Dvv and everyone else: Yes, to make a really exceptional power amp, you have to fulfill a checklist of requirements. The L/O amp did its best, for the time, (over 40 years ago) to fulfill this checklist. Virtually none of the other solid state commercial designs, at the time, made the same attempt to complete the list, and this rather humble measuring amp wound up to be sonically superior to virtually everything else at the time, and is even considered exceptional today.
However, IF the L/O amp topology had been optimized as an OP AMP, it would have failed to be exceptional in subjective, (and even in some objective) performance.
However, IF the L/O amp topology had been optimized as an OP AMP, it would have failed to be exceptional in subjective, (and even in some objective) performance.
Dejan. Just returned from a mini holiday is Dorset and Somerset . Dorset seems to have excellent ciders also. My goodness Dorset is beautiful as is Cheddar and Somerset in general ( Cheddar is still a 1970's place for the good and the bad) . Went Oxford to Hungerford, Salisbury, Blandford, Piddle Hinton. Returned via Sherbourne, Cheddar, Bath, Bradford on Avon, Marlborough, Lambourn, Oxford. Never touched a motorway.
Sad note was stopping in Chew Magna where Nytech was and the late Richard Hay. Very beautiful man made lake there.
Purbeck Cider Company- Dorset Cider at its best
Sad note was stopping in Chew Magna where Nytech was and the late Richard Hay. Very beautiful man made lake there.
Purbeck Cider Company- Dorset Cider at its best
For azimuth I see now. If all it uses is a 1kHz mono tone I don't see why any test record and an XY scope display can't extract the same info. I see they set stylus rake on a tiny needle to 1000'th of a degree.:roll eyes:
Last edited:
A long time ago Doug Sax make a sort run of test records with L, R, L+R and L-R tracks and the simple process was to optimize all 4 axis with a scope or even a meter. Adding a pink noise track with an extended frequency range would help to see what it happening in the ultrasonic range. A track encoded on the disk as a triangle wave with a manageable velocity would generate a square wave at the output of the cartridge that could highlight the stylus resonances. I would also add a high frequency or ultrasonic reference tone for looking at wow and flutter. The usual 3.16 KHz tone hides most HF flutter components. Arta could probably do a fine job of analyzing audio from such a disk.
Getting a disk like this made today would not be too hard. It could make for an interesting group buy? Since its not audio some flaws like noise could be ignored which would increase the yield enormously.
Getting a disk like this made today would not be too hard. It could make for an interesting group buy? Since its not audio some flaws like noise could be ignored which would increase the yield enormously.
Lucky sod!
But seriously, I'm glad you seem to have had a good time, Nige. In the meanwhile, I've been dealing with the builders and contemplating suicide. I'll hold on until Monday night, becaue of Tuesday early morn I'll honk my horn, throw myself and wife into the Chevy Cruze and hit the road heading sourt, to Greece, to the warm seas of the Aegean.
Exactly 750 km (app. 470 miles) from my home to my hotel, been there, done that. Lunch on the shores of a man made lake in Macedonia around 2 PM. Espresso just after crossing the Macedonia/Greece border, in the hotel just in time for dinner.
Ay hombre, I can't wait!
Quite so, I said so in my own words, but we obviously agree completely.
I'd just like to add that if one initially sets out for lowish GNFB, one has one's work cut out for him because he doesn't have much to fall back on. Has to get it right in the first pass, stage by stage, and the all together now, which is another way of saying one has to make it clean straight away. The GNFB should serve as a sort of topping on the cake, not as an universal panacea.
Bluntly stated, the amp has to be able to offer very good sound without any GNFB and no gross stability issues. And that's a tall order, but is quite possible and has been since the day of the O/L amp. And as you say, that was 42 years ago.
Not too many, I suspect, at least not in its original form. You have to understand that is the harder way to go, which requires considerably more work, and development time is expensive.
But it has generally reduced the overall amount og NFB from really mad numbers they used to have to more reasonable numbers today. Unfortunately, the industry is still under pressure to produce ultra low THD figures, so generally they still use more than they really have to to get those numbers.
I'm sure there must be companies and people out there who do it as it was done then, some go further, for example Densen had some power amps which used no global NFB at all, they did it all locally. I can't actually name any, but I don't follow the market too much, I did that and own what I wish to own.
Yes, the original Otalla-Lohstrom had a lot of forward gain and a lot of feedback.
The early Electrocompaniet was not that way.
It had only 20dB of negative feedback and quite high distortion numbers.
I have Sandströms later schematic somewhere.
I can post it as comparison maybe tomorrow.
I also have the earlier circuit.
What i understand that could be an advantage of high open loop bandwidth is that the distortion spectrum is more consistent over bandwidth and level.
On the other hand a classical high open loop gain low open loop bandwidth amp has usually less distortion overall although it may rise in the treble, still to unconsequential numbers though.
All this has beeing beaten to death many times here already but for some latecommers it may still be of interest.
I myself could not care less.
When the amp measures competent ( say less then 0.1% harmonic distortion and no crossover distortion ) i am fine.
I hear diffences in amps though.
I think many amps in use clip on typical high end speakers with maybe 85dB efficiency and drop to 2 Ohm somewhere.
Yes, this is not rare, many so called High End speakers eat power for breakfast.
