BillFitzmaurice.info

This is based on pond's question about long extension cords and using generators instead and whatnot...

I've heard talk that Class D amps are a little more sensitive to generators with less than perfect output. Any experience with this? I'm trying to understand why, and under what circumstances this could be true. Obviously the more efficient power draw is a plus, but it's probably not quite that simple.

I don't think AC waveforms are all that important, as with either Class D, or linear PSUs and Class A/B transistor outputs, the AC is converted to DC and smoothed before becoming an internal power rail. So square, sine, stepped sine, who cares -- right? Though, I think that's more of a UPS thing anyway -- I'm not sure if any generators actually have to reconstruct the waveform. It could still be noisy or the frequency could vary from the ideal 50/60Hz.

If I understand, a switched PSU will rectify and smooth to DC, then pulse it through a transformer at high frequency to step up to high voltage before converting it back to steady, lower-voltage, high current DC again. If that's correct, then because of the high pulse frequency, the filter caps really need to charge adequately on the (much less frequent) incoming AC peaks. This could potentially be an issue with poorly-regulated or "slow" waveforms. Does that make any sense?

From my own experience, I had a small PA gig (announcements, background music, low-volume stuff) off a dinky 1.5kW gasoline generator last year that had horrible regulation. (It's not mine, it's just what they had..) With no load, my Furman power strip refused to turn on because the voltage was near 130v. With an Old Iron amp plugged in, it would sag to nearly 80. It appeard to be cyclicle, and worse on songs with heavy kick or bass lines. The amps didn't seem to care -- they would (in theory) just clip at a lower output level, right? My DriveRack however, took exception a couple times, making a horrible racket as the PSU would brown-out, then come back to life. I was very close to packing it up when that happened the second time. The gig definitely wasn't paying enough to replace it.

So, for those using Class Ds and/or linear amps with a switching PSU ... any troubles with generators? Or is everyone just using a nice beefy diesel with plenty of reserve capacity and good line regulation? I imagine built-like-a-tank amps are more resilient regardless, so this is aimed more at the budget level IPR and XTi stuff, where they're economizing every last component.

I have Class D amps and have had issues with generators.
The issue is just as you stated, unregulated voltage. Amp manufactures give specs of min and max input voltage. Getting near either number is not going to make anything happy. A steady healthy supply of 110v-118v will make for the best.

Class D or not shite power is shite power.

Haysus wrote:
Class D or not shite power is shite power.

+1

The better (and bigger) the generator you have the better results you will have, but except for extreme cases (ie: far distance or not enuf juice), land power is still the better option...

I was thinking of starting a thread on this to the other day, good someone already did

I´ve had some real issues with Class D amps with switched powersupplys on generators.

First with the XTi 4000 I had last summer, I used it two times with generators, cheap 2.5kW gasoline gensets, not very good regulation in them, specially the one I owned, I measured it at 75Hz. 250V for a while, and the XTi sure didnt like that, it sounded horrible and couldnt get any volume, but when I got the genset to run a bit slower, at 50hz 230V, it worked alright. The we were only powering a single T48 and two 1x15 tops.
Next time I used it with a genset, a different one, more stable frequency and voltage, but still 2.5Kw, we ran with 2 T48s, and the same tops. It worked, no clipping issue, but the amps was sagging for lack of a better word, what happened was this: In the breaks of some songs, with no real bass, or only like a background rumbling, we got one certain level of output i whole system, first bassbeat comes, still that level, and then it dropped about 6db or more, both on the subs and tops (running mono, subs on one channel tops the other).
This wasent very fun.. But we found that when our small 600W smokemachine was warming up, the soundlevel didnt drop with the bass:P

The genset I had and the XTi and the voltmeter I had that measured frequency got stolen last summer, so never got to test it more than that..

And just a few days ago I tried my new amprack with the same genset I used on the second gig last summer, the rack consists of a Driverrack 260 DSP, one Lab Gruppen 1600, one 1300C and one 300 amp, so a tri-amp system (bi-amped tops)
First startup, the driverack indicates signals out as normal, the Lab 300 amp drove the compressiondrivers as normal, one channel on the 1300 drove the woofer in one of the tops, and the remaining three channels didnt make a sound, the amps indicated that they where on, and had input signal, but no output..
Restarted, then one of the 1600´s channels started running one sub, for maybe 1 minute, then nothing. The 1300 switched channels so to say.

