Who The Feck is Ohm?

[Chienmortbb]

Resistance, Impedance and Ohms Law

 

On another thread the old chestnut of amplifier power, valve amp magic etc. I thought it might help if I explained (I used the term explained loosely) what Watts and Ohms are and why Ohms Law matters. My qualifications 40 plus years in Electronics, 7 years study, an HND and had letter after my name  but let lapse when fees were too high. I should say that as I try to simplify this, I realise how many other things are involved, so please forgive me for errors or mistakes. I will correct them as soon

 

Ohms are the unit of measurement of resistance and impedance. The difference is that most electrical components, especially loudspeakers, exhibit resistance (DC) and Impedance (AC) and often capacitance. However, we will ignore capacitance here.

 

Ohm's Law states V=I x R Or the voltage V = the current I x the resistance R (Z).

 

If you have ever tried to measure a loudspeaker driver with a multimeter on the resistance range, it will read below the rated impedance.  A Fane 12-500 has an Re of 5.6 ohms and a nominal impedance Z of 8 ohms. The Impedance is frequency dependant, and 8 ohms is usually the lowest load that the driver will present to the amplifier. If there is a passive crossover in the cabinet, this becomes a part of the impedance too.

 

 

 

Edited Monday at 21:02 by Chienmortbb

[Chienmortbb]

DC Power

 

The Power Law, as applied to electronics, takes the form

 

P = V x I where I is current in Amperes or Amps, V = voltage in Volts

 

P=I2 x R where I is current in Amperes or Amps, R is Resistance

 

P= V2 / R where V = voltage in Volts, R is Resistance

 

By using the Power Law and Ohm's Law, we can manipulate to find other values.

 

Easy, so if P=V2 X R then if the voltage is 25 and the resistance is 8, or 78 watts. Not watts RMS or peak, just watts. 

Edited Monday at 15:41 by Chienmortbb

[Chienmortbb]

AC Power

 

Can of worms time. In the old days, or should I say ye olde days, they needed a way to compare AC power with DC power. The way it was done was to rub two knights together  They took a heating element and worked out how much AC voltage was required to raise the temperature of the heating element to that of a DC voltage. From that it was found, taking the Root Mean Squared of a number of voltage samples. If you like maths (math for our colonial cousins) look HERE.

 

For the rest of us, the numbers 0.707 and 1.414 are the magic numbers. So if we know the peak value of a sine wave, either divide it by 1.414 or multiply by 0.707 to get the RMS voltage. So what does the mean?

 

Let's look at European main voltage. Nominally 220 and nominally a sine wave.

 

220 x 1.414 = 311 volts peak.

 

Of course, an AC voltage moved above and below zero volts. If you wanted to impress the opposite sex, you could quote the peak to peak value 311+311=622 volts.

 

So peak to peak voltage is almost 3 times the RMS voltage. Now somewhere along the way some marketing "executive" found this out, opened the window and out went the truth.  If you also take the peak to peak current and multiply the two together, you can get a power 9 times larger.

 

Hi-Fi amps usually strive to keep the signal as clean as possible, with distortion figures of 0.01% or better at full power. Our pesky marketing guy also found that by quoting a distortion figure of 10% he got a higher power. I like to work at 1% THD+N for a power amp but

 

Bugera Veyron quotes a power output of 2000 watts peak into 4 ohms. This approximates to just short of 700 watts RMS, so a loud amp but nowhere near 2000 watts.

 

What if I am  fed up listening to a sine wave?

 

Firstly, what does a sine wave sound like? A flute or acoustic feedback/howlround.

 

However, most of us do not play flute and a monotonous tone is no fun. Music is almost never a perfect sine wave. So you cannot use the shortcuts shown above. Luckily for Electronic Engineers, we can measure RMS voltages either on a True RMS voltmeter or Digital Oscilloscope. Otherwise, it is the long-winded equations seen at the link above.

 

Edited Thursday at 15:23 by Chienmortbb

[bremen]

While you're here, could you help bury the myth that a four ohm speaker is twice as loud as an eight, with the same amplifier 😍

[Chienmortbb]

4 minutes ago, bremen said:

While you're here, could you help bury the myth that a four ohm speaker is twice as loud as an eight, with the same amplifier 😍

I will leave that until later if you do not mind.

[nige1968]

https://en.wikipedia.org/wiki/Georg_Ohm

[agedhorse]

Just a quick note... a nominal 8 ohm speaker can have different definitions but under IEC-268-5, will have an average impedance of 8 ohms with a minimum impedance of no less than 20% below the nominal value within the defined pass band.  In reality, many manufacturers play fast and loose (or highly creative) with this definition, something that amp designers must anticipate (and that regulatory safety test labs will test for).

[agedhorse]

15 hours ago, bremen said:

While you're here, could you help bury the myth that a four ohm speaker is twice as loud as an eight, with the same amplifier 😍

In theory, it will be 3dB louder BUT in practice, it's very difficult to design a 4 ohm version of an 8 ohm speaker with the same sensitivity (based on power). Typically, the 4 ohm version will be from 1-3dB less sensitive, which offsets the increase in power. Additionally, if a 4 ohm speaker is being driven closer to maximum rated power than the 8 ohm version, there will be some additional losses due to thermal and electro-magnetic power compression.

[Downunderwonder]

Aka no free lunch.

[Chienmortbb]

12 hours ago, agedhorse said:

In theory, it will be 3dB louder BUT in practice, it's very difficult to design a 4 ohm version of an 8 ohm speaker with the same sensitivity (based on power). Typically, the 4 ohm version will be from 1-3dB less sensitive, which offsets the increase in power. Additionally, if a 4 ohm speaker is being driven closer to maximum rated power than the 8 ohm version, there will be some additional losses due to thermal and electro-magnetic power compression.

Very close to a reply that I was going to post, The way to create a 4 ohm load is with two 8 ohm speakers in parallel. I would have gone on to explain that despite looking like a simple machine, the drivers we use are complex. 
 

When Beyma published power compression graphs, the loss in power stated at way below rated thermal power.  If my memory serves correctly, most LF drivers at least in the Beyma range were approximately 1dB down on power by the time they had been driven to 1/3 of their right thermal power. At rated  power, the compression had risen to over 3 dB.

The 4 ohms us that many amplifiers cannot supply the extra current needed to achieve the extra power. If you look at an amplifier spec, and it’s 4 ohm power output is not double the 8 ohm power output, then it is likely that the power supply is not able to supply the required current. 
 

Taken all these factors into account, There is very little to gain, if any, in going for a 4 ohm driver/cabinet. Two 8 ohm cabs, stacked vertically, with a decent amp is much better. 

Edited Tuesday at 16:17 by Chienmortbb

[bremen]

^^^^

I love that the above exchange is between an old horse and a dead dog.

 

Both speaking wise words though!

[Chienmortbb]

1 minute ago, bremen said:

^^^^

I love that the above exchange is between an old horse and a dead dog.

 

Both speaking wise words though!

I am not fit to exist in such exalted company.

[agedhorse]

4 hours ago, Chienmortbb said:

Very close to a reply that I was going to post, The way to create a 4 ohm load is with two 8 ohm speakers in Parallel. 
 

I would hAve gone on to explain that despite looking like a simple machine, the drivers we use are complex. 
 

When Beyma published power compression graphs, the loss in power stated at way below rated thermal power.  If my memory serves correctly, most LF drivers at least in the Beyma range were approximately 1dB down on power by the time they had been driven to 1/3 of their right thermal power. At rated  power the compression had risen to over 3 dB.

The 4 ohms us that many amplifiers cannot supply the extra current needed to achieve the extra power. If you look at an amplifier spec and it’s 4 ohm power output is not double the 8 ohm power output then it is likely that tbr power supply is not able to supply the required current. 
 

Taken all these factors into account, There is very little to gain, if any, in going for a 4 ohm driver/cabinet. Two 8 ohm cabs, stacked vertically, with a decent amp is much better. 

Agreed, a 4 ohm load from 2 x 8 ohm drivers does. It suffer from the driver design sensitivity challenges that most 4 ohm drivers face. 
 

It’s also not impossible to design an equally sensitive 4 ohm driver, it’s just much more difficult and may trade off against other important parameters.

[SimonK]

[Chienmortbb]

38 minutes ago, SimonK said:

Much better than the water pipes analogy that they taught us a college.

[Downunderwonder]

6 hours ago, agedhorse said:

Agreed, a 4 ohm load from 2 x 8 ohm drivers does. It suffer from the driver design sensitivity challenges that most 4 ohm drivers face. 
 

It’s also not impossible to design an equally sensitive 4 ohm driver, it’s just much more difficult and may trade off against other important parameters.

And still not even free dessert because although it could get louder with a tiny amp, most people have an excess of amp power.

[tauzero]

16 hours ago, Chienmortbb said:

The 4 ohms us that many amplifiers cannot supply the extra current needed to achieve the extra power. If you look at an amplifier spec, and it’s 4 ohm power output is not double the 8 ohm power output, then it is likely that the power supply is not able to supply the required current. 

 

Won't it be a little less than double because the internal impedance of the amp will form a potential divider with the speaker - while the internal impedance will be low, AIUI it won't be zero. The ratio of internal to speaker impedance will be higher for the 4 ohm speaker than the 8 ohm.

[Bill Fitzmaurice]

You usually get double the power at small signal levels. In most cases you don't get it at full power due to power supply sag.

[bremen]

I'm enjoying this bs-free thread and I'm wary of digressing into subjectivity. But.

 

Conventional wisdom has it that you need a power supply with little sag (eg overrated transformer and caps or regulated smps) for a "solid" (subjectivity alarm) bass. Ask any audiophile/phool. But the renowned GK 400 has tiny reservoir caps and transformer and is much loved for its weighty heft, and for being much louder than it should be for its power. Bob Gallien stated that the sag was deliberate, and a means to this end.

 

Valve amps have saggier supplies and are also generally thought to be "loud". Their clipping behavior is part of this but maybe the sag is important?

 

I wish I still had my V4.

 

Sorry, a bit OT.

[Chienmortbb]

I will be honest that my training did not include valves and I am  ready to be shot down by @agedhorse or any real expert. 
 

As I understand is, Sag in power supplies adds to the natural compression of valve amps that may or may not continue in the pre-amp stages and is certainly true of the output stages.HiFi would demand that an amplifier is not used in its non-linear region to maintain adequate distortion figures (THD or THD+n). Any type of compression, overdrive or fuzz increases the distortion. 
 

Of course sag  occurs at higher power levels and also happens to some degree on solid state amps using an unregulated power supply. Sag is a form of current/voltage limiting. 

[Bill Fitzmaurice]

3 hours ago, Chienmortbb said:

As I understand is, Sag in power supplies adds to the natural compression of valve amps that may or may not continue in the pre-amp stages and is certainly true of the output stages.

True. One reason why earlier Fender amps are preferred is they have valve rectifiers, which have better sag characteristics than those from the mid 60s and later that have solid state diode bridges. That additional bit of sag contributes to the soft knee compression of those amps.

[bremen]

Got that, yes.

 

But you could also see an amp with saggy rails as an expander with a slow attack; it lets the transient through, but as the small caps discharge it's unable to maintain the rail voltage, so the power drops. Or maybe what happens is the transient gets through unscathed, but the following note is clipped by the falling rail, and the consequent distortion is pleasing to the ear - maybe perceived as louder. Bob Gallien designed this into the GK400.

[agedhorse]

Sag is both measurable and subjective. It can be responsible for perceptions that are at odds with what’s really happening. Some of it is related to “perceived compression” and some is due to distortion harmonics that are generated. For some players this is good, for others it’s undesirable. 

[Bill Fitzmaurice]

And that's why Mesa Dual Rectifier amps give you the choice. 😉

[SimonK]

Here's another helpful picture (from https://en.wikipedia.org/wiki/Ohm's_law😞