A stiff power supply is needed to get deep and tight bass from a power amp. But there is another parameter that seems equally important.
It all started when I had my 100W Class AB power amp PCB´s on break-in with a 12 Volt PSU. I got curious how it would sound as a headphone amp. Said and done, I hooked up the headphones and put on some music. I was stunned by the immence power and control in the bass range compared to what I´ve heard from stand-alone headphone amps. OK, 16 pcs of 15A Sankens for a stereo headphone amp is quite an overkill, but I was totally surprised that the sense of power and control was still there even if just milliamps was delivered by the amp. Damping factor should not matter when driving 300 ohms, right ?
Theory says that a tiny amp will perform equally well when driving an easy load like a pair of headphones, but this is obviously not the case. If amplifier A sounds more powerful on speakers than amplifier B does, it will sound equally more powerful with headphones. This is counter-intuitive since damping factor really should be a non-issue when driving 300 ohms with a power amp. But the difference over headphones is real and very easy to detect - if fact just as easy as when listening to the speakers. Interesting indeed.
I have a range of power amps I´ve designed and built over the years. They all have a very stiff power supply (typically the transformer is 8 times rated power and the electrolytics are 180.000 uF or more). But they all have different "power" in the bass range. As can be guessed the smallest amp (with the smallest output devices) sounds smallest. All of them have a very tight bass, only the sense of "power" differs.
The observations has led me to the following conclusions:
- The stiff power supply gives bass control, i.e. the amp will be able to control what is there. If the output stage is small the sound will be slender, but still, very well controlled. I recall that builders of the Hiraga 8W "Monster" like the bass. Surely the control comes from the stiff power supply.
- The sense of bass "power" is a function of the output devices and is, it seems, INDEPENDENT OF THE RATED POWER OF THE AMP. My headphone experiment confirms this.
Class A amps are often praised for their powerful bass. I think that these amps would have the same powerful bass even if biased in Class B. The reason for the deep tight bass is probably the stiff power supply and the output stage, not the bias mode.
These findings suggest that a stiff power supply is simply not enough to achieve powerful bass. There is an absolute need to use powerful output devices. And it also suggests that a 10W power amp can outperform a 100W amp if the latter has a "smaller" output section. The rule is simple: use more (or bigger) output devices and the sense of bass "power" will improve (all of course given that everything else in the design is done right).
Kind regards
Syl
It all started when I had my 100W Class AB power amp PCB´s on break-in with a 12 Volt PSU. I got curious how it would sound as a headphone amp. Said and done, I hooked up the headphones and put on some music. I was stunned by the immence power and control in the bass range compared to what I´ve heard from stand-alone headphone amps. OK, 16 pcs of 15A Sankens for a stereo headphone amp is quite an overkill, but I was totally surprised that the sense of power and control was still there even if just milliamps was delivered by the amp. Damping factor should not matter when driving 300 ohms, right ?
Theory says that a tiny amp will perform equally well when driving an easy load like a pair of headphones, but this is obviously not the case. If amplifier A sounds more powerful on speakers than amplifier B does, it will sound equally more powerful with headphones. This is counter-intuitive since damping factor really should be a non-issue when driving 300 ohms with a power amp. But the difference over headphones is real and very easy to detect - if fact just as easy as when listening to the speakers. Interesting indeed.
I have a range of power amps I´ve designed and built over the years. They all have a very stiff power supply (typically the transformer is 8 times rated power and the electrolytics are 180.000 uF or more). But they all have different "power" in the bass range. As can be guessed the smallest amp (with the smallest output devices) sounds smallest. All of them have a very tight bass, only the sense of "power" differs.
The observations has led me to the following conclusions:
- The stiff power supply gives bass control, i.e. the amp will be able to control what is there. If the output stage is small the sound will be slender, but still, very well controlled. I recall that builders of the Hiraga 8W "Monster" like the bass. Surely the control comes from the stiff power supply.
- The sense of bass "power" is a function of the output devices and is, it seems, INDEPENDENT OF THE RATED POWER OF THE AMP. My headphone experiment confirms this.
Class A amps are often praised for their powerful bass. I think that these amps would have the same powerful bass even if biased in Class B. The reason for the deep tight bass is probably the stiff power supply and the output stage, not the bias mode.
These findings suggest that a stiff power supply is simply not enough to achieve powerful bass. There is an absolute need to use powerful output devices. And it also suggests that a 10W power amp can outperform a 100W amp if the latter has a "smaller" output section. The rule is simple: use more (or bigger) output devices and the sense of bass "power" will improve (all of course given that everything else in the design is done right).
Kind regards
Syl
How come an Ampzilla sounds like a brick house in the bass then? Only two 16A outputs per rail per channel, and they are in series.
talking about power suplyes and their influence in the bass.
what about a regulated power suply. If the transformer is "man" enough for the job and enough capacitance before and after the
regulator circuit it wil end up with a very low output impedance´...but as wallways life is that not simple...
regulator circuits have troubles taking the EMF from the woofer
because they are engeneared for suplying corrent not absorving it...
