Uh-oh, we're in for another round of speaker bashing! To those who wish to participate, I would suggest an experiment: get a cheap and nasty speaker, and hook up the most powerful, most competent, possibly the most expensive, amplifier you're familiar with, and see what you get - if it sounds pretty damn good, it's not because the extremely capable amplifier is "compensating" for the speaker's flaws; it's because the amplifier is no longer the weak link, thereby allowing the speaker to do its job properly ...
Frank, that opens the door to "the wife in the kitchen" knocking here, but I'll try anyway ...Hi,
Ever tried to prove it to someone other than yourself, Frank?
Sorry, open door....
Cheers,
One of my early test CDs was Status Quo, because normally it was a "thick" album - used it often, my wife hated it, it irritated her intensely, it was "rubbish sound". One day, the system was truly on song, Status Quo were powering superbly, my wife came in, uninvited and said "Gee, I really like this!! What group is it?' and I say, "Well, you remember saying I play stuff that really bugs you ... "
Another time, a keen audiophile came, and didn't get it at first - he said afterwards it took him an hour or two to realise what was happening - the system was "spectacularly" non-hifi in nature, and this didn't "compute", it sounded too "normal"; the sound didn't hit you over the head, it was just "there" ...
Yeeaahh! ... why not!?Hi,
....... A pair of cheap PC speakers O.K. ?
Cheers,
Of course, you have to bypass the inbuilt amps, you realise ...?
Hi,
Gravity took care of that already. These are the same ones that fell off the shelf a few months ago so all I need to do now is to solder some zip cord to the speaker terminals.
Status Quo isn't all that bad if and when the system's O.K. (Not that I like the music all that much but I see your point).
Cheers,
Gravity took care of that already. These are the same ones that fell off the shelf a few months ago so all I need to do now is to solder some zip cord to the speaker terminals.
Status Quo isn't all that bad if and when the system's O.K. (Not that I like the music all that much but I see your point).
Cheers,
If you do want to do this seriously, one suggestion I would make: drive the speaker hard, really hard, for an hour or so before appraising it more closely - I find the suspensions are the main weakness, and need to be conditioned by strongly exercising them; typically, the longer and harder you drive them the better they end up sounding ...
I replaced the surround of a full range dayton 5" driver with a chamois surround I white glued in. It sounded wonderful. Looked bad. I think the leather helped damp the cone.
I suspect that the surrounds, and spider attenuate low level movements riding on top of major movement, when not in an optimum working state. IOW, the low level detail is lost - the sound is not as "transparent" as it could be, after heavy conditioning has made these parts more "flexible".
Of course I do John, I was merely pointing out that in a DIY project, this will not be possible without some experimenting on the side of the maker. Not possible to give one-suits-all schematic.
That's 100 % correct. The distortions we hear via SE amps are drive power and modulation of frequency due to output impedance. The transformer also is an expensive device which might be embarrassed by a $10 capacitor in terms of degradation if added to a transistor amp.
I find the bad nature of speakers affects all amplifiers. The Audio-Lab amp has never been my cup of tea until recently. It loves simple speakers. When designing a speaker it is distressing to hear SE type distortion introduced by trying to correct baffle step. Think about it a minute. The inductor in series with the drive unit is no good thing. Add a capacitor to make it ring is the next bad step. In my limited experience first order or no crossover is best. I recently built a speaker like this and found the problems a brick wall like supersonic flight must have been. Very best was a first order bass filter and then use simple tone controls to get it correct. A semi active approach where simplicity is still a virtue. Kenwood added tone controls to the power amp feedback loop. Seems the ideal way to do it. I have tried it and it does work. Nearly all power amps are op amps so easy to do. Texas Filter Pro download is good to use. I used the linear phase option. My OB are + 16 db at 30 Hz. No other filter used. I will add a super tweeter one day as they roll at 15 kHz. 12 inch with mechanical crossed mid. They will not tolerate low damping factor. An SE amp needs 12 dB feedback to work with them as my best guess. If my amp that sets input at 2.5 V rms which is OK and THD at 0.25%. The graphs are not bad at 12 db although better at 0 db feedback if using Hiraga criteria. If looking carefully 3 rd and 5 th slightly more dominant with feedback. However not at all like a typical feedback amp . When valves feedback amps often look ugly. Even with +16 dB 30 Hz the 8 watts is enough to damage the ears after a few hours with my OB's. 1 watt 1 kHz is seriously loud ( 100 db typical in my listening seat). At 1 watt the THD is 0.2% . It is mostly second harmonic. I would guess all other harmonics > - 80dB. I say guess as I didn't really look when it goes off the graph. At 5 watts the Hiraga preference comes in. What this gives me is acceptable distortion on transients. At 8 watts 3rd is larger than 2 nd slightly. Then it clips.
If you add all the distortions of the chain together 4 % THD seems a likely minimum if domestically approved of speaker types. Microphone, recording, amplifier, speakers. To introduce a 0% distortion amplifier might give us 3 % THD. I fully understand the complexity of distortion. My hunch is the ideal sound chosen by natural selection ( customer choice ) will be like this. Thus to look at a well designed SE amp of 1% THD and say intolerable is not good science.
1% THD and 1 % THD might describe two amplifiers. One is dreadful and one is sublime. On the oscilloscope the sublime one looks pure. Only when superimposed with the reference input signal can the distortion be seen as a wider top to the sine wave. The other just looks bad and is statistically not as bad as it looks. That's the rub, statistics are not ears.
Feedback is the easier way to get results. There are other ways and they happen to be better. They take months to perfect and are not in books. One distortion most do not know is the same as speaker to amp output. All stages of amps have this problem. If the complexity is increased the stability often suffers. Even when cured it is not cured,fast signals can find it. There is a sweet spot and it can be found.
Dejan and I have often argued the same point from very different perspectives. Dejan says if I understand correctly to make an amplifier which is by SE amp standards near perfect without feedback. He then adds about 20 dB feedback. His argument is it is too low to affect stability. On questioning 15 dB is as low as he would go. He uses plenty of output current to ensure any crossover distortion comes in above 2 watts. This all seems very reasonable to me. Dejan argues I build an amplifier by removing parts until it stops working. That is almost true. They usually still work when 3 more components could be removed. I stop where I feel the music suffers.
My power supplies often have more components than my amps. I totally dislike shunt PSU's as it seems to sell a false ideal. People say as it is not in the series signal path it is less harmful. That would be true if stability can be guaranteed. Always try the very simple amplified zener diode you were told not to use at University. I was told even 40 years ago not to . It is hard to beat. I asked recently if a shunt regulator could work at 30 MHz. No I was told. That seems to me the reason not to be obsessed with an idea that is not easy to perfect. The truth is all such functions are by local decoupling which works to 1000 MHz if asked.
The advantage of a simple amp is the layers set in the recording jump from the speakers. I don't think the engineer wanted this. It is so nice to hear it. Music like Earth Wind and Fire is very well recorded and the layers seems 10 or more. Like the View-master 3 D viewers the objects have black edges.
Joachim said my Magneplanars have a black sound. I suspect it was not a positive comment? I understand that and feel the amplifier is being heard and that is all. My OB speakers are great fun and in some ways better. Both eat 90% of box speakers for breakfast. It's the layers and the sheer grunt.
It is much easier than you think. Big technical problem is selling it to your wife. Colleen has learn't the hard way big is beautiful. In England the house you need is the big problem. $1000 000 a minimum around here. France ( Limoges ) would be ideal. Colleen doesn't like France. North of England might be the answer.
Careful now, Nige, a few feet and you might end up as an emigree, if the Scots so vote.
Well said Dejan. Sri Lanka has charged my batteries. Cost of a house about $20 000 there. A teacher gets about $220 a month I was told. Fuel and beer almost impossible dreams for a teacher to afford. Rice is about UK price which was a surprise to me. 50% if in bulk. A Tuk-tuk taxi about $4000 as the import duty is high. That means the Tuk-tuk man has spent as much as materials to make a small house. A Morris Minor is highly prized as it will never depreciate and is considered superior to Japanese cars ( A Buddhist way of thinking must operate I guess). $1400 for an excellent example. Many have been fitted with diesel engines. That is a refreshing point of view I feel. A teacher in the UK gets $50 000 and buys a VW with 7 months salary. The Sri Lanka teacher invests 7 months salary into a Moggie and is the winner as it will never be worth less.
You were right about Scotland Dejan. Morpeth is as near as I dare go. I don't like the idea of buying a house by sealed bid. We buy it by me driving a hard bargain or we don't bother. In Scotland the house is sold at about 30% above the quoted price. The one who offers most gets the house. No way Jose.
You were right about Scotland Dejan. Morpeth is as near as I dare go. I don't like the idea of buying a house by sealed bid. We buy it by me driving a hard bargain or we don't bother. In Scotland the house is sold at about 30% above the quoted price. The one who offers most gets the house. No way Jose.
Best in term of what?? It's like the feedback: a necessary evil. If it is not necessary then of course it is better be removed.
