Recently bought some big PA Mosfet amps off a mate of mine.

Hey everyone,

I recently bought four big PA Mosfet amps plus two 8 Ohm 400 Watt Precision Audio PA speakers off a mate of mine, two of the PA amps I bought are a couple of Perreaux 6000C dual channel professional power mosfet amps which are rated to put out at least 300 watts, the other two amps are a couple of ARX SS1200VC professional series mosfet power amps, the two Perreaux amps seem to work fine, as do the two Precision Audio PA speakers, on the other hand, the two ARX amps need some work to get them working properly, but I always like a challenge.


Anyway, I'm also planning on retro-fitting Speakons to the four amps once the two ARX amps are working, the reason why I bought the amps was so I could build myself a small PA system.


I'll post some pics of the amps and speakers later on today so you can see what they look like, should make for an interesting diy/servicing project for next year.


Here you go, some pics:


Attachment 33037


A pic of the two ARX amps..


Attachment 33038

A pic of one of the Perreaux amps.


Attachment 33039

The two 8 Ohm 400 Watt Precision Audio PA speakers stacked one on top of the other.

I'll tell you something, my golly are those Perreaux amps heavy things to lug-around, I was just able to lift one up by one of it's carrying handles, I'm about 60kg in weight so I'm estimating that the Perreaux must weigh something like 50kg or so, the bulk of the weight is probably the dirty great big power transformer lurking inside.


All I need now is a good mixing desk that has stereo line-level outs, plus a stereo EQ, and some decent mics, and foldbacks, and I've got a full PA system.


One thing I will have to do is figure-out how to stop two furry felines from using my Precision Audio PA speakers as scratching-posts, without hurting or getting rid of said furry felines.

Ahhh... the good old days....

Back in the '80's the big PA's of the day would have racks loaded with four of those sized amps.... I know the River Stage at Expo '88 had five racks of four per side(L+R) of Jands 920's. Each 920 weighed in at a little over 35kg..... And why would you have 20 x 450W RMS per channel amps per side you ask, well it was an active crossover 4 way system where they fed 16 three way Concord cabs a side slung in the air plus 4 Concord sub cabs on the ground each side, all plugged together using XLR's. I'll let you do the math on the total output of the system... Suffice to say that on the 5th night of Expo on Brisbane's South Bank they had Jimi Barnes playing, and they got noise complaints from Indooropilly the next day...lol

A side note for you Doc, they chose the 920's due to the way they go into clipping. Much softer clipping than the equivalent Perreaux or Yamaha's of the day. I had a bench full of all of those top high power amps and drove all to hard clipping with my Cro and N&D set attached, and a team of experts watching and making the final decisions on what amp would be chosen for the Expo contract..... I walked away with the thought in my head that a small sacrifice in THD at lower levels meant softer clipping at higher levels and hence less chance to kill speakers at full volume. And Hifi THD more often than not means hard and nasty speaker killing clipping..... Solid state and Tube guitar amps fit this thinking perfectly.....

Put your Perreaux into a dummy load and attach your Cro and see how it clips at high output.... A re-wired spiral resistance wire porcelain jug element in a cup of water makes an ideal 8 ohm 750W continuous dummy load. It will even quietly sing for you with over 200W into it and without a magnet at 1kHz, and make hot water for coffee at the same time ....lol

Quote:

---

Originally Posted by Marcel

Ahhh... the good old days....

Back in the '80's the big PA's of the day would have racks loaded with four of those sized amps.... I know the River Stage at Expo '88 had five racks of four per side(L+R) of Jands 920's. Each 920 weighed in at a little over 35kg..... And why would you have 20 x 450W RMS per channel amps per side you ask, well it was an active crossover 4 way system where they fed 16 three way Concord cabs a side slung in the air plus 4 Concord sub cabs on the ground each side, all plugged together using XLR's. I'll let you do the math on the total output of the system... Suffice to say that on the 5th night of Expo on Brisbane's South Bank they had Jimi Barnes playing, and they got noise complaints from Indooropilly the next day...lol

A side note for you Doc, they chose the 920's due to the way they go into clipping. Much softer clipping than the equivalent Perreaux or Yamaha's of the day. I had a bench full of all of those top high power amps and drove all to hard clipping with my Cro and N&D set attached, and a team of experts watching and making the final decisions on what amp would be chosen for the Expo contract..... I walked away with the thought in my head that a small sacrifice in THD at lower levels meant softer clipping at higher levels and hence less chance to kill speakers at full volume. And Hifi THD more often than not means hard and nasty speaker killing clipping..... Solid state and Tube guitar amps fit this thinking perfectly.....

Put your Perreaux into a dummy load and attach your Cro and see how it clips at high output.... A re-wired spiral resistance wire porcelain jug element in a cup of water makes an ideal 8 ohm 750W continuous dummy load. It will even quietly sing for you with over 200W into it and without a magnet at 1kHz, and make hot water for coffee at the same time ....lol

---

I might actually do that one day, would be an interesting exercise, apparently Mosfets are supposed to behave in a similar way to tubes when driven into clipping, a mate of mine borrowed one of my Perreaux amps for a show in Humpty doo that he was sound-teching for, and he said that the Perreaux had a nice midrange to it.

If I wanted to, I could probably hook a Perreaux 6000C up to two Marshall quad boxes and then plug say a Marshall JMP-1 preamp into the line-ins of the Perreaux, and then I'd have a pretty good stereo 300 Watt guitar rig, makes my ears ring thinking about that.

Sounds like that Jimmy Barnes gig must have been a good one, must have been one of his post Cold Chisel bands.


Here's a bit of trivia for those who don't know what a Mosfet is:


Okay, the word "Mosfet" is short for Metal-Oxide Semiconductor Field Effect Transistor, think of it as a solid-state equivalent to a Tube.


Oh and thanks for reminding me about building a dummy load Marcel, cheers mate.


More trivia:


Pretty much any amp will go into clipping if you drive it hard enough, clipping is a form of distortion, when an amp is driven to the point where it starts clipping it can no longer reproduce the input signal faithfully because the output is limited by the power supply and the signal top and bottom is chopped-off, this form of clipping is known as hard-clipping and sounds very harsh to the human ear, the sound will have a "shattered-glass" quality to it, when a Tube amp is overdriven to the point where it starts to clip, it generates soft-clipping and the effect is less harsh or noticeable because the resulting waveform has more rounded corners to it, rather than the sharp corners generated by hard-clipping, the rounding of the corners sounds smoother.

Here's a pic of the insides of one of the ARX SS1200VC Power Mosfet amps:

Attachment 33044


If you look at the lower left-hand corner of the ARX amp, you'll notice a small circuit board, this board has the two amp volume control pots soldered directly to it, I am very strongly suspecting that a broken circuit-board track on the underside of the board is responsible for the amp not working, it powers up but no sound comes out no matter where you set the volume controls.

And yes, that huge hulking great piece of iron in the middle is the power transformer....no guesses as to why the amp is so heavy.


And here's a pic of the insides of one of the Perreaux 6000C power Mosfet amps, spot the huge power transformer:

Attachment 33045

I haven’t stumbled across Perreauxs for at least 15 years. They were practically indestructible, unless you are doing an outdoor gig on the Central Coast over the lake from one of the power stations........

Its a funny thing, there are more strange myths about clipping than just about any other subject in music technology... you can be sure if you hear someone mention both square waves and DC in the same or adjacent sentences then they have heard the myths and have been misinformed by them.
What really happens when a signal is clipped is that it artificially generates a lot of extra high frequency content, hence the harsh sound. Its exactly the same thing that a distortion pedal does, except that a typical distortion pedal has a tone circuits/filters in it as well to fine tune the end result.

Quote:

---

Originally Posted by JimC

Its a funny thing, there are more strange myths about clipping than just about any other subject in music technology... you can be sure if you hear someone mention both square waves and DC in the same or adjacent sentences then they have heard the myths and have been misinformed by them.
What really happens when a signal is clipped is that it artificially generates a lot of extra high frequency content, hence the harsh sound. Its exactly the same thing that a distortion pedal does, except that a typical distortion pedal has a tone circuits/filters in it as well to fine tune the end result.

---

That is exactly what the "team of experts" were looking for when selecting the high powered amps for the Expo 88 contract. Most of the team had many years of experience in sound reinforcement at entertainment centers and very large venues around the country, and while there may have been some company allegiances in play there was consensus among the group as what was good and bad.

