diyAudio (
-   Class D (
-   -   non-isolated amplifiers for active loudspeakers (

Davey T 28th October 2008 09:29 AM

non-isolated amplifiers for active loudspeakers
I was thinking...

You can rectify anf filter 240V AC to get a 320Vdc rail which could possible go straight into an half-bridge output stage to give you a nice 1-3k amplifier for a high power active PA subwoofer.

If this amplifier is (hermetically?) sealed INSIDE a loudspeaker, and there are no metal parts or connections to the outside world. AND the signal is electrically isolated too...AND the speaker is not visible (horn loaded) AND there is no user servicing or access...

Can you live without an isolating transformer? You would save on weight, cost and efficiency. Electrially its simple. but the big question here is health and safety.

dimmers and CLFs dont have transformers, neither do some old tvs. So what do you think?

acid_k2 28th October 2008 12:33 PM

hi Davey T

that amp is still really dangerous

how do you think to connect input signals?
how do you think to insulate the whole speaker? (do you imagine a fault on the voice coil?)

check this thread:

richie00boy 28th October 2008 01:27 PM

He did say the signal would be isolated, presumably by a transformer would to mains isolation specifications (don't know if normal signal transformers are).

Yes I think it would be viable for a home use only if it is impossible to touch the drive unit. But just think about your own testing and electrical measurements of the unit. If it is for PA application I would not dare as you have lots of other people to think about and things get abused in the PA world, also some water or other liquid might get on/into the cabinet which could be touched.

Eva 28th October 2008 01:45 PM

This is a very demanding project and should not be considered DIY. I'm working on similar designs and all the development and testing has to be done isolated anyway.

The main problems are:

- Input signal must be mains-isolated with litttle or no degradation. This is quite a difficult task, particularly because a huge CMRR of 140dB or so is required in order to reject those 230V all the time across the isolation barrier and still get humm and buzz close to or below the noise floor. In my experience, neither analog optocouplers nor signal transformers provide satisfactory CMRR, at least not without going to complex application circuits. Input to output capacitance is really evil.

- The speaker driver and its mounting hardware should be designed and certified to provide mains isolation.

- Those 350V DC are way too high for any linear amplifier topology, and obtaining voltage taps for class G or class H would require a transformer anyway.

- Those 350V DC are quite on the high side for class D too. MOSFET with Vds ratings over 200V are not suitable for hard switching applications due to body diode charge storage. Losses in the output inductor at these voltage levels become a problem too, and in general, the higher voltage results in an high idle power consumption. Special solutions are required whose drawbacks may be higher conduction losses, higher distortion, higher propagation delays and losing the ability for full range operation.

Davey T 28th October 2008 02:24 PM

Thanks for the replies

Ideally i would like to stream the signal wireless-ly. We're talking subwoofers so thats very low frequency and therefore low bitrate/bandwidth stuff. Failing that, digital IR would be the way to go perhaps.

The driver for my application is horn loaded sub so the driver would sit in the middle of the cab away from fingers. The front of the cab would have to have a grill and foam to avoid moisture/drinks inside.

Of course the output stage would be class-d. because the bandwidth is 200hz max, you can have a lower switching frequency and less switching losses with high voltage mosfets. Perhaps a PFC buck converter prior to the output stage to step down to 150vdc or something. In essence - the whole amp would be a 2-stage buck i guess.

As i said, electrically i think it can be done without too much difficulty. The issue is Health and Safety.

someone once told me that BOSE tried to do this but couldn't pass HSE due to moisture/condensation.

Any standards or literature you can point be towards regarding H&S and mains isolation?

Of course all testing/development will be done with an isolation transformer (1:1) prior to the unit. Only once its sealed up with that be removed.

OH, and i don't recommend anyone does this and i'm not planning to either. I'm just wanting to discuss whether it COULD be done with proper care taken.

Workhorse 28th October 2008 02:39 PM

Using new generation HI-voltage600V HI-speed IGBTs with inbuilt FRED, One can think for Class-D amplifier without any problem with SWFreq around 120kHz, if the application is dedicated towards Subwoofer or LF arena. One can also add PFC to regulate the DC voltage rail and get optimum effciency from mains.;)

Its would be nice powerpak arsenal of LF Thunderbolt, lets say 4 channels of 6KW each Class-D amps in 1U rack space. But norms would be a nightmare to pass, but imagination certainly has no limits.:D

Pafi 28th October 2008 02:48 PM


This is the key:

losing the ability for full range operation
It is for low freq only, so you can use lower freq, and IGBT. It works perfectly. 10 W idle loss in a quite simple 1.2 kW amp, fsw=160 kHz. Efficiency is around 94 %.

lumanauw 28th October 2008 02:52 PM

Hi, Davey T,

We close threads about powering amps direct from live line. This is dangerous and lethal practice, so it is not discussable in here in DIYaudio.

Even if you don't get electric shock, if anything happen inside sealed loudspeaker box (Wood or MDF or plastic box), you are making a fire furnace from inside. ClassD and SMPS can make fire and smoke in miliseconds.

For now this thread is closed

anatech 29th October 2008 09:50 PM

Closed for good, gone soon
This is deadly dangerous. This practice is not to be encouraged ever, and so this thread is closed permanently.

Don't even consider building anything like this. You must use an isolation transformer of some kind. Switching power supply, or even a 1:1 mains isolation transformer would make this a safe project as long as there is complete isolation from the AC mains. Otherwise the shock hazard to too high.

One other thing to consider here. Your house insurance may be rendered null and void if they were to find out a DIY device like this existed on your premises.

All threads using a direct AC mains connection to circuitry will be removed. This is nothing new here. DO NOT attempt to continue this discussion anywhere here on this site.



All times are GMT. The time now is 12:10 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio