Tube preamp & Class D Power Amp. Is it best of both worlds?

[IndianEars]

Just my 2 cents on this debate.... I agree with Fantastic.

Yes, Kanwar, as you said

It is always stated in specs usually [Max voltage swing 30V into 600 ohms<<< like this]

Such qualified specs are the norm rather than the exception for tube pre amps.

Tube Preamps rarely operate well into impedances below 10K ohms, (they are often tested with 100K loads) even though their output impedances are typically around 300 Ohms ( at 1 KHz ).

The consequence of driving low impedance loads is typically an increased distortion of several hundred percentage ( 400% or 12dB increase in distortion is a 'low' increase in distortion for such cases).

You said

No other linear amplifier class comes closer to the reactive current drive feature of Class-D amps. They effortlessly give reactive current SINK-SOURCE and again recycle them back to the power supply.

I was unaware of this and find this a very good argument for Class D power amps.

I would GENUINELY like to understand and learn more on this.

Can you please elaborate or point me to links and papers that are informative on this ?

Many Thanks

P.S: PLEASE Do Not mis-understand this post. My interest in this topic is Genuine, and certainly not any attempt to contradict or needle you.

 

[captrajesh]

Lot of technical stuff going on here. Guys, keep it coming, arguments and counter arguments so that we can draw conclusions.

 

[asliarun]

Hi Kanwar

you said


I was unaware of this and find this a very good argument for Class D power amps.

I would GENUINELY like to understand and learn more on this.

Can you please elaborate or point me to links and papers that are informative on this ?

Many Thanks

P.S: PLEASE Do Not mis-understand this post. My interest in this topic is Genuine, and certainly not any attempt to contradict or needle you.

I second this opinion.

Kanwar, it would be a huge deal if you could share some insights into this? Why and how are class d amps superior? How are they immune to speaker impedance changes?

What do your class d amps and other amps do to handle power supply issues?

 

[Fantastic]

There is lots of info on classD but you could start with this
http://www.irf.com/technical-info/appnotes/an-1071.pdf

Note that though classD is placed under 'amplifiers' it's actual function is like a modulated switching power supply with a filter at the end. The amps output stage has signals going from rail to rail as a square wave. You can see that on page 3. It's only after filtering this signal to remove all the HF signals that you get back the audio signal. It's a very crucial filter!
Note what they said on that page " A Class D amplifier works in very much the same way as a PWM power supply ".

It's a completely different animal compared to regular linear amplifiers !
Enjoy !

 

[IndianEars]

Fantastic, Thanks for the pointer. ( IRF Application Note). It is an Excellent primer.

I am well aware of the basic operation of Class D amplifiers.

My current interest to to learn what makes the Class D amplifiers / output stage particularly suitable / capable for reactive loads.

Yes, as mentioned by you, I am also aware of the output Filter ( usually an Inductor). Its value is usually fairly critical. This fact (specific inductor value) would seem to be a concern when connected to a Loudspeaker who's impedance can vary from Capacitive to resistive to inductive ! It would therefore seem to me that a Class D amplifier will not take to well driving non resistive loads and still meeting its specs.

These are the Technical papers I am interested in. Any pointers that Elaborate on this mechanism ( rather than state it in the passing ) would be Most Appreciated.

Thanks

 

[omishra]

there is zobel network (shunted RC) at the output of LPF. LPF is LC filter and coupled with zobel it gives stable output load characteristics.

 

[Fantastic]

It would therefore seem to me that a Class D amplifier will not take to well driving non resistive loads and still meeting its specs.

Generally speaking, yes. But it's effect can be reduced a bit. Ideally it has flat response only at one impedance ( resistive). The newer ones operate at much higher switching frequencies and using smaller filter components.

Here is a paper on the output filter.
http://www.ti.com/lit/an/sloa119a/sloa119a.pdf

There are several more but this should a good place to look at the complexity of the filter performance. The inductor has to be a power inductor as the full speaker current flows through it. Trying to keep it small is tough.
There are other papers on making such inductors............and so on !

 

[saikatbiswas82]

Some says internet is a blessing. Some says its evil.
When audio related gadgets are concerned, for me its evil I believe. I browse through lot of things and that is sufficient to confuse myself which way should i go.

First of all, I would like to thank each and every one for their contribution to this thread. 'Class D' is talk of the town these days and I am very much surprised that this thread did not get passed 3 pages (or 27 replies only) so far.

When DIY is concerned, I consider myself below novice level. The reason is - I did nothing DIY in audio related areas so far. But I have plans.

To start with, I have few queries.