Interestingly by design the Jands 920 was flat to 12kHz and rolled off by 3dB at 20kHz, and 30dB at 50kHz.... Most of the others were flat to near 100kHz. It was truly weird putting 400W of super sonic audio (50kHz) into my jug element dummy loads in water and watching them boil in only a few minutes.

Also when back at 10kHz, 1kHz and 60Hz looking at the amplifiers output waveform when being overdriven by 6dB or 10dB and how the various amplifiers clipped. In some cases definite ringing could be clearly seen after the start of each sine wave cycle as it went into clipping indicating power supply inadequacies or basic amplifier design flaws. Some amps even had sharp kW sized spikes at the commencement of the flat top portion of the wave which came as a real surprise to many witnessing the tests. The Jands 920 was a simple all transistor amp that went into clipping much like any tube amp does with a gentle transition from sine wave to flat top with nil ringing.

For me it was just fun getting to blast the buggery and try and cause 'meltdowns' out of what was then the top range PA gear of the day, and still be able to walk away with my hearing intact. We boiled a lot of water.!!!

Pretty much all of these high-powered amps used negative feedback to reduce THD (Total Harmonic Distortion) to very low levels, and pretty much any amp has negative feedback included as standard, if the feedback was made positive instead of negative you would end up with a very powerful oscillator instead of an amplifier, hence the reason why it's very important to get the primary windings of the output transformer of a tube amp connected the correct way around for it to operate properly.

I've seen some power amp schematics where there's a network called a Zobel Network tacked on to the output where the speaker is connected, the Zobel Network acts as a snubber to stop high-frequency oscillations from occurring when the amp is driven into clipping.


I think my local Bunnings shop might have some of those electric kettle elements in stock, so next year I'll pay them a visit and see if I can buy enough to make at least two dummy loads.


For solid-state amps, driving an amp to the point where it is driven into clipping means a risk of burning-out the voice coil of a speaker, since when the amp is clipping on the top or bottom of the signal waveform, the power transistors will be in a state where they are fully conducting, which means that pure DC is being applied to the voice coil, usually at several amps if it's a high-power amplifier, it doesn't take long for the voice coil to burn-out and become open-circuited.

Quote:

---

Originally Posted by Marcel

Ahhh... the good old days....

Back in the '80's the big PA's of the day would have racks loaded with four of those sized amps.... I know the River Stage at Expo '88 had five racks of four per side(L+R) of Jands 920's. Each 920 weighed in at a little over 35kg..... And why would you have 20 x 450W RMS per channel amps per side you ask, well it was an active crossover 4 way system where they fed 16 three way Concord cabs a side slung in the air plus 4 Concord sub cabs on the ground each side, all plugged together using XLR's. I'll let you do the math on the total output of the system... Suffice to say that on the 5th night of Expo on Brisbane's South Bank they had Jimi Barnes playing, and they got noise complaints from Indooropilly the next day...lol

A side note for you Doc, they chose the 920's due to the way they go into clipping. Much softer clipping than the equivalent Perreaux or Yamaha's of the day. I had a bench full of all of those top high power amps and drove all to hard clipping with my Cro and N&D set attached, and a team of experts watching and making the final decisions on what amp would be chosen for the Expo contract..... I walked away with the thought in my head that a small sacrifice in THD at lower levels meant softer clipping at higher levels and hence less chance to kill speakers at full volume. And Hifi THD more often than not means hard and nasty speaker killing clipping..... Solid state and Tube guitar amps fit this thinking perfectly.....

Put your Perreaux into a dummy load and attach your Cro and see how it clips at high output.... A re-wired spiral resistance wire porcelain jug element in a cup of water makes an ideal 8 ohm 750W continuous dummy load. It will even quietly sing for you with over 200W into it and without a magnet at 1kHz, and make hot water for coffee at the same time ....lol

---

Hey Marcel, would you know the specific THD of the Jands 920 and it’s specs? I’ve got two of these and was wanting to use them in my home cinema setup. My speakers are ar9’s and rated at 4ohms and need heaps of power! Thanks heaps Marcel!

Cheers
Hoa

Quote:

---

Originally Posted by DrNomis_44

pure DC is being applied to the voice coil, usually at several amps if it's a high-power amplifier, it doesn't take long for the voice coil to burn-out and become open-circuited.