Plugged over to line power, everything working as normal... The Lab 1600 and 1300 are both Class D amps with switched PSU, and not soo lightweight at around 8kg each. The 300 are Class AB (I think) and has a linear PSU.

I research this a bit, and I think much of the problem is that switched powersupplys simply dont store very much energy, and the generetors cant speed up and down near fast enough to be able to deliver the power the amp needs, when it needs it, and the regulators in the gensets cant really cope with such fast and large changes in powerusage.
With a linear powersupply you get alot of stored energy, and that smooths it all out, I guess, atleast I´ve never had a problem with linears.

And I think the reason why it worked better with the smokemachine running in the second example, was because the dynamic portion of the total load on the genset got a lot smaller when we had 600w constat resistive load on it. It did not however work at all the same when we tried hooking up our 2 scanners with 200W discharge lamps and maybe 100W motors. I dont really know why, I thought the ballast for a discharge light was basically a type of transformer, but perhaps it doesnt give a very resistive load?


One idea I had the other day was if it might work if one hooks up a say 2-3 kW transformer, 230-230V or 380-380, between the genset and the amps? I dont have one lying around to test with sadly

//Robin

Your problems are not in the amp, they are in the generator. Class D amps work great with the honda EU series generators and those with with power inverters which are typicaly required to further stabilize the the dsp function of the amp.

Many users on the forum (including myself) have used these with class d amps and had teriffic results.

Netwerks, thanks man! Fantastic to hear that there are generators that work good with Class D amps, with switched PSU! I still belive that is has more to do with the PSU rather than the amp itself, class AB or D.
I guess you run your XTi´s on Honda EU generators then? which size generators if I may ask?
Even with the Honda generators, shouldn´t the problem of the generator not being able to deliver increased power in big burst still exist? or how do they deal with that? If load varies the generator must still change its fuel injection, to be able to keep the same rpm, doesnt it?

Any other generator models that are known to work good with Class D amps with swiched PSU´s? I mean, there are Class D amps with ironcore transformer PSU´s also

Last Saturday, I ran SM12Vs in stereo off a Crown XLS 202, and a single Tuba 30 off one channel of a XLS 802 (ran out of pack space for both subs). Power was provided by a Honda EU 2000i. That thing r-o-c-k-e-d it. Absolutely no problems whatsoever -- until it ran out of fuel... but even then, while it was starting to sputter a little, the voltage was a perfect, consistent 125vAC according to my Furman strip.

I had serious doubts about whether a dinky 2kW generator would have enough guts, but the ONLY hesitation I saw was on power-up while the amps were charging the filter caps. A little "WHUMWhumwhumwhuhuhuh" from the generator and it was good to go. Even when we brought the levels up a bit and played some bass-heavy tracks, the regulation stayed put.

So, if you're shopping for a (relatively) inexpensive generator, get a Honda EU. Starts like a dream, too.

Robin_Larsson wrote:Netwerks, thanks man! Fantastic to hear that there are generators that work good with Class D amps, with switched PSU! I still belive that is has more to do with the PSU rather than the amp itself, class AB or D.
I guess you run your XTi´s on Honda EU generators then? which size generators if I may ask?
Even with the Honda generators, shouldn´t the problem of the generator not being able to deliver increased power in big burst still exist? or how do they deal with that? If load varies the generator must still change its fuel injection, to be able to keep the same rpm, doesnt it?

Any other generator models that are known to work good with Class D amps with swiched PSU´s? I mean, there are Class D amps with ironcore transformer PSU´s also

honda eu 2000i on a float. You can also chain these generators together for more power.

http://www.youtube.com/watch?v=-7JOTYI2RGc

WOAH. Parallel capability... That is COOL.

So, it looks like this thing doesn't necessarily use an AC alternator, but instead generates DC and then has a built-in inverter to create 120vAC. That's why the load regulation is so good, without having to maintain tight RPM control based on load.

Looks like Yamaha has an EF2000i with a nearly identical design, too.

SirNickity wrote:WOAH. Parallel capability... That is COOL.

So, it looks like this thing doesn't necessarily use an AC alternator, but instead generates DC and then has a built-in inverter to create 120vAC. That's why the load regulation is so good, without having to maintain tight RPM control based on load.

Looks like Yamaha has an EF2000i with a nearly identical design, too.

Generac does as well for >$600 usd. YMMV