• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

A question of dynamics: Amplifier design?

Status
Not open for further replies.
I'm afraid you don't know what your talking about. First, you quote me out of context than blather on about digital recording. What part of what I said is wrong? Do you know what 0dbfs means and how it relates to 0 db VU ( or reference level in digital recorders) and the relation between these and dynamic range and head room?
And aren't you the one who thinks 48volts isnt enough for phantom power. I think Neumann knows better.
 
A question of resolution?

The technology and history of DAT (DAT from 1987, digital from 1972 on...)

And here :

tapers.jpg


Article from 1999 on PC recording...

The only thing that never changes is the monumental stupidity of the RIAA and its friends:-

"In the late 1980s, the Recording Industry Association of America (RIAA) unsuccessfully lobbied against the introduction of DAT devices into the U.S.

Initially, the organization threatened legal action against any manufacturer attempting to sell DAT machines in the country (!)
It later sought to impose restrictions on DAT recorders to prevent them from being used to copy LPs, CDs, and prerecorded cassettes.

This opposition by CBS softened after Sony, a DAT manufacturer, bought CBS Records in January 1988. By June 1989, an agreement was reached, and the only concession the RIAA would receive was a more practical recommendation from manufacturers to Congress that legislation be enacted to require that recorders have a Serial Copy Management System to prevent digital copying..."


"
Professional recording industry

DAT was used professionally in the 1990s by the professional audio recording industry as part of an emerging all-digital production chain also including digital multi-track recorders and digital mixing consoles that was used to create a fully digital recording.

In this configuration, it is possible for the audio to remain digital from the first AD converter after the mic preamp until it is in a CD player."


Amateur and home use

"DAT was envisaged by proponents as the successor format to analogue audio cassettes in the way that the compact disc was the successor to vinyl-based recordings.


It sold well in Japan, where high-end consumer audio stores stocked DAT recorders and tapes into the 2010s and second-hand stores generally continued to offer a wide selection of mint condition machines.


However, there and in other nations, the technology was never as commercially popular as CD or cassette. DAT recorders proved to be comparatively expensive and few commercial recordings were available. Globally, DAT remained popular, for a time, for making and trading recordings of live music (see bootleg recording), since available DAT recorders predated affordable CD recorders."


Currently the only equivalents to emerge for the portable DAT are from people like Nagra. Which has a much higher price tag and does not provide phantom power, and uses the "orrible" 3.5mm jack format again for connectic, or the
Olympus LS-100 which does.



Then there is this stupidly expense toy from Sony for playback only, the Walkman ZX2.
 
And aren't you the one who thinks 48volts isnt enough for phantom power. I think Neumann knows better.

It's you that doesn't know what you are talking about.

I use Neumann microphones on a daily basis.
So what? They all have their uses as do Schoeps.

They are less than 1/2 the price of B + K or DPA.
The popular KM18x series is a budget microphone widely used for the kind of recordings you are listening to on a daily basis for such things as Jazz and backing bands.

I was astonished to see the output from a DPA omni was 14dB higher than my reference KM100 series mic...and you are lecturing me about mV levels not mattering?
In that case it was the difference between clipping on my recording after a back to back mic swop, and "normal' levels.

Everyone knows in the professional audio industry, Bruel and Kjaer (or DPA), are much better microphones than Neumann, but even they made a concession to the standardization of phantom power to 48V for mixing consoles.
I doubt Neumann would argue with that.

WHY?

Because the industry is widely using both dynamic and digital COMPRESSION for dynamic range & reduction of the artefacts & the defects of digital audio.

I had interesting conversation about it with several people including R Burwen who confessed to me he can't hear anything above 12khz, and is way down above 6khz...

I doubt if you are middle aged and male you can do any better.
You are probably -20dB at 8khz, which means compared with a teenager you are deaf.

hearing4.jpg


The B + K 130V (& DPA) 4000 series uses 130V phantom for the precise reason that for SPL above 100dB (eg. industrial jet engine noise analysis) 48V phantom power is woefully inadequate for the required headroom.

Those mics use metal diaphragms for this reason.

Here is a short debate on gearslutz about it.

Read up a little before trying to knock me:-

DPA Microphones :: Products

To give you an even better idea, the media don't even know what they're talking about, never mind the audiophiles who repeat hearsay....

The most famous "Percussions de Strasbourg" recording using B+K 130V recording was claimed by the music media "not to have used any compression",-

...whereas I know from the person that recorded it, that it was impossible to have made such a recording with such enormous dynamic range WITHOUT compression.

It was a private joke to watch people's eulogies of the recording based on nonsense, when the reality was totally different. 😀

dpa_4003.jpg
 
Last edited:
48 volts is not enough for a mic that puts out a volt? Do you have any idea what gain structure is in a recording chain? Can you show me a mic pre that will take 50 volts input signal ?You say things that have nothing to do with the conversation. If you know about recording why don't you answer my questions? Because you don't know much and try to cover it by regurgitating other peoples comments. If you know so much answer the few questions.
 
Do you have any idea what gain structure is in a recording chain?
Can you show me a mic pre that will take 50 volts input signal ?

oh dear! 🙄
Do you know anything about microphone circuits at all and the requirements for symmetric lines at 48V?

Can you not see the presence of transformers in this design? 🙄

What makes me laugh is the insistence of the utility of valve amplifiers and "dynamics", even OTL valve amps for so called "purity" when the whole recording chain is commonly being done with op-amps as below, coupled of course by transformers.

Jensen_Twin_Servo_Microphone_Preamp_1984_small.JPG


Jensen_Twin_Servo_1997.jpg
 
Last edited:
It's really time you guys got some kind of handle on what you are supposed to know about.

May I recommend an excellent link.
If you input 1V into the nice little online calculator and ask it output a level of -40dB you will find it equates to 10mV, yet in the quietest concert halls we can't reliably get below about a noise floor of about -55dB.

If I ask it to show the level for 1mV that is now down at -60dB.
I can't go down there, it's below silence.

This is something like 5 x lower than the typical output of a magnetic cartridge, and it means my uncompressed 24-96 output from my concert recording (without the public adding further racket) is going to go from about 1.8mV>1V.

Someone on here claimed my FM tuner must be rubbish because their's is reliably able to output from 0.290uV>1V (a dynamic range of 70dB) in reliable conditions, yet you are not listening to my uncompressed source material which doesn't even hit a dynamic range of 60dB, and is typical of what we record in the concert hall BEFORE compression to broadcast.

How on earth does that work?

I would just like to know how this is all supposed to happen when the signal to noise ratio of most FM radio is about 25-30dB on a good day.
The frequency response is also strongly affected as shown here, making a strong cut off at 50hz and another at about 12khz.

FM broadcasting - Wikipedia, the free encyclopedia

QED.

An externally hosted image should be here but it was not working when we last tested it.


NB:-
Converting the multiplex signal back into left and right audio signals is performed by a decoder, built into stereo receivers.

In order to preserve stereo separation and signal-to-noise parameters, it is normal practice to apply pre-emphasis to the left and right channels before encoding, and to apply de-emphasis at the receiver after decoding.

Stereo FM signals are more susceptible to noise and multipath distortion than are mono FM signals.

In addition, for a given RF level at the receiver, the signal-to-noise ratio for the stereo signal will be worse than for the mono receiver.

For this reason many stereo FM receivers include a stereo/mono switch to allow listening in mono when reception conditions are less than ideal, and most car radios are arranged to reduce the separation as the signal-to-noise ratio worsens, eventually going to mono while still indicating a stereo signal is being received.
 
Last edited:
Here is the spec sheet of a typical Digigram PC audio card.
(These are very widely used for broadcast applications, and made by a leading French company)

As you can see, they exceed the wildest most exotic specs of any audio amplifier let alone the speaker system by a large order of magnitude.

So there is someone that's going to pipe up, and say MY mega-buck DAC, or 30 000USD amplifier linked to 100 000 USD horn speakers is better?

Give me a break!

Audio specifications
Sampling frequencies From 7 - 50 kHz, 100 Hz steps
Audio processing 32 bit floating point
A/D and D/A converter resolutions 24 bit
Frequency response at 48 kHz (record + play) ± 0.2 dB
Signal-to-noise ratio (unweighted) > +93 dB (line)
Distortion + noise at 1KHz (record + play) 0.004% (<-88 dB)
Channel phase difference : 20Hz to 20kHz 0.2° to 2°
Analog channel crosstalk at 1 KHz < -105 dB

Some 15 years ago I was invited to do some audits of local FM radio stations, with the transmitter distance at only about 3-4kms.

