Hey guys. I have a Cambridge Audio amplifier which I would like to know more about. It's the 640A version 2. All I know at this point is that it's a class-AB which does 75wpc into 8 ohms and 120wpc into 4; frequency response 10-30kHz +/-1dB and it can slew 30V/uS.
But I've read contrasting reviews on the net about this amp. (Before anyone chimes in about the original 640A, I realize they weren't so hot, and yes I searched here and read the review about it being a gainclone in disguise....)
The manufacturer provides very little info on the design, and I would like to know more.
It does sound quite nice with my Vandersteen 2C's and Grado 325i cans. (got a small power amp for cans built in, rather than dropdown resistors)
But my concern is that it uses 2200uF times 4 per channel. While the toroidal PSU is very hefty, and has dual secondaries for a pseudo-dual mono config, I think 8800uF per side is chintzy. What' the deal with that, besides cost-cutting? Is this thing 'tuned' or 'voiced' somehow?
If you can't tell already, I'm really interested in audio and would love any info you guys have on this amp.
🙂
But I've read contrasting reviews on the net about this amp. (Before anyone chimes in about the original 640A, I realize they weren't so hot, and yes I searched here and read the review about it being a gainclone in disguise....)
The manufacturer provides very little info on the design, and I would like to know more.
It does sound quite nice with my Vandersteen 2C's and Grado 325i cans. (got a small power amp for cans built in, rather than dropdown resistors)
But my concern is that it uses 2200uF times 4 per channel. While the toroidal PSU is very hefty, and has dual secondaries for a pseudo-dual mono config, I think 8800uF per side is chintzy. What' the deal with that, besides cost-cutting? Is this thing 'tuned' or 'voiced' somehow?
If you can't tell already, I'm really interested in audio and would love any info you guys have on this amp.
🙂
chintz⋅y
/ˈtʃɪntsi/ Show Spelled Pronunciation [chint-see] Show IPA
–adjective, chintz⋅i⋅er, chintz⋅i⋅est.
1. of, like, or decorated with chintz.
2. cheap, inferior, or gaudy.
3. stingy; miserly: a chintzy way to entertain guests.
Hi,
It is not at all. 4,700uF to 10,000uF is a typical class aB range.
4,400uF per rail per channel is not stingy, splitting the capacitors
can give higher ripple ratings and lower hf impedance.
🙂/sreten.
/ˈtʃɪntsi/ Show Spelled Pronunciation [chint-see] Show IPA
–adjective, chintz⋅i⋅er, chintz⋅i⋅est.
1. of, like, or decorated with chintz.
2. cheap, inferior, or gaudy.
3. stingy; miserly: a chintzy way to entertain guests.
Hi,
It is not at all. 4,700uF to 10,000uF is a typical class aB range.
4,400uF per rail per channel is not stingy, splitting the capacitors
can give higher ripple ratings and lower hf impedance.
🙂/sreten.
Well that's good to know, since it does sound good indeed.... Just that my vintage Onkyo amp was rated for 45wpc at 8 ohm and it had 14,000uF per side. Class-AB, too.
with ONLY +-4400uF on each channel it will not sound like a -1dB amplifier at 10Hz, even if it measures as that on a constant sinewave test signal @~1W.
A full bodied amplifier capable of driving a 4ohm speaker to extended and strong bass (but not exaggerated bass) requires ~+-40000uF per channel. That gives excellent response down to below 20Hz by having a -1dB ~=3Hz. reducing the capacitance to 4400uF raises the -1dB frequency to ~27Hz and this will be easily audible if you were to compare PSUs on like amplifiers.
A full bodied amplifier capable of driving a 4ohm speaker to extended and strong bass (but not exaggerated bass) requires ~+-40000uF per channel. That gives excellent response down to below 20Hz by having a -1dB ~=3Hz. reducing the capacitance to 4400uF raises the -1dB frequency to ~27Hz and this will be easily audible if you were to compare PSUs on like amplifiers.
I have self drawn schematic of the amp section only, nothing wrong with the design, not sure if its version 2. Its not a gainclone, and its based on the Self designs. Its a simple design and has a vas design Ive never seen used before in any amp, its good and linear. Its a cheap amp and a lot of cost cutting and cheap components will most probably have been used. Good candidate for upgrading by just using better quality parts, those supply caps would be the first Id replace, 4 X 2200f is lowish if one is looking for better sound quality.
Parts quality is usually the reason a diy design of even simpler design most times outperform this kind of amp although this amp is well designed and could be made to sound very good indeed. It would have pushed the price into a higher bracket with better quality parts.
Parts quality is usually the reason a diy design of even simpler design most times outperform this kind of amp although this amp is well designed and could be made to sound very good indeed. It would have pushed the price into a higher bracket with better quality parts.
I would LOVE to see that schematic, if at all possible.
Regarding the capacitance, and the rolloff at roughly 27Hz, this isn't much of a concern since my speakers are only flat to 30Hz anyway.
Regarding the capacitance, and the rolloff at roughly 27Hz, this isn't much of a concern since my speakers are only flat to 30Hz anyway.
Send me an Email via the link, and Ill email a copy.
AndrewT has a good point, 30Khz might be the rolloff of your speakers but things dont quite function that way when it comes to sonics. This amp is said to sound somewhat lightweight with poor dinamics compared its price competion and even to cheaper units. I know some of the cheaper units that this amp is compaired to and they all have somewhat bigger caps.
AndrewT has a good point, 30Khz might be the rolloff of your speakers but things dont quite function that way when it comes to sonics. This amp is said to sound somewhat lightweight with poor dinamics compared its price competion and even to cheaper units. I know some of the cheaper units that this amp is compaired to and they all have somewhat bigger caps.
yeah I figured as much......Dynamic response isn't too bad but at high volume the clipping indicator begins to light here and there, and the gain reduction system turns back the volume pot a bit.
