Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Charles Johnson]

Hi folks,

It appears that I am (finally) going to be able to put a Yorkville A4.4 up against a QSC PLX3402 for both (rudimentary) bench testing as well as a more comprehensive listening test. Here are my plans wrt to setup:

For each amplifier:

Ch 1 - drives (1) EAW SB250
Ch 2 - drives (1) Yorkville Unity U15 in passive mode

We'll use the (Ashly 4.24C) processor to handle crossover duties as well as EQ duties (I'm thinking I'll take (and save) a transfer response curve in Smaart of the tri-amped performance of the system (it's technically a 4-way rig, but only 3-way active, since the MF and HF are passively crossed over) and then use that as a reference when EQ'ing the U15s in passive mode. Once we are satisfied with the sound, we'll then use a test tone (I'm open for suggestions here as to what frequency to use - note: both amps use switching power supplies) to match output levels and then commence with the listening portion of the test.

As for the bench portion of the test, since the amps either a) operate in the sub-500Hz region or b) would operate in the sub-500Hz region, my main interest is in their power levels down to, say, 30Hz. My question (and the prime reason for this post) concerns how to conduct those tests. A fellow LABster (Phil Ouellette), who will be helping me and participating in this A/B comparison, has a signal generator that will let us pick what frequency to send to the amp. The natural question then is, What do we use as an indicator that we've reached max output power? The amplifier clip lights? AFAIK, we've no way to measure THD+N Also, I should note that Phil has (as I recollect) (2) 400W, 8-Ohm resistors and (1) 800W, 4-Ohm resistor, so max power output tests at different frequencies will have to be taken at 8-Ohms, though I do planning on wiring up the resistors for a 2-Ohm load and connecting it up to 1ch of the A4.4 (which is only rated to 4-Ohms) and seeing just how it responds

FYI, the bench tests are primarily for 2 reasons:

1) To see how the PLX3402 responds in the lower frequency ranges - it's younger brother, the PLX1602, didn't fair all that well, as the 20Hz output was ~65% of the 1KHz output...

2) To see how the Yorkville A4.4 handles a 2-Ohm load, since you never know when an amp might die on you (*knocking on wood*) and you need to "double up" on an amplifier's load.

One other note, we plan on using a VOM to measure output voltages (though Phil might bring along his o'scope), so if the recommendation is to use pink noise (I have Bink's Audio Test CD), would averaging the measurements yield fairly good test results?

I appreciate your thoughts/comments/opinions/suggestions! Cheers!

Charles Johnson

PS A question that has been nagging me every since Bink's Amp Shootout is, assuming switching power supplies were being used, what causes an amp, when pushed towards the limits of its output rails, to produce more power at higher frequencies than at lower frequencies? Perhaps Bob Lee, John Roberts, and some others can chime in? Thanks!

[John Roberts {JR}]

Charles Johnson wrote on Wed, 23 February 2005 09:07

Hi folks, Charles Johnson PS A question that has been nagging me every since Bink's Amp Shootout is, assuming switching power supplies were being used, what causes an amp, when pushed towards the limits of its output rails, to produce more power at higher frequencies than at lower frequencies? Perhaps Bob Lee, John Roberts, and some others can chime in? Thanks!

The power supplies are charged at 2x the mains frequency (120 or 100 times per second). If you visualize sinewaves at higher frequencies they only pull from the + or - power supply 1/2 the time. At signal frequencies lower than this mains limited recharging rate. The + or - power supplies must alternately deliver more than half the current, thus the lower output.

An O'scope is the obvious choice for detecting clipping, some folks hang a piezo across the output to enhance audibility.

For evaluating an amplifiers bass behavior, you need to be pretty honest with yourself about how you plan to use the amps. There are significant differences in how amplifiers act, after they are hard clipped. Too often amplifiers are operated this way so it is not inappropriate to perform a listening test for this characteristic.

For the record, I do not endorse such operation, only I can't ignore that it goes on.

Finally, you might want to go back and read the threads before and after Bink's last amp shootout. There was a great deal of discussion surrounding methodology. The obvious blinding of tests, etc.

JR

[Michael 'Bink' Knowles]

For monitoring distortion you can use SmaartLive to watch the frequency spectrum. If the 30Hz tone begins to show the presence of 60 and/or 90 Hz rising up out of the noise then you've broken over into distortion. Unfortunately for you it's a visual determination, not a hard number.

I would pick a different low freq such as 27 Hz or 33 Hz just so that the possible presence of 60Hz AC line freq artifacts doesn't become a factor.

To tap the speaker line for Smaart you can use a speaker-level DI into a mixer or preamp which drives Smaartlive. Parallel the DI with the speaker you are driving.

When I tested the Yorkville A4.4 I noticed that it relied on the circuit breaker for amplifier protection more than some other amps that dealt with hot signals with internal circuitry that reduced output but didn't take the amp off line. If you hit your A4.4 really hard it will shut off. I didn't test a PLX3402 but the littler PLX amp I checked out didn't shut off under hot signal conditions.

Good luck with your test!

-Bink

[Michael 'Bink' Knowles]

Quote:

To tap the speaker line for Smaart you can use a speaker-level DI into a mixer or preamp which drives Smaartlive. Parallel the DI with the speaker you are driving.

I recommend using the electronic output of the amp instead of trying to measure the speaker output with a test mic because the speaker's own disortion will very likely be much higher than the amp's distortion.

One more way to safely test ~70 volts of AC speaker power is to use a resistor network in parallel with the speaker. If you choose the right values for resistors you will get a cleaner representation of the amplifier signal vs. using the DI. You want an amp test, not a DI test.  

-Bink

[John Roberts {JR}]

I hope the irony of your advice isn't lost on the testers. Almost everything associated with a system will exhibit higher distortion than the amplifiers being singled out for testing   .

Perhaps the only other element of the signal path even more linear is the mixer, contrary to sundry opinions and marketing claims

For example make sure the sine wave source is appropriately low distortion, and don't assume that a loudspeaker's few percent of THD will mask a function generator's .5% since different types of distortion will be more or less audible.

Cheers,

JR

[Michael 'Bink' Knowles]

Quote:

...we'll then use a test tone ...to match output levels...

How about freqs that are near the middle of each bandpass? Maybe 75Hz for the subs and 175 Hz for the LF on the Yorkies. Don't use pink noise for VOM power measurements -- the readout will bounce around too much.

-Bink

[Bob Lee (QSC)]

Charles Johnson wrote on Wed, 23 February 2005 07:07

PS A question that has been nagging me every since Bink's Amp Shootout is, assuming switching power supplies were being used, what causes an amp, when pushed towards the limits of its output rails, to produce more power at higher frequencies than at lower frequencies?

Other than a filter, I don't know. Let's see if that actually happens first before trying to find a cause.

[Bob Lee (QSC)]

And to echo Bink's comment, you don't use pink noise for power measurements or to set gains; use a sine wave.

[ProSoundWeb Community]