A Simple 10W Direct-Coupled Class-A Power Amplifier

[JR.]

By switching artifacts are you talking about crossover distortion (spikey looking in THD product)?

Yes. And it was absent in the gritty one.

Grunge is also what I mean by grit.
Just dirty - like HF leakage through a pot.
Grunge is probably the better word for it.
(Though I tend to think of "Grunge" more as a genre.
Maybe "Grungre" would be the grunge genre. Did I invent a word?)

Yup, IMD typically sounds like muddy LF crap present beneath strong HF content caused by poor HF linearity. Back in the day a simple listening test for IMD was to jingle some keys in front if a mic.. Keys have all kinds of >20kHz energy that stresses slow circuitry.

Your output stage not only goes ohmic, but probably is ohmic all the time.

I think that it is. While in the linear region maybe approximated by Re/2 plus some change?
What's also invisible are the series base resistors, 50 Ohms, hiding inside the DRV134.

The base resistance is divided down by the power transistor's Hfe, but still significant in the context of 0.22 emitter degeneration. Beta is actually rising with higher current and higher temperature.

The apparent curved "soft" clipping could come from one output device hard clipping so the summed output signal drops 6 dB, until the other output clips. I think I saw something like that before but can't remember where just now.

Of course it should sound fine as is, and the extra complexity will detract from it's elegance that such customers will embrace.

Thanks for the compliment.
It seems ultra simple to make from something I already had.
People seem to like the transparency of the headphone amp.
radiance was one of the first to build one that I'm aware of besides me.
I dunno how many people that have bought the boards have actually used one for a line amp but I do sometimes hear from those that have used it for 'phones.
Roger uses it in the FCS Class-A to drive transformers.

The bigger output devices kick up the power up a few notches.
Glad we can get TO-3P devices with good gain and BW.
BTW it slews around 10V/uS with the DRV134 which is a little more than I get with the THAT1646 at around 8V/uS.

I used TO-220 output devices in my Peavey HP amp, but they were silly cheap so I didn't mind the overkill. Even though I drove speakers with it, I discouraged customers from doing so.

JR

[ricardo]

I like to characterize amp recovery from clipping, like how a race car handles after bumping the wall at speed. Important to race care drivers, not so much for careful drivers. At Peavey and in the sound reinforcement industry clipping power amps is more normal than not. That's why pretty much 100% of Peavey amps were designed with clip limiters that managed the behavior when overdriven to sound relatively benign (like a little extra compression).

JR, could you post a sketch of your favourite clip limiter please.

IMHO, the 'sound' of a 'good' 50W amp is entirely dominated by its overload & recovery behaviour. Many (all?) Golden Pinnae amps have appalling overload & recovery behaviour .. especially into real speaker loads.

I've only seriously investigated (with DBLT bla bla) a simple diode 'soft clipper'. It was preferred but the published THD spec would have been quite blah.

The '50W/channel' stereo amp is interesting cos if your cost constraints point to eg 200VA transformer, rewinding that 200VA item to give slightly less voltage (and hence power) actually sounds better cos the better regulation possible at the lower voltage on clipping means the PSU crud that is superimposed on the clipped signal is less.

I hasten to add, this was tried on a single design and I wouldn't expect the same behaviour with a 'good' 200W/channel amp.

[ricardo]

Back in the day a simple listening test for IMD was to jingle some keys in front if a mic.. Keys have all kinds of >20kHz energy that stresses slow circuitry.

In da 21st century with all music sources via evil digital, this test should be benign cos anti-aliasing filters.

Alas, for da zillion MHz sampling fans, Sample Rate Converters often have atrocious downsampling filters. The ones in the expensive DAWs are particularly poor. There are loadsa zillion MHz sampling recorders that are like that too.

Poor anti-aliasing/downsampling sounds very similar to HF IMD from Jurassic times and indeed sorta like HF mistracking on vinyl ... shing of the triangle rather than ting

[JR.]

Back in the day a simple listening test for IMD was to jingle some keys in front if a mic.. Keys have all kinds of >20kHz energy that stresses slow circuitry.

In da 21st century with all music sources via evil digital, this test should be benign cos anti-aliasing filters.

Alas, for da zillion MHz sampling fans, Sample Rate Converters often have atrocious downsampling filters. The ones in the expensive DAWs are particularly poor. There are loadsa zillion MHz sampling recorders that are like that too.

