I don't know if someone had this ideea before or if will really work, but is nice to think about it.
I'd like to make some input first. Everyone knows that class D amplifiers are best energy efficient, and have a pretty good response in frequency too and they don't heat much.
On the other give a "nice" saturated, compressed "round" sound.
!!! I wonder what if we combine class D circuit (4ex. LM4651 ) as first stage with power tubes as finals?
Practically we combine the oldest technology with the newest. It will be really nice to have someone who would want to try it! We may have all a big surprise...
ps:interesting compromise, isn't it?
I'd like to make some input first. Everyone knows that class D amplifiers are best energy efficient, and have a pretty good response in frequency too and they don't heat much.
On the other give a "nice" saturated, compressed "round" sound.
!!! I wonder what if we combine class D circuit (4ex. LM4651 ) as first stage with power tubes as finals?
Practically we combine the oldest technology with the newest. It will be really nice to have someone who would want to try it! We may have all a big surprise...
ps:interesting compromise, isn't it?
fdegrove
Maybe u're right, but I stil don't find any power class D amplifiers, and all of them (I know few) are pretty low in power handeling.
I forgot to mention that class D amplifiers compress sound a little bit, just almost like the tubes and so they have this common feature. I would ask u to tell me if there are tubes capable of switching?
Thanks, we should think about this; I'll just wait for a professional opinion.
Maybe u're right, but I stil don't find any power class D amplifiers, and all of them (I know few) are pretty low in power handeling.
I forgot to mention that class D amplifiers compress sound a little bit, just almost like the tubes and so they have this common feature. I would ask u to tell me if there are tubes capable of switching?
Thanks, we should think about this; I'll just wait for a professional opinion.
I think it could be done without transformers quite easily using same tubes that are used in an OTL amp but the benefits would not be the same as in a SS circuit.
Also I fail to see the benefits of building class D amplifiers from a sonic perspective, class D main benefits are small size and high efficiency, there are no benefits from a sonic perspective but in many cases instead sound is suffering due to improper implementation.
Of course there are people arguing that class D amplifiers sound so much better than anything else but that is the same exaggurated sales talk that was used to sell the first transistor amplifiers, "tubes are obsolete! buy our new modern transistor amplifier" We all know what happened, when most of the transistor hype had calmed down some people who where not afraid to say what they really thought said "Hey, my old tube amp sound better than this modern stuff that shoul have been so more advanced" and people started to be interested in tube amps again.
I predict that the same thing will happen with class D, in many applications they will be dominant as non-switching SS technology is today but in those cases where the ultimate in sound production is important there will be hard battle between tubes, non-switching SS and class D.
I also predict that there always be people that fall for all the hype and sales talk for any new tecdhnology and fool themselves into believing that everything new and modern is superior, that is OK , it would be very difficult to introduce new technology without these kind of people driving the market.
Regards Hans
Bah. Tubes are only useful for quality class A or AB sound and the occasional kW class C radio using them purdy 3-500Zs or such. 
But since you want a design challenge...
The later sweep tubes certainly are good at switching, that being their designed purpose after all. With a saturation voltage as low as 30V while passing 500mA one gets the same performance as a transistor whose Vsat might be around 1.5V and 10A. Ran from +300V, the tube gets a max. of 300-30 = 270V across a load, drawing .5A, putting 135W into the load and only 15W in the plate. Said transistor, from +15V, would get 15-1.5 = 13.5V at 10A = 135W into the load and 1.5*10 = 15W collector dissipation. Unfortunately the heater dissipation is equal to plate dissipation here, ruining total efficiency, and let's not even touch peak screen current.
But you run into the problem of coupling the high voltage signal to the speakers. 1/2A isn't any good to an 8 ohm speaker. You can't just transform it because the supposed advantage of class D PWM is small size -- the carrier will easily transmit through a HF transformer but the audio will be lost as operating current because the transformer ignores anything so slow. To do it this way you'd have to use a regular OPT for impedance matching and do the carrier filtering before it. Although you could count it as part of the speaker's weight, that's not fair.
