Are very low impedances really that hard to drive?

toroechado

Hi to all!

I was bench testing an amp that was capable of delievering 33 V into 4 ohm load. According to my calculations, that is 272.25 watts (P=V^2/R). There should be 8.25 amps (I=V/R). The amp is rated for 4 ohms minimum. Nice.

Now the question. What if I had a nasty speaker thad drops to 1 ohm at certain frequency but is very efficient, for example 100 dB / W at 1m. As per Crown Audio calculator, such speaker should need 16 watts to listen at 85dB at 4m and with a headroom of 15 dB, that should be plenty.

16 watts at 1 ohm should be 4V and 4A (V= (P*R) and I=P/V)... So, if this amp is capable of doing 33V and 8.25A, why won't it be able to put out 4V and 4A? Or can it?

Am I missing something here?

Thanks!

Attachments

• sc.jpg
25.7 KB · Views: 40
Last edited:
1 user

jean-paul

Nasty loudspeakers that drop to 1 Ohm at a certain frequency tend to not be of the high efficiency type.

3 users

rayma

Yes, such dips are due to eccentric crossover design, cough Wilson Audio cough.

2 users

jtgofish

One of the great crimes of audio design is using multiple 8 ohm drivers wired in parallel.If that resulted in a 4 ohm load that would probably be fine but it usually means more like 3 or even 2.5 ohms.I have owned speakers like that and they are a pain because they severely restrict your choice of amps.You have to use something overbuilt so it can deliver lots of current but that undermines value for money .
And the reason?Well they can use cheaper ,lower value inductors in the crossovers.So to save say \$50 in parts they create an unreasonably fussy and demanding speaker.

Citizen124032

16 watts at 1 ohm should be 4V and 4A (V= (P*R) and I=P/V)... So, if this amp is capable of doing 33V and 8.25A, why won't it be able to put out 4V and 4A? Or can it?

Am I missing something here?
No, you're right and not missing something.
Specifications, topology and expectations.
Knowing that a loudspeaker is a complex impedance, consisting of an electrical environment, a mechanical and an acoustical, it should be of no surprise that with transient signals this contraption can generate energy trying to feed its energy back to the supposed ideal voltage source, aka the amplifier.
But that thing refuses to swallow external energy, as it is not permitted by its topology (flyback diodes 1N400x... sure). It just cuts off.
What can the power supply really deliver? What amount of current can the amplifier process (as an ideal voltage source)?
Most specifications are 'sine-wave' peak-power consumer-home-party yells, the sophisticated high end (and DIY-audio fanatics!) are not representative.
I've posted this some time ago here, but given the net supply frequency of 50/60 Hz, the 'refill' rate for the power supply can keep up pace to 100/120Hz at best, which is 10/8.3msec, and that is very slow. Three-phase or SMPS has other issues (promising though!).
Delivering 33V into 4Ω is indeed 272W / 8^A, but at what frequency? Is that 33V peak? Is that 8A peak? Can the power supply deliver both channels 8A, given that the most energy demanding signals are in the 200-300Hz (5/3msec) range, way above that refill rate, where both (stereo) channels are in phase?
I doubt. Seriously.
Getting only 10W out (or into) an perfect voltage source amplifier is a very difficult challange.
Your amplifier might be able to push 2 x 272W into two 4Ω heat dissipating non-reactive loads with a 1kHz sine wave.
There is a wide gap between reality and expectation. Or perception. Or imagination. Or illusion.

1 user

toroechado

I totally get your point. To answer your questions about my numbers, I measured it at 1KHz, 33V was RMS continuous with no visible clipping in the scope and using a purely resistive dummy load. There is no other channel, the amp is a monoblock.

planet10

Paid Member
Nasty loudspeakers that drop to 1 Ohm at a certain frequency tend to not be of the high efficiency type.

The dip is usually caused by the XI and the reactive components add signicant phase angle response.

Power = Volts x Amps x cos(phase angle).

You require lots of excess current capability if you do not want the amp to be current limited.

dave

1 user

steveu

Forgive me for not doing a lot of work to illustrate but remember the amp dissipation is a function of the current times the DIFFERENCE between the output voltage and the power supply voltage. So, 4 Amps into 1 Ohm is much harder on an amp than 4 Amps into 4 Ohms. The amp dissipation peaks at about half the clipping voltage, above which the transistors drop less voltage from the supply to the speaker, so the dissipation falls off.

Some amps have VI protection that allows more current as the voltage rises because that means the OP transistor voltage drop is less. VI protection is not popular because it often behaves badly into a transformer or other reactive load.
https://sound-au.com/vi.htm#:~:text...e voltage across and the current through them.