Judgement on how much series inductance needed is at the end affected by our priority in choosing what "characters" to have in the system.
When complex filter cannot be avoided due to the nature of the driver (and our ear sensitivity to breakups) then the amplifier must support the speaker by having perfect damping as compensation.
Why ideal??
Since Gainclone popularity with short feedback path, I have looked at this kind of feedback (tone control in FB path) suspiciously.
Not only Kenwood, almost all Japanese vintage amps have employed it, but only in cheaper models. Never seen one in Accuphase (as far as I remember), but many in Marantz.
Only one amp that sounded okay and is using that kind of feedback/TC. But the PCB track is very compact and has become my standard in designing a PCB. Wait till I implemented a replacement circuit for the FB, I think no TC in FB path will be better.
In series or not, or whatever the technical reason, it really doesn't matter for those who have good ears. It sounds good. About stability, I don't understand why a stable regulator should be difficult to design.
For regular class B amps, I haven't used this kind of regulator for the output stage, but only for the input stage. For class A amps, It has been a standard for up to 15W amps.
I think this reaches an area where each person has his own priority and preference of a "good" sounding system.
All I can say is the journey I have been on is a very long one. The answers I sort came very recently. I have thrown the rule book away and asked what are the most basic problems. The rule book mostly is concerned with acceptability to a customer. This often means to his wife mostly. As I said one has to be a dollar millionaire to own a house where I do to have the first hi fi requirement. That is a medium sized listening room.
Negative feedback in the loop sense is 100 % not required. Even local feedback does not exist on my amp. 1% THD is fine for the speaker I use as mostly 0.2% second is all there is from the amp. To avoid unnecessary extra info the distortion spectrum modes not raise in the odd harmonics and all are of the near exponential decay type.
A simple shunt regulator was given over more than 1000 entries and did suffer many problems here at DIY.
OB speakers once heard working correctly can not be forgotten. Sadly the box of a speaker is always heard after that. It makes all other colourations seem trivial by comparison. Also because OB have to be EQ'ed that means one feels free to do it. Forget conventional drivers and pet devices. They simply will not work. A JBL driver might. Sometimes an OB will hit the buffers. That is exciting as in truth ear danger is not far away. It is like hitting the engine red line. Sometimes I am so obvious to volume only the maths suggests it is loud. I have to be very strict and not go too far. Organ music is very good on OB. Each note is a note. Box speakers always seem to have one note after that.
I spent last night listening to T Rex. When that lost it's charm I spent the rest comparing Bellini's Caste Diva from Norma. Callas is by my prejudice my favourite. Renee Fleming least. Joan Sutherland weird and wonderful. Renata Tebaldi stunning . The list is endless. Despite my prejudice I would love to hear Renee live. Ave Maria perhaps. OB speakers only ask me to listen to the music. The recording styles are different, it matters very little. My hi fi lets me listen to music. That has taken me 45 years.
Back to feedback and THD. Hiraga principles. At 1 watt virtually nothing with no feedback when my amp . If music is 30 % sinusoidal and 70 % triangle wave we might weight a curve as a square when transients. Thus if Hiraga rules apply the actual distortion of my amp is zero with respect to all external factors which must include the listener.
My amp is SE class A.Thus at no time does the amp stop conducting or enter the lower BH curve of the output transformer. It is fast and with no feedback back overlap or time delay. The amp will maintain useful power between 20 Hz and 64 kHz. 15 Hz is available albeit at 5 % 5 watts all second.
I spent about 2 months researching the repeatability of the design by using worn out valves. Not only it worked but also gave life back to the device of choice as it is run at 20 % full power. The valves are common and new production at low cost. The combination is never seen in other amps and seems wrong at first glance. The technique I call East West UL. The damping factor is about 4 so needs care in speaker matching. My speaker is very stiff and uses a gathered edge of treated paper. It seems the speakers gains more from the lack of box than lost in the stiff edge. It seldom bottoms even with +16 db at 30 Hz. The driver will handle 450 watts transient in a box. The cost of the whole system about $1000 and 2 years work. I don't count sources as some I have cost silly money. Garrard 401 being the cheapest.
Negative feedback in the loop sense is 100 % not required. Even local feedback does not exist on my amp. 1% THD is fine for the speaker I use as mostly 0.2% second is all there is from the amp. To avoid unnecessary extra info the distortion spectrum modes not raise in the odd harmonics and all are of the near exponential decay type.
A simple shunt regulator was given over more than 1000 entries and did suffer many problems here at DIY.
