**See inside an example of what can be achieved, the Behringer NU6000dsp.
formatting link
This amp weighs 5.5kgs, has four PWM power stages wired in bridge mode pairs. There are only 4 MOSFETS in each channel delivering 2.5kW of sine wave power into 4ohms with acceptably low THD (under 0.3%).
The case is aluminium with a 2mm thick doubler plate on the bottom acting as the heatsink - thought it barely gets warm on music programme.
The retail price is tad over US$600, though you regularly see them heavily discounted to nearly half that. I kid you not.
I had one on my test bench recently and could not fault it.
There are two PCBs, PSU and power amps plus another inside the front panel. These are easily replaced but Behringer will not sell them to third parties.
The larger, through hole parts are all replaceable and available. The SMPS & PWM driver ICs are common numbers. There are fans, an NTC thermistor and a relay that could fail and are easy to replace.
The DSP stuff inside the tin box and would be impossible to service.
FYI:
I have a QSC PLX3402 here now ( similar mixed construction, but on one PCB) where one power stage failed and wrecking a lot of parts and shorting the main PSU rails resulting in massive destruction to the two IGBTs and drive circuitry.
They've been around for years. I built one in 1980 or so (with a high-voltage PNP transistor as a switch), but could only get relatively high output impedance to work. The big problem, was my simple design (just PWM by feeding a comparator with a triangle wave and a sound input, and using that to drive the transistor). To make that kind of thing work with a resistor pullup output stage, required either limiting the power output severely, or getting some VERY expensive NPN or PNP power switch transistors.
Those lines are a bit hard to interpret. According to Behringer home page, this is a 6 kW (2 x 3.1 kW @4 ohms) stereo amplifier.
Those must be some kind of PMPO figures, since according to the video, the power supply connector on the back of the amplifier is ordinary IEC device connector, which are usually rated at 10 A 250 (= 2500 W continuous power) quite a lot less than the rated 6 kW audio output power.
For 6 kW continues output, you would need at least 3 x 10/16 A (230/400 V) three phase feed (with the red IEC 60309 connector).
** Behringer sometimes use their own definition of a watt instead of the usual "watts rms" number.
My figure is actual measurement, under sine wave drive and resistive load.
** Nope, the amp can deliver 2.5kW from one channel for long periods. I did not try both at once cos it would have tripped my 20A supply circuit breaker.
The amplifier, like all audio amplifiers is sold to be used for AUDIO - not driving an incandescent lamp or motor for hours on end.
** What do you think happens if you exceed the 10 amp rating ??
A massive explosion maybe ???
Get real, the pins just heat up a bit more than usual.
The nominal rating is for a lifetime of service and allows for hot plugging and unplugging.
llowed PWM frequencies of 250kHz or more with >very low losses."
o drive them. With MOSFETs all you have to do is effectively charge and dis charge a capacitor.
It can be a fairly large capacitor, and you can want to do it quite frequen tly. Driving a big MOSFET hard and fast can take quite a bit of current.
ich negated some of the efficiency. And you still had the capacitance probl em. You also had to go negative to make sure it turned off fast enough. So yeah, MOSFETs were a nice development.
somewhat smaller PSU. If the amp has a SMPS as >well, then there is also a major reduction in weight. "
ules, hundred watt modules and who knows what else. No heatsinks at all. Th e only heatsink is the copper on the board. Icepower modules are made by B& O which is considered a high end audio company.
ssue and it is rare to see a class D >amplifier with AM tuner in the same b ox."
emember, in this country, that if you had a color TV you had to keep AM rad ios away from them if you wanted to hear anything useful. The sound and col or subcarriers combined to interfere with the IF frequency of 455 KHz. If y ou switch at 250 KHz, the second harmonic is pretty close. (in more ways th an one)
. Ideally the PWM would be done with a comparator fed by a sawtooth wave fr om an oscillator.
That's one way of doing it. A fully digital scheme to generate the pulse-wi dth modulated waveform offers more options.
I see in the diagram they're showing the output pair as being of the same polarity for some reason. But it's a useful chip for this kind of application nevertheless.
I think the audiophiles even don't consider a switchmode powersupply acceptable for their amplifiers... and of course the amplifiers they love are designed with such bad powersupplies that they need to use a separate fuse group with audio-rated fuses and special connectors to avoid any influence of the power on the output signal.
If you're going to use MOSFETs instead of BJTs for the output stage, though, as I think we all would these days, would I be right in thinking you don't have to worry about obtaining a closely-matched pair? After all, there's no need for the transition region to be predictably linear, is there? Sorry if I didn't put that very well but I'm still low on caffeine at present.
Your wordfing was not clear if you are taking about a stereo amplifier consisting of two bridged stages. If four 1600 W stages bridged into 2 ohm load would produce a bridged stereo signal into 4 ohms at 3200 W
Unfortunately. your figures do not add up.
Two channels into 8 ohms or something else ?
At least in Europe, you will more likelu find a 3 x 16A sockt than a
320VDC supply, buck configuration. Bottom is a 38HE7 beam tetrode (2nd-to-the-bottom div is GND, 50V/div, so it's saturating under 25V, not bad at all. Top is the damper diode section of the same tube, clamping it against +320V. Can't really see the divs on the top of the screen, I guess, but the saturation voltage is similarly low.
With modulation, it looks thus:
formatting link
Notice the switch AND diode saturation voltages increase with increased load current (and increased duty cycle, visible by the change in trace intensity).
The plate efficiency was around 80%, not bad at all for tubes.
But considering it's buck configuration -- that means the load current flows through a resistor, and the signal is capacitor-coupled to the load. So, it's Class D-A: Class D switching, but Class A load and output coupling. Overall efficiency 20-ish percent (less heaters), which is more what you'd expect for a single-ended amplifier. :^)
Hahah, some rookie mistakes in that one. Here's a better one:
formatting link
I haven't built it yet. The biggest problem is that, in saturation, the cathode current doesn't track plate current ... so the remaining current finds its way into the screen grid. Meltdown quickly follows.
The easiest solution is simply supplying the screens with an RC, so they ride it out with limited power (this causes saturation voltage to suffer slightly, but that's okay).
Notice because the output transformer has to handle audio frequencies, there are no advantages to be gained there. You should still use a ferrite cored inductor (two winding, differential mode, for PP outputs) to filter the switching waveform away from the OPT, but otherwise it's very conventional.
More recently (last year), I got as far as planning a layout,
formatting link
which should be good for 200W output at more than 60% _overall_ (including heaters) efficiency. Hopefully the switching frequency will be over 200kHz, using some unusual drive circuitry.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.