Did you like the PMG pre better than the Krell?
They are both excellent preamps.
I plan to listen to the Krell Illusion again with the M400’s to make a final decision.
Just wanted to clarify that aren’t class-D modules. This is a class A/B amp with high class A bias. Under the covers on the M800 picture are dual 1.2 kw power supplies.
Exactly, Class D is not the same as, say, an SMPS (switch mode power supply).
Thanks Chris, that’s what I understood. Twice 1.2kW is a lot and, if driven continuously, would explain a large heat dissipation. Still, there’s a reserve considered to get to 800 W/4 ohm. And those watts will mostly be peak. That’s why I still don’t quite get the reason behind the spacious design, comparing the interior layout of other amps. 
The required size of the heatsink is driving the size of the chassis. The amount quiescent (idle) power dissipation in a class A/B amp is defined by:
Pquiescent ≈ 2Vrail x Ibias
So, it is a function of both the rail voltage (setting the maximum voltage output of the amp before clipping) and the amount of bias. For a given heatsink size, you can have more class-A bias (helping sound quality by extending the the most ultra-linear range of amp) if the maximum output power is lower. The effect of voltage rails is pretty large. Going from the S200 to M400 (same amount of heatsink and total watts but lower voltage rails for the stereo), we go from 10 watts to 25 watts of Class-A.
The main reason for the size of the M800 is this heatsink area, not how much “stuff” going into it.
Switching power supplies are more space efficient because they work at a much higher frequency (not 60 Hz from the wall) and chop AC at much higher frequency so the magnetics (transformer) is much smaller. Additionally they recharge at hundreds of thousands of times per second, so the bulk capacitance to hold the voltage steady is much lower. It’s not very useful to compare a linear supply (large 50/60 hz transformer, bridge rectifier and caps) to a switching supply. More “stuff” doesn’t necessarily mean more expensive or better.
Out of interest (or not) I’ve two types of old Crown CE series amps for bass, one is a CE2000 and the others are CE4000. They also made a CE1000.
The CE1000 weighs 32.6 lbs and the CE2000 40.3 lbs (these feel heavy), both have conventional power supplies. The CE 4000 weighs 33.3 lbs but uses an internal switch mode power supply. Into 2 ohms each of the two channels produces 560w, 975w or in the case of the CE4000 with a 230v supply 1800w.
I once played through a borrowed Crown for a gig (between 1965 -68) but don’t recall the model. Tons of clean power. I eventually went with an SVT (tube) stack but lost track of the head; still have the SVT cabs (8 10’s in each) now powered by two SVT heads, one tube (bridge pu) the other solid state (neck pu). Plenty of power with these. The solid state head is about 1100 w into 4 ohms, the tube head I think is around 300 w. No longer gigging, at my age moving this setup is life threatening.
These Crown amps were all bought used. The last one had been used by a drummer (I think for the bass drum). These amps are very reliable and there other versions that are often used in clubs to produce powerful bass for hour after hour. They’re bridgeable, have sensitivity selection, very flexible input and output connectors etc. I use BHK250s for my mid and treble.
What preamp controls the amps, and speakers?
Thanks so much for clarifying. Now it makes perfect sense.
Hello, I am not sure if this question has already been asked/answered. Please accept my apologies if it has and point me to the right place.
Okay, PS Audio, please explain why there is a 10W Class A bias per channel on the PMG Sig. S200 and only 5W Class A bias per channel on the S400?
Not certain but I believe it was answered by Paul in this thread.
Well, I previous posted this - so ,for the same heatstink size, let’s say that the S200 has 30V rails and the 400 has 42V rails (they may be very slightly different from this but you get the idea).
The idle power (heating up the heatsinks) is 2 times this voltage and what’s left is for the Class-A bias. So, if you have a higher power amp, there is less thermal headroom on the heatsinks for class-A bias, all other things being equal.
Some lower sensitivity speaker (like Maggies, for instance) may benefit more from the higher rail voltage of the S400 but others with higher sensitivity may be better off with the S200 which has a bit more Class-A range for the most linearity and sweetness at more moderate power.
Thank you, Chris.
So where do FR20 fall in this scenario?
Listening preferences matter if you need the extra 3 db of output. I’d probably do an s200
Hi Paul…
I’m familiar with the trade-in discounts you offer directly on your product pages, but what’s this TMR option? I’m over in Montrose (the best slope 
).
When they (PS Audio) have trade-ins for new equipment, your old gear would go to The Music Room (tmraudio.com) rather than to PS Audio.
Ah, thanks.