Life is not easy for making the "ultimate amplifier"...
but we never give up!!!!
Regards
Jorge
what about a regulated power suply. If the transformer is "man" enough for the job and enough capacitance before and after the
regulator circuit it wil end up with a very low output impedance´...but as wallways life is that not simple...
regulator circuits have troubles taking the EMF from the woofer
because they are engeneared for suplying corrent not absorving it...
Life is not easy for making the "ultimate amplifier"...
but we never give up!!!!
Regards
Jorge
"A shunt regulator supplies and absorbs. " I'd like to see a high voltage shunt regulator for an amp like the Ampzilla. 30A?
Hello Djk,
I do not share your opinion on the Ampzilla so I think the findings still apply.
I have listened to the Ampzilla a couple of times, and in my opinion the bass is not very capable. At least it is not capable of playing musical bass on a pair of B&W 801´s. Even my tiny 20W amp outperforms the Ampzilla on the 801´s and those are not my words, it comes from the owner of the Ampzilla. The 20W amp starts and stops the woofer immediately all the way down to 30 Hz. The Ampzilla doesn´t. Maybe things would be more acceptable with book shelf speakers, we did not try that.
High rated power, yes. Brick house bass, not to my ears. For its age it is good though, most amps designed back then was worse.
Have you compared the Ampzilla with the LATEST models from Krell, for example FPB300 ? That´s the kind of bass I want. Unfortunately there is no shortcut available to get killer bass. You need big output devices, and a lot of them.
Regards
Syl
I do not share your opinion on the Ampzilla so I think the findings still apply.
I have listened to the Ampzilla a couple of times, and in my opinion the bass is not very capable. At least it is not capable of playing musical bass on a pair of B&W 801´s. Even my tiny 20W amp outperforms the Ampzilla on the 801´s and those are not my words, it comes from the owner of the Ampzilla. The 20W amp starts and stops the woofer immediately all the way down to 30 Hz. The Ampzilla doesn´t. Maybe things would be more acceptable with book shelf speakers, we did not try that.
High rated power, yes. Brick house bass, not to my ears. For its age it is good though, most amps designed back then was worse.
Have you compared the Ampzilla with the LATEST models from Krell, for example FPB300 ? That´s the kind of bass I want. Unfortunately there is no shortcut available to get killer bass. You need big output devices, and a lot of them.
Regards
Syl
djk said:
A bit of a design challenge, but i expect doable (it would no doubt be bigger & bader than the poweramp it feeds).
My needs are on the order of 300V, 500 mA max and i can mail order that shunt reg off the internet -- those aren't my bass amps thou.
dave
"My needs are on the order of 300V, 500 mA max and i can mail order that shunt reg off the internet -- those aren't my bass amps thou. " This makes sense to me. Rowland uses a shunt regulator in the voltage gain stages of the Model 7 amplifier, with the outputs on a brute force unregulated supply. "I do not share your opinion on the Ampzilla so I think the findings still apply. " I should have mentioned that I add power supply bypass caps to every amplifier I get, and film bypass to all the coupling caps in the signal path. An Ampzilla/Dynaco 400/SAE 2400 with these mods will have tighter bass and more slam than an unmodified Mark Levinson ML3 with its 20 parallel TO3 outputs per channel. The ML3 after my mods had more control and slam than the Rowland 7, which has a regulated front end and 32 TO3 per channel.
Hello Djk,
OK, so your Ampzilla is not the original design. That makes a difference. Adding caps to the PSU gives better bass control, such modifications are well spent money.
The differences you describe relative to the ML3 and Rowland 7 highlight the real issue with many commercial high-end amps. They use a lot of output devices but they don´t spend money on the PSU. Remember I wrote "all of course given that everything else in the design is done right". Just using a lot of output devices does not guarantee good bass. That´s why your mods are so successful.
I am not the least surprised that your modified Ampzilla beats the original ML3.
Regards
Syl
OK, so your Ampzilla is not the original design. That makes a difference. Adding caps to the PSU gives better bass control, such modifications are well spent money.
The differences you describe relative to the ML3 and Rowland 7 highlight the real issue with many commercial high-end amps. They use a lot of output devices but they don´t spend money on the PSU. Remember I wrote "all of course given that everything else in the design is done right". Just using a lot of output devices does not guarantee good bass. That´s why your mods are so successful.
I am not the least surprised that your modified Ampzilla beats the original ML3.
Regards
Syl
Bass slamm
Hi DJK,
As I am a happy owner of a GAS ApzillaII too I am eager to respond.
What makes the good sound and bass slamm is <B>THE</B> subject of this forum.
I can add that the old Ampzilla of mine outbeats the latest Krell 700 with 1 square meter floor space. I say floor as I can't even lift <B>ONE</B> of these monsters. This was my experience at my friends house using Wilsons Grand Slamm. I am not quite sure what makes the difference. I suspect it has to do with ample drive to the outputtransistors. I added some 2x10000µF to the powersupply but that did not make much of a difference.
Some 20 years back I had a interesting experience after selling a pair of Klipsh Heresy spaekers and a amplifier. The customer had already overrun his budget by choosing the Klipsch and asked me if it was OK using his old two times 15 watt receiver as a FM-tuner. The only problem was the cheap receiver did not have recoutputs so we connected the speaker outputs to the tuner inputs of the amplfier (easy to do as the old receiver used cinch for speakeroutputs)
The customer and I me were stunned by the good audioquality from this configuration especially the power and bass-slamm.
Also there once excisted a Krell preamp KRS1 that used a outputstage that more looked like a small poweramp. Again the power level of the sound striked me in comparison with he MarkLevinson JC-2 I owned at the time....
Did you see my mods for the Ampzilla on
http://groups.yahoo.com/group/SAE_Talk/message/2660 ?
I was burned to the the bone by the designer Mr. James Bongiorno. 🙂
Hi DJK,
As I am a happy owner of a GAS ApzillaII too I am eager to respond.
What makes the good sound and bass slamm is <B>THE</B> subject of this forum.
I can add that the old Ampzilla of mine outbeats the latest Krell 700 with 1 square meter floor space. I say floor as I can't even lift <B>ONE</B> of these monsters. This was my experience at my friends house using Wilsons Grand Slamm. I am not quite sure what makes the difference. I suspect it has to do with ample drive to the outputtransistors. I added some 2x10000µF to the powersupply but that did not make much of a difference.
Some 20 years back I had a interesting experience after selling a pair of Klipsh Heresy spaekers and a amplifier. The customer had already overrun his budget by choosing the Klipsch and asked me if it was OK using his old two times 15 watt receiver as a FM-tuner. The only problem was the cheap receiver did not have recoutputs so we connected the speaker outputs to the tuner inputs of the amplfier (easy to do as the old receiver used cinch for speakeroutputs)
The customer and I me were stunned by the good audioquality from this configuration especially the power and bass-slamm.
Also there once excisted a Krell preamp KRS1 that used a outputstage that more looked like a small poweramp. Again the power level of the sound striked me in comparison with he MarkLevinson JC-2 I owned at the time....
Did you see my mods for the Ampzilla on
http://groups.yahoo.com/group/SAE_Talk/message/2660 ?
I was burned to the the bone by the designer Mr. James Bongiorno. 🙂
So Bongo burned you at the stake, huh? No one could ever say James was unbalanced, he has a chip on each shoulder. James has always used polarized elecrolytics in the signal path, and inexpensive ceramics elsewhere. Doesn't seem to have a clue about servo 'feed through' either. I would like to continue this later, compare notes, right now I really need to get some sleep.
djk said:
and
This is a reply from a very old thread.
I have two ML-3's (very old ones 0107 & 0324), one with a biascircuit in one channel which did not fire up when I switch the amplifier on. Discoverd that one of the drivertransistors: 2N3585 only worked well when heated or cooled down with coolingspray. I soldered it through and fastened the base and it works for now at least.
This oldest one has been upgraded from the 18.000 uF silver Callins to 56.000 uF blue Sprague powerlytics nad has been serviced/refurbished by Madrigal in 1995, all transistors etc.. complete bag full of components that were replaced. New relays and two black boxes connected with the line (white cables) connection with the AirPax rockerswitch attached to the frontplate. Are they linefilters? The Avel Lindbergh toroids hum anyway.
The other one smokes the Dale limiting inrushcurrent resistors. The white Guardian relay doesn't release it's connection so that the current keeps flowing through the resistors till it burns.
Could not solve that problem yet. Maybe a connected condensator?
Ordered 10 Dale's 15 ohm 50 watts, 2 Airpax rockerswitches and three Guardian relays probablt for nothing the last two items.
I'm interested in the mods you performed with your ML-3 if you still posses it. And maybe you can relate to some of my questions.
Thanks in advance
After reading all of this thread I think I have something to add. I have either had or listened to a lot of upper end amps, and my conclusions are as follows. Outputs that are not bridged have slightly more bass slam, control is about equal. Power supplies are the largest factor. My conclusion is the DC resistance is a major factor. You can lower the AC impedance with caps but the DC is controlled by the amperage capability (output resistance) of the transformer. Output stage capability is also important as is the driver stages ability to control the outputs. Possibly from a subjective viewpoint, how the amp is controlled or rolled of in the bass may be another large factor. Good reliable performance without servos seem's most consistent. Servos either control too much or too little most of the time.
Just remember there are exceptions to every rule!
However if you have a weak PS you are not going anywhere.
Don
Just remember there are exceptions to every rule!
However if you have a weak PS you are not going anywhere.
Don
- Status
- Not open for further replies.