I can tell, doesn't make clear which is Berry, and which is Kam) Feel free to give your opinion on sound quality (perhaps compared to the QX?), complete with disclaimer

Yes sorry for that, I'll re-read and make it clearer.

My listening tests were a bit flawed due to the venue but I may still give an opinion, maybe 

Muckerbarnes1 wrote: Good work David.   With such high current variation I would not like such a device in any rack with any brand name. Also miffed at the linearity of the amps. Very poor.   The Behringer's storage is somwhat less at operating voltage. 10.6F x 84V = 890.3 Coulombs x 2 (rails) = 1780.8 Coulombs = 74.8 Joules. This is 17.9 calories... They are deffo LOW CALORIE

Already covered the current thing but can you elaborate on your disappointment with the linearity. I guess you mean the THD? I didn't think it was too bad but certainly no decent class-AB!

The energy figures I used included the primary side capacitor storage as that is very relevant for switch-mode supplies, that energy is just as accessible from the secondary side as from the secondary caps directly. Yes there are some small losses in the transformer and switching devices but that energy is still available to the amp.

How it compares to a 4kW class-AB is a relevant point but I dont really have one I can take figures from. If it were to have say 40,000uF per rail and a rail voltage of 150V then the class-AB would have an energy storage of 900J. From that perspective the Class-D's are a bit lightweight and hence cannot generate the same power at 31Hz as they can at 1kHz. They're not disastrous by any means though.

kedwardsleisure wrote: im confused about the storage figures,surely ripple is a better measure than capacitance?

Ultimately for the period of time between refreshes (100Hz for our 50Hz mains) its the energy stored that will be used to fill the gaps. A smaller stored energy will lead to a faster lowering of voltage, or more ripple, hence the two are related. Its just easier for people to comprehend I thought, ripple is of no concern to the end user, just the available undistorted output power.

as suggested a few more "compare-the-power-amp.com"s would be very welcome by one and all.

you are a true soldier to the cause and i wish to thank you for all your hard work and the right up has been a damn good read.... gotta love those thermal pics too!!

keep up the good work and we all thank you for your time.

Hi Mik, Yes you're right, the current probe is fine for basic AC input current analysis but won't show the harmonics of the typical amp/smps switching frequencies. My scope is 4 channel so i'll definitely be having a go at the inductive load testing to see what's going on.

The scope also does FFT but doesn't have any low-pass filtering so any signals above the Nyquist point (which is timebase dependant) get aliased down into the spectrum and can give false readings. I'll have a look at it though as it could be interesting.

vibez wrote: Having read this thread yesterday i decided to try my inuke (that usually run monitors ) on sub just for the fun ofit. I must confess that i was quite suprised by the result. i hung two 1850 loaded scoops off each channel and this little budget amp spanked them pretty well. I must admit that i wouldn't feel absolutely comfortable using this on sub long term but for emergency situations i wouldn't hesitate to use this to get me out of trouble. The only thing i did notice was that when it hit clip level the limiting was very audible but once limiter on dcx was set to prevent this it sounded pretty good and certainly delivered more bang per buck than i have ever heard from an amp in this price range. Haven't tried the dsp yet but will have a play at the weekend when i have more time. Top review and suprisingly good budget amp

Was that an NU6000 DSP then? If so then I'm not surprised it sounded OK. To really give people confidence it needs someone like yourself to actually run it at a gig and spank it into 4 Ohms for a few hours, that's something I can't do. I think it will be a cool running and reliable amp so as long as it sounds good too then we have a winner!

MessyM2k8 wrote: Nice report.  Out of interest, what test gear are you using?

The audio performance side is done with an Audio Precision "Portable One Plus" which is interfaced to a computer with GPIB. A custom Visual Basic script then runs in excel to allow me to make graphs etc. The Portable One Plus was never intended as an R&D instrument really so its difficult to get data off of it but this little set-up works really well and the analogue performance of the AP is outstanding.