Poor anti-aliasing/downsampling sounds very similar to HF IMD from Jurassic times and indeed sorta like HF mistracking on vinyl ... shing of the triangle rather than ting

Yes pre-recorded sources should be free of out of band content, but I was talking about real keys in front of a mic preamp which generally doesn't have anti-alias filters (while I advocate band limiting mic preamps early and often).

JR

[JR.]

I like to characterize amp recovery from clipping, like how a race car handles after bumping the wall at speed. Important to race care drivers, not so much for careful drivers. At Peavey and in the sound reinforcement industry clipping power amps is more normal than not. That's why pretty much 100% of Peavey amps were designed with clip limiters that managed the behavior when overdriven to sound relatively benign (like a little extra compression).

JR, could you post a sketch of your favourite clip limiter please.

Not sure I have a favorite, I do not advocate clipping power amps, but Peavey's DDT design was quite mature by the time I started working there. I'm sure there are many published examples as Peavey used DDT in most power amp channels. As I recall the limiter was performed by a series input resistor shunted by an OTA. The OTA we (they) used was a CA3080 but selected for a guaranteed control voltage feedthrough so there was no thumpy feedthrough when transients were limited. The limiting was triggered by monitoring the amplifier feedback node, so whenever the amp lost NF typically from voltage clipping but also triggered by current limiting, the gain would drop until negative feedback was back in control. I do not know the time constants off the top of my head but as i recall they were relatively fast attack and fast release. The actual times were the result of years (decades) of listening to the circuit in use, and listening to customer feedback. Most were blissfully unaware that they were in limiting unless they saw the DDT LED.

IMHO, the 'sound' of a 'good' 50W amp is entirely dominated by its overload & recovery behaviour. Many (all?) Golden Pinnae amps have appalling overload & recovery behaviour .. especially into real speaker loads.

Not just audio-phool amps... I recall one very popular value amp from a Peavey competitor that had a bad habit of clipping/current-limiting asymmetrically. The net effect of this was to put DC into the speaker voice coil. Peavey ended up making many warranty replacements for speakers killed by this series of rouge amps, until the manufacturer cleaned up their act. The customers were more willing to believe that they overpowered the speakers with the new amp, than that the amps were faulty.

I've only seriously investigated (with DBLT bla bla) a simple diode 'soft clipper'. It was preferred but the published THD spec would have been quite blah.

The '50W/channel' stereo amp is interesting cos if your cost constraints point to eg 200VA transformer, rewinding that 200VA item to give slightly less voltage (and hence power) actually sounds better cos the better regulation possible at the lower voltage on clipping means the PSU crud that is superimposed on the clipped signal is less.

I hasten to add, this was tried on a single design and I wouldn't expect the same behaviour with a 'good' 200W/channel amp.

A good limiter, sounds indistinguishable from limiting in the audio chain elsewhere. It just more precisely determines a threshold based an when the amp runs out of poop. I was routinely disappointed by customers who were deaf dumb and blind in listening tests who preferred the exact same amps with DDT defeated because A) they made noticeably more power on LF bass events, and B) low bass clipping does not sound "bad" to many deaf, dumb, and blind listeners.

I even had one disappointing listening trial where a 1500W amp without DDT was considered better/louder than a 1800W amp with DDT. FWIW these tests were performed decades ago and the amount of clipping tolerated in recorded media has increased dramatically since then.

JR

[edit] In fact with clip limiting the amps do make less power and this protected many Peavey customers from themselves [/edit]

[mediatechnology]

Selecting the "Best" Power Transformer

My primary requirement for the "best" power transformer for the Simple 10W Class-A was any transformer that was available.

My first stop was Hammond where I found toroidal units having 24, 30 and 36V CT secondaries.
30 volts center-tapped seemed the only reasonable option since 24V wouldn't make 10W and 36V would produce too much heat.
Mouser stock some of the Hammonds locally.

But. it was Sunday.
My first stop was the junk box where I found an ancient Stancor RT-402 transformer.
It has multiple primary taps and is a so-called "universal rectifier" transformer.
Unfortunately it only has a 2A rating - I need at least 4 amps.

I also found in the MCI repair stock some 10,000 uF 75 computer-grade caps and a BR252 25A 200V bridge.
With some clip leads I made my test power supply.

I was able, between setting taps and using the Variac, dial it in to 30V.
The measured rms current (using a Fluke current transformer) was about 3.6A-4A.
So a 30Vrms CT secondary looks about right.