2 users

WhiteDragon

Actually yeah.
Very good point a impedance curve with 1 ohm dip be rather horrible.
and very much likely to trigger a VI.
VI clamps send all kinds of flyback voltage bursts and highly likely
to cause stability issues as well.

Even with just 4 ohms amplifier distortion rises quickly.

Modern high quality speakers and computer aided design.
roller coaster impedance dips are very much avoided easily.

have no interest in even 4 ohm loads.

having seen live sound designs with 3 ohm dips in the tweeter
section. Easily overheated and shutdown class D at high frequency.
Low frequency even worse. A amp tolerating 1 ohm is pumping
the heck out of a power supply. specially garbage switch mode.

1 users

davidsrsb

Popular output transistors like the 2SC5198 are 10A headline rating. Into low impedances, where 40V may appear across the transistor at peak current, the SOAR graph limits you to 3A at 100ms, so the actual power output into a 1 Ohm load from a single pair would be around 5W only.
Audio power transistors are designed in Vceo vs Ic for the 8 Ohm world

1 user

wg_ski

Many times those impedance dips are used for amplitude equalization. It’s common to see midrange drivers that are only 84 dB/W, and the way to bring them up to the level of the woofer is to let the impedance drop on the bandpass. With a dual driver WMTMW system, 2 ohms in the low midrange, right where there is a lot of peak power demand, can happen.

wg_ski

Audio power transistors are designed in Vceo vs Ic for the 8 Ohm world
One pair per 8 ohm load. To drive 2 ohms you just need 4 pair. To go down to 1 ohm, you probably don’t need all 8, but you sure as heck could use 6.

6A3sUMMER

You should think more creatively.

Design and build a very special transformer:
4 Ohms input to 1 Ohm output that can handle 300Watts from 20Hz to 20kHz, has less than 0.5dB insertion loss, and be sure to purchase hernia insurance.

By the way, even if your solid state amplifier blows up, there will never be DC sent to the loudspeaker.
Sounds like a super vacuum tube amplifier to me.

Never connect any amplifier to a loudspeaker, until you have made an accurate measurement of the loudspeaker DC resistance.
At least one frequency will have an impedance that is equal to DCR; trust me.

Just my opinions.

Mister Audio

The way I delt with the possibility of driving low impedances NOW & FUTURE was to invest in
a high current - high power amplifier, capable of easily driving 2 ohm loads.
I found a 'sale' opportunity to buy at nearly half price and just 'bit the bullet' and got what I wanted.
I just love having all the current & headroom I could want - along with excellently high damping factor.
https://www.zzounds.com/productreview--macm1400i

jean-paul

Series filters in the high efficiency loudspeakers and a low power amplifier that can deliver current. Designing a system with attention to every part of it and adapting the parts to eachother and to real world living room requirements works out OK in many attractive aspects.

One may loose some but also may gain more than that what was lost.

Last edited:

Mister Audio

davidsrsb said:
Audio power transistors are designed in Vceo vs Ic for the 8 Ohm world

In the 1980's a massive transformation of "Nominal Impedance" changed from 8 ohms to 6 ohms.

1 user

6A3sUMMER

Circuit City was full of receivers that were rated for 8 Ohms. Only the \$1200 receiver was rated for 8/4 Ohms.
And they had 6 and 4 Ohm loudspeakers, and No 8 Ohm loudspeakers.

Circuit Who? out of business.

Reminds me of Borders, with 5 platter CD players and headphones, you could listen to all or part of 5 CDs, then move on to the next 5 CD station.
That sold CDs.
Then, they installed some kind of streaming system that played part of a song, part of another song, and you had to switch to another CD.
The streaming sound was so horrible, that compared to the original low resolution MP3 recordings, MP3 sounded like a \$10,000 music system.
I could no longer listen to Borders crap song playback.

davidsrsb

One pair per 8 ohm load. To drive 2 ohms you just need 4 pair. To go down to 1 ohm, you probably don’t need all 8, but you sure as heck could use 6.
With all the RF stability, current sharing. space taken issues this causes.
There is no real justification for very inefficient drivers

1 user

JMFahey

davidsrsb said:
Audio power transistors are designed in Vceo vs Ic for the 8 Ohm world

In the 1980's a massive transformation of "Nominal Impedance" changed from 8 ohms to 6 ohms.
Funny thing is that they actually didn´t change the designs at all; simply some Marketing guy was having a beer or two with a real Engineer and learned that (as a safety margin), 8 ohm rated amps were designed to properly drive 6 ohm DCR* speakers so in a stroke of genius he started quoting "RMS" power based on that "new rating"

Techbabble snake oil, no real advantage but it certainly offered an 8/6=33% "free increase in advertised power.
Worst is they could prove that on a bench ... on 6 ohm resistive loads that is.

Once somebody started quoting that half-truth, everybody else followed.

Remember PMPO? .... it´s still with us.

1 users