DIY Class A Amp

[LTig]

I don't know which model the OP has, but the Magnepan 3.6/R have 2 fuses - one for the tweeter (2.5 amp), one for the midrange (4 amp).

I had the 1.6 which had one fuse for the tweeter. The woofer did not have a fuse.

 

[captainbeefheart]

The fuse is not going to tell you much

It would tell him what he wanted to know which is how much current the amp takes from the mains supply. I explained to him about program material and how the amp stores energy to be used for short duration transients of high power demands.

 

[captainbeefheart]

Crossover distortion in output stages has been a solved problem for some time now.

While technically true in theory it's in execution where the problem still is often found in mass produced amplifiers. Too often I have found when testing higher powered Class AB amplifiers at lower power levels the distortion in much higher than when at full output. Typically increasing the forward bias will get rid of it but why was it there to begin with? My guess is they make so many that due to the output devices having varying hfe the fixed bias types may not provide enough bias current for good low power performance. If there is an adjustment it may have just gone out that way and was overlooked because they may not look at distortion at 1 watt, they very well could just do a full output power test and if it's within spec it gets approved and sent out for sales. This is because the quality control most likely has a bias reading that's acceptable although on the lower scale it's not ideal but passes their QC.

 

[dlaloum]

I've seen the question asked before, and it appears you know the answer. How did you switch the supply voltage input? Is it internal? What did you have to do?

It was internal - and it was probably simple - but it was also around 10 years ago, and I don't remember!!!

I'm quite sure it was just a jumper or something similar - but I recall needing to open the case - the procedure was documented on a web posting somewhere....

 

[dlaloum]

I've seen the question asked before, and it appears you know the answer. How did you switch the supply voltage input? Is it internal? What did you have to do?

Just searching through old postings...:

"a quick update, it seem that cutting the jumper Z901 which is located on the main board by the power switch does the trick.
By cutting that jumper so it’s open the amp operate on 230/240VAC."

Sounds like this is probably what I did.... but like I said, 10 years later, I honestly don't remember!

I also believe some versions (later) had auto-switching power supplies....

 

[Nobleman]

Thank you all for sharing the wisdom. I understand now.

 

[Nobleman]

The new amp arrived today. Hypex Nc250 dual mono based. Transients are more full bodied now. With the 6000a they slim and slightly in the back but now they are powerful. But I’ve lost some of that black magic, ……that note you hear clearly out of nowhere in complete blackness which always put a grin on my face listening to the 6000a…Has lost some softness too…Or maybe my ears are readjusting…let’s see.

 

[orangejello]

I’ve never had a serious class A amp so wanting to hear the difference. Currently using an Audiolab 6000A and love it. It drives my Maggie’s quite well but it’s only 75wpc so have to turn up the volume quite high. Which is why I’m considering a class a build. As a self taught technical person I’m worried about adjusting the bias Of class a amps. Is that difficult? Should I just get a Hypex NCxxx instead?

Depends on what you want to do. I’ve had a couple of Class A amps, one very highly regarded. I got a smokin deal on NAD M22 v2 (nc400 based) and never looked back. I prefer it to my class A amps and my various tube amps. It is clean, powerful, and efficient. Checks all of my boxes. I thoroughly enjoy it. You could build one.

 

[bfburkejr]

Been thinking of building a class A amp for quite sometime. Leaning towards some kits based on Krell KSA50 and Accuphase E270 based kits. Has anyone build any of these? What was your experience and result? Any recommendations? Thanks.

Try building one of the Nelson Pass ALEPHs, he released all the copyrights for them and the PCBs can be found on line. You will need match or purchased matched MOSFETS, and have a some experience with building electronic devices. The original transistors are no longer available, but there are perfectly good substitutes. They sound wonderful when built correctly, and there are various wattage outputs available.

 

[bfburkejr]

IMV it's just hype from 50 years ago. In a proper controlled double blind listening test people fail to hear differences between decent power amps as long as they are not driven into clipping. Hence the best one can do is to make sure that the power amp never clips which means lots of power.

I can hear a difference in some power amps, the differences are more apparent the better the speakers are that you are driving as well as you hearing ability and source material.

 

[bfburkejr]

Wattage is the same on input & output only when efficiency is 100%. Class A amps are only 25% efficient; let's assume class AB is 50% efficient. 1200 watts to the speaker must draw 2400 from the wall, which is 20 amps at 120 volts, and that's assuming a power factor of 1.0 (ideal conditions), so it would probably have to draw more than that from the wall.

That is not exactly correct when talking about amplifying music. Think about the difference between a average switching power supply which has little capacitive reserve when compared to a high end linear power supply with a huge capacitive reserve. For short periods it can provide far more wattage than it draws from the wall. Consider the nature of a music waveform compared to Direct Current. You can blow a tweeter made to take 1,000 watt transients with relatively little DC wattage, for instance.

 

[SIY]

I can hear a difference in some power amps, the differences are more apparent the better the speakers are that you are driving as well as you hearing ability and source material.

What were your controls used and (if your listening had basic controls) what were some specific sonic differences?

 

[Jim Taylor]

I can hear a difference in some power amps, the differences are more apparent the better the speakers are that you are driving as well as you hearing ability and source material.

In the first place, welcome to ASR.

May I ask you the nature of the amps that allowed you to hear differences? Were any of them tube or hybrid designs? Were the outputs equal (matched)? Were any of the amps DIY?
And as @SIY asked ... were there any controls? Was this sighted or blind? Were the outputs checked with a voltmeter? (etc, etc. ...)

IME, differences that are commensurate with "better" speakers, source material and which depend on hearing ability (imaginary or otherwise) have actually been due to unmatched power levels. Other differences have not shown me the same consistency as unmatched power levels.

Jim

 

[Doodski]

The way I discovered Neurochrome was the statement by Sigfried Linkwitz , on his webpage, "I don't know how to measure such low distortion." Tom C's amps have like three zeros to the right of the decimal. they start with a good clean chip (the LM3886) and then add a servo corrector in the feedback loop. Every bit as good as class A.

It is important to have enough power. I my case the class A amps are hooked directly to drivers in a multi-amp (active crossover) set up. Most passive crossover networks eat a lot of your power.

When I hear or read that a system "drops to two ohms" that tells me the crossover network is parallel (as most are) and that it is a too many way system to have a passive crossover.

I tried to help someone bi or tri amp an Infinity system from back in the day. It was like a five way plus it had a rearward firing tweeter for "ambience" which I would have unhooked, which impedance wise made it like a six way system. It is no wonder that the impedance gyrated. Instead of finding an amp that is comfortable driving a below two ohm(and very reactive) load, dissect it apart and multi amp it. You get lots more headroom and better performance in a lot of ways. And the benefit of a four or five way speaker system.

Regarding the Kappa Series Infinity speakers that you active crossover connected dedicated amps to... A man after my own heart! LoL...

 

[Doodski]

@Nobleman. A real life hardcore class A amp that does a large power output figure into 2 Ohms is expensive and will burn heaps of electricity. I have been there and done it and if I where doing it again I would step up in technology to a higher end class D design. It will sound fantastic, eat less electricity and not heat your room.

 

[john65b]

I built a Krell KSA-50 and Aleph 5 Clones 17 years ago with PCBs off DIYAudio. The Aleph was a Brian Bell PCB (sanctioned by Nelson) and the Krell was a Pinkmouse PCB (I know this likely means nothing to many of you). The Krell was true to original circuit, but with modern output transistors (3 pairs Toshiba SA1943 / SC5200), while the Aleph was true throughout. The Aleph is an easier circuit than the Krell. I also would not trust any of these clone boards off ebay....fake transistors, etc...

Both amps were biased as originals and were quite hot - the Aleph heatsinks sat at 140F, while the Krell had a much bigger heatsinks and never exceeded 120F...

Anyway, both amps drove difficult speakers with no issues - including Acoustat 2+2, Apogee Scintilla and Duetta Sigs, Magnepan Tympani T-IVa, MG20.1...

To me, the 2+2 was the toughest load - the only speaker to drive my rock solid Class D amp (ICEPower 1000ASP) to protection. The Aleph and Krell had zero issues on all of them.

The Krell sounded a fair bit better than the Aleph so I sold it. I later realized a minor (DIYAudio member recommended) modification to the input may have been the reason. I also changed the power supply from CRC to CLC on the Krell and it seemed to have lost some of its magic (albeit quieter). I will be switching bit back when I have the time.

I like to use the Krell in Wintertime as it heats up my office nicely while the Hypex UCD and NCore Class D amps are used in summer months. I am also driving DIY open baffle speakers that are much easier amp load...

Its a tossup on the Hypex Class D vs Krell Class A on my current speakers. They both sound great.

I have plans to bring in Martin Logan CLS Electrostatic speakers - most Class D amps have issues with Capacitive loads from ESLs. I only use the Hypex UCD400 instead of the NCore NC400 on ESL speakers, as the UCD are easier to fix while blown NCores are not (replace amp modules - $$), but only really an issue when driving ESLs hard..

 

[SIY]

most Class D amps have issues with Capacitive loads from ESLs

I've run my Quads from Nilai, NCx, Purifi, GaN Sytems... zero problems.

 

[john65b]

I've run my Quads from Nilai, NCx, Purifi, GaN Sytems... zero problems.

Me too....but just don't push it...the Quads will likely fail before the Class D amp, and will be far more expensive to fix.