Power metering is courtesy of a Hameg NM8115-2 Programmable Power Meter.

Burst signals were generated on the computer using the MTPM signal generator program, which is brill!

Scope is a Tektronix TDS2024B 4-channel 200MHz

VentureSound wrote: Teslaman, thanks for all your tests, it was a very interesting read indeed.  I would be keen to see some similar results on a Peavey IPR series, as I guess they are somewhat comparable being class D, 'new generation' and cheap.  I've heard an IPR on subs and was suprised as well, sounded pretty good to me! Cannot remember the model, but it was one of the lower ones like 1-2KW.

Yes the Peavey class-D's would be more than welcome to visit the bench. I suspect they are the same chip-set as the Behringer but different layouts etc. can make big differences to how things sound. As someone mentioned in this thread Abletronics has already tested the 1600 so you can see the results of that here  http://www.abeltronics.co.uk/amptesting.php?z=peavey_IPR-1600" rel="nofollow - http://www.abeltronics.co.uk/amptesting.php?z=peavey_IPR-1600

Timebomb wrote: Id be interested to read that if you've got a link BitSmasher? I would be very keen to see measurements of the Peavey IPR and the Crest Prolite,  it seems only the smaller 2 models are available though, the larger ones have been forthcoming for several years now, and in the UK we haven't even got the IPR3000 yet i dont think.  They are defiantly on the short list for budget lightweights though.

I'd be very interested in a link to that comparison too please BitSmasher!

If someone has an IPR3000 or Crest Prolite they want testing then drop me a PM.

Cheers,

Dave.

These figures also tally with my measurements for output power vs input voltage (which I didn't publish as there not really of interest!)

It sounds fine to my ears, not as smooth and seductive as the best class-AB amps but certainly not harsh, better than you would expect for its power and price.

U.Viktor, have you even listened to an NU6000 amplifier? I gather from your posts that you have a certain preference for PKN and little else is any good for you. There is room in the universe for half decent budget amps for those of us who can't afford the likes of PKN etc.

My advice is go listen to one for yourself, just like with any product, and make up your own mind.

Maybe someone would like to offer a PKN amplifier for the test bench! 

+1

I'd also be interested in knowing how the d4-500 performs. I have been eyeing them up as suitable compact monitor amps and backups for the more meaty amps.

Potential for great value for money!

Cheers,

John

i get the feeling they havent given the 1850s any where near enuff power to hear what they can do!!!

I'd also suspect that any failures are pretty catastrophic, taking out far more surface mount components than would be practical to consider repairing.

 

So what is the real-world peak output of the Behringer?  I realize that might not be a spec that the reviewer here is comfortable necessarily using, but it's one a lot of speaker and sub specs have.  I'm thinking in terms of matching amp to speaker capabilities, or vice versa, and also in terms of voltage limiting to prevent driver damage assuming proper filter and crossover settings are used. 

 

If that's a tough spec to actually quantify for some reason, what's the lowest power handling rating for a properly-rated speaker you would drive each channel of this with to full power, and what's the highest power handling rating speaker you'd want to pair these with but still get the most (in your opinion) out of the speaker?

 

Do you mean peak in the sense of peak voltage or do you mean peak in the sense of power for short bursts like in a music signal? As per my report the music-like output ratings are:

If you wanted peak figures in a "peak voltage" sense then you can double these power figures or convert it to voltage. Peak voltage for full range is 148V/8 Ohms or 128V/4 Ohms, and for bass duties 136V/8 Ohms and 120V/4 Ohms.

In terms of matching an amp to a speaker then I believe conventional wisdom is to have the amplifier rms power equal to double the continuous rms rating of the loudspeaker, or equal to the "Programme power" rating of the speaker. Provided you don't abuse the amplifier (lots of clipping) or use an LMS with the limiters set appropriately then this approach is considered OK. Speakers are damaged through the long-term average input power being too high (overheating) or for mechanical reasons (inappropriate high-pass filter settings or none at all). 