My email is soundofcloud21@gmail.com
Couldn't send you and email because I am 'under moderation' (??) But just want to say thanks!
My email is soundofcloud21@gmail.com
Couldn't send you and email because I am 'under moderation' (??) But just want to say thanks!
my speakers start to roll off at ~55Hz and I can hear the difference.AndrewT said:
I know which I prefer.
AndrewT said:
Hmmm.......
you seem to be bandying around irrelevant numbers.
As if you can hear -1dB at 10Hz or 3Hz ...
Your implying the supply capacitance is 10 times too low,
and this somehow affects the amplifiers bass response,
well it affects the response but not in the above manner.
Your design criteria seems obscure and ill-defined.
It seems an excuse for overdoing it, which cannot be "bad",
but is not realistic for "budget" commercial amplifier design.
🙂/sreten.
Especially considering the caps are being charged at 120hz (assuming full-wave rectification).
Even if your amp has such terrible PSSR that these things are any problem at all, hum will be a problem before bass roll-off is.
Even if your amp has such terrible PSSR that these things are any problem at all, hum will be a problem before bass roll-off is.
The size of the PSU caps only affects bass response when the amp is CLIPPING. When you do clip at low frequency - and it happens more often than you think if you don't have indicators - the bigger the caps the flatter the flat tops of the waveforms. This distortion is less objectionable than if it doesn't clip cleanly. If it clips clean, the further into clip you can drive it before it starts sounding raunchy.
You can mitigate this effect somewhat by pole staggering - to have the feedback pole and input pole higher in frequency than the effective pole from the speaker Z and the reservoir cap. THAT will affect bass response below clipping. But not all amps are designed this way (budget PA gear tends to be, budget hi-fi usually isn't).
You can mitigate this effect somewhat by pole staggering - to have the feedback pole and input pole higher in frequency than the effective pole from the speaker Z and the reservoir cap. THAT will affect bass response below clipping. But not all amps are designed this way (budget PA gear tends to be, budget hi-fi usually isn't).
Tim__x said:
PSRR is zero at clipping. What you hear first isn't hum, but the intermods between the driving signal and whatever junk is on the rails. Which includes 120Hz hum and modulation of the power supply by the audio signal envelope (dc voltage jumping up and down with the beat - remember old incandescent FM dial lamps?)Bigger caps = less "junk". Of course, if you're never clipping and have decent PSRR you don't hear any of that.
Well there certainly isn't any hum, but bass response could be a bit stronger.
Honestly though, I knew this was a budget component when I purchased it.
Say I was to increase the rail caps, how much could I theoretically over-do it before inrush current gets too to be too much, or the poer supply rectifiers need updating?
Honestly though, I knew this was a budget component when I purchased it.
Say I was to increase the rail caps, how much could I theoretically over-do it before inrush current gets too to be too much, or the poer supply rectifiers need updating?
Hi Sreten,
your response
I say again, it is audible.
your response
seems to indicate you have never tried to size the PSU RC with the NFB RC and finally sizing the input RC to hear what effect that has on the way a power amp delivers it's signal to the speaker.
I say again, it is audible.
Yes, exactly, I'm well aware of that. I agree with everything you said.
From an engineering standpoint it makes a lot more sense to take the money that might have gone towards caps oversized by a factor of 10, and build in some (dynamic) headroom. Sagging rails aren't necessarily a sign of bad design, likely as not it's a sign of well balanced compromise between cost and fidelity. Generally a better bang-for-the-buck, even at the same price, than the lower peak power, built-like-a-tank rails-never-sag-1% designs.
An amp with little rail capacitance that never clips will almost always sound much better than an amp with even full farads of rail capacitance that clips on every beat.
My early post assumed no clipping, as surly everyone agrees that that is of primary concern and the importance of sizing rail capacitors is secondary.
IMO, financially efficient designs are very good things, clipping is very bad, and high dynamic headroom is actually a fairly elegant solution to try and combine the two.
IMO, financially efficient designs are very good things, clipping is very bad, and high dynamic headroom is actually a fairly elegant solution to try and combine the two.
wg_ski said:
So what would you suggest under the given scenario?
I own twelve Peavey CS 800s (4 rack & 3 rack versions) that are all bridged. All are loaded into a 4-ohm nominal load with the lowest impedance dip of 3.4 ohms @ 34 Hertz. They are only used for bass ranging from 100 – 25 Hz. More 50 - 25 Hz due to the music material. At the moment, I updated all twelve amplifiers from 10,000 & 15,000 MFD per rail to 25,000 MFD per rail.
While I don’t have any clipping issues, I would like to have the amplifiers prepared for the worst. Should I increase the capacitance to 50,000 per rail or, should I leave them at 25,000?
Cheers!
Schematic?
Of course would be interesting for all of us.
Pictures?
Inside pictures?
regards,
Carlos
..................................................................
Model 340 looks alike a Gainclone..this one you are talking about i am deeply curious to know something about the circuit... Doctor Self's circuit i think.
Of course would be interesting for all of us.
Pictures?
Inside pictures?
regards,
Carlos
..................................................................
Model 340 looks alike a Gainclone..this one you are talking about i am deeply curious to know something about the circuit... Doctor Self's circuit i think.
Attachments
tomman888 said:
If you increase the rail caps, you won't notice any difference until you drive it to clip. If the amp also has small coupling caps for intentionall roll-off making these larger will be a bigger bang for the buck. As long as you stay away from clipping.....
Unless you're putting more than 20,000 uf per rail, don't worry too much about the turn on surge. Below that, the trafo itself dominates the surge.
- Status
- Not open for further replies.