The frequency response, signal to noise ratio and stereo seperation figures returned were a pretty salutary lesson.
One of them was used for regional classical music radio, and I was shocked how bad it was.

Another national one I visited confessed to doing ALL their music broadcast from files held on their server held in mp3 format.

Both of the above ones I originally did the audit for, went to this format shortly afterwards...as does 99% of the industry.
(I was very strongly against this for classical music, but they refused to comply).

This became the subject of a breach of contract on their part, for which they were successfully take to court and sued
(
which they never paid).

So please, please don't try to claim things which are manifestly untrue. OK?

 
Last edited:
Thanks for the responses.

I should mention that I have two amplifiers that most dramatically exhibit this dynamic affect. Both were built by Jef Larsen of Abraxas Audio. One is a tube rectified SET/SEP KT88 and the other is a SS rectified PP 6V6. The KT88 is the most "dynamic". I guess that if the PSU is key to this then I must say that Jef has the power supply part of the design down very well.

Compression would actually best describe how most amps /that I've heard/ behave compared to these two.

There is a side issue. If a small adjustment is made to your amplifier it should work OK with EL34. It might be changing 470 to 560 R if using a cathode resistor. Up to the clipping point you should find it a difficult choice. Even at clipping the output is at best to worst 8 to 5 watts if simple SE. To be very different it needs to be 30 watts. I shouldn't say the next part as it might spoil the fun. I suspect the EL 34 will produce more distortion. This distortion might be of a slightly nicer type harmonic to harmonic when EL34. That is it looks like a suspension bridge. You might find this sounds less distorted and louder in the dynamics! In a sense it is less distorted if Jean Hiraga is right. If the system distortion resembles that of the ear it seems to have mysterious qualities. Not really a mystery except to perhaps 90% at DIY Audio, they seems to get worked up about - 80 dB not being good enough. I spoke years ago to part of Oxford University about this after supplying 2 x KEF T27 tested to 40 kHz by KEF. They said in the analogue part the ear has about 30 % THD of the suspesion bridge type. They described the ear as nearly a digital process that knows it's own signature. They were working on a signal that comes back from the brain in the MHz responce time. A servo was the conjecture. If the total hi fi system distortion is 3% and of the suspension bridge type it will be heard as zero distortion. A small change in the spectrum will not go unnoticed. The ear-brain seems set up for is things that obey natural laws, the twang of a string, beat of a drum and song of a bird. It also likes transients as that would be the snap of a twig when being hunted. The liking is we know there is no real danger. As people who lived in forests we need these skills. Stereo helps at close range. Why we tollerate phase distortion is due to having lived in forests. However phase distortion makes us work hard at listening, it should not be accepted as OK. Damping factors also come into it. Doubless a speaker able to move sounds more dynamic until it sounds bloated. Transitor guys never think of the crossover when so interested in this. What of the inductor and it's additional resistance? This is often why active speakers sound worse. Most likely they were design slightly by ear and assuming an inductor.

Talking of the forest effects. My Ex came to stay. I have just made some very large OB speakers that have enough EQ to have 30 Hz ( tested, flat to 40 Hz ). Unlike box speakers this is a very different bass end. Many more notes and seldom obvious bass. 90 % of the time it is slightly strange as the usual box bass is not there. When bass is in the mix it is very deep. As she watched a film she became restless. " There is a very strange sound in the house ". It took me a minute to work it out. It was near subsonics she was hearing. I said " don't you hear that in real life " ? She then caught on and smilled. " The sound comes from nowhere " she said.

If Hiraga is right a weighting system might give a very different view of hi fi. It might explain why so many questions go unanswered. To do that the overal distortion of LP's and CD would need to be added to that of amplifer and speakers. All that are on the exponential curve or whatever and below a total of 3% might be hi fi ? Anything other than that even if lower in distortion might be thought to be coloured ? I suspect the systems with correct distortion colour also will have greater dynamics? There is an analogy with black and white TV when CRT. Often the linearity was not perfect and shadows could be green or blue depending on the level. This became more obvious as the tube aged. When the tube was replaced the seemingly awfaw picture dynamics were bightingly good again. Up to the point of doing that the colour cast was the obvious problem. The other analogy is black. Black and white is never balck and white as the TV can not do black. It does it well enough that the eye sees black. I dare say even that is 30 % away from real black ?
 
mullered said:
Someone on here claimed my FM tuner must be rubbish because their's is reliably able to output from 0.290uV>1V (a dynamic range of 70dB) in reliable conditions
0.29uV to 1V is a ratio of 130.75dB. Perhaps you meant 0.29mV?