OB speakers once heard working correctly can not be forgotten. Sadly the box of a speaker is always heard after that. It makes all other colourations seem trivial by comparison. Also because OB have to be EQ'ed that means one feels free to do it. Forget conventional drivers and pet devices. They simply will not work. A JBL driver might. Sometimes an OB will hit the buffers. That is exciting as in truth ear danger is not far away. It is like hitting the engine red line. Sometimes I am so obvious to volume only the maths suggests it is loud. I have to be very strict and not go too far. Organ music is very good on OB. Each note is a note. Box speakers always seem to have one note after that.
I spent last night listening to T Rex. When that lost it's charm I spent the rest comparing Bellini's Caste Diva from Norma. Callas is by my prejudice my favourite. Renee Fleming least. Joan Sutherland weird and wonderful. Renata Tebaldi stunning . The list is endless. Despite my prejudice I would love to hear Renee live. Ave Maria perhaps. OB speakers only ask me to listen to the music. The recording styles are different, it matters very little. My hi fi lets me listen to music. That has taken me 45 years.
Back to feedback and THD. Hiraga principles. At 1 watt virtually nothing with no feedback when my amp . If music is 30 % sinusoidal and 70 % triangle wave we might weight a curve as a square when transients. Thus if Hiraga rules apply the actual distortion of my amp is zero with respect to all external factors which must include the listener.
My amp is SE class A.Thus at no time does the amp stop conducting or enter the lower BH curve of the output transformer. It is fast and with no feedback back overlap or time delay. The amp will maintain useful power between 20 Hz and 64 kHz. 15 Hz is available albeit at 5 % 5 watts all second.
I spent about 2 months researching the repeatability of the design by using worn out valves. Not only it worked but also gave life back to the device of choice as it is run at 20 % full power. The valves are common and new production at low cost. The combination is never seen in other amps and seems wrong at first glance. The technique I call East West UL. The damping factor is about 4 so needs care in speaker matching. My speaker is very stiff and uses a gathered edge of treated paper. It seems the speakers gains more from the lack of box than lost in the stiff edge. It seldom bottoms even with +16 db at 30 Hz. The driver will handle 450 watts transient in a box. The cost of the whole system about $1000 and 2 years work. I don't count sources as some I have cost silly money. Garrard 401 being the cheapest.
Dejan. I know Ivor. I must go see him again.
Linn factory is built where Rudolf Hess landed. No accident that.
I still have an LP12. I was a little lad when I met Ivor. I phoned him from down the road and begged a visit. Best thing I ever did. He never once told me anything that didn't hold up to investigation. He told me most people are stupid so he will play games with them.
Alas Linn is a company that grew too large without a continuous true path to follow. Sad and the fate of all who believe their own BS. I suffer that as badly as any. I try hard not to overly believe what I say. As I said Dejan " I never believe anything especially if I said it ". I was no different at 20 , just faster. I think Ivor got that very quickly and never BSed me. The only BS was the factory. The submarine engine was not expected sitting on the Castlemilk floor. Ivor said " very nice , but a one off job".
Linn factory is built where Rudolf Hess landed. No accident that.
I still have an LP12. I was a little lad when I met Ivor. I phoned him from down the road and begged a visit. Best thing I ever did. He never once told me anything that didn't hold up to investigation. He told me most people are stupid so he will play games with them.
Alas Linn is a company that grew too large without a continuous true path to follow. Sad and the fate of all who believe their own BS. I suffer that as badly as any. I try hard not to overly believe what I say. As I said Dejan " I never believe anything especially if I said it ". I was no different at 20 , just faster. I think Ivor got that very quickly and never BSed me. The only BS was the factory. The submarine engine was not expected sitting on the Castlemilk floor. Ivor said " very nice , but a one off job".
The building in the first picture is a stack of old sea containers.
The occupant lives there for free with his Dominican g/f, in exchange for keeping an eye on the docked vessels, one is a party boat.
He has no car, spends all the money he has on food (and beer), but has all the privacy, a private beach, and one hell of a view from his con-do.
200 yards to his left is the location of the terminal for the largest cruise ships.
At one click to his right, is the only private island in the bay, 1 acre size, think $25 million if it would be offered for sale.
I talked to the guy about the possibility of docking a boat there.
(There's only one commercial marina that can handle yachts over 70ft. Private docking starts at 3/4 of a million dollars for a rundown 5000 sqft waterfront property, a million US with all expenditures included)
The Brits spend their money and time on/in their Castle, but little on food
Attachments
- Status
- Not open for further replies.