Current Drive VS Voltsge Drive

[GlennGregor]

ACT (Audio Current Transfer) was used in several Sony integrateds (and some sources like tuners) in the 1980s. It was short lived. The amplifiers were pretty good, although rather ugly with the volume control on the 'wrong' side.

The power stage is realtively conventional, and I really am not a fan of tone controls in the negative feedback loop.

The block diagram:
View attachment 349307

View attachment 349312

I've got a TA-F444es (1986) in the storeroom. It was comprehensively bettered by the TA-F333esR from 1989.

I run the TA-F555ES direct so I bypass the tone controls. This also gives me a perfectly flat response .

 

[GlennGregor]

The amplifier you linked does not provide current drive to the speakers, just the low-level signal inside from preamp stage to amplifier stage. I think the output (driving the speakers) is a conventional class AB amplifier. Which by no means detracts from its great sound! Just not current drive to the speakers.

MP3 offers a wide variety of compression levels.

@restorer-john -- Do you know what the Sony TA-F555ES output stage is like? My memory of these Sony units is pretty vague at this point, despite having seen and even owned a few back then.

This is from the operating instructions. Legato Linear Power Amplifier Stage

 

[Chrispy]

So is current drive different from current dumping?

 

[DonH56]

This is from the operating instructions. Legato Linear Power Amplifier Stage

OK, so a class B amplifier, voltage-mode (not current drive), with "legato linear" whatever that is. Reads like they used high-frequency (high fT) transistors to switch more quickly through the zero-crossing point to reduce crossover distortion (crossing over from one side to the other of the output device, e.g. plus half to minus half of the waveform, not related to the crossover in the speakers). Feedback reduces that distortion, and presumably higher-bandwidth transistors may help produce higher gain-bandwidth product for greater feedback to reduce distortion. Or something like that. In any event, current drive or not, what is most important is that it sounds great to you.

Edit: As @restorer-john already said -- depend on him to know the goods!

 

[DonH56]

So is current drive different from current dumping?

Yes. Current drive as most folk (or maybe just I) consider it in this context means the output stage is a current source, not a voltage source. Current dumping has several meanings that depend upon the context (and the circuit) but to me (for audio amplifiers) implies an additional high-current stage to supply additional power on demand. That allows a low-power output stage (still voltage mode though could be current) to provide power to the speakers most of the time, but when a large peak comes along the high-current ("dumping") stage augments the low-power stage to provide greater output on demand. More common today (I think) are class G or H amplifiers that use a higher-voltage output supply voltage to provide additional power for peaks. The higher voltage is only applied as needed.

 

[welwynnick]

The usual expression is current source or voltage source, rather than current drive or voltage drive.

Analogue audio connections are generally voltage source. That means the output signal is the voltage, and the output current is whatever it needs to be. The source impedance is low (ideally close to zero) and the load impedance is high. Because the source impedance is low, the output voltage doesn't vary much when a current is drawn by the load. A perfect voltage source can supply any current into the load, and still supply the correct voltage.

With a curent source, the output signal is the current, and the output voltage is whatever it needs to be. The source impedance is very high, and the load impedance is low. The output current doesn't vary much when a voltage is demanded by the load. A perfect current source will supply any voltage, and still supply the correct current into the load.

Line level connections, such as DAC to pre-amp or pre to power amp, are usually quite close to a perfect voltage source, with much higher load impedance than source impedance. Its a medium level voltage (say up to 1, 2 or 4V) and a very low current. That means that very little power is transfered in the connection.

Loudspeaker or headphone connections are still voltage sources, but a bit different. The load impedance is medium and is variable, around 8 ohms for speakers and higher for 'phones. The output is still a voltage, but the amp's ability to supply current is usually set by real world limits. A good amplifier typically has it's output voltage and current and hence power set for 8 ohm loads, but may be able to double the current (and hence power) into a 4 ohm load. That's doubling down, and that's usually as far as power amps get to being perfect voltage sources; any better would be expensive and wasteful. The ratio of load impedance to source impedance is the damping factor.

The speaker connection does transmit a lot of power. In high frequency electronics, power transfer is usually optimised by having the same source and load impedance, That's the middle ground between current and voltage sources. But with audio equipment, that would mean the amplifier dissipates the same power internally as it does into the speaker, which would be inefficient. Therefore a voltage source is used instead.

I don't think it would be practicable to have current source with speakers as so much power would be dissipatedd in the power amp.

It could however be used with line level connections instead, as they transmit so little power, it wouldn't matter. However, the global hifi industry is set up around voltage source connections, and a true current source would be proprietary and exclusive.

Internally, within the equipment its different , and a manufacturer can do what it wants. DAC architecture is usually designed around current sources, and an I to V convertor and a buffer are required to generate a voltage output.