I would just like to know how this is all supposed to happen when the signal to noise ratio of most FM radio is about 25-30dB on a good day.
This is utter nonsense. As before, I suspect you are confusing signal-to-noise ratio with useful dynamic range. Any competent FM tuner has a S/N ratio, given enough RF signal, of around 70dB. You really should learn what words mean before using them.

The frequency response is also strongly affected as shown here, making a strong cut off at 50hz and another at about 12khz.
FM radio in the UK goes from around 20-30Hz up to 15kHz. I suspect most other places are similar.

Stereo FM signals are more susceptible to noise and multipath distortion than are mono FM signals.
Of course. That is why an FM tuner needs around 1mV RF for good low noise stereo, but only 100uV for mono.

We sometimes get people on here saying things like "I am a recording engineer, so of course I know what I am talking about - unlike you DIY ignoramuses". Often it turns out that their grasp on real engineering facts is somewhat tenuous, so they have to bolster their weak arguments with bluster and insults.
 
0.29uV to 1V is a ratio of 130.75dB. Perhaps you meant 0.29mV?
.Any competent FM tuner has a S/N ratio, given enough RF signal, of around 70dB. You really should learn what words mean before using them.

FM radio in the UK goes from around 20-30Hz up to 15kHz. I suspect most other places are similar.

Often it turns out that their grasp on real engineering facts is somewhat tenuous, so they have to bolster their weak arguments with bluster and insults.
It's a shame you are making comments about yourself.
Where did I even mention RF signal?

Did I mention the S:N of the FM tuner CIRCUITRY??
EVER?
I suggest you come and show me how to do my job, as I suspect you don't earn a living from it.
FM radio frequency response 20-30Hz up to 15kHz dream on!

As I just told you, what you are listening to on your wonderful FM radio is badly distorted Mp3, because that is the files they are playing back over the air.

If you don't believe me, may I suggest going to visit one?

The dynamic range of FM radio is invariably determined by the LIMITER which they stick on the end of the Audio.

This also introduces a hard cut, and substantially affects the entire frequency spectrum, never mind the redundancy which has all be forced out of the sound by conversion to Mpeg audio.

You really will get F-all above 8khz, and nothing much below 100hz, so people like Classic FM (Virgin), screw with the equalisation like crazy and pump up the bass and treble to compensate, so that their target audience sitting in traffic jams on the M25 can say their "smooth classics" sound great.

You really should wake up and stop dreaming.
FM radio is with just a few rare exceptions completely screwed & broken.

The only noteable exceptions I can think of just now are German (Eg. SWF), French (France-Musique) and British (BBC) live broadcast, but even then, they are messed up thanks to the haste to make them compatible with that awful disaster called DAB.

The national Radio Classique in Paris (owned by the French biilionaire Bernard Arnaud the boss of LVMH) is ENTIRELY Mp3.

I know because I asked, when I called there. 😱
 
Last edited:
I have a FM transmitter that cost £5 from the local petrol station. After I removed it's switch it works very well. Of all the bits of hi fi I have it is one of the best. I listen to it via my mobile phone that is a give away at £20 including an OK camera. Also £5 JVC earphones. It is genuine hi fi. It's just a way of using TV at night. I do have Stax phones and 4 very nice FM tuners ( Quad FM3 , Sony STA3950 inc SW, Armstrong 625, Kenwood KR750, Leak in bits, a valve type, Quad AM also made in Oxford at Horns ). I still think the £30 group is hi fi albeit mono. The FM transmitter uses the output of most modern things to best advantage, nicely worked out. It has 5 output frequencies in 0.1 MHz steps. Even these are correct. As far as I can see it has no ic's. Slope modulation at a guess? When played via my downstaries hi fi hum is the only problem. Even so not awful. I guess distortion must hit 10% at times. Mostly it is far lower. If I were to carefully adjust the input level even the 10 % would not happen. Where this little FM transmitter is so very good is treble detail. This is a big surprise. The Nokia phone perhaps will be good as it was designed by an RF expert ( team ). As much as the FM will be low on the list these people are not idiots.
 
