Initially, I considered placing the Classe CP-65 preamplifer review under the "Projects" section, but after upgrading that unit, I though it deserved its own page, rather than a quick five bullet point summary.  

 

Here it is:

​

At a quick glance, the Classe CP-65 is very impressive, with the following features:

·         An external, over-sized power supply with a generous power transformer.

·         A double voltage regulation topology, and even triple in some areas. (one regulation stage in external PSU and one or two inside the main chassis)

·         Real dual mono power supply and circuit arrangement.

·         A separate power supply for digital duty (remote control, display, input relays switching).

·         It boasts more total capacitance than many power amplifiers. This thing is quiet.

·         Dual XLR and RCA inputs and outputs, so it’s very versatile.

·         An elaborate (but obsolete) discrete circuit driving an XLR output on a separate PCB board. This output driver circuit has plenty of power to drive headphones. Nice!

·         DC coupling. (no capacitor in the signal path, rare for this era)

·         A nice aesthetic, industrial and timeless appearance.

·         Silent relay switching of input and muting relay circuit. (nothing to color the sound)

·         Very high quality connectors we don’t typically see anymore.

 

Having a fully balanced topology to reduce noise and distortion, it tends to deliver very good performance versus a single ended circuit (RCA). Classe is aware that many users still want to have RCA inputs and outputs, and as a result, have opted to add this feature in order to maximize market value.

 

 

Evidently, the RCA input is not required due to internal fully balanced topology of the XLR connectors. Right off the bat, the RCA input is internally changed to balanced signal thru an Op-Amp stage. Not any opamp mind you, but a very inexpensive and lame sounding TL072!!! I can’t even make this crap up. A $10K preamp using a 30-year-old, $0.42 opamp for processing all single-ended (RCA) audio input signals. No wonder the CP-65 sounds better when fed from the XLR inputs, avoiding passing through the lame TL072. On a positive side, Classe mounted the TL072 opamp on a socket so the end user could upgrade it.  I have to ask. How many buyers of statement products actually open them up to tweak? Most obviously don’t, and the Classe CP-65 bottleneck… in RCA mode… is the TL72, which dictates the sound quality of the pre-amp. Unacceptable, if you ask me.

  

The power supply is massive for a pre-amp. However, if you pay attention to the inner detail… not so much anymore. Everybody understands companies rely on bulk orders to increase profits. This is business. Classe applied that same mind set with the CP-65’s external power supply. They used 4700uf @ 63volt capacitors for a circuit working at less than 28volts. Inside the pre-amp chassis, voltage dropped to less than 18volts, and after the internal regulation, it dropped to 15volts or 5volts in some areas. Yet, those same 4700uF @ 63volts capacitors are used all around. I reviewed a CAM200 power amplifier for the same owner, and it used the same capacitors in its power supply as the CP-65 does. (Same era product) Bottom line is that when Classe designed that product line, it seems they maximized on bulk orders, and chose to order a boatload of 4700uf at 63volt capacitors. This mandated them to use these capacitors, regardless of what little sense it made using them in specific locations. Buyers will likely be impressed by their size and quantity. Upon recapping the pre-amp, I changed those 4700uf at 63volts to 10,000uF at 35volts, which is a 112% improvement for the same physical size. Inside the main chassis, I could have used either 20volt or 25volt capacitors of bigger value, but I ordered all 35volts capacitors to make my client benefit from bulk ordering as well.

  

Volume control was often a physical potentiometer in older products, which is basically a friction device. Noble and Alp were popular brand due to their tighter tolerances and good perceived sound among audiophiles. Some products even used motor driven potentiometers for allowing remote control operation. Classe used a chip to do the job. The chip was the Cirrus Logic CS3310. Back then it made sense, and these chips were popular for in CP-65 era. The only issue is that this chip isn’t spectacular, providing at best 0.001% distortion @ 116 dB’s worth of dynamic range. Good… but not great. The best solution would have been a pure R2R DAC to control the volume (Krell does that in some products). However, that option is more costly to implement and Classe didn’t choose to follow that route. Perhaps in their newer products they will. I sure hope so. The issue with those CS3310 chips is not unlike a R2R DAC (considered like a passive device for the signal path). These CS3310 have built in 31.5 dB worth of gain and are somewhat of a complete preamplifier in a chip. The sonic "gain stage" inside the chip is unsure and they absolutely color the sound. The CP-65 preamp takes the CS3310 sonic signature resulting in the end user having to live with it. Based on positive review, it seems like the CS3310 isn’t too bad.

 

Under the hood of the main chassis, the user sees a massive discrete output stage board on each side, and a massive number of capacitors in the middle. The elevated output stage PCB hides the sad truth under them. Once you remove the output PCB board, you clearly see the budget volume control chip (CS3310) and the buffer opamp around it. Before the signal is sent to the volume control chip, it passes through an OPA2134 buffer opamp. After the signal is processed by the volume control chip, it passes through another OPA2134 buffer… yet again. The output of the volume control chip already has an internal output opamp buffer built-in, and it’s able to drive 600 ohms. Why on earth would Classe use an external buffer fed by the CS3310 internal buffer? One too many for sure. Listening tests prove that by removing the OA2134 the output buffer was subjectively beneficial.

Do not forget to short the proper pins in the opamp socket or you'll have no sound :)

 

Before the signal reaches the discrete output stage, it already passed through three opamps (two OPA2134 buffers and the internal buffer of the Cirrus Logic volume control chip). If the user used the RCA input, the signal passed through the extra RCA/XLR conversion stage, done by the lamentable TL072. This worse case scenario involves four solid state buffer opamp stages before reaching the discrete main Class AB pre-amplifier circuit. Not exactly a great design in today’s world.

 

The main output stage circuit is decent, but it’s a miniature Class AB amplifier. The input stage is cascode JFet, with current source, transistor cascode VAS and mosfet push-pull output device. And while the output bias is set at a low 12mA, this is enough to ensure pure Class A operation to most, if not all power amplifiers. TL431 are a top-notch voltage reference and ensure tight tolerance and low noise, and they provide regulation inside the CP-65.

 

The main flaw is that the preamp circuit has no built-in DC servo. And while this makes for transparent sounding product, it will amplify any DC signal fed through it and possibly send several mV’s worth of DC offset downstream to the power amplifier. This is a grave potential flaw that could create serious “domino effect” damage, should a single fault be introduced upstream. The CAM200 amplifier of that same line, uses similar topology as the CP-65, and has no DC servo either. WOW! If a user owns both and uses a defective or poorly designed tuner for example, providing a few mV’s of DC offset, he could literally blow his CAM200 amp and/or the speakers.

The system works most of the time, but it’s not designed to withstand a worse case scenario.

 

I know most newer products today employ DC protection by way of capacitor coupling, DC offset or protection shutdown, and in the best-case scenario, a properly built in DC servo.  I find it rather strange that older Classe products didn’t.

​

Here are a few pictures of the CP-65 below: