Clean and inexpensive- Cambridge Audio 651W Power Amplifier

[JonMarsh]

OnTheBench.key by Jon Hancock, on Flickr



The Cambridge Audio 651W is now my new go-to amplifier for small and portable for speaker measurements. I got spoiled by the Benchmark AHB2- the 651W isn't truly an AHB2 replacement, but it's as close as you have any right to expect for 1/3 the price! The 651W is rated for 100W per channel at 8 Ohms, 150W/ch at 4 ohms. I picked the 651W to try because of the positive experience I'd had with it's big (and much heavier) brother, the 851W, and because I was curious if it would really match the quoted specifications.

• Description
• Features
• Specs
• Manual
• Video
• Warranty

• Power output: 100 W per channel into 8 Ohms; 150 W per channel into 4 Ohms
• THD (unweighted): < 0.001% 1 kHz; < 0.005% 20 Hz - 20 kHz
• Frequency response: 5 Hz – 80 kHz +-1 dB
• S/N ratio (ref 1W / 8 Ohm): > 90 dB (unweighted)
• Sensitivity: 1.5V rms unbalanced; 1.5 + 1.5V rms balanced
• Input impedances: Balanced input 38 k Ohm; Unbalanced input 68 k Ohm
• Power amp damping factor: > 125 at 1 kHz
• Trigger in: 5 – 12V AC or DC
• Trigger out: 12V DC @ 100mA
• Maximum power consumption: 750 W
• Idle power consumption (no signal): 45W
• Standby power consumption: < 0.5W
• Dimensions: 16.9" wide, 4.5" high, 13.4" deep
• Weight: 23.5 lbs.

651w-si by Jon Hancock, on Flickr


651w-bl by Jon Hancock, on Flickr


651w-rear by Jon Hancock, on Flickr


Both balanced and unbalanced inputs are provided; binding posts are EU style with insulators blocking the banana jack receptacle, but they are easily removed.


Interior layout is fairly clean and well organized, with the power supply on the left, output circuitry in the center, and inputs and protection more towards the right.


651W_offen_t by Jon Hancock, on Flickr


Figure 1: Frequency response and channel Balance

FreuencyResponse+10dBV by Jon Hancock, on Flickr


No surprises here, extended flat response to beyond 50 kHz on the top end; at 80 kHz protection kicks in. (not shown).




Figure 2: THD+Noise @ 1kHz from 200mW to 100W using Balanced Input

CA651W_LevelSweepTHD+N_1kHz by Jon Hancock, on Flickr


Note, with this kind of steadily decreasing slope on the THD+N, this indicates that the major component is noise- note that the 651 drops steadily from a low 0.005% at 1W output to about 0.0007% at 50W output- this is very good performance, indicating a very quiet front end (including the balanced amp receiver) and low distortion in the entire signal path.


Figure 3: THD+Noise @ 20W from 20Hz to 20kHz using Balanced Input

CA651W_FrequecySweepTHD+N by Jon Hancock, on Flickr

Distortion stays low over a wide frequency range, rising slightly as is typical at high frequencies.


Figure 4: FFT 1Hz to 1kHz with 50W 1kHz output



1kHz_50W_LF_FFT by Jon Hancock, on Flickr


This is a good way to look at power supply noise in the output when driven by a moderate output signal; here, we can see that the maximum influence of power supply harmonics lies at -130 dB below the output signal- yes, this is a good clean number!



Shifting the measurement point of the FFT to the high frequency band, and looking at the signal harmonics,


Figure 5: FFT 1kHz to 30kHz with 20W 1kHz output

1kHz_50W_HF_FFT by Jon Hancock, on Flickr


We can see that the highest level is the 2nd harmonic at 2 kHz, which is over 110dB down in level 3rd harmonic just a bit below that, and higher harmonics all 130dB down or more. These values are getting close to the residual level of the instrument.



