Setting limiters for a Behringer EP2500: Reality check

[whines]

I'm setting up a preset on the DCX to use my old EP2500 amp to run the T39's (3012lf) off of, one per channel in mono (not bridged) mode.

The weirdness I'm seeing is that when I hooked up the voltmeter and ran a 50hz sine through it, then opened up the gain to full, it hit 100v. This is higher than I was expecting...the lookup tables in the sticky on this subject suggest that the 550w into 8 ohms should give me 67v. (http://speakeressentials.com/pages/8-oh ... okup-table)

I suppose it's not a huge deal, since at worst I'd be over-limiting it--assuming the 100v is spurious and 60v is reality when using an actual speaker--but I'm confused by what I'm seeing. I don't need to put any sort of dummy load across the output (aside from the voltmeter, set to measure AC) to get an accurate reading, do I?

[Bas Gooiker]

I don't need to put any sort of dummy load across the output (aside from the voltmeter, set to measure AC) to get an accurate reading, do I?

Yes, to get an accurate reading you would need to use a dummy load. But Bill advises you to do it with nothing connected as an extra safety margin i guess...

[Rune Bivrin]

If it was clipping the average voltage would be higher. What your volt meter is showing depends on what it's actually reading - peak, average, RMS.

And within the amps linear region it won't matter much if you have a load or not. However, the lower the load impedance gets, the smaller the linear region.

[byacey]

In bridge mode, 100V RMS doesn't sound unusual for that particular amp; why do you want to run in bridge mode? If I recall correctly, the EP2500 will run up to at least 60V RMS on a single channel.

Edit: Sorry, I misread your original post; disregard my question about bridging.

The linear operating region of the output devices has little to do with setting limit voltage, as long as it's not running into clip. The reason for using a dummy load is to take into account power supply sag when the amp is under load.

[SoundInMotionDJ]

I'm setting up a preset on the DCX to use my old EP2500 amp to run the T39's (3012lf) off of, one per channel in mono (not bridged) mode.

The weirdness I'm seeing is that when I hooked up the voltmeter and ran a 50hz sine through it, then opened up the gain to full, it hit 100v. This is higher than I was expecting...the lookup tables in the sticky on this subject suggest that the 550w into 8 ohms should give me 67v. (http://speakeressentials.com/pages/8-oh ... okup-table)

Try with a 1kHz sine wave. That is what the amp is rated with. You can NOT use that to set limiters for the sub...I use 60Hz....but this is to just measure the amp and get a "gut check." I get around 70v when I measure my EP2500's with a 1kHz signal.

Keep in mind that the amp has an input sensitivity. 1.23V (+4dBu) will produce a "full" signal. How large an input signal are you sending to the amp. You can measure that with the same volt meter.

Keep in mind that the amps rating is more a function of the amount of current it can provide and the capability of the heat sinks to keep up with the load on the amp.

I don't need to put any sort of dummy load across the output (aside from the voltmeter, set to measure AC) to get an accurate reading, do I?

Solid state amps do NOT require a dummy load.

If it was clipping the average voltage would be higher. What your volt meter is showing depends on what it's actually reading - peak, average, RMS.

The specific setting of the volt meter can make a difference. RMS is 0.707 of the peak value. So, for a peak of 100V, the RMS would be 70V. This is "close enough" to make me wonder what your volt meter was really showing.

Find another volt meter and double check the reading.

[byacey]

I checked my schematics of the EP2500, but they don't indicate what the rail voltages are; I suspect somewhere around 86 volts. In my mind I thought the EP2500 was a beefier amp than that.

Extrapolating from the published power output specs, it looks like that amp will just reach 60V RMS before clip across an 8 ohm load.

[byacey]

Solid state amps do NOT require a dummy load.

They don't require it, but it will give a much more accurate indication of output voltage when properly loaded. If you measure the output of an amp, note the reading, and then apply a load under the same signal conditions it will be found that the voltage will drop a considerable amount.

Some of the newer amps that use a switch mode power supply will probably vary a lot less between full load and no load, due to the power supply regulation.

[Tom Smit]

Somewhat related....a year and a half ago, I put pink noise through an EP4000 with the gains wide open (no load) and got wall voltage (120v). Weird. But, there was no current flowing, so.... I just thought it was interesting. Anyway, I just used the DCX to limit down to 50v for the Lab12.

[SoundInMotionDJ]

Solid state amps do NOT require a dummy load.

They don't require it, but it will give a much more accurate indication of output voltage when properly loaded. If you measure the output of an amp, note the reading, and then apply a load under the same signal conditions it will be found that the voltage will drop a considerable amount.

That solid state amps are constant voltage devices is not really up for debate.

If the input is held constant, and the load and resulting power draw is within the rating of the amp, and the load is linear with time....the amp should continue to produce the same voltage into any load. If not, the amp is defective and should be replaced.

For purposes of setting a hard limiter, a solid state amp can be accurately measured without a load.

[byacey]

I beg to differ. If you don't believe me, try it yourself. The EP2500 in particular is a typical unregulated supply.

I've repaired literally hundreds of power amps of all makes and sizes, so I'm speaking from experience. I load test and measure every one of them as a final test.

Edit: I should add that this sag is much more pronounced near the limits of the amp, when it's drawing the most current. When the rail voltages drop due to high current demand, the clip point also drops by the same amount. If you are trying to set your limiters at 60V, and that's near full output of the amps capability without a load, you may find the amp is already clipping at 56 volts under load, before the limiters even start to clamp the signal.

[Rune Bivrin]

I beg to differ. If you don't believe me, try it yourself. The EP2500 in particular is a typical unregulated supply.

I've repaired literally hundreds of power amps of all makes and sizes, so I'm speaking from experience. I load test and measure every one of them as a final test.

Edit: I should add that this sag is much more pronounced near the limits of the amp, when it's drawing the most current. When the rail voltages drop due to high current demand, the clip point also drops by the same amount. If you are trying to set your limiters at 60V, and that's near full output of the amps capability without a load, you may find the amp is already clipping at 56 volts under load, before the limiters even start to clamp the signal.

Well, there's the rub. If you want to set the limiters and be absolutely sure about what you're measuring you need an oscilloscope. A DVM just won't cut it. But fortunately that's only an issue if you're running the amp close to capacity, where the load of a speaker might result in the power supply sagging enough for the amp to start clipping. As long as there's a reasonable margin there should be no difference between output voltage open circuit and under load. That's the simple result of the negative feedback in the circuit employed to reduce distortion.

But load testing after repair is absolutely vital, that is true. In that situation open circuit fails to expose a number of failure modes.

[SoundInMotionDJ]

Edit: I should add that this sag is much more pronounced near the limits of the amp, when it's drawing the most current. When the rail voltages drop due to high current demand, the clip point also drops by the same amount. If you are trying to set your limiters at 60V, and that's near full output of the amps capability without a load, you may find the amp is already clipping at 56 volts under load, before the limiters even start to clamp the signal.

Following a repair, amps do need to be tested with a load....and an O-scope. But any RMS capable volt meter is sufficient for setting a limiter for a normally functioning amp.

Even so, inside the rated limits of an amp (both in terms of voltage and load impedance)...the voltage should be constant. Even at maximum output AND minimum load....a competent amp should maintain constant voltage...otherwise the THD (Total Harmonic Distortion) of the amp will be high.

Setting an external limiter to control the amps output voltage to a desired maximum...and then running the total system BELOW the maximum output level or ABOVE the minimum load impedance....should be safe for any competent amp that is working within its limits.