Hi Guys,
Sorry to ask such a basic question, but I'm getting myself confused about something and I've not found an answer through searching previous posts:
At my local indian restaurant, the music coming out of their ceiling speakers is very distorted, even when on quietly (most noticeably at the highest frequencies). It is also quite 'tinny' - Lacking anything much at the low end.
Last night while I was there I mentioned this to the owner and then offered to take a look.
There is an unbranded 4 channel PA head, and 11 speaker cables going down behind it. I'll get a chance to pull all this out in a couple of weekends time to find out exactly what is there.
I have a nasty feeling that he has got 11 full range ceiling speakers (Guessing 8ohms each) wired in parallel!!!
I am assuming that the speakers and Amp he is using are not 100V line.
I've read loads of stuff about Impedence and Amps and seen many seemingly conflicting stories about what does what. Below is an idea of what I am thinking - I'd love to know if I've got the right idea or not:
If we take eleven 8ohm speakers an wire them in parallel we get an approximate impedance of:
(1/R)*n = 1/R
(1/8)*11= 1/1.375
R = 0.727 ohms
The wire run is quite long so perhaps we get a bit of extra impedance from that (It's standard thin cheap speaker cable running up to about 15metres per speaker I reckon).
So if we take a typical and forgiving 8ohm P.A. Amp, and only load it with 1ohm - I think from what I have read that we are going to try and exceed the capabilities of the Amp's PSU and get nasty clipping and non-harmonic distortion (and possibly burnt the amp out).
How does the maths work?
I was thinking about V=IR and P=IV but if a 100W amp will run at 40v peak on an 8ohm load for example, if I then drop the Resistance/Impedance the voltage can't go up because of the PSU - Do we then pump more current out? I assume not, as the voice coils might have burnt out if we'd done that...
Any Ideas/Suggestions you all have would be nice to hear.
Many Thanks
Jonathan
Sorry to ask such a basic question, but I'm getting myself confused about something and I've not found an answer through searching previous posts:
At my local indian restaurant, the music coming out of their ceiling speakers is very distorted, even when on quietly (most noticeably at the highest frequencies). It is also quite 'tinny' - Lacking anything much at the low end.
Last night while I was there I mentioned this to the owner and then offered to take a look.
There is an unbranded 4 channel PA head, and 11 speaker cables going down behind it. I'll get a chance to pull all this out in a couple of weekends time to find out exactly what is there.
I have a nasty feeling that he has got 11 full range ceiling speakers (Guessing 8ohms each) wired in parallel!!!
I am assuming that the speakers and Amp he is using are not 100V line.
I've read loads of stuff about Impedence and Amps and seen many seemingly conflicting stories about what does what. Below is an idea of what I am thinking - I'd love to know if I've got the right idea or not:
If we take eleven 8ohm speakers an wire them in parallel we get an approximate impedance of:
(1/R)*n = 1/R
(1/8)*11= 1/1.375
R = 0.727 ohms
The wire run is quite long so perhaps we get a bit of extra impedance from that (It's standard thin cheap speaker cable running up to about 15metres per speaker I reckon).
So if we take a typical and forgiving 8ohm P.A. Amp, and only load it with 1ohm - I think from what I have read that we are going to try and exceed the capabilities of the Amp's PSU and get nasty clipping and non-harmonic distortion (and possibly burnt the amp out).
How does the maths work?
I was thinking about V=IR and P=IV but if a 100W amp will run at 40v peak on an 8ohm load for example, if I then drop the Resistance/Impedance the voltage can't go up because of the PSU - Do we then pump more current out? I assume not, as the voice coils might have burnt out if we'd done that...
Any Ideas/Suggestions you all have would be nice to hear.
Many Thanks
Jonathan
Are you sure that the speakers are not '100 Volt Line' types? (75V line I believe in the USA).
In this case each speaker has its own step-down transformer with the ability to select different ratios by tappings. The amplifier has an output step-up transformer so that it can put a signal of 100V rms out. Basically you select the tapping on the transformer at each speaker to set the maximum power that is required at that particular speaker. In this system you can have many speaker/transformer units all connected in parallel and (provided that the tappings are set properly) you won't overload the amp.
In this case each speaker has its own step-down transformer with the ability to select different ratios by tappings. The amplifier has an output step-up transformer so that it can put a signal of 100V rms out. Basically you select the tapping on the transformer at each speaker to set the maximum power that is required at that particular speaker. In this system you can have many speaker/transformer units all connected in parallel and (provided that the tappings are set properly) you won't overload the amp.
Hi Ouroboros,
I'm open to the possibility that these speakers could be 100V line types - BUT that wouldn't explain why they sound so bad..
Apparently the system was installed by builders, so there's not much chance that they had a clue what they were doing.
The Amp looks similar to some nasty Adastra P.A. heads that I've got, which have a output transformer with taps off for 4ohm, 8ohm, 16ohm, 100v (and I think 75v too!)
I would assume that if you took 100V line speakers and connected them to the 8ohm output by mistake, you would not get any sound, and if you attached 8ohm speakers to the 100V line you'd blow them up?
I'll know all this for sure when I get a chance to examine the back of the Amp, but I've just got a feeling that they may have badly installed this equipment rather than having broken something.
Cheers
Jonathan
I'm open to the possibility that these speakers could be 100V line types - BUT that wouldn't explain why they sound so bad..
Apparently the system was installed by builders, so there's not much chance that they had a clue what they were doing.
The Amp looks similar to some nasty Adastra P.A. heads that I've got, which have a output transformer with taps off for 4ohm, 8ohm, 16ohm, 100v (and I think 75v too!)
I would assume that if you took 100V line speakers and connected them to the 8ohm output by mistake, you would not get any sound, and if you attached 8ohm speakers to the 100V line you'd blow them up?
I'll know all this for sure when I get a chance to examine the back of the Amp, but I've just got a feeling that they may have badly installed this equipment rather than having broken something.
Cheers
Jonathan
Thanks Myhrrhleine,
That's very similar to what I was planning. My plan of action is to look at the back of the amp - confirm that the output is 8ohms and then pull all speakers out except one, and test it alone.
If it works fine on its own, I'll then arrange the speakers in a Parallel/Series combination that results in a sensible load on the Amp.
One of my main reasons for this post is actually to see if anyone can explain to me the basic principles of why a really low impedence causes the Amp to hit its PSU limits and clip the tops of the signal.
Anyone up for giving me a quick maths/physics lesson about this?
Cheers
Jonathan
That's very similar to what I was planning. My plan of action is to look at the back of the amp - confirm that the output is 8ohms and then pull all speakers out except one, and test it alone.
If it works fine on its own, I'll then arrange the speakers in a Parallel/Series combination that results in a sensible load on the Amp.
One of my main reasons for this post is actually to see if anyone can explain to me the basic principles of why a really low impedence causes the Amp to hit its PSU limits and clip the tops of the signal.
Anyone up for giving me a quick maths/physics lesson about this?
Cheers
Jonathan
Hi,
to try and simplify the analysis, I'm going to suggest we think in terms of output current from the amp.
Into 8ohm; a 100W amp will push 5Apk. and then develop 5*8 =40Vpk
Into 4ohm the same 5Apk will develop only 20Vpk and into 1ohm only 5Vpk (=9.1W into 11 speakers = 0.83W/speaker. The missing 3.4W is lost in the cable going from 0.7ohms to 1ohm).
If the amp can only deliver 5Apk then as soon as you ask it to produce more then it will distort, whether due to current or IV limiting or a bottleneck in the current coming from the PSU.
In reality the amp is probably capable of delivering more than 5Apk but the example gives you some arithmetic basis for extracting some real numbers.
Solutions; increase the load impedance by using a transformer or rewiring the speakers or changing the speaker impedance.
to try and simplify the analysis, I'm going to suggest we think in terms of output current from the amp.
Into 8ohm; a 100W amp will push 5Apk. and then develop 5*8 =40Vpk
Into 4ohm the same 5Apk will develop only 20Vpk and into 1ohm only 5Vpk (=9.1W into 11 speakers = 0.83W/speaker. The missing 3.4W is lost in the cable going from 0.7ohms to 1ohm).
If the amp can only deliver 5Apk then as soon as you ask it to produce more then it will distort, whether due to current or IV limiting or a bottleneck in the current coming from the PSU.
In reality the amp is probably capable of delivering more than 5Apk but the example gives you some arithmetic basis for extracting some real numbers.
Solutions; increase the load impedance by using a transformer or rewiring the speakers or changing the speaker impedance.
You are probably hearing a good amount of cross-over distortion.
If you overload an amp stage with a lower impedance. It will distort because the amp is biased to sound clear driving lower currents such as an 8 ohm speaker. When using under 1 ohm, the amp is nearly running in pure class B and does not have enough bias current in the output stage to smooth the crossover between NPN to PNP when driving such a hard load.
If you overload an amp stage with a lower impedance. It will distort because the amp is biased to sound clear driving lower currents such as an 8 ohm speaker. When using under 1 ohm, the amp is nearly running in pure class B and does not have enough bias current in the output stage to smooth the crossover between NPN to PNP when driving such a hard load.
Thanks for that explanation Andrew & EWorkshop1708 - That helps me understand this a bit better.
One thing I want to get clear in my mind (and I know I could probably test this when I get a chance):
When you increase the output volume of a power amp into a stable load - What increases? Voltage or Current (or both)?
The reason I ask is that I think you're saying that current increases in your last example, but I managed to blow up a piezo tweeter the other day (KSN-1141 Protected) and I believe it is because we increased the volume too much and exceeded its 35v limit (I am told it is voltage that kills piezos - we've since added a crossover to the replacement to protect it!).
Thanks
Jonathan
One thing I want to get clear in my mind (and I know I could probably test this when I get a chance):
When you increase the output volume of a power amp into a stable load - What increases? Voltage or Current (or both)?
The reason I ask is that I think you're saying that current increases in your last example, but I managed to blow up a piezo tweeter the other day (KSN-1141 Protected) and I believe it is because we increased the volume too much and exceeded its 35v limit (I am told it is voltage that kills piezos - we've since added a crossover to the replacement to protect it!).
Thanks
Jonathan
Ovation said:
Hi Jonathan,
Ohm's law says that Voltage = Current x Resistance (V=I.R). Assuming that the resistance (the piezo tweeter) is constant, if you increase the level which increases the voltage, it cannot but also increase the current in the same ratio, otherwise the equation wouldn't hold any more.
Example: You have 8V output voltage across 8 ohms then that 8 ohms load draws 1 amp: 8 = 1 x 8.
Now increase level to 16 V: Now the current is 2 amps: 16 = 2 . 8.
Jan Didden
Hi Ovation,
I am not sure since I have not read the specification sheet, but that 35V limit may be a PEAK voltage limit.
This equates to a 24.75Vac limit which in turn is about 76W into 8r.
75W into a piezo tweeter is a very high loading. I realise that the piezo tweeter is a cap equivalent but piezo manufacturers assume that their customers understand power ratings and tend to use this in describing limit conditions.
If you assume that the 35V limit is an rms voltage, then the power limit jumps to 150W into 8r and this seems unreasonable for a treble unit.
I am not sure since I have not read the specification sheet, but that 35V limit may be a PEAK voltage limit.
This equates to a 24.75Vac limit which in turn is about 76W into 8r.
75W into a piezo tweeter is a very high loading. I realise that the piezo tweeter is a cap equivalent but piezo manufacturers assume that their customers understand power ratings and tend to use this in describing limit conditions.
If you assume that the 35V limit is an rms voltage, then the power limit jumps to 150W into 8r and this seems unreasonable for a treble unit.
I guess we shouldn't go too off topic here, but what you are saying seems to tally up with what I've read/heard before.
The CTS Powerline Tweeters (KSN-1141 , KSN-1061) have a very high impedence (approx. 1000ohms so they claim) and as such claim that you can wire them directly in parallel with the woofer. They state 300W Power Handling (Which I since found out is into 4ohms, so it's actually 150W into 8ohms). These powerline devices have a built in resistor to stop the amp from oscillating and a tungsten lamp inside as a protection circuit.
As I have now learnt, it is still better to use a crossover network with these tweeters, primarily to limit the amount of voltage they are hit with. IIRC I've got a 4.7uF cap in series with a 22ohm resistor from Hot to Cold after that and then the driver. I'm told the X-Over Frequency is about 1500Hz which is fine as the Piezo goes down to 1800Hz quite happily and then drops off very rapidly without any noticeable distortion.
While I have digressed to this (and I can post in the Loudspeakers forum if necessary) - Is there a simply formula for working out speaker power requirements? So I'm going to buy a 400W 12 " Woofer - Do I simply want any Tweeter (Perhaps only 50W) that has a similar SPL to the woofer as I obviously can't buy a 400W tweeter?
Anyway - Thanks guys for the notes about Ohms law... It's a long time since I studied it and I had in my head that there are 2x Constants and one Variable! Of course, there is ONE contstant (In this case Resistance/Impedance) and TWO variables. Silly me!
So I think I'm all set to go and play with the Indian Restaurant's speakers on Saturday - I've got a decent TrueRMS Multimeter that I will take with me - All I need to find now is that it's actually 100V line and my theory about impedance is wrong
I'll post back and let you all know.
Hi Ovation,
you can have speakers of matched sensitivity. This almost makes unecessary any volume adjustments to the crossover components, although a small adjustment may still be required.
It is more normal to select units where the treble unit is much more sensitive than the other drivers and this definately requires padding down or a large series resistor added to the tweeter.
For sensitivity you require a specification that states spl/volt input (more usually 2.83Vac). If the driver is exactly 8ohms then 2.828Vac into 8ohms draws 1Watt.
Often low impedance speakers are quoted in sensitivity because this method makes the low impedance draw more power in turn making the sensitivity look better. Whereas for high impedance speakers you will tend to find spl/watt (efficiency), because that suits the manufacturer. However a manufacturer with nothing to hide and a pride in their expertise and product will quote all three parameters (efficiency, sensitivity and impedance)
you can have speakers of matched sensitivity. This almost makes unecessary any volume adjustments to the crossover components, although a small adjustment may still be required.
It is more normal to select units where the treble unit is much more sensitive than the other drivers and this definately requires padding down or a large series resistor added to the tweeter.
For sensitivity you require a specification that states spl/volt input (more usually 2.83Vac). If the driver is exactly 8ohms then 2.828Vac into 8ohms draws 1Watt.
Often low impedance speakers are quoted in sensitivity because this method makes the low impedance draw more power in turn making the sensitivity look better. Whereas for high impedance speakers you will tend to find spl/watt (efficiency), because that suits the manufacturer. However a manufacturer with nothing to hide and a pride in their expertise and product will quote all three parameters (efficiency, sensitivity and impedance)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.