 

[kschmit2]

Here's an article by Nelson Pass titled "Current Source Amplifiers and Sensitive / Full-Range Drivers": https://web.archive.org/web/20051104070915/https://www.passdiy.com/pdf/cs-amps-speakers.pdf

The conlusion reads:

So here you have it – a re-examination of some things that have been known for a long time but were here tested with currently available drivers. It’s easy to dismiss the older ways.
Somewhere in the 1960s, the speaker/amplifier interface took a left turn and headed off to high-power voltage source amplifiers and speakers designed to lean on them for performance. A small group of iconoclastic cranks has stayed interested in these fossils and in the end, we have to recognize full-range high-efficiency speakers as offering elegance, charm and a different sort of quality.

Current sources and amplifiers with low damping offer interesting possibilities for improvement with these drivers but they require considerable work to get the enclosures, electrical networks and acoustic environment just right. Remember, you don’t have to own a current-source amplifier to put this information to work. Placing R0 in series with the output of a powerful voltage-source amplifier instead of in parallel with a current source will give similar results. If R0 is 47 ohms or higher, you are going to want a big amp, say 300 watts, but in any case, be certain to use a high power resistor of 50 watts or more. That said, it’s all just entertainment and I hope it inspires some of you to have some fun.

The article also contains a simulation summary on pages 26-30.

 

[ShadowFiend]

Current drive has advantage that if frequency of input signal greater than 2-4 times resonance frequency of speaker driver in system, it can reduce the 3rd harmonic distortion from 8dB-15dB, depends on drivers, while having no impact on 2nd harmonic distortion.

But it has the several problem
Problem 1: In normal cone driver, the resonace frequency of driver has very high impedance (50 Ohm-60 Ohm), and since the current driving is by definition constant current then the voltage which amplifier needs to output at resonance frequency is very high, V = R*I. It will easily clipping amplifier, and the distortion is so high that negates any advantage of current drive.

Problem 2: Ok, so if we uses impedance compensation circuit to flatten the impedance curve, then the problem 1 is not exists? No, because in most speaker, the resonance frequency of of driver inside the enclosure will be effect by air compression or resonance of port as well, so the resonance frequency we need to flatten is not constant, and make the impedance compensation circuit pretty useless.

Problem 3: For any speaker driver, the impedance will raise again in high frequency, and induction also lower so current drive is not particular effective for treble region.

With all those disadvantages, the current drive is not a universal solution, and that is why it is not popular. It is very effective only in two situations:
1. For midrange from 300-2000 Hz with resonance frequency well below the operation frequency with high order filter crossover. That is what Kii 3 does with its midrange driver.
2. For planar/AMT/ribbon driver because of its flat frequency response, and low impedance at resonance frequency. Example is my Mundorf AMT 25D1.1 have nominal impedance of 6 Ohm, and raise to on 6.3 Ohm at resonance frequency and constant raising to 6.8 Ohm at 40 kHz.

 

[GlennGregor]

I just would like to thank everyone for the responses I got. I appreciate it.

 

[DrCWO]

I like to report my listening experience using a NCx500 amplifier with my horn spekaers.
The horn speakers only transport sound above 130Hz and are accompanied by two woofers.

At the beginning I used a Krell 2250e to drive the horns. Was quite happy with it. As the NCx500 was praised so much here I got one and was disappointed as the sound was not as pristine and clean as with the Krell.

In the past I owned a Firstwatt F1 (not enough power) and remembered that the impedance of the amplifier (nearly zero in case of NCx500) may cause the issue. I also remembered an article of Nelson Pass regarding current drive for Sensitive Full Range Speakers claiming less distortion. I also read here about the effect of increasing amplifier impedance.

After some experimenting I ended up inserting a 15Ohm resistor in the speaker cable. As the speakers are very sensitive (96db) there is not much current flowing through the resistor even at very high volume. Screwed on a cooler they never get hot.

This means I have no current drive amplifier and I burn half of the power in the resistor. What I got is an amplifier with 15Ohm impedance instead of zero. But I really don‘t care about that as the NCx500 delivers much more power as I need for the horn plus resistor.

The result is absolutely satisfying. Much better as with the Krell absolutely transparent and clean. The NCx500 is directly driven by a RME ADI-2/4 which delivers enough voltage to run the NCx500 without an amplifying buffer.

Sure this is a very special setup but driving the speaker with less „Voltage“ and more „Current“ gives such a good result that anybody I had here for listening was euphoric about what he heard.

EDIT:
The new active horns speakers from Avantgarde now market iTron which also claim to use current drive. I listened to them at the HighEnd Show in Munich this year and was satisfied. Good as mine now

 

[egellings]

Problem with current drive is that speakers in general are designed to be driven by voltage sources, not current sources. A voltage source is characterized as having a low source impedance, which can have a braking or damping effect on unwanted cone motion, while current sources have a very high source impedance, which allows the cone to flap in the breeze more, undamped after the signal is gone. Of course, sound quality is in the ear of the beholder, and some may prefer the, um, more relaxed sound of an underdamped speaker. The serious audiophile, however, sniffs at such goings on.

 

[DVDdoug]

Sure this is a very special setup but driving the speaker with less „Voltage“ and more „Current“ gives such a good result that anybody I had here for listening was euphoric about what he heard.

Ohm's Law* says current is proportional to voltage with constant resistance (or impedance) so you can't have one without the other.

Audio amplifiers are supposed to be "constant voltage" (voltage that's independent of the load) and speakers (and headphones) are specified & tested with voltage sources.

The current can change at different frequencies where the impedance is higher or lower, but the frequency response should (ideally) be flat regardless of how the current varies over the frequency range.

This also means the speaker is getting less power (wattage) at frequencies where the impedance is higher, and more power where the impedance is lower. But again a good speaker will be flat as long as the voltage doesn't vary over the frequency range.

Sometimes people prefer the way a speaker or headphone sounds with a higher impedance source... For example you might get a bass-bump that you like. But it's the "wrong way" to get bass boost.



* It's a law of nature with man-made units-of-measure.

 

[pma]

Speaker distortion under voltage drive and current drive

Please let me contribute to not so often discussed theme of speaker distortion. I have done some research measurements on this topic in past years and would like to present it here. The simplest way is to put some links to my web page, so in case you are interested in the subject, please have a...

www.audiosciencereview.com

 

[egellings]

Ohm's Law* says current is proportional to voltage with constant resistance (or impedance) so you can't have one without the other.

Audio amplifiers are supposed to be "constant voltage" (voltage that's independent of the load) and speakers (and headphones) are specified & tested with voltage sources.

The current can change at different frequencies where the impedance is higher or lower, but the frequency response should (ideally) be flat regardless of how the current varies over the frequency range.

This also means the speaker is getting less power (wattage) at frequencies where the impedance is higher, and more power where the impedance is lower. But again a good speaker will be flat as long as the voltage doesn't vary over the frequency range.

Sometimes people prefer the way a speaker or headphone sounds with a higher impedance source... For example you might get a bass-bump that you like. But it's the "wrong way" to get bass boost.



* It's a law of nature with man-made units-of-measure.

If it tickles your nun-handles just right, then that's what counts. Might not tickle someone else's, however.

 

[DrCWO]

Problem with current drive is that speakers in general are designed to be driven by voltage sources, not current sources. A voltage source is characterized as having a low source impedance, which can have a braking or damping effect on unwanted cone motion, while current sources have a very high source impedance, which allows the cone to flap in the breeze more, undamped after the signal is gone. Of course, sound quality is in the ear of the beholder, and some may prefer the, um, more relaxed sound of an underdamped speaker. The serious audiophile, however, sniffs at such goings on.

In my example the speakers have their resonance frequency at 35Hz and are only used down to 130hz. So damping is no issue at all.

 

[DrCWO]

Sometimes people prefer the way a speaker or headphone sounds with a higher impedance source... For example you might get a bass-bump that you like. But it's the "wrong way" to get bass boost.

Forgot to mention that the system was calibrated in both scenarios to the same target curve using Acourate. So impedance changes did not lead to a different amplitude response

 

[DrCWO]

Ohm's Law* says current is proportional to voltage with constant resistance (or impedance) so you can't have one without the other

Sorry, but as Ph.D. In electrical engineering I know Ohm‘s law
What I wanted to say - and there you seem to have misunderstood me - is, that I can have current drive (source impedance endless) on one side and voltage drive (source impedance zero) on the other side. But in-between there is a region where the source impedance is not zero nor endless but has a certain resistance. This I meant with more „Current drive“ than „ Voltage drive“.
Hope this makes things clearer. As I‘m not a native speaker I might have expressed this in a wrong way…

 

[egellings]

In my example the speakers have their resonance frequency at 35Hz and are only used down to 130hz. So damping is no issue at all.

True. Don't tickle it where it giggles, and you'll be fine. A resonance as low as 35Hz could actually help if the room does not allow sounds of that low a frequency to be heard.

 

[MaxwellsEq]

Does current drive subtend to an infinite source resistance

 

[restorer-john]

@DrCWO What are you doing below 130Hz? Handing off to a powered subwoofer?

I can see how so-called "current drive" at frequencies above 130Hz, with certain content could be perfectly fine, but if you're passing the buck to the sub, what exactly are you proving?

Cheers.