A soft compressing tube amp may have a subjective advantage there.
The best amp i heard in my system is a high power transistor amp though.
My 2 x 30W tube amp simply does not sound as clean, composed or dynamic.
The early Electrocompaniet was not that way.
It had only 20dB of negative feedback and quite high distortion numbers.
I have Sandströms later schematic somewhere.
I can post it as comparison maybe tomorrow.
I also have the earlier circuit.
What i understand that could be an advantage of high open loop bandwidth is that the distortion spectrum is more consistent over bandwidth and level.
On the other hand a classical high open loop gain low open loop bandwidth amp has usually less distortion overall although it may rise in the treble, still to unconsequential numbers though.
All this has beeing beaten to death many times here already but for some latecommers it may still be of interest.
I myself could not care less.
When the amp measures competent ( say less then 0.1% harmonic distortion and no crossover distortion ) i am fine.
I hear diffences in amps though.
I think many amps in use clip on typical high end speakers with maybe 85dB efficiency and drop to 2 Ohm somewhere.
Yes, this is not rare, many so called High End speakers eat power for breakfast.
A soft compressing tube amp may have a subjective advantage there.
The best amp i heard in my system is a high power transistor amp though.
My 2 x 30W tube amp simply does not sound as clean, composed or dynamic.
The Otala/Lohstroth amp paper is using the purported TIM reduction of Otala's "flat loop gain" formulation as the starting point
but as we keep pointing out that initial proposition is flawed
proven in theory and hardware - including custom built measurement hardware - Cordell's work is a model of objective analysis, respecting his "opponent's" position, demonstrating the reasoning flaws, offering corrected formulation and practical demonstration - careful reading of Bob's Mosfet Power amp paper will show he references Otala's loudspeaker peak output current work, tests his amp to that recommendation too - Bob had no problem with suggestions just because they come from Otala
the next claim of it being good to "equalize the distortion contributions" is again logically weak - if reducing all distortions reduces some distortion more than others we should refrain altogether?
and the high output stage bias isn't Otala's contribution - read the paper's ref section already
making sure the prior driver stages can drive the output Q is not an "innovation" as I understand what engineers are expected to do daily
cheap or ignorantly designed amp's of the era flaws aren't relevant to the engineering understanding
but as we keep pointing out that initial proposition is flawed
proven in theory and hardware - including custom built measurement hardware - Cordell's work is a model of objective analysis, respecting his "opponent's" position, demonstrating the reasoning flaws, offering corrected formulation and practical demonstration - careful reading of Bob's Mosfet Power amp paper will show he references Otala's loudspeaker peak output current work, tests his amp to that recommendation too - Bob had no problem with suggestions just because they come from Otala
the next claim of it being good to "equalize the distortion contributions" is again logically weak - if reducing all distortions reduces some distortion more than others we should refrain altogether?
and the high output stage bias isn't Otala's contribution - read the paper's ref section already
making sure the prior driver stages can drive the output Q is not an "innovation" as I understand what engineers are expected to do daily
cheap or ignorantly designed amp's of the era flaws aren't relevant to the engineering understanding
Last edited:
That was for adjusting VTA to get SRA. We’d better forget it. Even the different thickness of each vinyl record will set you off.
And who will tell the optimum gross VTA for each record? Depending on production year and continent or label, recommended VTA varied as much as 5 to 8 degrees.
True. A phasescope on your recording SW is even better.
If one is to adjust the cartridge azimuth on a radial arm, whatever such method he uses, he will precisely adjust for a particular modulation velocity only.
For all the other mod velocities, adjustment will be off.
This is because the skating force is a function of modulation velocity too .
Alas the arm's antiskating compensation caters only for radial position changes and not for this dynamic variation.
George
Last edited:
Wide dynamic true to scale speakers will eat power and will not be of high sensitivity if properly balanced ..
Usually the case for me and this includes high up the totem pole big dollar toobs. Toob pre's on the other hand can hold their own ...
Find Ron Quan's follow-up article virtually nothing modern has any issues except maybe valve amps. I ran my simple op-amp from the open design thread, 80 PICO-sec of so called "PIM" at 20kHz and gain of 20dB over a full 20V p-p to 1V p-p output.
Last edited:
I checked on a having a disk made.
Basic numbers (1000 disk run)
Disk audiophile pressing $3.00
Jacket $2.50
Labels $.16
Cut two sides one pass ea. $1000
Plating $500
And then shipping etc.
So around $8.00 per disk real cost. Or $8,000 to do such a project. Probably not as good an idea after all. No wonder new vinyl is so expensive.
Basic numbers (1000 disk run)
Disk audiophile pressing $3.00
Jacket $2.50
Labels $.16
Cut two sides one pass ea. $1000
Plating $500
And then shipping etc.
So around $8.00 per disk real cost. Or $8,000 to do such a project. Probably not as good an idea after all. No wonder new vinyl is so expensive.
I have heard that 5000 to 10000 disks can be pulled from a stamper, depending on the desired quality level. I'm sure in the heyday of LP's they would get far more from a set of stampers. It would be a daunting production task to manage new stampers that often for a blockbuster hit. That would have been a more complex process with masters, mothers etc.
- Status
- Not open for further replies.