Last edited:
mullered said:
You really will get F-all above 8khz, and nothing much below 100hz, so people like Classic FM (Virgin), screw with the equalisation like crazy and pump up the bass and treble to compensate, so that their target audience sitting in traffic jams on the M25 can say their "smooth classics" sound great.
I don't listen to Classic FM except in the car. I don't listen to Virgin. At home I sometimes listen to BBC Radio 3, as that still has some semblance of competent sound engineering. As far as I am aware, R3 does not use MP3. I don't listen to DAB because it is awful.

I was talking about the FM system, not some modern misuses of it carried out by your professional colleagues.
 
I knew Michael. I wish I had been a better friend. I often ask myself what Michael would have said when thinking of things.

I suspect the feed is MP3/DAB via BBC as it sounds so similar. This wasn't true in the past. It makes sense to do it that way. That is send DAB to the transmitter and then do FM from it. I like MP3. It has a simplicity about it. Just paper thin at times. The OB speakers have helped it work. That is looking for things that were always needed and now even more so.

Michael must have lived about 2 minutes from Percy Wilson. Wish I had asked. He is called Oxford's unknown Stephen Hawking.
 
I must tell you the unusual story of the Quad AM tuner as best I know it. When Horns hi fi shop was sold the skeleton of a horse was found. Mr Horn knew instantly why. Horn's was a horse driven taxi firm. It went over to cars and as a service charged lead acid batteries. From this old Mr Horn started to repair and make radios. Sent his son to college and his son did the same studies as myself. When Quad wanted an AM tuner old Mr Horn offered. 1.5 units per day. I met the man who helped old Mr Horn at a car boot sale. He left Horn's when betting shops became legal as his father was in that trade. Funny world ? If I remember rightly most AM tuners went to people living in Paris who worked for the UK government as typical customers. BBC Home service that was at 200 kHz or 1500 metres. Now 198 kHz Radio 4. Personally I would go back to the old names. Home, World , Light, And Radio 1 as that is it's name. It is a reflection on our standards that the newby was called Radio One. Dr Beeching comes into it ( familly member alas , My Mum is a Beeching, not a common name, Grand dads cousin I believe ) .

My Sony SW is not bad. In theory 2 kHz bandwidth ( text books said ). In reality far more. It was surprising at how much good music was on SW. To educate rather than impress. I haven't listened for the Lincolnshire Poacher or Number Counters recently. That was facinating. Albania was stated. Wiki thinks differently. Radio is a vast subject. Poor Mr Armstrong how he was duped.

Lincolnshire Poacher (numbers station) - Wikipedia, the free encyclopedia
 
I was born and raised in Oxfordshire and Radio 1 was my favourite. Of course during evenings with the girlfriend it became Radio Luxembourg, which in reality was a pirate radio station operating from a ship in the English Channel - or so I heard. It was an AM channel with dubious reception quality and yet it gave more pleasure than most any radio station I've heard since. There is no requirement for 'hi-fi' to achieve enjoyment.
 
I suspect the feed is MP3/DAB via BBC as it sounds so similar.

That is send DAB to the transmitter and then do FM from it. I like MP3..

'fraid that is totally impossible. (fortunately!)
DAB uses a totally different topology.

FM stereo has always used a PCM distribution network, so any idea that BBC radio was some how "analog" in the good old days, has that myth destroyed in one sentence.

DAB was originally developed from musicam which is another lossy coding system, substantially inferior to Fraunhofer's horrible version.

If you like masking and all the effects, 90% of all the information missing that should be there, then you will love mp3.
It suits anyone over 45, because they can't hear the artefacts anyway as shown above, in the hearing loss graphs.
As you see, women have much better hearing than men, especially as they age.

As an aside:-
Gerzon was interesting in the analog days,-

He coped admirably well with all the horrible bodges that introduced, struggled unequally with the noise, distortion, phase problems, non linear freq response and Calrec's really noisy microphones to do experiments with surround, matrixing and the soundfield mic.

I don't doubt the awful B format UHJ/Matrix H surround stuff from Nimbus records directly resulted from his work.

I don't think that was his fault.
I must be one of the few actually to have the proper discrete decoder for it.
 
Last edited:
Status
Not open for further replies.