Test Signal for max power test of Class D?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
A

AP2

Banned
Joined 2009
if you have an oscilloscope, put a probe on the output (on load) and ch2 on rail voltage. (this help you to understand some things)
Use 33-66ms burst 1Khz. if the amplifier is created along with the psu for guitar, is probabbile that has been designed for the frequency range suitable.
33ms (duration of sine 1K) , 66ms pause.
 
E

eem2am

Banned
Joined 2011
Thanks.........I note that some diy'ers are choosing 30Hz , and some 1KHz, and some want burst operation, and some not.

Is there any concensus of opinion.

Also, why 30Hz....?.....why not 10Hz?

And why 1000 Hz, ...why not 500Hz?

Does any reader know of a graph of input sine wave frequency and power draw from the power supply powering the class d amplifier?

For example, would a 30Hz sine wave into a class d amplifier make the power supply deliver more power than a 1000 Hz sinewave into a class d amplifier?

Also, why sine waves?....we are testing the power supply, not the amplifier.......why not put a square wave signal into the class d amp.......?

...surely that will make the power supply deliver the maximum power, and then we can really see if the power supply is fit to supply the class d amplifier under all worst case conditions?
 
E

eem2am

Banned
Joined 2011
Basically i am asking about testing an SMPS which is designed to supply a 60W class D amplifier.

What signal would you put into the class d amplifier to test that the power supply was well capable of doing the job reliably for a long time.?

I would say that the highest power would be drawn from the smps when a square wave signal of any frequency was fed to the class d amplifier....am i correct?
 
Saturnus

Saturnus

Member
Joined 2005
eem2am said:
I would say that the highest power would be drawn from the smps when a square wave signal of any frequency was fed to the class d amplifier....am i correct?
Click to expand...

Correct but is that a good representation of the signal that will be fed to the amp? No. It's also not that interesting to test peak output beyond a burst test as described above. It's much more interesting to test with actual representative signals as you will then test the amp and power supply's ability to react to real life signals and especially if the power supply exhibit lag in rise and fall times. Another thing you will test with real signals is the actual efficiency of the system which especially for the power supply will be vastly different from the efficiency at maximum RMS output.
 
E

eem2am

Banned
Joined 2011
Guitar amplifiers are run on average, at such low power that efficiency is not of much interest to us.

We want to be sure that the SMPS is not going to go "pop" after a few weeks of heavy use with the customer.......noone can define what is "heavy use" so we must test with the worst case square wave signal......do you agree?
 
Saturnus

Saturnus

Member
Joined 2005
eem2am said:
Guitar amplifiers are run on average, at such low power that efficiency is not of much interest to us.

We want to be sure that the SMPS is not going to go "pop" after a few weeks of heavy use with the customer.......noone can define what is "heavy use" so we must test with the worst case square wave signal......do you agree?
Click to expand...

No.

Guitar amps are generally constructed with a power supply that exhibits "sag" when overdriven.
 
B

BlacK_Chicken

Member
Joined 2005
Not Class D apmplifiers though
Click to expand...

Why not? Every PSU sags (linear and switch mode), it depends on the topology and circuit how 'deep' the rails drop.

For amplifer testing: Use a bursted Sinwave, i.e. 1kHz 10 cycles full power (near Clipping) and 50-150 cycles a very small Amplitude - ore none at all. You can adjust the burst-duration and frequency to your needs, if you know exactly what you need.

Don't forget audio- and guitar- signals have a huge crest-factor. The latter a bit less than the former - depends on the artist ;)
 
E

eem2am

Banned
Joined 2011
You are right but i wasnt aware that that was the subject of the discussion at this point.

I basically want to test if a given power supply can power a given class d amplifier....what signal do you feed the class d in doing this....to garantee reliability and quality of sound
 
E

eem2am

Banned
Joined 2011
please see this LTspice simulation of a class d amplifier......its a 4khz sine wave being "played" by a class d amp...

Here is the .txt file which needs to be in the same folder as the ".asc" simulation file for the simulation to run.....
sin1.txt download - 2shared



By the way, here is the LTspice simulation as a .txt file also...and you can simply change it back to .asc to run it in the simulator

CLASS D _80V 100W.txt download - 2shared



anyway, you can see that the load is the 8R resistor, which is imitating the speaker.......the average power over several seconds in this load is 100W, however, at the peaks and troughs of the sine wave the power is 200W......so it looks to me that to allow the class d amplifier to play low frequency notes...the power supply is either going to have to have ENORMOUS output capacitors....or be sized to supply double the actual average power of the amplifier when on maximum volume.

is this true.

So anyway, for a guitar amplifier, i would have thought that to test a power supply for a class d amplifier, you have to feed the class d amp a signal corresponding to low E on the guitar....i.e. 67Hz

is this true.?
 

Attachments

  • CLASS D _80V 100W.asc
    3 KB · Views: 56
Last edited:
A

audiomik

Member
Joined 2010
try a test frequency of 60Hz, continuous Triangle wave just below clipping at the output of your Amplifier. Monitor the Amplifier and PSU Rails with an Oscilloscope and suitable Voltmeters. Run this for the expected duration of a full session time that you will expect the musician (user) to play for. Monitor your Power Supply Voltages and very importantly the heatsink temperatures of the Power Supply as the first posts in this thread relates to PSU capability. Be very wary of high DC Voltages in SMPSU's You also have to be aware of the actual Impedance curve of the Loudspeaker Load which the Amplifier is driving. If you have a single 8 Ohm Loudspeaker for your 60 Watt Amplifier, then use a 6 Ohm resistive Load for your tests. Pro-rata for other nominal Loudspeaker Load impedances. There is a further issue with Loudspeaker loads which is the Power Factor, this reflects Power back into the Amplifier and with Class D more often into the PSU Rails. Testing at the lower Load Resistance to a certain extent can compensate for this as well in terms of PSU dissipation. This will provide you with a reasonable 'worst-case' test situation to properly evaluate both your Amplifier and it's Power Supply. You may also find short circuit tests of the Amplifier Output of use as well. When I used to build Guitar/Bass Amplifiers many years ago, I used this type of 'soak test' before any left the workshops. hope this assists Mik
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom