The Class TD Amplifiers
- Thread starter opacheco
- Start date
Help Support GroupDIY Audio Forum:
I have never heard that term. Any references to where it is used? Is it a marketing term that one particular company came up with?
ccaudle Hello,
The Lab Gruppen company did this Technology reality in its famous Lab Gruppen FP-6400 and the same unit done for L Acoustic LA48, the same amplifier with minors differences; both TD class amplifier using a switching power amp in order to get more efficient amp and get a sound like the B class amplifier….but look like there aren’t more info in the web.
Thanks
opacheco
I don’t know if “Class TD” is a real thing. Class D certainly is. For while, Tripath (a company that has made a number of nice Class D amplifiers) used the term “Class T” in their marketing, but it was purely marketing - all the “Class T” amps were Class D.
I’m guessing that got further confused and turned into “Class TD” somewhere along the way.
More info here: https://en.wikipedia.org/wiki/Class-T_amplifier
I’m guessing that got further confused and turned into “Class TD” somewhere along the way.
More info here: https://en.wikipedia.org/wiki/Class-T_amplifier
OK, I think I see what they are doing.
It is like class H described on this page:
class G and H amplifiers
The power dissipated in the output stage is determined by the current through the output stage, and the voltage across the output devices.
A class D amplifier reduces the product of voltage and current by having the output devices either low impedance, fully on, so the voltage is low, or high impedance/off so the current is low, but the full power supply voltage is across the device.
Class H or TD gets the reduced power dissipation by reducing the power supply voltage when the signal amplitude is low, and raising the power supply voltage as the signal increases, so that the output device always has say 5V across it no matter the output voltage or current.
If the power supply voltage is changed using an analog regulator design, then the entire amplifier dissipation is still high, because now the power supply is dissipating power when the output stage is not.
Lab Gruppen just took the obvious approach of using a variable regulated SMPS for the power supply in a class H design to reduce the total power dissipation, then tasked their marketing department with coming up with something clever to call it.
![Electronics Soldering Iron Kit, [Upgraded] Soldering Iron 110V 90W LCD Digital Portable Soldering Kit 180-480℃(356-896℉), Welding Tool with ON/OFF Switch, Auto-sleep, Thermostatic Design](https://m.media-amazon.com/images/I/41gRDnlyfJS._SL500_.jpg)
Class TD mean Transition D class type, its PSU are continuously variable but no into discrete steps.
opacheco
The only descriptive info that I found about was the Lab Gruppen company:
“Class TD Output Stage
Delivering the High Efficiency of a Class D Design With the Sonic Purity of a Class B Design Amplifier.
The foundation of Lab Gruppen’s prowess in power amplification, Class TD is a breakthrough design that occupied Lab Gruppen’s co-founder Kenneth Andersson for nearly two years. Class TD combines the high efficiency of a digital Class D with the sonic purity of classic Class B designs.
The power supply rails precisely track the audio signal at all times, providing the required voltage plus additional headroom. This process is highly efficient, with only a tiny portion of the supplied voltage emitted as heat in the output stage.
The audio path itself remains entirely in the analog domain. And because the audio path remains outside the switching portion of the circuit, the signal is free from any filtering ripple effects.
Class TD is bridgeable, highly reliable, and maintains a flat response with complex loads as low as 2 ohms nominal. Also, it does not interfere with RF equipment such as wireless (radio) microphones. Though treated separately here, R.SMPS and Class TD were conceived and engineered as an integral design within the amplifier platform. Working together, this new generation of proprietary circuits produces more power from a smaller package while at the same time maintaining Lab Gruppen’s peerless reputation for sonic excellence. The highs stay crisp and transparent. The mids are defined, warm and natural. And the tight low end delivers breathtaking visceral impact.”
(https://www.labgruppen.com/technologies/class-tc-output-stage.html)
Class H designs boost efficiency by modulating the power supply voltage with the input signal, “tracking” the input in order to provide the instantaneous voltage needed for optimum operation of the output devices. However, maximum efficiency with class H is achieved only within a relatively limited dynamic range. The TD stands for “tracking class D,” which means that the power supply tracks the audio signal at all frequencies, supplying the required rail voltage while at the same time reserving additional headroom. The high-speed switching principles of class D are employed as well, though the final output stage remains a proven class AB component. (https://mediadl.musictribe.com/download/webservice/labgruppen/C-Series_Technology_Brief.pdf)
Thanks
opacheco
“Class TD Output Stage
Delivering the High Efficiency of a Class D Design With the Sonic Purity of a Class B Design Amplifier.
The foundation of Lab Gruppen’s prowess in power amplification, Class TD is a breakthrough design that occupied Lab Gruppen’s co-founder Kenneth Andersson for nearly two years. Class TD combines the high efficiency of a digital Class D with the sonic purity of classic Class B designs.
The power supply rails precisely track the audio signal at all times, providing the required voltage plus additional headroom. This process is highly efficient, with only a tiny portion of the supplied voltage emitted as heat in the output stage.
The audio path itself remains entirely in the analog domain. And because the audio path remains outside the switching portion of the circuit, the signal is free from any filtering ripple effects.
Class TD is bridgeable, highly reliable, and maintains a flat response with complex loads as low as 2 ohms nominal. Also, it does not interfere with RF equipment such as wireless (radio) microphones. Though treated separately here, R.SMPS and Class TD were conceived and engineered as an integral design within the amplifier platform. Working together, this new generation of proprietary circuits produces more power from a smaller package while at the same time maintaining Lab Gruppen’s peerless reputation for sonic excellence. The highs stay crisp and transparent. The mids are defined, warm and natural. And the tight low end delivers breathtaking visceral impact.”
(https://www.labgruppen.com/technologies/class-tc-output-stage.html)
Class H designs boost efficiency by modulating the power supply voltage with the input signal, “tracking” the input in order to provide the instantaneous voltage needed for optimum operation of the output devices. However, maximum efficiency with class H is achieved only within a relatively limited dynamic range. The TD stands for “tracking class D,” which means that the power supply tracks the audio signal at all frequencies, supplying the required rail voltage while at the same time reserving additional headroom. The high-speed switching principles of class D are employed as well, though the final output stage remains a proven class AB component. (https://mediadl.musictribe.com/download/webservice/labgruppen/C-Series_Technology_Brief.pdf)
Thanks
opacheco
Last edited:
For decades amplifier manufacturers have been trying to claim ownership new classes of amplification for marketing purposes.
Most amplifiers made today are variants on a small handful of amp classes; A, B, AB, G/H, D. I consider class G and class H similar enough to lump them together). There are too many variants on class D to list them all.
JR
Most amplifiers made today are variants on a small handful of amp classes; A, B, AB, G/H, D. I consider class G and class H similar enough to lump them together). There are too many variants on class D to list them all.
JR
moamps
Well-known member
The amplifier in the TD class consists of a conventional amplifier in the AB/B class with an SMPS +/- power supply whose output voltages are amplitude modulated by the input audio signal. That AM must be defined in such a way that the output voltage is always higher by 5-10V than the output voltage of the amplifier itself in the A/AB class. From a design point of view, the SMPS power supply is actually a class D amplifier and has high efficiency, so it can be concluded that the TD class does not give better performance compared to the A/AB class except for reduced heating, which gives it the best application in active or passive PA systems.
Imagine an amplifier with an IC 5534 that is powered by a +/- power supply of 317/337 that do not have a constant output voltage, but the voltage on the ADJ pin is determined by the input/output voltage of the 5534 using a PCM modulator.
I written about this here before (perhaps search PMA70+). Back in the 80s I designed a small bedroom recording monitor amp that was only capable of 35W continuous, but used a variant capacitor doubler on each amplifier voltage rail so it could momentarily deliver well over 100W peaks on brief transients. I wasn't so arrogant as to invent a new amplifier class name to call it
but it effectively used a modulated rail voltage to deliver high short term peak power. There are any number of modulated rail amps that enjoyed some success. IIRC Yamaha had one that delivered cost effective power for their topbox powered mixers. I don't remember if they tried to give it a unique name.
JR
Disco Volante
Well-known member
At what frequency? music is complex...
Reminds me of the NAD Power Envelope amps..
It's not trivial. In my design I used a conventional class A/B (TO3 devices) output stage with the modulated power rails (glorified cap doubler topology). I quickly discovered that too fast of an edge rate on the power supply rail, coupled backwards into the output stage (thru the power transistor's Cob) causing distortion. My solution was to slew rate limit the PS boost stage and add some anticipation in the circuit managing PS rail boost drivers, so I could get there in time to support anticipated signal crest factors. This cost me some efficiency but was a fair trade off to reduce distortion.
I also power managed the boost stage because I was still using a 35W continuous heat sink and power transformer. I used a PTC fuse in series with my boost stages so it would deliver 120W+ for very narrow transient peaks (I think I specified only 100W because that transient power is not for long, in theory it should have made 4x 35W but didn't long enough to claim on a data sheet. Then the PTC current limiter was sized to deliver about 60 watts for around 15 seconds before getting hot and collapsing back to the basic 35W continuous capability.
This is all academic now that pure class D amps can deliver lots of power at high efficiency and with modest cost.
JR
And of course SMPS always have (relatively) fast edges at the switching devices. It seems like it might have been tricky to have enough filtering at the SMPS output that high frequency noise did not feed through the AB output stage, but also enough bandwidth in the control loop that the power supply changes could stay ahead of the audio signal. Possibly helped by the amplitude of the signal being primarily controlled by low frequency amplitude (assuming full range signals, i.e. not using one of the amps for just the tweeter signal of a really big tri-amped PA).
Yeah, I wonder if Lab Gruppen even bothers with that topology now. It sounds like it was an attempt to work around high distortion and out of band noise in early class D output stage designs.
I thought about that some years ago (80s) but didn't attempt it. It seems you would need to delay audio to get the PS voltage you need there first. Did you do that? If so, how? Thanks
No, a quality audio delay was not practical in an inexpensive analog amplifier back in the 80s. Still not practical today with the advent of low cost class D.
There was a finite rise time for the boosted rails so the leading edge of above band transients larger than 35W could get momentarily clipped. After the initial leading edge, the boosted rail didn't decay immediately and supported clean playback. I forget the published spec but I think I made it fast enough to cleanly reproduce 10kHz tone bursts.
This was not an expensive audiophile power amp but a small 35W amp that kicked like a 100W amp. As usual there was no free lunch. It was well received by the target market (-10dBV bedroom recording).
JR
Thanks. I'm still not sold in D amps but with today's processing that concept could work really well. No output inductor needed in the audio path with a low nfb analog ckt. A real light weight high power analog amp could be had.
Maybe, depends on how much noise feed-through you get from a high bandwidth SMPS powering your output devices. Presumably less than when your output devices themselves are a high bandwidth SMPS (i.e. class D amp), but you would have to verify.
Have you ever tried Bruno Putzey's class D designs (sold under Hypex)? He has cleverly incorporated the class D output filter inside the NF loop for very good performance.
I use several Hypex amp modules in my home theater system, they sound good to me but what would I know.
When I first came up with the PMA70+ boosted rail design I was all excited about extrapolating the concept (2x voltage = 4x power) to the full line of larger amp models BUT reality reached out and bit me in the butt. While the cap doubler boosted rail allowed me to extract more peak power from the same size mains power transformer and same size heat sink, the 4x higher power still required 2x current output which meant the bipolar drivers and output devices needed to be sized similar to the continuous power versions. In other words the cost savings just weren't that significant. The low cost class D made all this moot.
The modern low cost class D is especially painful to me because I spent 15 years inside Peavey waiting for cost effective class D to use in powered mixers, and sound reinforcement amps. Peavey made and sold many thousands of amps. The low cost in-house class D amp module project was always right around the corner. I noticed some very cost effective class D Peavey amp models only a few years after I left.
JR
PS; Regarding the short term transient response of the boosted rail PMA70+ keep in mind that the 35W power supply rails were always up so at least the first half step of any transient didn't have to wait for boost to come up. In addition to that the highest slew rate of a sine wave is while passing through zero so even a high frequency tone burst was already slowing down by the time it reaches 1/2 the rail voltage. I am pretty sure I bench tested this with dynamic samples and it didn't suck, while it was a clever trick amp design for 30+ years ago, not very clever today.
