Citation :
QAA4125
An Unusual Amp That Leaves Good Impressions
TEXT. BOB NORVELLE
TESTING: BEDROCK ELECTRONICS
PHOTOGRAPHY: CASEY THORSON
Sure, I've dabbled with MB Quart over the years: some home speakers, some car audio speakers. My experience has been that they have delivered better than average sound for a higher than average price. Now, like many "speaker" companies, they are making amplifiers. Is it just something to sell with their speakers or does the MB Quart QAA4250 "Quad Mono" amplifier stand on its own merits?
The QAA4250 is a regulated Class A/B amplifier with a digital R.E.A.D. display on top that tells you about battery voltage, temperature, gain, crossover frequencies and so forth. MB Quart has some rather curious power ratings for the QAA4250: 4 x 250 watts at 4 ohms, 4 x 250 watts
at 2 ohms, 1 x 500 watts bridged (rear only) at 4 ohms and 1 x 500 watts bridged (rear only) 8 ohms. Yeah, I know what you are thinking, you're supposed to get twice the power at 2 ohms—I'll explain more about this later. For crossovers, it has 50Hz to 5kHz highpass and lowpass selection and a front-end (preamp) bypass switch in case you have a really clean 2.4-volt input and don't want to use the internal crossovers.
COSMETICS
This amplifier, well, doesn't really look much like an amplifier. I can't quite figure out what it reminds me of, but it will come to me later I'm sure. My description may not be very complimentary, but it looks like a large, metal box with rounded corners and a digital display on top looking up at you. The outside cover is a large cast piece of aluminum that runs across the top and down the sides to meet the bottom "tray" of the same material. Cast endcaps cover the fans on one end and the connections at the other end. Everything is powder coated a dark gray, almost charcoal color, which contrasts nicely with the natural-finished, brushed extruded aluminum heatsink fins that you can see through large slots along the edges of the amp.
I like the fact that all of the connections are at one end of the amplifier. On top of the amp, right in the middle, is a square cover that can be removed by taking out a single Phillips-head screw. Beneath it are the crossover, bypass and load impedance switches. The switches are a variation of a toggle switch and provide a very pleasant, high-end look. The crossover frequency and the front and rear gain are adjusted by way of four thumbwheels -very nice, indeed.
CIRCUIT DESIGN
If you're a regular CA&E reader, you probably already know my thoughts on amplifiers—keep the big signals away from the small signals and MPPBS the power supply away from the inputs. The engineers at MB Quart have used possibly the best way to accomplish thisthree separate circuit boards stacked up with the input and crossovers (tiny signals) on top, then the amplifier itself (bigger signals) in the middle and then the power supply (big signals and lots of noise) upside down on the bottom. The amplifier and the power supply boards are suspended solder-side to solder-side about 1/2" apart and held against two shared heatsinks running the length of the amp on each side. This internal arrangement reminds me more than a little of the Precision Power 2500F1 ($15,000 and 2,000 watts) from a few years back.
The circuit boards are double-sided fiberglass with both surface mount and through-hole components. The input/crossover PCB uses Bournes potentiometers for gain and crossover frequency controls, which feel very smooth and weighted with their molded, thumb wheels. High-quality toggle switches and film caps are used to keep the signal path clean. One of the toggle switches accomplishes a bypass of the gain and crossover controls, sending your input signal directly to the amplifier. A flat, multi-strand computer cable connects this PCB to the amplifier board. There is flash-programmable chip and a digital display on this PCB, but the amplifier does not perform any digital processing on the signal itself. What it does is read the positions of the switches and potentiometers and displays what the status of the circuits should be according to the switch and pot positions. I say should because the display has no way of knowing for sure. The crossovers and gains are analog, not digital, and therefore prone to differences due to part tolerances or even incorrect or defective parts. As an example, let's say you set the crossover so that the display reads 2,OOOHz. The resistance through the pot, combined with a capacitors(s) and an op amp determines the actual measurable crossover frequency. The display shows 2,OOOHz based on the position of the potentiometer and the program in the chip. If the other components in the crossover circuit are incorrect or the op amp is dead, you won't end up with a 2,OOOHz crossover. So the accuracy of the display depends on the consistency of the production parts as well as the programming of the chip. Don't be discouraged, though. This is way more accurate than trying to read the numbers printed on the end plate or counting clicks on the pot. It also gives you voltage and temperature readings, reminds you that the inputs are bypassed and alerts you of protection modes'. It is cool, but it's not DSP.
The power supply board is all business. Imagine opening an amplifier and seeing nothing but power supply. From the bottom side, that's exactly what you have here. Actually, there are two complete power supplies—one for the front channels and one for the rear, mirror images of each other down the length of the amp. B+ (+12 volts) enters the amplifier through a large split-pin arrangement, where a separate terminal block accommodates up 2awg cable. There are six 2,200uF/16 volt caps on the primary side for a total of 40,800uF (almost 1/2 farad). The power feeds two litz-wound transformers switched by two pair of Fairchild 285 watt Mosfets (each). Total dissipation capability is 2,280 watts. The switchers are driven at a frequency of 31kHz by a pair of Motorola MC34025P PWM (pulse width modulation) controllers, which are the brains behind the protection circuitry and the regulation of the power supply as well.
