Hello all,
It seems that bridging Class-D amplifiers is on the list of several members here.
There are some posts where members have said that the amplifier works well as a mono channel but in bridge mode it does not work as well.
White papers on UCD also mention that there are no patents on the bridge topology as it is a complex subject.
I thought of an old fashioned approach that could sort things out...and if it does, will help in a lot of a applications.
What if we employ an HF transformer at the output stage and follow that with the LPF?
Transformer output will do two things: 1. Impedance matching 2. Isolation
Unlike linear amplifiers, the output transformer can be small in Class-D topology.
Not only will the isolated output ease bridging but the output voltage can be adjusted to give 100/70V line in applications that require it.
I am not an expert in power switching or magnetics and so would like to invite the members here to give an opinion. I hope that I have not missed anything obvious, if so, my apologies in advance.
regards,
bimbla.
It seems that bridging Class-D amplifiers is on the list of several members here.
There are some posts where members have said that the amplifier works well as a mono channel but in bridge mode it does not work as well.
White papers on UCD also mention that there are no patents on the bridge topology as it is a complex subject.
I thought of an old fashioned approach that could sort things out...and if it does, will help in a lot of a applications.
What if we employ an HF transformer at the output stage and follow that with the LPF?
Transformer output will do two things: 1. Impedance matching 2. Isolation
Unlike linear amplifiers, the output transformer can be small in Class-D topology.
Not only will the isolated output ease bridging but the output voltage can be adjusted to give 100/70V line in applications that require it.
I am not an expert in power switching or magnetics and so would like to invite the members here to give an opinion. I hope that I have not missed anything obvious, if so, my apologies in advance.
regards,
bimbla.
One very obvious error in your thought. Class D is a rail to rail switching PWM amplifier. A transformer will only see the rail to rail frequency and level that doesn't vary. The PWM will not cause any difference in the flux of the core and therefore will not transform any audio frequency modulation. If you wish to bridge the output from two amplifiers, ensure the frequencies are 100% locked to avoid distortion.
The perfect opposite is true. The PWM definitely changes the flux, since flux is the integral of the voltage. Flux changes so much that a small core will saturate.
Well, actually the interference caused by cross-coupling 2 ClassD amps has 2 forms: when frequencies are different, then it's whistling, and if the frequencies are locked, then noise and/or harmonic distortion. This is the real challenge in bridging 2 ClassD amps. The interference should be kept very low even without synchronisation, this way THD will also remain low when synchronised.
I was looking for a way to reduce the size of the output transformer for 100V line output application. But then that is not possible.
Theoretically I could eliminate it completely if I have a bridge amplifier operating on single supply of say 142Vdc. I then land up with an amplifier that should be able to deliver around 1250W/ 8E
Theoretically I could eliminate it completely if I have a bridge amplifier operating on single supply of say 142Vdc. I then land up with an amplifier that should be able to deliver around 1250W/ 8E
More popular and easily available are unbridged designs running an 8R load from rails in the category of +/-65V...+/-85V.
In bridged mode they could serve 16R with voltage levels of a 100V system.
Some of these types are even fine for 4R unbridged - allowing 8R bridged.
I did a class D circuit in LTspice that uses just one amplifier to drive two outputs that it has . It drives them 180° out-of-phase.
I had gotten an early-era car amplifier that fed the speakers through output transformers. It was most likely done because the designer was familiar with that now-suboptimal method just to get some output, and much bigger transformers would have been needed to get high power output.
I did simulating of tracking power to an amplifier and came to conclude a good way to do it chops already rectified and filtered power rails to output transistors in an amplifier .
I did get a true ampliverter to work, but it was a bit complicated and not as good as the other way I mention above.
I had gotten an early-era car amplifier that fed the speakers through output transformers. It was most likely done because the designer was familiar with that now-suboptimal method just to get some output, and much bigger transformers would have been needed to get high power output.
I did simulating of tracking power to an amplifier and came to conclude a good way to do it chops already rectified and filtered power rails to output transistors in an amplifier .
I did get a true ampliverter to work, but it was a bit complicated and not as good as the other way I mention above.
Hello Chocoholic,
Some background is needed here from my side:
I have 20W and 80W composite current drive amplifiers based on LM1875 and TDA7293. These are fully analog power stages and use a linear unregulated power supply. I use it with a sigmaDSP processor and these are JUST AWESOME!
I am very satisfied with these, so much so that I have little motivation to do anything further in linear amplifiers.
I want to jump power levels and hence this NEW interest in Class-D. So I admit I am a newbie in Class-D (and SMPS)
I will continue to use linear power supply till I develop a stable Class-D module and then turn to SMPS for the Class-D.
So for high power (say 1250W/ 4E) Class-D amp with linear split supply, the 50/60Hz transformer cost, bulk and its utilization factor become critical.
Using a single supply helps the power supply transformer but I then need a bridge amplifier.
I am looking around in this forum to check what topology and what key components I should consider.
I have looked at your SystemD_LiteAmp and SystemD_2KW designs and have been very impressed.
I wonder if you have any plans to take up to design a full-bridge single supply amplifier.
Here's what I came up with, based on the famous UcD circuit. It's food for thought maybe at least.
View attachment ucd-H-bridge.asc
You're welcome!Sorry, there is nothing like this in the pipeline.
Played over the last years too much with classD amps, now going
to some other playgrounds for a while...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.