Studiomaster MOSFET 1000
Ben Duncan gives this power amplifier a thorough going over before telling you that he quite likes it.
Studiomaster is a US brand name, used by RSD, whose initials stand for 'Recording Studio Design.' RSD has its origins in a partnership between ex-Led Zep sound engineer Phil Dudderidge, and Paul Dobson. Over a decade after an amicable splitting to setup his own empire, Phil has become successful as Mr. Soundcraft. In the meanwhile, Paul Dobson has built up RSD's reputation as one of the most cost-effective and streetwise makers of mixers and amps for studio, stage PA, clubs and discos.
In the realm of power amps, RSD have not had to release many models. Their first serious power amps, the 800B and 800C (introduced in 1977 and 1979 respectively) became classic Rock 'n' Roll workhorses accepted by professional users, who wouldn't normally have any use for RSD mixers! They were bipolar amps, a development of the Phase Linear 700, an esoteric US Hi-fi amp, which sounded good and had lots of welly, but was also expensive, and needed extensive modifications to keep it in one piece on the road. The RSD amp's roadworthiness was proved to my own satisfaction back in the late 70s, when I worked for a band called The Jam (remember?), whose PA had 42 RSD 800CS heating up the air backstage. Despite trying out more expensive models from RSD's arch-competitors (2 of whom no longer exist... Can you guess why?), the 800C was the only unit which fulfilled the most basic criteria: it didn't blow up or fall apart and it sounded good, even when wildly overdriven on numbers like 'Eton Rifles'. In those days, you had to be brave to use anything other than a 'name' amp like Amcron or BGW, yet the gamble paid off. This month, we're reviewing the MOSFET successor to the trusty 800C. Housed in a 3U enclosure, MOSFET 1000 sources a nominal 500 watts into 4 ohms, or 300 watts into 8 ohm speakers. Bridging into 8 ohms is also available, but the power output is unspecified.
I liked the BGW-lookalike handle up-front, but found the ones at the back more comfortable for carrying. Nevertheless they all protect the panel hardware. The gain controls are pseudo-stepped attenuators, in other words, ordinary control pots, with a series of detents. The result is some wild mistracking when the two channels are compared, as table 1 reveals. This shouldn't be enough to stop you tuning the stereo balance, but any gain adjustment deserves a balance check.
For more mundane PA work, you'll be interested in the dB scaling, which divides the pot into 4 equal 10dB sections, giving a 40dB control range overall. Full shutdown is on the last click-stop; it can prove handy when tracing buzzes, and you need to mute the amp but not necessarily at the console. Between shutdown and full gain, however, 40dB is way too big a range. A 30dB attenuation would be ample, for even this represents a thousand-fold reduction in power output. A 40dB attenuation implies turning 500 watts down to 50 milliwatts!
Below each gain control, there's a yellow thermal LED. These flash if the temperature exceeds 80 degrees, while the output relays ditch the speaker connections. More subtly, the LEDs flash immediately after switch-off in one of two distinct patterns. To understand this, you'll need to know that the relays also open up if there's a DC fault. The trouble is, RSD have omitted to provide a dedicated 'DCF' LED to indicate why the sound has abruptly halted. That's until our Dr. Marlowe discovered the secret LED flashing code, not mentioned in the manual. If the amp is working normally at switch off, the two yellow LEDs flash once or twice, in tandem. However if there's been a DC fault, and one or both of the relays have dropped out, you can confirm this by switching off and observing the two LEDs flashing alternately, 3 or 4 times before they fade. This will save you scrabbling round the back of the rack with a testmeter in your mouth, should the speakers go silent and you're not quite sure what's going on.
Moving up the panel, a further pair of LEDs remind us that we've selected the power bridging mode, and also confirm either 'mono' or stereo operation. 'Mono' in this case means an internal link between channel 1 and 2 inputs. For PA, this cuts down on leads where amps are driven in tandem, into adjacent speakers working in the same frequency range. Reminder LEDs are needed for both functions because when the bridge LED is on, the stereo LED has to be off for correct operation. Put into colours, you can have yellow, red or neither (not bridged, mono mode), but definitely not both on together.