The unloaded voltage regulation was poor because my transformer was seriously overloaded.

Mouser had stock on the Hammond 1182P15 30VCT 5.33A for $66.15. http://www.hammondmfg.com/pdf2/1182P15.pdf
Mouser also had the Hammond 1182H15 30VCT 7.5A for $67.50 http://www.hammondmfg.com/pdf2/1182H15.pdf
I spent the added $1.35 for an extra 2 amps.
(Digi-Key is considerably cheaper but not in stock. Owing to Mouser being 15 miles away I'll have my in-stock parts tomorrow.)

Depending on the regulation of the Hammond toroid I may be able to Zener-regulate the driver board.
If there isn't enough power supply headroom I may use a separate +/-18V regulated sub-supply for the driver.

I think where I'm headed with this is a box that combines a dedicated headphone amp using one dual Class-A board having it's own level control and a second dual Class-A as the 10W X2 pre-driver.
The +/-18V low-current supply can then be used for both the headphone amp and power amp pre-driver.

Efficiency?

The dual class-A draws about 120 VA just sitting there.
Deliver 10W X2 into a load and it might get to 160-180 VA.
The heatsink temperatures with +/-17V rails are about 185 deg F.
At normal listening level I'm likely using 2-5W, or 50% of available power.

The 35W/Ch Class-A/B amp I measured earlier is drawing 88 VA just sitting there.
And, I never turn it off.
Its heatsinks, much larger than the dual Class-A's heatsinks, have device temperatures of 125 deg F.
Deliver 10W X2 into a load and I may add another 30-40 VA making the total about 120-130 VA.
The heatsinks might reach 140-150 deg F.
At normal listening levels I'm using 15% of the available power.

So though it may seem I'm creating a lot more waste heat am I really?

Seems like if I just turn off the class-A amp when I'm not using it versus letting the "efficient' Class-A/B run all the time I actually could come out ahead.

[JR.]

It seems a little silly to worry about efficiency when pursuing a class A design.

As I mentioned, lots ofd work has been done by Japanese in decades past to make a class A amp where the class A bias is modulated by the signal to be just enough to never cut off, the criteria for class A. Of course this modulation would (probably) add distortion to your open loop design. Hypothetically you might be able to tweak the class A bias just right, but that is not trivial.

JR

[mediatechnology]

It seems a little silly to worry about efficiency when pursuing a class A design.

Who's' worrying?
My point is that based on the way I use either amp they're not much different in terms of overall heat and line current.
As you often say YMMV.

The class-A/B sitting there doing nothin' generates a fair amount of heat and consumes a fair amount of VA.
More than I would have thought.
Almost as much has the hopelessly inefficient class-A with 1/3rd the power sittin' there doin' nothin'.

But yes pursuing efficiency in a class-A is silly.
The biggest issues are having a big enough transformer and heatsink.
And not having too high a Vcc/Vee with a stock transformer...
That has a huge effect on heatsink size.
I can always back off idle current.

[ricardo]

IMHO, a Class A amp will benefit greatly from using regulated SMPS.

The huge standing current means the crud on the PSU rails is also huge (and adds to the inefficiency).

A regulated supply definitely makes a small Class A amp sound better cos overload doesn't superimpose the sawtooth of a conventional supply on the signal.

... part of my Jurassic DBLTs on a 50W commercial prototype. The idea was to use the same circuit and metalwork but up the standing current & derate the amp to 15W (IIRC) so we could put a "Class A" label on a second product with only 2 new parts, the transformer & PSU caps.

I think Nelson Pass uses laptop PSUs as inexpensive regulated SMPS for his DIY designs.

[mediatechnology]

A regulated supply definitely makes a small Class A amp sound better cos overload doesn't superimpose the sawtooth of a conventional supply on the signal.

I think Nelson Pass uses laptop PSUs as inexpensive regulated SMPS for his DIY designs.

I had thought about that.
One of the big problems with the JLH is that it has poor PSU rejection even when not in overload.
Some people use huge filter banks and/or ripple eaters to keep the hum down.

Most laptop supplies I see are 18V which is perfect.
The Zen are AC-coupled and use a single supply.
With a dual/split supply it becomes a little more of a challenge but still do-able.
For the first one I went with a big hunk of iron.
I'll regulate the pre-driver and see how it sounds when clipped.