This would be a good time to address the power versus impedance trick built into this amplifier. Remember the power ratings of 250 watts x 4 at 4 ohms and at 2 ohms? There is a "load impedance" switch in the control panel of the amplifier allowing you to select either 4 ohms or 2 ohms. The 4-ohm setting gives you rail voltages of +53 volts and switching to the 2-ohm position changes the rail voltage to +43 volts, which will result in the same power output as the 4-ohm setting with the higher rail. Simply put, the switch changes the input going back into the PWM chips, which then adjust the duty cycle of the pulses to raise or lower the rail voltage.
I know what you're thinking—put the switch in the 4-ohm position and use a 2-ohm load! That's what I thought, too. Forget it, it doesn't work. The amp actually puts out less power this way. I know it's a drag to follow the instructions, but use the load switch the way it's labeled for max power.
Where was I? Oh yeah, the voltage rails go through a pair of huge litz-wound inductors and past a bank of huge capacitors containing eight, 6800µF/63 volt secondary stabilizers (54,400µF, or just over 1/2 Farad if you're playing along at home). It's amazing what you can do with all that room and MB Quart's definitely done it. The rails are transferred to the amplifier board by way of brass rods at the far end of the RGBs. Screws go through the boards into the brass rods, which also act as spacers to keep the RGBs separated. There are a total of 13 of these spacer/power conductors.
OK, we've got signal and we've got power, let's combine the two in the amplifier section and send a big, powerful signal out to the speakers. The signal comes from the input board through the computer cable up at the end of the RGB nearest the RCA connectors. There is a stereo/mono switch for the rear channel, which inverts the left rear channel allowing you to bridge the rear channels, but this board is mainly just an amplifier.
MB Quart is using National Semiconductor "transconductance" op amps, and yes, that is a word. What it really means is that the low-level gain stages are current controlled rather than voltage controlled. These current-controlled op amps directly drive the output devices, with no other devices in the signal path, which should result in much higher fidelity from this amplifier than with more conventional topologies. There are 2 pairs of complementary Fairchild Mosfets per channel, each capable of dissipating 227 watts of power. That's over 900 watts per channel, giving the QAA4250 the prize for highest margin of safety in the output section. It also means that the amp should be able to deliver impressive transient performance, pretty much limited only by the PWM chips in the power supplies.
Another unique feature of this amplifier has to do with how the "hang-off" devices (the high-power outputs and switchers) are coupled to the heatsink. All of the usual hang-offs are actually soldered flat to one of 4 daughter boards, which are then bolted to the heatsink itself using plenty of heat transfer grease—you know, the white stuff that won't come off! It's a little like surface-mount, only the devices are huge TO-247 sized Mosfets. The daughter boards are not standard PCBs of course, since most PCBs are made with insulating materials. These actually appear to be thin aluminum plates, about 3/32" thick. There are also some circuit traces on the boards and surface mount components attached between the larger devices, presumably parts of the temperature-monitoring circuitry. I can see how the heat transfer from the output devices and the switchers into the heatsink would be very fast indeed, making this amp even more bulletproof
PERFORMANCE
The measured frequency response correlates well with MB Quart's figures. They state 0.25Hz to 100kHz as the -3dB response. I measured the low end at 7Hz (-1dB) and the high end was well past 48kHz on my Spectralab setup. Making voltage measurements with a signal generator and a scope, I estimate the top end to be in the upper 90kHz range.
The test results were very impressive, with the power measuring well above the ratings in all categories, topping out at 276 watts per channel into 4 ohms and 294 watts per channel into 2 ohms at 1 %THD. The amp is very regulated, showing virtually no change in power output between 14.4-volt and 12.5-volk inputs. The crossovers were very accurate with picture perfect curves on the graph. The protection circuits all worked well, even though the amp got frighteningly hot to the touch before shutting down thermally. I have found that a smooth surface tends to feel hotter than the fins of a heatsink, and I think it has to do with the amount of surface that your hand is making contact with. In any case, it seems to transfer heat away from the electronics very well indeed.
I was disappointed to hear a very slight turn-off pop from the amplifier through my front speakers. I stack the deck against the test subject (hey, it's my job!) by turning the gain way up for the noise test and there was no turn-on noise at all.
MANUAL
The owner's manual was downright scary because it's really fat. Opening up the novel, I realized that I only had to deal with 1/6th of it. Besides English, the manual is written in French, Spanish, German, Italian and Russian! I can only speak for the English instructions, of course, but they were good. After the obligatory "Congratulations you're a genius for buying our amplifier" spiel and the cautions/warnings, it gets right to the controls with clear pictures and enough information that you can hook it right up. Speaking of hooking it up, you need three different Allen wrenches to connect the power and speaker wires—not my favorite thing. Following the hookup is a section that clearly tells how you should set all of the switches before firing it up, and over a page explaining how to set the gains correctly, which is my favorite thing. The manual winds up with a couple pages of troubleshooting and then the warranty page explaining the 1 -year (3-year if installed by a MB Quart dealer) warranty and how to go about it. It's a pretty good book, especially since you can stop before you get to the French version.
LISTENING TEST
My favorite part of all these reviews is getting to sit in the driveway and listening to all my favorite CDs while the neighbors shake their heads and roll their eyes. This time was no exception, with 1,000 watts of MB Quart power distributed as follows: 250 watts to the front left and front right separates, and 500
watts mono to my 12" DVC sealed cabinet sub-woofer. I set the front crossover to 125Hz highpass, the rear crossover to 10OHz lowpass and both gains at a fairly low level to start out. I started with Donald Fagen's Kamakiriad to get the gains and crossovers dialed in, and then climbed in for a listen. The first thing that struck me was that my front speakers were putting out quite a bit more volume than usual before starting to pop. With 250 watts available I had anticipated either turning the gain down or raising the highpass crossover point to avoid blowing them out. The second thing was that the sub seemed to be doing the same thing, playing somewhat louder than usual before popping. It was exhilarating to be able to play the separates low enough to actually blend into the sub frequencies.
The low Fender Rhodes notes in "Snowbound" were really clear as a result of the lower frequency setting on the separates, while the cymbals in "Tomorrow's Girls" were very live as well. I played Mary Chapin Carpenter's vocals, which seemed a little muddy with the guitars in "Rhythm Of The Blues," but a slight cut of the bass control fixed it right up. Whenever I want to trip up a system with mid-bass weakness, I always pull out my old test CD with Michael Ruff's "Eyes of Love." There is low drum that just sounds terrible if your system is out of balance. Too much midbass and you can't hear anything else, and if you try to equalize it to fix the drums, you always screw up the vocals. All that said, there was no problem at all on this track. OK, enough of the cheesy old guy sound quality stuff, where's my Bass Mekanik? Actually couldn't find it, maybe I loaned it to my mom? I did find a CD with Yello on it, Resistor being my favorite cut—an incessant driving bass, with the trademark yells in a foreign language that Yello is known for. What a great massage and all the neighbors went inside. I also found my old Extreme CD with lots of tight kick drums, bass guitar and distorted electrics. What can I say? I was very pleased with myself. I spent a lot of-time in the driveway with this amp, just enjoying listening to my CDs and several times hearing new material on them. To me, that says it all. To listen to a familiar cut, but hear something new means that this amplifier is not hiding anything.
CONCLUSION
I really, really like the sound of this amplifier. I have to go back to the Audison HV-Thesis review to remember anything like this sound quality. However, there are a couple of things about the amp that I don't like—the need for three Allen wrenches and that the connectors and switches on the end panel are difficult to access or even see after the amp is installed. So have an authorized MB Quart dealer install it and you don't have to worry about it. Another really good thing—this amp seems to be very overbuilt. I would gamble that it will be very reliable over the long term simply because it operates so far within it's capability. This kind of quality comes at a price, however. The retail price for the QAA4250 is $1,795.00; at a measured total output of 1176 watts, that's about $1.53 per watt-definitely on the high side for this size amplifier. But then again, this amplifier is perfect for a system with really nice front separates and a high-end dual 10" or dual 12" sealed box sub cabinet. If you're into sound quality, you won't be disappointed.
Output power @ t%THD, 1kHz, 14,4 volts
Stereo @ 4 ohms .............................4 x 276 watts
Stereo @ 2 ohms .............................4 x 294 watts
Bridged © 4 ohms ......................1 x 529 watts (Rear Channels)
Output power @ 1%THD, 1KHz, 12,5
Stereo @ 4 ohms .............................4 x 270 watts
Stereo @ 2 ohms .............................4 x 288 watts
Bridged @ 4 ohms ................1 x 522 watts (Rear Channels)
Distortion at rated power, 1kHz, 14.4 volts.........0.45% © 4 ohms
Input sensitivity, "Low" setting ................260mV to 2.4 volts
Input sensitivity, "High" setting ..................2,3V to 21 volts
Input sensitivity, Front End Bypassed ...................2.4 volts
Frequency response (+ 1dB) ................7Hz - beyond 48kHz
S/N ratio (A weighted, below clipping, min gain) ............>89dB
Slew rate........................................18V/mS
Damping Factor & 100Hz, 4 ohms .......................>100
Idle current ......................................6 amps
Maximum current consumption, undipped . 104 amps @ 1,053 watts
Efficiency at 1/3rd power lowest impedance.................32%
Efficiency at full power, 1%THD, lowest impedance ...........67%
Crossover slope................................12dB/octave
Crossover range, lowpass .....................55Hz - 5,156Hz
Crossover range, highpass .....................52Hz - 4.875HZ
Dimensions ....................151/41x11 3/4"Wx37/8"H
|
) par sa frappe et sa "rapidité".