---

I think you'll find the tales about DC are pretty much a myth. This is an AC signal like any other. It's not DC at all, even if the signal, by the time it got through the wiring, any crossover and anything else, were still a pure square wave. What it is is a very high energy signal, and any sufficiently high energy signal will burn out the coil. Most likely what causes the confusion is that a nominal 1KW amp will deliver 1KW as an undistorted sine wave, but getting on for double that if you actually drove it into the unlistenable level of distortion a pure square wave would imply. Real DC does have an extra risk, since the voice coil isn't moving, which deprives it of cooling, but an AC square wave has the cone moving just as much as any other signal, and provides the same air movement and cooling.

For all the myths around square waves, we should remember they're one of the main building blocks of additive synthesis, and there's no risk at all in reproducing them provided that the drivers aren't overloaded.

Oh wow, calling on detailed memory from 32y ago....

The Jands 920 was a brute at 450W RMS per channel into 8 ohms with 0.1% THD. It will do 4 ohms but not that well. Due analogue (non-switchmode) power supply constraints of the 920 and transistor based output finals design the best I measured back then at 4 ohms was 550W with 2% THD.

Modern switchmode amps with MOSFET output stages will do 4 or 2 ohms with relative ease with more or less doubling of output power with each halving of load impedance and virtually nil change in THD. The biggest difference and swaying factor for the committee in the selection of the 920 for Expo 88 was the soft clipping in the 920 at max output as opposed to the very square hard clipping of the Yamaha, Crown and Perreaux competition despite their better THD figures at typically better than 0.01%

All the PA's that used 920's back then in the late '80's were configured with 8 ohm or 16 ohm combined speaker loads. It was rare for any 920 system to use 4 ohm combined loads. And a fast way to blow 920 amps was to mistakenly wire up 2 ohm loads as the amp truly doesn't like feeding 2 ohms or less. A daily check of each drivers functionality at Expo 88 was done by pulling the XLR connectors on the patch panel behind each amp. Incorrect re-insertion of the XLR could cause a momentary short across the amp outputs instigating instant failure of the amp final output stage. I repaired dozens of failed channels across the duration of the 6 month Expo 88. The indication of a channel failed in this way is the "DC" alarm lamp being lit.

Through a neat set of circumstances I got a 920 cheap. Had it in my home stereo for a number of years beside my CD player and dual 15 band EQ. Burnt out quite a few 12" drivers at various parties until my partner convinced me to sell it and buy a new TV.

Hope this helps...

Quote:

---

Originally Posted by JimC

I think you'll find the tales about DC are pretty much a myth. This is an AC signal like any other. It's not DC at all, even if the signal, by the time it got through the wiring, any crossover and anything else, were still a pure square wave. What it is is a very high energy signal, and any sufficiently high energy signal will burn out the coil. Most likely what causes the confusion is that a nominal 1KW amp will deliver 1KW as an undistorted sine wave, but getting on for double that if you actually drove it into the unlistenable level of distortion a pure square wave would imply. Real DC does have an extra risk, since the voice coil isn't moving, which deprives it of cooling, but an AC square wave has the cone moving just as much as any other signal, and provides the same air movement and cooling.

For all the myths around square waves, we should remember they're one of the main building blocks of additive synthesis, and there's no risk at all in reproducing them provided that the drivers aren't overloaded.

---

Some qualifications needed there, Jim. A square wave is a ‘high energy signal’ in a relative sense to say a sine wave or a triangle wave of the same amplitude, but a 1mW square wave is far less powerful and has far less energy than a 10W sine wave. A constant square wave contains the most % energy you can transmit for a given signal amplitude.

At very low frequencies, a square wave will approximate to DC being passed through a speaker coil e.g. at 1Hz, you’ll get alternating patterns of 1 second of positive DC signal and then 1 second of negative DC signal. The speaker coil will move to a position, stay in that position with the DC current passing through it, and then move to the next position and stay in that position with the DC current passing through it.