Tim
But since you want a design challenge...The later sweep tubes certainly are good at switching, that being their designed purpose after all. With a saturation voltage as low as 30V while passing 500mA one gets the same performance as a transistor whose Vsat might be around 1.5V and 10A. Ran from +300V, the tube gets a max. of 300-30 = 270V across a load, drawing .5A, putting 135W into the load and only 15W in the plate. Said transistor, from +15V, would get 15-1.5 = 13.5V at 10A = 135W into the load and 1.5*10 = 15W collector dissipation. Unfortunately the heater dissipation is equal to plate dissipation here, ruining total efficiency, and let's not even touch peak screen current.
But you run into the problem of coupling the high voltage signal to the speakers. 1/2A isn't any good to an 8 ohm speaker. You can't just transform it because the supposed advantage of class D PWM is small size -- the carrier will easily transmit through a HF transformer but the audio will be lost as operating current because the transformer ignores anything so slow. To do it this way you'd have to use a regular OPT for impedance matching and do the carrier filtering before it. Although you could count it as part of the speaker's weight, that's not fair.
Tim
distortion
My knowlege base says that the main problem with class D is that they distort much more (especially @ HF) than other classes of amps. Now this defective effect it's because of the speed of switching. But the switching is done by the final active parts in such amplifier. When I say class D I refer basiclly to the comparator (4 ex. LM4652) and with this comparator who gives actually the benefical effects of this class type and I think few of the negative ones, replacing the MOSFETs with the tubes as final stage will only sound better than some MOSFETs, because tubes are more self controlled than transistors, especially @ high frequencyes where the problems can appear. All that I want now to findout is a serial no for some tubes who have comparable switching speeds with traditional silicon devices.
Any other opinions?
My knowlege base says that the main problem with class D is that they distort much more (especially @ HF) than other classes of amps. Now this defective effect it's because of the speed of switching. But the switching is done by the final active parts in such amplifier. When I say class D I refer basiclly to the comparator (4 ex. LM4652) and with this comparator who gives actually the benefical effects of this class type and I think few of the negative ones, replacing the MOSFETs with the tubes as final stage will only sound better than some MOSFETs, because tubes are more self controlled than transistors, especially @ high frequencyes where the problems can appear. All that I want now to findout is a serial no for some tubes who have comparable switching speeds with traditional silicon devices.
Any other opinions?
lol... I forgot to add:
So the circuit drives the sitching tubes who now can operate more efficiently than in pure tube amps. And there we have it. If I'm assured that tubes can lower the distortion level (or keep distorting but in their own way - the good one "the ordered harmonics"), the results will be more than great!
So the circuit drives the sitching tubes who now can operate more efficiently than in pure tube amps. And there we have it. If I'm assured that tubes can lower the distortion level (or keep distorting but in their own way - the good one "the ordered harmonics"), the results will be more than great!
The transformer can be as small as in a typical PC supply. You only need to transform the carrier that is pulse width modulated by the audio.
Compare with the PC supply, you create a carrier by chopping 300V DC and pulse width modulate it with the DC level you want at the output.
DC is slow, very slow
I know it's silly to design a class D with tubes but I see it as a revolt against technology changes that are not always for the better.
Hans is right but the difference now compared to the 60's is that we now know that older technologies will survive. We declared the death of the tubes in the late 60's. We are not going to declare the death of linear semiconductors this time.
My prediction, boomboxes and mainstream audio like HT systems will have class D in a near future.
Compare with the PC supply, you create a carrier by chopping 300V DC and pulse width modulate it with the DC level you want at the output.
DC is slow, very slow
I know it's silly to design a class D with tubes but I see it as a revolt against technology changes that are not always for the better.
Hans is right but the difference now compared to the 60's is that we now know that older technologies will survive. We declared the death of the tubes in the late 60's. We are not going to declare the death of linear semiconductors this time.
