diyAudio (http://www.diyaudio.com/forums/index.php)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   10 KHz Square wave output? (http://www.diyaudio.com/forums/solid-state/58069-10-khz-square-wave-output.html)

 mura 28th May 2005 05:02 AM

10 KHz Square wave output?

Square wave output at 2.83Vrms at 10Khz into 8 ohms - Why is this parameter mesured?:confused:

What do you mean by "output at 2.83Vrms " ,

Is this the Vrms at the output or the output with 2.83Vrms input?

Why is getting a perfect Square wave important , what effect does it have audiably?

Thanks

Murarie.

 ashok 28th May 2005 05:40 AM

2.83 volts rms is generally used to specify speaker sensitivity ( not using square waves). 2.83 volts gives 1 watt into 8 ohms ( 2.83x2.83/8). But you will also see spl ratings specified with a 1 watt input. The difference is because 2.83 volts need not produce 1 watt into the speaker because the impedance need not be 8 ohms. So to get 1 watt into the speaker the voltage will be something else , if the speaker is not 8 ohms at the measuring frequency or band of frequencies.

This 2.83Volts has nothing to do with amplifier ratings.

Square wave testing is generally done with amplifiers as the square wave consists of the fundamental sine wave and its harmonics extending to very high frequencies . You can see the amplifiers low frequency and high frequency performance by looking at it's square wave response. Additionally some amps are unstable with high frequencies and with a square wave signal you can see that easily - often not seen with a simple sine wave.
That is a simple explanation. Hope it is clear enough.

 AndrewT 28th May 2005 07:34 AM

Hi,
you would also look at the 1kHz square wave and 100Hz square wave outputs at various output voltages checking to ensure you do not exceed the max output voltage.
These three square waves tell the designer a lot about frequency response and phase linearity as well as stability.

Q. is there any advantage driving the amp into clipping when displaying a square wave output? i.e. trying to exceed maximum output voltage.

 ashok 28th May 2005 12:34 PM

You should be careful while using square waves to test the amp.
Generally the amplitudes used are not very large. Full amplitude square waves can cause excessive heating of some devices especially the resistor in the Zobel network ( if it is there) and burn it . You could blow some amps running full amplitude square waves. Blow it pretty fast ! Especially if it has HF instability problems.
Be careful.

 fab 28th May 2005 01:51 PM

Hi Mura

Of course when you test a new amp you need to be careful not to blow the components but the final goal - once the design is completed - would be to have a high voltage (close to sin wave max rating and considering the power amp capability)10 KHz (or more) square wave not showing HF instability. However, when you test 10 KHz square wave into capacitive load (parallel to 8 ohms) the current driven from the amp will become excessive and that is why no one test at higher voltage than let say about 10V (depending on the current capability and ability of amp to drive high current).

 fab 28th May 2005 02:08 PM

Quote:
 Originally posted by AndrewT ...Q. is there any advantage driving the amp into clipping when displaying a square wave output? i.e. trying to exceed maximum output voltage.
Hi AndrewT

Any amp can possibly be exposed to exceed max output voltage because input voltage is too high and/or the amp current limit has been reached. When you listen to high level of sound it is hard to know if the max level is reached or not due to the constantly changing level of sound in music. Thus, the amp can clip and when it does can sound worse or better than another amp or maybe have problems (maybe latch-up of active devices). On my side I test it to ensure the design has no flaw.

 mura 29th May 2005 07:06 AM

Thanks all for your valuble inputs,

So Sq wave testing is to find out the hi frequency stability.

I use Cooledit Pro2.0 to generate sine waves and Sq waves,is this a reliable method?

In my Amp i can get a 1kHz Sq wave perfectly but a 10K Sq wave looks almost like a sine.

Any ideas why?

It is a Leach with a few Mods.....

110 V rails,Choke Regulated , Dual Mono with 2 X 1KVA transformers,

Output stage MJ21195 and MJ21196 8 pairs per channel.

Driver stage MJE15034 NPN,MJE15035 PNP.

 EchoWars 29th May 2005 07:23 AM

Quote:
 In my Amp i can get a 1kHz Sq wave perfectly but a 10K Sq wave looks almost like a sine.
What's the input wave look like? PC sound cards aren't known for their ability to generate square waves...the amp is likely amplifying an already slew-distorted signal.

 Duo 29th May 2005 07:24 AM

Interesting, 10K square should be pretty square in a good amplifier.

I was able to push my design to 200K without much degradation (certainly didn't look like a sine.)

Perhaps the computer is not outputting a good square? Try putting your scope right on the computer and see what you get, it could just be that.

Otherwise, I'd imagine something's up with the amplifier.

 paologatto 29th May 2005 08:54 AM

Quote:
 Originally posted by mura I use Cooledit Pro2.0 to generate sine waves and Sq waves,is this a reliable method? In my Amp i can get a 1kHz Sq wave perfectly but a 10K Sq wave looks almost like a sine. Any ideas why? [/B]

If you use for example a sample rate of 48 kHz, it's quite normal that a 10 KHz Sq wave looks almost like a sine because you don't have any harmonic inside the Nyquist frequency (half of sample rate).
The first odd harmonic would be at 30 kHz that is greater than 24 kHz.
So a soundcard, also a pro model, cannot generate a Sq wave of that frequency. You should use a frequency generator with some MHz bandwith for this task.

For example there are a lot of projects on the net that use the IC MAX038 for generating signals sine/tri/sq from 0.1 Hz to 20 MHz. I builded one and is perfect for testing audio gear..

All times are GMT. The time now is 12:25 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.