Having an amp capable of double the thermal rating of a loudspeaker sounds dangerous but when you consider the "crest factor" of music it makes perfect sense. You actually need a significantly higher power amp to be able to deliver anything near the thermal ability of the speaker. That gives you extra headroom also allowing a very clean sound even at high levels. Even the most heavily compressed music will have a crest factor of 6dB or more, meaning the average power will be one quarter of the power during the musical peaks, hence, provided you are not clipping, your 2kW amp will deliver an average power of no more than 500W...

I hope my explanation makes sense 

Of course if you further compress the music, or drive any stage of your signal chain into distortion (which decreases crest factor) then the safety margin gets eroded and eventually you will burn out the speaker.

Not sure on the nu3000 but it was possible to make the nu6000 shut down when driving two 4 Ohm loads very hard. I'm hoping they've rectified this issue by bringing in the limiter before the power supply locks out, after all you want the show to continue and not lose sound mid-gig!

I think I read that the nu3000 does this correctly, making sure the limiter acts first, so you may get away with 3.3 Ohms on that but I'd be hesitant with the 6000.

Only one way to find out of course  I don't think you will damage anything trying...

As I reported in another thread the amplifier I used for the test suffered from this exact failure, confirming its credibility somewhat. Said amp was replaced without question and its replacement is still going strong. I'm still of the opinion this is a good value amp and its a shame such an issue spoiled the party.

I currently have an inuke 6000.  I use it only to run one dynacord dlite sub 115.

It's not as loud or as deep sounding as the internal dynacord amp on my other sub (the matching power sub 115) which is 1000 dynacord watts bridged into the 15 inch sub. 

I'm wondering would the Kam do a better job than the inuke if I bridged it.....

Any thoughts?

Try a known-good sounding external amp instead to confirm this point if you can but the inuke 6000 is a good bass amp and has around 1000 "real watts" into an 8 Ohm load per channel.

Bridging an amp always increases the stress on it so its unlikely to sound better unless you really do need more power (unlikely).

If the dynacord has EQ and filtering built in then you could mimic that with an external speaker management system (e.g. ultradrive etc.) or use an inuke 6000 DSP unit and set the filters accordingly.

No issues with the amp starting up when the diode was OK.

With the DSP section of the amp, it has a setting for the peak limiter (in the configuration screen via laptop connectivity).

Does the "W" stand for watts and if so, when one moves the knob to the approximate wattage of the speaker (say 300Wrms), will this feature limit the volume/ output given to the speaker even if the amp gain knobs are turned to maximum.

 

I am wanting to use the NU6000 to power a pair of midtop (300Wrms each) on one channel and a pair of bass bins (700Wrms each) on the other. This will be used for mobile dj application.

 

I do know how to set the cross over points on the unit, just the peak limiter I am a bit puzzled about

 

The first time I used it was outdoors in very high heat and out in the sun. I ran it agressively without a single hiccup. This was with canned music. The second was with a small band for a wedding under an event tent. Overall volume wasn't super loud and everthing went well. Last Saturday, my band played for a large outdoor gathering. Everything mic'd with an experienced guy running FOH. I couldn't tell from behind the monitors but the sound guy told me the system would periodically "drop out" and immediately come back.

 

One side ran the subs (4ohms) and the other ran the highs (also a 4 ohm load.) This wasn't thermal shutdown and the amp wasn't being abused. The clip light would ocassionally blip, but nothing pegged. During the breaks we played dance stuff with a lot of sub bass, etc. with no problems. With the band it was different.

 

I've been trying to figure out if it's the venue's power that was the problem or the amp just not handling a 4 ohm load well.

Need some advice guys, 

Got pair of PD2450 8Ohm and Im considering the option to obtain Nuke 6000 dsp to fuel them and setup it for sub only.

is this combo gonna work well together?

tks