I think this will greatly depend on how hard does the amp have to work for its living.
If it drives something like my speakers, an easy, very well beaved load, at relatively low power levels because it's also fairly efficient, then increasing the caps will have to substantial to be preceived at all (aassuming the initial values were not ridiculously low).
The more it has to work, the easier it will be to detect the differences.
This apsect, how hard it has to work, will also determine how much bigger they should be. If they were smaller intially, and have to drive seriously misbehaved loudspeakers, then even doubling or more may well in play. The problem you may run into is how much more can the rectifier handle, meaning that at some point, you may have to start changing the rectifier diodes as well.
In some cases, you could get lucky. My Marantz amps happened to have used damn oversize rectifiers right from the start, so upping the caps from 15,000 uF to 22,000 uF, an increase of 50%, was quite easy. But in other cases, this may not be so peachy. And you can just about forget it with entry level products, theirs was caesarian birth to start with.
tvrgeek
Part number for the tweeters you are using? Which tweeter was the first to not be pleasant for your wife?
Dave
Part number for the tweeters you are using? Which tweeter was the first to not be pleasant for your wife?
Dave
Guys, let me reveal you one secret. "Oversizing" is kind of stupid term invented by ignorant people. Actually, the proper term is "undersizing", against optimal values, in order to cut costs. All sizes above optimum are not worth increasing, so Oversizing is a bad term that means waste of materials (designer's failure, actually).
Undersizing is the result of voluntary consideration of so called "nominal" power that is below clipping level. Who undersizes less (like Marantz, John Curl, etc...) win.
Undersizing is the result of voluntary consideration of so called "nominal" power that is below clipping level. Who undersizes less (like Marantz, John Curl, etc...) win.
Parasound makes amps with varying power supply capacitance. You pay your money, and you get your choice. Usually, I don't happily accept credit for the design, unless at least 15,000 uF caps are used, 12,000 for 5-6 channel amps. More later.
Examples: 4 X 33,000 / channel JC-1
2 X 33,000 / channel HCA-3500
2 X 15,000 / channel HCA-1200
2 X 6,800 / channel HCA-1000 Bass is somewhat weak, but OK for small system.
Examples: 4 X 33,000 / channel JC-1
2 X 33,000 / channel HCA-3500
2 X 15,000 / channel HCA-1200
2 X 6,800 / channel HCA-1000 Bass is somewhat weak, but OK for small system.
Last edited:
Where can i find these "great " Marantz amplifiers ....
I did not say they are great. They win against those that are more "undersized". To check how "undersized" it is, run all channels and see what fraction of power from all of them simultaneously can be obtained before it goes out of specs, or simply dies in a smoke cloud.
John; what about transformer and heatsinks? Are they upgradable as well, or they are undersized in capacitance department only?
I recall how much those late '70s Marantz models cost new.
Four times that amount now (~35 year inflation) buys a much better amp.
Is called progress.
No disrespect here either, but I find that their last "Designed in the U.S.A., manufactured in Japan" series, 1978-1980, sonically beats the pants off most modern attempts, in their relative class, of course.
Not perfect, of course, but dating back to the quickly fading days when those people REALLY tried on their standard series. Unfortunately, that was the last of Marantz worth mentioning for quite a while after that.
It was the last of the "whole, grand picture" design objective time, after that, it was all about being "analytical", "subtly detailed", etc, and the whole picture, top to bottom, was left out in the cold. The entire concept changed.
Not to even mention the quality of build. By today's standards, their middle of the range model 1152 DC would be classed as top of the range with most manufacturers except for the nominal power.
Take a look at them inside, and you'll see what I mean. Take a look at their schematics and you will be surprised; of course they show their age in some respects, but are surprisingly up-to-date in other aspects.
Not to mince wirds here, my Polish cousin, but "oversized" as I use it refers to an average of the day.
In absolute terms, "oversized" would be when you add another cap and nothing happens, because you already had enough. And as far as I am concerned, you have "enough" when your amp can deliver the specified power output into the lowest load you specify it for without any appreciable degradation of the sound. Obviously linked to nominal power output.
Personally, I am wholesale into John's reasoning, although I get there using a different route. When I asked him about what he'd consider a job reasonably well done for a nominal 100/200W amp into 8/4 Ohms, he said 22,000 uF minimum; my calculations told me a total of 32,000 uF was required for what I wanted. That's 32,000 uF per line, per channel, a total of 128,000 uF for a stereo amp.
DNM-designed Capacitors
DNM found many years ago that putting slits in the foil reduced eddy currents . This helps high frequency response . The risk is , in increasing the cap size it also reduces the HF quality . Many have found adding small caps is not the complete answer . Nor is multiple caps as defining the exact star point is then more difficult . Going back to the use of bigger transformers ( more iron ) . Has anyone ever calculated the resonant frequency of a power supply ? I imagine it must be of some significance . One should also look at the 50. 60 , 100,120 Hz input of the rectification . It must have a sonic signature . Perhaps this is why increasing cap size works . It swamps the time signatures ? My tests with this indicated 40 Hz is better . I had by chance a 40 Hz supply . That was the limit , below that the amplifier's transformer was unhappy . This is contrary to expectations , people think a higher frequency better . Specifically the mid range was more open . This was using a big re-generator I built .
DNM found many years ago that putting slits in the foil reduced eddy currents . This helps high frequency response . The risk is , in increasing the cap size it also reduces the HF quality . Many have found adding small caps is not the complete answer . Nor is multiple caps as defining the exact star point is then more difficult . Going back to the use of bigger transformers ( more iron ) . Has anyone ever calculated the resonant frequency of a power supply ? I imagine it must be of some significance . One should also look at the 50. 60 , 100,120 Hz input of the rectification . It must have a sonic signature . Perhaps this is why increasing cap size works . It swamps the time signatures ? My tests with this indicated 40 Hz is better . I had by chance a 40 Hz supply . That was the limit , below that the amplifier's transformer was unhappy . This is contrary to expectations , people think a higher frequency better . Specifically the mid range was more open . This was using a big re-generator I built .
Last edited:
Progress is one of the most misused words of today. That's what they usually call cost cutting.
Please define "a much better amp", I'd really like to know, since I have quite a few vintage and current "High End" amplification (assuming you think of an integrated amp delivering 180/250 W into 8/4 Ohms and costing around € 6,000 as "High End").
Right. Neither cap, transformer, transistors, heatsink. Waste of material if to increase above optimal value, or even undesirable effect when some parasitic parameters increase with added value, like one more pair of output transistors increase capacitive load to driver.
Right, and nominal power equals to peak power, like in all my power amps.
There, you see, Wave, we CAN agree on something! Now, was that so hard?
But seriously, we have touched upon a subject not often mentioned, that of using wrong terms for various ideas. Ever since the marketing departments took over from the engineering people, that effect has been on the rise.
Sometimes, it can be even a bit of fun - for example, calling the rate of signal exchange "slew rate". Damn, I hope my amp isn't drinking secretly so it slews.
But we all like abbreviated versions, so "slew rate" is easier than "rate of exchange". Whoever thought of "slew rate" surely had a good sense of humor, and that is (almost) never a bad thing.- Status
- Not open for further replies.