My prediction, boomboxes and mainstream audio like HT systems will have class D in a near future.
Jax said:
Please check out the effect PWM has on an "AC" signal then get back to me on this statement.
Compare with the PC supply, you create a carrier by chopping 300V DC and pulse width modulate it with the DC level you want at the output.
DC is slow, very slow
It is also demodulated with a rectifier and due to the capacitors and load, any attempt at coupling audio through the system would be incredibly inefficient.
AC and DC are different beasts. They cross over at some points but this isn't one of them.
No doubts here. There are already kits with little 8 pin chips that give 10W output, IIRC.
Tim
Sch3mat1c said:
A constant amplitude square wave with varying duty cycle. We are chopping a DC source here.
Yes, the easiest way to demodulate the PWM is a diode. Choose a low pass filter after the diode with a suitable time constant and you will see the audio. You will have to remove the DC component of course.
Efficient? Not really but that was not my point.
You could also just feed the PWM directly to the speakers and use the mechanical mass as filter.
Not recommended though.My point was not to come up with an efficient class D amplifier with tubes, just that it is possible and with a large dose of humour.
Yes, I think that it would be possible to build a class D amp using tube that in some ways is better than when using SS, (as an example tubes have no reverse recovery charge problem so reverse recovery is more or less instant and not really dependant on the driving circuit as in a bipolar SS cirtcuit, compare schottky devices).
but WHY? why do anyone want to build a class D amp if the goal is sonic pleasure? it is much easier to build a tube or even SS amp that sounds good than trying to build a class D amp giving the same sound quality, there is nothing in a class D amp that make it sound better than an analogue amp but there are a lot of problems with not ideal switches and filtering problems in a class D amp that affect the sound in a negative way so why build one if not needed in the first place?
Regards Hans
That's half the point of DIY!! It's like sitting by an open fire is not the best way to keep warm but it can be enjoyable. It's an end in itself. A class D tube amp is definitely not a me-too, cookie-cutter, plastic, commodity, seen-one-seen-them-all piece of stuff.Jax said:My point was not to come up with an efficient class D amplifier with tubes, just that it is possible and with a large dose of humour.
I say do it!
tubetvr
quote: "why do anyone want to build a class D amp if the goal is sonic pleasure? it is much easier to build a tube or even SS amp that sounds good than trying to build a class D amp giving the same sound quality[...]"
Because if this way (class d+tubes) if it can sound at least at same quality, we can obtain some important economical benefits. Anyway all the world is trying now to make things as small and low energy as it can.
In other order of ideea, i found some switching tubes that could do the job. They're called "Trochotron" VS10G. The bad part is that they're used in nuclear bombs too
, so I think there could be a dificulty in buyng one of those. Maybe one of you heard about this kind of tubes, but here's a link 4 the others: http://www.tubecollector.org/vs10g.htm .
4 Profi: "can this work?"
quote: "why do anyone want to build a class D amp if the goal is sonic pleasure? it is much easier to build a tube or even SS amp that sounds good than trying to build a class D amp giving the same sound quality[...]"
Because if this way (class d+tubes) if it can sound at least at same quality, we can obtain some important economical benefits. Anyway all the world is trying now to make things as small and low energy as it can.
In other order of ideea, i found some switching tubes that could do the job. They're called "Trochotron" VS10G. The bad part is that they're used in nuclear bombs too
, so I think there could be a dificulty in buyng one of those. Maybe one of you heard about this kind of tubes, but here's a link 4 the others: http://www.tubecollector.org/vs10g.htm .4 Profi: "can this work?"
more details about "trochotron"
here's a little more about VS10G:
www.mif.pg.gda.pl/homepages/frank/sheets/022/v/VS10G.pdf
here's a little more about VS10G:
www.mif.pg.gda.pl/homepages/frank/sheets/022/v/VS10G.pdf
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.