Going down to 50Hz for an output harmonic test, at about 50W output,

Figure 6: FFT 1Hz to 1kHz with 50W 50Hz output

50Hz_50W_LF_FFT by Jon Hancock, on Flickr


Here we see a mixture of harmonic distortion and power supply line frequency related noise, generally below -120 dB.


Last test is the combined 19kHz + 20kHz CCIF high frequency intermodulation distortion- peak level is correlating with a 50W sine wave, where the crests of the individual frequencies add up.

Figure 7: Scope 19kHz+20kHz CCIF output

CCIF_19+20kHzCA651W_Output by Jon Hancock, on Flickr


Figure 8: FFT 19kHz+20kHz CCIF output

CCIF_19+20kHzCA651W_FFT by Jon Hancock, on Flickr


The highest sidebands are down by about 105dB; the highest difference product (1 kHz) is down by about -115dB - both are very good numbers for a power amplifier.


In fact, the results on this test compare favorably with the $2,999 Benchmark AHB2...


Figure 8: Benchmark AHB2 Bridged Mono 100W peak output FFT 19kHz+20kHz CCIF output

FFT Spectrum 19+20kHz 100W Pk AHB2 by Jon Hancock, on Flickr


Now, if this isn't good enough for your $999, and you've got some change rattling around in your pocket, there's a German firm that will do mods and update a lot of parts in the power supply and signal path. Oddly enough, they don't offer on the site any before and after measurements that I could find.... :W

http://www.hoer-wege.de/cam_651w.htm

Positives:

• Very low noise, very low distortion by all test measures
• Includes balanced and unbalanced inputs
• Compact size, good ventilation and good built quality for the price point
• control options for system power up and down, programmable via Standby control
• Low cost relative to performance
• Both Silver and Black finish options

Concerns:

• Binding posts are rather light weight and wiggly- banana jack inputs for binding posts are initially covered (but cover is removable)
• Won't impress your friends who are looking for Audio Jewelry with massive tooling

[wkhanna]

impressive numbers for an affordable amp.

i remember earlier in my days in pursuit & learning of good sounding home audio, i felt 100 watts of solid state power was not considered 'manly' enough for a truly hi-fi (whatever that is?) system.

after many years of listening to as many systems & different equipment as possible, i now know that is not always the case.
one product that really brought this to life is the Odyssey Khartago.
even lower power amps can provide quite satisfying listening experiences without shattering the neighbors windows.

one of the most impressive characteristics of this and any of Klaus' amps, regardless of the rated power, is it transient response.
that is, its ability to handle a big, fast swing from low level to high level output without giving the impression there is a wheelbarrow full of bricks tied to the amp.
quick, accurate transient response helps to maintain the flow & timing of the music, most notable in a lot of orchestral works like Mahler, for example.
however, this character, when executed well, shows up in all music, even at low power levels, and provides the listener with a sense of speed, detail and reality that can not help but pull the listener in with great satisfaction and involvement.


is there any test or information in the measurements you are performing that can quantify this ability of an amp to provide fast power response relative to time?

[madmac]

impressive numbers for an affordable amp.

i remember earlier in my days in pursuit & learning of good sounding home audio, i felt 100 watts of solid state power was not considered 'manly' enough for a truly hi-fi (whatever that is?) system.

after many years of listening to as many systems & different equipment as possible, i now know that is not always the case.
one product that really brought this to life is the Odyssey Khartago.
even lower power amps can provide quite satisfying listening experiences without shattering the neighbors windows.

one of the most impressive characteristics of this and any of Klaus' amps, regardless of the rated power, is it transient response.
that is, its ability to handle a big, fast swing from low level to high level output without giving the impression there is a wheelbarrow full of bricks tied to the amp.
quick, accurate transient response helps to maintain the flow & timing of the music, most notable in a lot of orchestral works like Mahler, for example.
however, this character, when executed well, shows up in all music, even at low power levels, and provides the listener with a sense of speed, detail and reality that can not help but pull the listener in with great satisfaction and involvement.


is there any test or information in the measurements you are performing that can quantify this ability of an amp to provide fast power response relative to time?

Well said Bill !!!

[JonMarsh]

impressive numbers for an affordable amp.

i remember earlier in my days in pursuit & learning of good sounding home audio, i felt 100 watts of solid state power was not considered 'manly' enough for a truly hi-fi (whatever that is?) system.

after many years of listening to as many systems & different equipment as possible, i now know that is not always the case.
one product that really brought this to life is the Odyssey Khartago.
even lower power amps can provide quite satisfying listening experiences without shattering the neighbors windows.

one of the most impressive characteristics of this and any of Klaus' amps, regardless of the rated power, is it transient response.
that is, its ability to handle a big, fast swing from low level to high level output without giving the impression there is a wheelbarrow full of bricks tied to the amp.
quick, accurate transient response helps to maintain the flow & timing of the music, most notable in a lot of orchestral works like Mahler, for example.
however, this character, when executed well, shows up in all music, even at low power levels, and provides the listener with a sense of speed, detail and reality that can not help but pull the listener in with great satisfaction and involvement.


is there any test or information in the measurements you are performing that can quantify this ability of an amp to provide fast power response relative to time?

This is one of those times where I think the "popular" way of describing what people hear and what they think is causing it is not physically correct. My own experience regarding correlating naturalness of transient response and definition has alway come down to the energy storage in the power supply and having very stiff rails... for high and low frequencies. Short burst and sustained.

Any amp that can deliver a full power 20kHz tone burst is ipso facto "fast" in delivering power, but this is not a test signal having correlation with musical transients- they are actually much slower in most cases, piano at realistic volumes being one of the few exceptions, or at least more demanding cases. So, while the sprint may be willing (the active circuitry) the flesh may be weak (the power supply that has to deliver now, 1 msec from now, and 10 seconds from now!)

Passive power supply impedance is often very important; the more capacitance, the better, in general. This I figured out in 1977, building a modified Dynaco ST400 with 100,000uF of supply capacitance in electrolytic, and 100uF in polycarbonate film capacitors. (polypropylene caps didn't exist back then... stone age, I know... that's how old I am... Its a wonder I can find the on-off switch on my iPad Pro... )


Now, that Odyssey amp you link to is an interesting case...

$ 995 USD
2 x 110 Watts RMS @ 8 Ohms
2 Ohm load stable
Class A/AB
2 - 400,000 Hz frequency range
60,000µF memory
40 amps current delivery
<0.04% THD (not audible)
>500 continuous damping factor
Input impedance 22K ohms
DC offset <1 mV
RCA inputs
400 VA Plitron transformer
Sanken Epitaxial Planar Transistors 2SA1216 & 2SC2922
Anti vibration dual thickness PCB
Power consumption when idle ±30Watts
4 internal fuses 250V/6.3A, 5x20mm, fast blow
Additional electrical protection fuse

Now, I probably shouldn't even write any of this, but as both a former professional amplifier designer and a DIY constructor, it's hard for me not to comment- I know Odyssey has a fair reputation in the community, and has built some popular products. That said... let's look at this through the curmudgeon's eye that has read a lot of audio babble in his day...

2 x 110 Watts RMS @ 8 Ohms Ok, this I can live with... no 4 ohm power specs mean it may not be able to pass 1/3 power run in prior to 4 ohm testing, or it may protect or something else. Might be nice to know. Otherwise, we should assume this is an 8 ohm design.

2 - 400,000 Hz frequency range. (+/- what, at what output level? Small signal bandwidth of amplifier to closed loop? What would happen if you tried to put 20VRMS out at 400kHz?) (I'm reminded of a phrase they use elsewhere in their specifications - "not audible" (certainly not the 400kHz. Or the 300kHz. Or the 200kHz. You get my drift... )

60,000µF memory Is this a translation problem? I'm presuming what they mean is 60,000 uF of capacitance in the main power supply. that's a good number at this $1K price point. A good value to have, though, in the greater scheme of things, would be the optional upgrade configuration that doubles that number which you'd really like to have. (that is what is actually pictured here- the upgrade model) This is called shipping a product at a price point to meet the price point.