The thing that’s different between a 1Hz square wave and a 100Hz square wave is the amount of movement of the speaker and its coil between those two positions. A speaker relies on movement of the cone and voice coil to provide most of its heat dissipation through the air. If it’s not moving much, then it can’t dissipate much heat and a 100W rated speaker may fail at 3 or 4 watts once enough heat builds up. It’s a similar thing with very high frequency signals – although the speaker is in theory moving very quickly and so should get well ventilated, a point beyond which the mass and inertia of a speaker simply can’t respond fast enough to the signal to move much at all. So, the cone/coil hardly moves whilst a large amount of energy can be passing through it.

The other thing DC does to a speaker is provide an offset to the normal rest position of the voice coil within the magnet, either moving it forwards or backwards in proportion to the level of the DC signal. Let’s say a voice coil normally has a maximum travel of ±4mm before the voice coil starts to move outside the magnet’s main field and so starts to lose control of its position. A DC signal may move the voice coil forwards by 2mm, so it doesn’t take as much power for it to reach that 4mm of forward travel, after which the voice coil isn’t held centrally by the magnetic files but can be pulled to one side, rubbing against the magnet and wearing away insulation or even catching on the edge of the magnet. Both situations spell a short and unhappy future life for the driver.

So mixing a very low frequency square- or near-square wave in with a normal music/guitar/bass signal isn’t a good idea, even though the speaker is still moving a lot and being cooled.

In fact, any very low frequency wave alone being passed through a speaker isn’t good for it, it’s just that a square wave is the worst sort of wave to do that with.

It’s why most amps have DC blocking capacitors in the signal path, and many bass amps have high-pass filtering to cut off sub-sonic frequencies (you are far less likely to generate any with a guitar).

But, messing about with synths is the most likely way to get very low frequency square waves in the real world, with low frequency oscillators capable of say 0.1Hz to 20Hz affecting output levels Some can be routed to audio outputs, and if not actually modulating an audible wave, the results will be too low for you to hear, so you won’t know what’s going on until the blue smoke of death starts to appear.

Just did a quick calculation and, if you have 20 X 450 Watt Power Amps all running together, that gives you a total combined power-output of 9,000 Watts.....yikes!!!!!

9kW is nothing these days. But with 2 channels per amp, that set up was probably 18kW

My small Yamaha PA system - 2 x DXR10s and a DXS12 sub was rated at around 3kW continuous and 5kW peak.

But it's the speaker efficiency that really determines the overall system loudness.

Quote:

---

Originally Posted by Simon Barden

9kW is nothing these days. But with 2 channels per amp, that set up was probably 18kW

My small Yamaha PA system - 2 x DXR10s and a DXS12 sub was rated at around 3kW continuous and 5kW peak.

But it's the speaker efficiency that really determines the overall system loudness.

---

Yep, basically how many dB per Watts SPL the speaker would put-out, taking into consideration how the load-impedance varies with frequency, speaker cab resonances, and the frequency response of the speakers, you're quite right, I forgot that each of the 20 450W amps were stereo so 450W per stereo channel.


So, a speaker that has an efficiency of say 105dB/W is going to be more efficient, and sound louder, than a similarly specced speaker with an efficiency rating of say 75dB/W, because it is more efficient at converting the electrical power (energy) output of the amp into acoustic power (energy).

Quote:

---

Originally Posted by DrNomis_44

Yep, basically how many dB per Watts SPL the speaker would put-out, taking into consideration how the load-impedance varies with frequency, speaker cab resonances, and the frequency response of the speakers, you're quite right, I forgot that each of the 20 450W amps were stereo so 450W per stereo channel.


So, a speaker that has an efficiency of say 105dB/W is going to be more efficient, and sound louder, than a similarly specced speaker with an efficiency rating of say 75dB/W, because it is more efficient at converting the electrical power (energy) output of the amp into acoustic power (energy).

---

Actually for the River Stage there were 23 Jands 920's per side.... 20 on the FOH each side (18kW + 18kW), and 3 more per side (2.7kW +2.7kW) on delayed stacks some 75m from the stage. (18 +18 + 2.7 + 2.7 = system total of 41400kW true RMS at onset of sine wave clipping)
As for speaker cabs do a search for "Concord PA cabinets" & "3way cabs" circa late '80's. There were 16 3 way cabs hung on motorized hoists in a 4x4 array per side on the stage with 2 of the 3 way cabs in each delay stack, and 4 sub cabs per side of stage for below 80Hz. ... I think the crossover points for the stereo 4 way system were 80Hz, 750Hz and 2.6kHz or somewhere near those frequencies.

As a side note....just days before the 6 month "show" started they had everything working including the 512 channel DMX controlled lights array.... PA was up making musical noise, and the lighting dude raised all his lighting channels and then the main dimmer on his console....
Next minit.... BZZZZ!!... the ground briefly shook at 50Hz, and all the lights suddenly went into uncontrollable auto disco mode with the PA making weird buzzing noises....within seconds everyone was scrambling to turn things off....
It tool a team of electricians over an hour to diagnose that there was nil Neutral connection from the river stage dedicated 250kVA supply transformer.... They were all baffled as to how a new power transformer of that size could literally burn out its Neutral link.... I laughed, and reminded them all how an SCR dimmer works.... and that there was over 200kW of incandescent lighting on that stage. 3 phase all on no problem, 3 phase all off no problem, 3 phase at 66% SCR dimming and damn huge Neutral problem. (and yeah there is another story on how I knew these things...lol)
A 750kVA transformer was fitted next day.... the next 6 months went just fine....

Quote:

---

Originally Posted by Marcel

Actually for the River Stage there were 23 Jands 920's per side.... 20 on the FOH each side (18kW + 18kW), and 3 more per side (2.7kW +2.7kW) on delayed stacks some 75m from the stage. (18 +18 + 2.7 + 2.7 = system total of 41400kW true RMS at onset of sine wave clipping)
As for speaker cabs do a search for "Concord PA cabinets" & "3way cabs" circa late '80's. There were 16 3 way cabs hung on motorized hoists in a 4x4 array per side on the stage with 2 of the 3 way cabs in each delay stack, and 4 sub cabs per side of stage for below 80Hz. ... I think the crossover points for the stereo 4 way system were 80Hz, 750Hz and 2.6kHz or somewhere near those frequencies.

As a side note....just days before the 6 month "show" started they had everything working including the 512 channel DMX controlled lights array.... PA was up making musical noise, and the lighting dude raised all his lighting channels and then the main dimmer on his console....
Next minit.... BZZZZ!!... the ground briefly shook at 50Hz, and all the lights suddenly went into uncontrollable auto disco mode with the PA making weird buzzing noises....within seconds everyone was scrambling to turn things off....
It tool a team of electricians over an hour to diagnose that there was nil Neutral connection from the river stage dedicated 250kVA supply transformer.... They were all baffled as to how a new power transformer of that size could literally burn out its Neutral link.... I laughed, and reminded them all how an SCR dimmer works.... and that there was over 200kW of incandescent lighting on that stage. 3 phase all on no problem, 3 phase all off no problem, 3 phase at 66% SCR dimming and damn huge Neutral problem. (and yeah there is another story on how I knew these things...lol)
A 750kVA transformer was fitted next day.... the next 6 months went just fine....

---

Sounds like it must have been a fun time getting that all sorted-out, I can just imagine how loud that 50Hz buzz was at the time, and 50Hz must have been the natural resonant frequency of those light fittings too, or close to it anyway....lol.


I seem to remember studying SCRs while doing some electronics courses at uni more than ten years ago, when they are used to control AC power, the waveform isn't exactly sinusoidal all the time, it's a spiky looking waveform, and I seem to remember that snubber-circuits are sometimes added to reduce high-frequency harmonics, or, I could be totally wrong, it has been quite a while.

Still, great story you have there mate.


Note for everyone else, SCR is short for Silicon Controlled Rectifier, it's a special type of power diode used to control AC power, most often used in light dimmers, although nowadays you do see some stage lights that use Leds (Light Emitting Diodes) rather than incandescent lights, I think Led stage lights are powered by DC if I'm not mistaken.


I'm really amazed at how far P.A. sound technology has evolved since it's early days, who knows what the future has in store for us.

A bit of searching and I was reminded that the speaker cabs at the entertainment venues at Expo 88 were predominately Concord TMS-3...

For a interesting read of Aussie PA and lighting history have a read of this.... it brought back plenty of memories for me of gear and people ....
http://www.australianroadcrew.com.au...production.pdf

Interesting stuff, Marcel.

That's a great read Marcel, a few names and fond memories in there