1. If a state-of-the art class d module which one should we go for? Keeping the budget as low as possible (e.g. around $500).
I have checked a few class-d modules which are famous amongst diyaudio members:
- a. Hypex UcD180HG with HxR x 2 with Connexelectronic SMPS300RE x 2
- b. LJM L15D Pro x 2 with Connexelectronic SMPS300RE x 2
- c. Connexelectronic IRS2092 Stereo Amplifier x 1 with Connexelectronic SMPS500RxE x 1
- d. Connexelectronic IRS2092 Modular Amplifier x 2 with Connexelectronic SMPS300RE x 2
2. If a tube buffer is introduced in chain, will this take care of the impedance matching? How good a MF X-10D clones (with 6DJ8/6922/6N11 tubes) available in ebay?
3. If a passive pre is intruduced for volume control, how good an ALPS pot compared to resistor ladder?
4. Last but not the least - how different this class-d plus tube buffer combo will sound when compared against a discrete class ab design/gainclones? e.g. Symasym, Rod Elliott P3a, MyRef, Linuxguru's Miniref, Akitika grainclone etc.

I am keeping 'CLASS-A' out of comparison for the sake of less confusion.

 

[Fantastic]

A very quick reply to some queries.
First , yes the Net is a blessing and an evil ! So beware .

I have a Hypex UcD 400. never switched it on for the past few years. Mainly as I didn't have the power supply and as time passed it was almost forgotten . But I will do it one of these days.
I have the L15D Pro kit. I assembled it but hasn't been switched on yet. The speaker protection and dedicated supply is still being constructed !
Don't know about the Connexelectronic unit. But I have another IRS 2092 unit. Didn't sound all that great.
Some classD amps have a fairly low input impedance. You need to check this out. My SMSL unit is quite high impedance but that could be due to a buffer.
Units without volume controls may just have the basic input circuit that often is low impedance. Any buffer, Tube or SS will help as the classD would need a low impedance source. Single tube units have fairly high output impedance and may not be an ideal match. Two stage tube preamps are better as the last stage is usually a cathode follower with low zout.
Ladder resistors are more accurate wrt to position and tracking of channels than a rotary pot . Do they sound better ? You better try them out yourself !

The last question has no clear answer and don't ever ask it again .... They all sound different in subtle ways and depends on the quality of the recording and speakers you use...and the power supply of the power amp ! So you have to try what you want as it will sound slightly different in each location due to many variables ! But adding a tube buffer at the input will severely degrade the measured specs of the classD amps ! That doesn't mean it will sound bad ! It is still nice but your measurements of the classD by itself will be lost. Say tube amp distortion at user levels about 0.05% and more! classD distortion 0.000X% or less ! Signal to noise ratio is much worse for tubes mainly due to supply hum ingress. But if kept low it will not affect you at listening levels.

Now I have to run. many things to get done.

 

[saikatbiswas82]

@fantastic...thanks a lot for taking your time and answering my queries.

I am planning to get my hands dirty with some gainclone first. Let's see where I end up with.

As per hypex specs, input impedence is 100k.

I am less concerned about the measurements. I am more focused towards the overall sonic quality.

I think I have to build different amps to figure out which does well with my current setup and ears.

 

[Fantastic]

I'm back. Hope that message of mine wasn't too mixed up. I wrote that in a hurry.
I have had a classD from 40Hz.com and they are Tripath chip based and did sound good. Hypex had a problem when I got it. The dealer sold me a different model and eventually exchanged it for what I have now. So due to the large time gap in getting it changed , it lost it's position on the 'to do list'.
With a 100K input impedance, a single tube is useable. But do remember that not all tube preamps sound great ! Some are better than others ! It depends on the tubes, brands, parts selection and pcb design AND the power supply. Try to read reviews of what you want to buy. Generally speaking they need not be 'tubey sounding'.
However I like a tube front end for several technical reasons. Without negative feedback they are a compromise on distortion and noise specifications. Like all circuitry , tubes amps come in different 'flavours' and you might develop a liking for a 'certain' configuration over others. There is no 'best' either.

It's like this. They do comparisons between different tube types on Youtube. Is it to pick the best ? No, it is to show case the sound of different tubes and it's up to you to pick what YOU like better ! Read the comments and you will find several people like different tubes according to their own taste ! There is no real 'best' ! But you could run a FFT on the tubes and pick ones with numerically better specs and call one the best. But that is what your measurement system 'likes' not 'your ear' !

 

[saikatbiswas82]

Re: Tube preamp &amp; Class D Power Amp. Is it best of both worlds?

Thanks for the explanation.
In my previous post I missed out B&O ICEPower amp completely.

I will try to read as much as I can regarding tubes. Fm bibin is already in process of making a new tube preamp. Waiting for his feedback too.

Edit: just checked bibin's thread and found your posts as well. [emoji3]

 

[saikatbiswas82]

Contributing to this thread after 2.5 years. Meanwhile I built a couple of class a amplifiers (ampcamp and jlh). Took a plunge into tube buffer and preamps. Recently tested some IRS2092 based amps (L15D Power by LJM). I loved the combination of 6J1 based tube pre+class d power.

 

[humblebee]

Tube preamp with class ab power amp.
I am also looking to buy something similar..have decided upon mcintosh ma252