The mains power switch is inadequately rated for long term reliability, but the MOSFET 1000's toroidal transformer is useful if we want to rack the amp up next door to magnetically sensitive gear like tape machines, effects and equalisers, because toroidal transformers reduce induced hum by about 10dB over a big standard laminated transformer. The magnetic circuit that cuts back stray fields and harmonic 'splatter' runs at low saturation, though, meaning the primary winding has a very low magnetising resistance. Therefore all toroidal transformers pull a big shot of juice at switch on; I have measured 150 Amp surges even on a lousy domestic ring-main. Much depends on the mains voltage at the instant of switching on. As it surges up and down fifty times a second, the lifespan of any particular switch must be the outcome of a thousand or more stabs on the 50 cycle roulette wheel!
Harking back to the articles on studio mains supplies, the MOSFET 1000's switch has just the sort of snubber we warned against: a 100nF capacitor is connected directly across each of the switch's poles, without the vital current limiting resistor. The eventual result outcome of heavy arcing is always a dead switch, in maybe a week, or maybe in seven months time. I hope that after reading this, RSD will reduce the value of this capacitor, or fit a proper RC snubber network, knowing that a 5 pence resistor can make a world of difference to reliability, as perceived by individual users, who cannot experience any statistical variation: one dead switch is a 100% failure if it's your only amplifier...
If you've seen any RSD mixers recently, the LED bars are the same type, but with a different, expanded scale around the zero level, the amp's clipping point. The meter ballistics are peak reading, with a slight hold, corresponding closely to the BBC PPM (Peak Programme Meter) specification. Expanded Scale means a scale of 1dB LED steps as we near 0dB, with a ½dB step just below 0dB, and another LED, marked Clip above. This is set to turn at a specific voltage, representing the overload threshold into a particular load impedance - in this instance 4 ohms. So it's not in fact a true clipping-indicator. For example, if the load impedance dips on peaks to below 4 ohms, actual overload will seemingly be 1dB premature, or more, relative to the Clip LED, which will end up reading what has already occurred. This is naughty in so far as stage speaker arrays frequently display lower impedances than anticipated. Thankfully for studio work, most impedances will be above 4 ohms, so the RSD's Clip LEDS will read conservatively, lighting up a fraction before the crunch.
Measurements proved the metering capable of responding up to 300kHz with fair accuracy. The practical outcome of this is that the read-outs will give some sort of indication if there's any inaudible ultrasonic nasties flying about. At the price, the meter accuracy (Table 2) is acceptable, but I'd rather have seen better accuracy on the top 4 readings, even at the expense of those lower down, not least because the ½dB discrepancy on the -0.5dB LED represents an uncertainty of 100%!
Twin fans are mounted on the back panel in a unique fashion, blowing both into the case, and laterally, along the external heatsinks. Unfortunately, the heatsink's profile creates a laminar airflow which dramatically reduces the cooling available.
Meanwhile the heat flow from the devices to the heatsink itself is through a relatively thin 12 gauge L-bracket, which is bolted onto the main heatsink member. What's more, the heatsink's inner 'U' section, which receives most of the cool air, is bolted to the main member, but neither surface is machined dead flat, which carries the risk of voids, (air gaps which prevent heat conduction).
The upshot of this doesn't cast doubts on the RSD's ability to keep cool in the studio. On tests, the MOSFET 1000 did better than some amps we could mention. However, it's all at a price. Power output per rack unit (or per cubic inch), heightened long-term thermal stresses and an inability to drive awkward loads without thermal trip-out are all consequences of inefficient heatsink design. Having said that, the RSD is unusually lightweight, and relatively small considering its output capabilities. So where's the tradeoff? Well, the metal can MOSFETs are being run at routinely higher temperatures than would be comfortable for Bi-polar devices. This leaves a tighter margin against the 80 degrees thermal trip out temperature, and prejudices the longevity of the MOSFETs in heavy duty stage usage.
Back on earth, we have the fan Stop/Run switch. Whether your RSD can live without fan cooling in practice depends on vague factors, like the ventilation in your rack or cupboard, the load impedance and the drive level. The easiest way to find out is to suck it and see. If the amp trips out thermally it looks like you're going to need the fans. But if the thermal tripout happened only after an extra long, and loud session, you may be able to live without them, provided you can liven up the local air movements or clear out any local airflow obstructions (or both). Before purchase, you should also listen to the two fans beating together: the ambient noise is high, though it needn't be bothersome in an enclosed rack or cupboard.
The terminations are mostly XLR and follow the sensible Rock 'n' Roll convention originally established by RSD, amongst others. This means a female input paralleled with a male output for daisy chain to the next amp, then a pair of paralleled output XLRs. For bridge-mode operation, we need to get at the two hot (+) outputs alone, which calls for binding posts (4mm sockets). These are marked with a red lightning - zap (like railway carriages), just to remind us that the RSD outputs over 100v AC (ouch!) in bridge mode when there's no load connected. On the other hand, the panel which is otherwise well labelled omits to tell us which of the two red terminals is hot in the bridge-mode. A + sign next to the left-hand channels red terminal would remedy this. Knowledge of absolute polarity is essential whenever two or more amps drive bass speakers in tandem viz. for PA, yet for practical monitoring, it's unimportant, save your monitor chain's absolute polarity is already a known and constant factor. Another moan: the position of the 'Bridged-Stereo' select switch is unclear, not by bad legend, but because the actuator is black, against a black panel. A white dot would make unintentional settings easier to spot. Just to recap, the stereo/mono mode switch on the front controls whether you drive each side of the amp with one signal applied to either input, or two signals (not necessary stereo), whereas the mode switch at the back, when set to bridged forces the amplifier into the mono configuration, driven from channel 1 input. However, unlike the majority of amplifiers, RSD's bridging switch doesn't isolate channel 2, so potential for abuse exists, and you'll need to wind down channel 2's attenuator to the stop. Otherwise the bridged output derived from channel 1 may be curtailed or freaked out. But this is good practice anyhow. One other facet of the bridge: like most, it's not designed for loads below 8 ohms - the manual makes this clear, so when we tried a 4 ohm load on a full power test tone, the relays promptly tripped out thereby proving the streetwise theory: RSD do know how to make gear that's foolproof.
Now back to the XLRs. These are riveted in place, which is a royal pain should emergency on-the-road surgery be called for, and is also unsightly. More annoying, the inputs are wired with pin 3 hot, according to the obsolete USA convention. I suspect RSD do this, again, to make their product especially abuse-proof. That's because the old US standard specifies pin 2 hot for speaker outlets, so with RSD'S unbalanced input socket (with input pin 2 being unused), it's impossible to plug the amp's own or any other speaker output up the input. If you're wired for pin 2 hot there are two options: you could cut and cross-link the PCB tracks, or if less keen, simply rewire your input cable. In the meanwhile, RSD should be asking themselves whether pin 3 hot serves any purpose with US manufacturers now changing over to the International standard.
The MOSFET 1000 is generously fused: there's a pair of DC supply rail fuses for each channel, and as is common with MOSFET amplifiers, removal or turnout of one of these won't shoot 50 or more volts of DC up your monitors. What's more, dead fuses can be identified by the soft glow of the red monitor LED beside each holder, so ending the boring business of unscrewing each fuse in turn and metering it. The holders are bayonet, for quick-release. Another nice touch is the inclusion of fuse values, tabulated on the panel, for high and low voltage mains.
The euroconnector socket has a combined mains fuseholder and this is a different kettle of fish altogether. The fuse carrier would be near impossible to remove in a confined rack space, and even in bright light and spacious surroundings, you'll have to look carefully and unearth some tools to change the fuse. 'But surely,' you're reasoning 'if the mains fuse goes, the amp already has a serious problem?' Wrong. All fuses age, so the need for periodic replacement exists alongside genuine faults. Last at the back, there's a highly accessible mains-voltage selector - you can adjust it with a coin. Oddly, the steps are alternately European and North American, viz. 130v is followed by 220v, then 120v and 240v. This could be nasty if you're anticipating the more logical voltage sequence. In fact, switching off the amp at the mains before making any voltage adjustments is essential.
Taking a look inside, access for servicing to the MOSFET 1000's constituent cards and components, is as you'd expect when heavy-duty circuitry is packed into relatively small space, without the benefit of modular assembly. Sure, all the parts are visible which is good; however, you would need to spend some time getting at them.
The parts most likely to expire in the short run are the output protection relays, especially if the amp suffers repeated abuse, such as thermal shutdowns when driving low impedances at high level. Under these conditions, heavy arcing will soon zap the relays' contacts. For routine work, it follows that the lifespan of the relays can be extended by always muting the MOSFET 1000's input before switching off. In other words, don't switch off (or on) with any sort of level going up the input. The same applies to any amp with output protection relays.
The toroid on our sample amplifier came wrapped in mu-metal shield, held in place by gaffer tape. This is fine in theory, because magnetic shields made of mu-metal don't take kindly to soldering or any sort of metalwork heat and mechanical shock destroy the magnetic properties of the material. Alas, the gaffer had come loose, but hoping that RSD can contrive a more permanent fixing technique, the mu-metal shield (rarely seen) is a good feature for studio use.
The toroidal transformer produced a fairly loud acoustic buzz. In practice, this would be a pain if working close up, but not with the amp racked up across the room. Also, the hum level will depend on the individual transformer, and also on the harmonics on your mains supply. Ergo, the buzz was the noisiest in the evening, with TV sets polluting the mains.
Electrical hum, over the speakers, was a little below par, at -80dB.
For big monitors, with a sensitivity of 95dB at 1 watt, this means the hum will be heard at around (108 - 80) = 28dB SPL, making it quite audible in a quiet room. The hum lessened as the gain controls were backed off: For example, at the -10dB setting, the acoustic level in the above example would fall to 13dB SPL, which is acceptable. You would then have to drive the inputs at around +10dBu for full output.
Listening tests were performed at The Studio Playground, by Duncan Bridgeman (Producer of I-Level). We began by fitting the RSD in place of the Quad 405, on the bass end of their tri-amped Andy Munro monitors, then up through the mid and top ranges. Overall, the MOSFET 1000 was punchier on bass, and all round clearer, and more transparent, when operated at levels comparable to the Quad. The ability to monitor at higher levels than hitherto was also appreciated. Duncan (no relation!) found the fans a bit troublesome' after long periods, but then the amp was sitting on the floor, and for much of the time, the fans could have been switched off. Trouble is, real people don't have much time to remember these little things, especially when the switch is hidden away on the rear panel. But when fitted up in the cupboard under the monitors, Duncan didn't anticipate any problems with the fans blowing full time.
Though designed principally for PA, the MOSFET 1000 deserves a place as a dual studio/road amp. However, some refinements wouldn't go amiss. In particular, more attention should be paid to curtailing acoustic hum and electrical hums, and fan noise. Also, in an age of electronic bass sounds, input balancing transformers are becoming increasingly unfashionable (good bass end on transformers comes at a high price), and a transformerless balanced input need be no more expensive.
The price per watt of the MOSFET 1000 compares favourably with many other professional amps, at around £1.17/watt into 8 ohm loads, but you should weigh up carefully the comparative facilities and aesthetics. As usual, the lower power spin-off, the MOSFET 500, is less cost effective, at £1.70/watt.
MOSFET 1000 - £699.00 ex-VAT £803.85 VAT inclusive
MOSFET 500 - £545.00 ex-VAT £626.75 VAT inclusive
Review by Ben Duncan
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!