40 amps current delivery Huh? At what voltage, for what time duration, and what frequency? This sounds kind of like the number that is likely to be the V/I limiter set point to protect the output transistors in the case of a fault. Doesn't have anything to do with musical load current delivery. note, they don't have a 4 ohm power spec, 8 ohm RMS current delivery at rated power is about 3.75ARMS, 5.3 Peak per channel. Doesn't sound very close to 40A. They don't specify phase angle (of current to voltage) and this is a biggie for linear amplifiers.

<0.04% THD (not audible) really? In whose system, with what speakers and source? At what level? I have a saying, shared with some other engineers at work: "In God we trust; all others we verify". THD is a very vague spec anyway, and it is usually specified as THD+N, and specified with regards to level, as it varies quite a bit. One other forum member was bringing me info about a Class D amplifier series from one company which is spec'd for distortion as "Less than 0.05% at -3dB to full power". I asked him, did he understand that this meant the THD spec was for half power output? At least they gave a hint about the level. (more about THD rating down below)

RCA inputs Oh geez, 'come on now, this is the 21st century- you're asking $1K for this, and only phono inputs? If you only offer one input, make it balanced, because you can always convert that to single ended. The other way around doesn't work.

400 VA Plitron transformer Nice brand, not the optimal rating for the application. Generally, because of the power factor issues with diode-rectifier capacitor power supplies (PF between 0.5 to 0.55) and transformer regulation, you want to at least double the VA rating over the output power. So, this transformer is OK for dual 100W output at 8 ohms, but is not consistent with driving 4 ohm or lower loads well- that should be a 600-800VA transformer.

Sanken Epitaxial Planar Transistors 2SA1216 & 2SC2922 Ok, now we've got some good stuff here- these are some of the best bipolar output transistors you can get, as regards their fundamental technology approach. (some say they ARE the best, and I wouldn't argue that point- they're in the top 3, in my opinion). (this is where I do NOT mention that I'm working on a hopefully reference grade 100w/channel class A design, using those same output transistors, but geez, it's not going to fit in a chassis like the Khartago, it will look more like this:



Yeah, it looks something more like a 500W/channel design, but that's what CLASS A does to things.... BTW, in small quantities, an empty chassis like that goes for nearly $400.

I figure I needed a little balance in life with all the Class D stuff lately, so here's where I go to the the opposite extreme. With a feasible target of 100 times lower THD than the Khartago based on circuit simulation and calculation. Whether it's inaudible or not, well, we'll just have to see. (gawd, I'm sounding awfully curmudgeonly even to me, and I had a really good day at work in the office, got a lot of stuff done- maybe it went to my head?)

Anti vibration dual thickness PCB Actually, I was thinking to myself, this is built to a price point, because the power supply and amplifier circuitry are all on the same board- often it's not a bad idea to separate these, including for vibration control and isolation, PARTICULARLY isolation of the half sine distorted power supply current from the gain stage and feedback circuitry.

This may account for the mediocre THD rating- back in the late 70's, before half of you guys were born, I designed Pro amps, and there was a little 150w/channel stereo amp I did in which I discovered JUST HOW IMPORTANT keeping the supply and amplifier circuit apart was- initial IM distortion numbers on that amplifier at full out were about 0.03-0.05%, but it wasn't due to the amplifier circuitry- it was magnetic field interference with the amplifier from power supply conductors. Relaying out the PCB and redressing how power was run in the chassis dropped the IMD to where I believed it was supposed to be, based on the amplifier circuitry capability, at or below 0.002%, which was the limit of my Crown IMA analyzer. (I still have that puppy, in case I need to open a museum some day... anyone want pictures? ) now, this is 2016, not 1976, and so for a G note of my money, I'd really like to see an amplifier that works as well as that old design from 40 years ago, made with the limitations of the mediocre Motorola output transistors available then. (OK, so I did cheat on that design, I used a cascoded matched differential JFET front end that was as quiet as a really good phono preamp, because I stole it from my phono preamp design I was custom building for a few individuals back then... )

OTOH, I could be full of it, because the Cambridge Audio design uses a single main PCB, (much larger one, though) and they manage to get down to 0.005% 20Hz to 20kHz, with very good performance even at just 1W output.



Bottom line, they don't have a very good marketing guy writing their specs, or just maybe the value proposition still needs a little work. Based on the specs, I would go for the Cambridge Audio, though generally I prefer to see buying US manufactured products.

It's a tough business to be in, but it's also a question of value received for your hard earned money. Unfortunately, with the exception of Schiit, doing good quality design and manufacturing in the US seems to normally results in Ayre Acoustics level pricing. ($7950 for a VX-5 amplifier, 175W/ch at 8 ohms, 350W/ch at 4 ohms, but Ayre's new diamond output stage circuit was copied from buffer circuits of the 70's, and used in a design I did for a pro audio company (while consulting) in the 90's. New is not always new. Specially not in this business, anymore.

[5th element]

I know that Douglas Self has done some design work for Cambridge Audio before so it stands to reason that some of their product lines measure really well.

[JonMarsh]

I know that Douglas Self has done some design work for Cambridge Audio before so it stands to reason that some of their product lines measure really well.

That is quite true, though not having the whiz bang class XD feature of it's big brother, it otherwise smells of "blameless amplifier". And that's a good thing.

[wkhanna]

thank you for the insightful analysis of the Odyssey specs.

regarding what i hear, how i describe it, and what i attributed it to......

it may well be the perfect case of the blind men & the elephant...as i mistakenly associated naturalness of transient response and definition with power response time variations between different amps.

this is in fact what makes this particular forum so valuable to me.
it is increasing my understanding of what certain design aspects of audio equipment effect what & how i hear.

this may be a question others are asking too?
how or even can we, in fact, infer any particular or specific audible characteristics from the data?

[JonMarsh]

this may be a question others are asking too?
how or even can we, in fact, infer any particular or specific audible characteristics from the data?

Now THAT is a topic for substantial discussion over time, and you will find men of principle and honor and good intentions disagreeing about that.. Especially if they conduct tests with "average" listeners from the population at large (who rarely hear live acoustic music) versus trained experienced listeners, such as classical or studio musicians, regular concert attendees, and audio professionals such as studio recording engineers.

There are some greater trends which are fairly easy to identify- low order even harmonics are rarely found objectionable by most listeners, even at higher levels, but intermodulation products or high order harmonic products are another matter. Then too, there is the notion of distinguishing between "euphonic" reproduction and really transparent reproduction. A little 2nd harmonic sweetener may have been part of the deal for a lot of older recordings- built in, or assumed to be imparted by the associated gear, and hearing them truly transparently may take some listeners aback. OTOH, if you know what mic feeds sound like, the improvements in mastering of older recordings/CD transfers can be very pleasing...

This may also be where the suggested topic of a tutorial comes in, and that means some days soon posting a thread comparing and explaining measurements "A" versus measurements "B" - discussing what it means with regards to the electrical performance of the gear, and what it may mean to your ears.

I hope you didn't find me excessively harsh about the Odessey amp specs- I do the same thing for fairly high end gear when my engineering senses get "offended". IF one is knowledgeable in this business, it's just not that hard to do a better job of creating specs that inform- not ones that obscure or distract.

[wkhanna]

I hope you didn't find me excessively harsh about the Odessey amp specs- I do the same thing for fairly high end gear when my engineering senses get "offended". IF one is knowledgeable in this business, it's just not that hard to do a better job of creating specs that inform- not ones that obscure or distract.

harsh?
what was it Al Pacino said?
no worries here, we can handle the “truth” :W
in fact, is that not why we are here?
the raison d'être for all this effort is to reveal the truth.

while i do like the way Klaus’ amps sound, i certainly am not here to defend his marketing strategies.
he is a great guy to hang out with at the shows & is V knowledgeable
he is also passionately anti –establishment when it comes to the industry as a whole.
being the self-made entrepreneur & outsider he is, his business & its resources are quite frugal.
his wife is basically the entire manufacturing dept all on her own as she does nearly all the board assy.
i highly doubt he has anyone advising or managing the marketing & spec writing, as he wears nearly all the hats within the business.

it would be quite interesting to me to see one of his units put through your ‘torture test’ though.

[wkhanna]

allow me to add how much i greatly appreciated the way in which you pull back the curtain to reveal the 'wizard' with regard to product specs.

so much marketing bs in many cases that, unfortunately, many consumers rely on to make 'informed' decisions.
everyone should keep in mind that there is no industry standard for the manner in which manufacturers are allowed to publish their product specifications.

[Renron]

Bill,
in a response to an old thread, I'm glad Jon's on our side and not on "theirs".
If you can't convince them with logic...............
Ron

[JonMarsh]

yeah, too many people really do believe in that "baffle them with BS", at all stages of life and professions...

A decision I made long ago is that I love music and audio too much to be trying to make a living at it, and doing some of the things that often attend to that endeavor. Necessitated by needing to pay the bills, of course, but still...

In the current industry, the figure I admire the most about how he goes about these things is Bruno Putsys. Very pragmatic and realistic, and a true analog guy, even though the corner stone of his business is switching amplifiers (done in a very analog focussed fashion, I can assure you). I'd like to hear some of the Mola Mola gear, and bench it, but as limited as the distribution is, doubt that will ever happen.

[Alaric]

one of the most impressive characteristics of this and any of Klaus' amps, regardless of the rated power, is it transient response.
that is, its ability to handle a big, fast swing from low level to high level output without giving the impression there is a wheelbarrow full of bricks tied to the amp.
quick, accurate transient response helps to maintain the flow & timing of the music, most notable in a lot of orchestral works like Mahler, for example.
however, this character, when executed well, shows up in all music, even at low power levels, and provides the listener with a sense of speed, detail and reality that can not help but pull the listener in with great satisfaction and involvement.

This statement really got my attention. It highlights one of what I feel are my amp's strong points and its major weakness. In Class A my amp makes a tight snare drum sound like a tight snare drum, Led Zeppelin's In My Time Of Dying from the Celebration Day album comes alive with the transitions from Plant's vocals to Jason Bonham's kick drum intro in the opening bits. Very lively and fast. But the amp gets tired fast with ongoing raucousness like The Immigrant Song, or almost anything from Metallica (or that style). My speakers were chosen to compliment/compensate for these characteristics. Since a lot of what I listen to isn't Led Zeppelin and Metallica I still feel like I won the bang/buck lottery. Sheryl Crow's I Shall Believe still sounds amazing on my little rig, and looking for music that compliments my rig has expanded my musical realm greatly. Jon's explanation as to why these things are the way they are makes my audition trips a lot more educational. Good stuff here on the 'Guide.:T

[wkhanna]

this unit is still available at Music Direct for $700

LINK

[JonMarsh]

That's a damn good price... oops! Not supposed to use language like that, but since it accurately reflects the matter at hand, I think I'll let it stand.

As mine went to a 3rd party as part of an overall system transaction, I just may get another one, as I trust it a lot more for speaker driver testing than I do the Beheringer A500 I've been using the last few times...

[JonMarsh]

Ordered.

[wkhanna]

i am getting V good at spending your $ lately! :B

also, is seems you got the last one as they are no longer available the site i linked to in the previous post :cry:

[JonMarsh]

Yeah, well, if I hadn't moved fast, it's likely ET would have gotten it- you did notice it was a black one left up for sale?

Now, that's exactly why I didn't go and buy up a bunch of those limited edition ScanSpeak woofers that Dan sent me to test, nice as they are! I could start getting a bad rep around here! :B