Can you make a really good hifi amp by copying a late 70s schematic + good parts

[Beamin]

? I asked something like this before and got a tons of responses most really good but many rather conflicting. And one person saying those amps that come in DIP packages sound just as good as any high end hifi  . So a better idea:

Find a service manual for a really good amp from the 70's. Pick one that's not too complex but was known for really good sound quality like a marantz. Hand select the parts to the exact value with better than original tolerances and match the final output transistors to each other. These will probably be the most expensive parts. But if everything is silicon and not germanium you should be able to get the same or equivalent. The transformer might be hard to get the exact one but you would pick an amp that runs off common voltages. One other consideration do these amps have really big chokes in them? If so couldn't you just wind your own making it out of a big toroid since it's just wire with one value you could measure and duplicate? Or do A/B amps not have a choke?

Lay it out on a nice hand made PCB and you would have a high end amp for the same price as a best buy amp... Maybe?  Maybe make it in sections so you could later add a tube preamp. I know in Macintosh amps the sound quality is mainly in their preamps. Matching tubes might be difficult.


When you match the transistors what are you matching, and what do you need to test; a curve tracer or scope? What about tubes? Can you measure tubes on an analog scope with regular probes even though they are high voltage and impedance? ?

[dmills]

Most of the 70s stuff was done with SLOW output devices, think 2N3055, TIP31 and friends, you can do far better today, but a direct sub would probably hurt stability. The mosfet stuff of the 80's was mostly lateral parts that are well obsolete, in fact finding good mosfets for linear as opposed to switching operation at power is problematic these days.

I would grab a copy of "Audio Power Amplifier Design" by Douglas Self, get a recent edition, and build his latest, likely to be as good as anything else out there, and he really doesn't do the tweaky bullshit, maybe also Bob Cordells book.

For a valve preamp you are only looking at a couple of hundred volts, so a ten times probe (get one fixed at ten times) and a standard scope is fine. 

Your major cost is likely in metalwork and power transformer and for a one off you are never going to beat the local box shifter, that is not the reason to play. 

Regards, Dan.

[Beamin]

Most of the 70s stuff was done with SLOW output devices, think 2N3055, TIP31 and friends, you can do far better today, but a direct sub would probably hurt stability. The mosfet stuff of the 80's was mostly lateral parts that are well obsolete, in fact finding good mosfets for linear as opposed to switching operation at power is problematic these days.

I would grab a copy of "Audio Power Amplifier Design" by Douglas Self, get a recent edition, and build his latest, likely to be as good as anything else out there, and he really doesn't do the tweaky bullshit, maybe also Bob Cordells book.

For a valve preamp you are only looking at a couple of hundred volts, so a ten times probe (get one fixed at ten times) and a standard scope is fine. 

Your major cost is likely in metalwork and power transformer and for a one off you are never going to beat the local box shifter, that is not the reason to play. 

Regards, Dan.

Ah that's really good to know. I actually came up with this idea when I was a teenager and bought a Marantz for $500 at 50 watts 2 channel. I found that 50 watts was better (both louder and cleaner) then any 100 watt amps including the top of line brands they sold at circuit city or best buy. Best buy didn't have decent amps besides maybe onkyo. So with this in mind I bought a pair of 100 watt  speakers. Problem is they wanted true 100 watts like a Marantz 100 watts amp. I tried powering them with a 100 watt amp I bought at circuit city but was even less powerful  then the Marantz's 50 watts. It was a Harmon Kardin I think, so I took it back. Plus the Marantz had 20x lower (THD) distortion on the specs the anything the box stores sold. Sony was 0.8%@100w Marantz was 0.03%@50w so big difference. Now I need a clean 100 watts to power them to get the bass/mid bass to come out and make them sound full. I will still build one of these amps at some point. If I had the funds I think they might be fun to build and sell for a small profit. I would totally buy one if I got to check it out and see it.

Do you risk damaging the speakers of the amp if the amp is under powered?

If the Marantz is Korean built in 2000 the caps won't be starting to go bad yet? I haven't used the thing in years. So it's a "low hour" set. I could see this in some Chinese amps since modern Chinese built TV's are starting to get bad caps five years latter.

[Benta]

From the 90s and until today, the output bipolars of choice for audio seem to be the Sanken 2SA1216 and 2SC2922.
The drivers vary widely.
With the Sanken pair, you have to watch stability, as someone mentioned. They are 50 MHz parts, which is way above 2N3055/MJ2955 and 2N3773/6609.

[NiHaoMike]

Nowadays, there are so many cheap analog input audio amplifier modules that there's little point building one from scratch. For a PSU, telecom "rectifiers" are very cheap on the surplus market.

[Alex Nikitin]

? I asked something like this before and got a tons of responses most really good but many rather conflicting. And one person saying those amps that come in DIP packages sound just as good as any high end hifi  . So a better idea:

Find a service manual for a really good amp from the 70's. Pick one that's not too complex but was known for really good sound quality like a marantz. Hand select the parts to the exact value with better than original tolerances and match the final output transistors to each other. These will probably be the most expensive parts. But if everything is silicon and not germanium you should be able to get the same or equivalent. The transformer might be hard to get the exact one but you would pick an amp that runs off common voltages. One other consideration do these amps have really big chokes in them? If so couldn't you just wind your own making it out of a big toroid since it's just wire with one value you could measure and duplicate? Or do A/B amps not have a choke?

Lay it out on a nice hand made PCB and you would have a high end amp for the same price as a best buy amp... Maybe?

No, most likely it would be a waste of time and money (though probably a lot of fun). To make a good amp from scratch requires much more than a "good" schematics and good components. More sense in buying a decent old amplifier and restoring it.  If you want something unusual from the schematics point of view, find the 4330 integrated amp from Creek Audio .

Cheers

Alex

[rdl]

Speaker wattage ratings are for the maximum power it can handle, not what it requires. However, speakers vary in efficiency, so there are considerable differences from one to another in how loud they will be at a given power level.

A Creek 4330 has been my main amp for nearly 20 years.



These days it's driven from a different source.

[chemelec]

Do you risk damaging the speakers of the amp if the amp is under powered?

Under-Powered Amps result in Square Wave Distortion, Especially at higher Volume levels.
Speaker Cones Can't respond fast enough to Square waves, so this tends to tear the Speaker coils off of the Coil form.
Eventually resulting in a Badly Damaged Speaker.

As to High Quality Amps, The System is only as good as the Worst Component in it.
And Your SPEAKERS are the WORST Part of any Audio System.

[Alex Nikitin]

A Creek 4330 has been my main amp for nearly 20 years.

Thank you, it is always nice to see my old design going strong after all these years.

 

Cheers

Alex

[dmills]

Audio power is a funny thing, because it typically has a HUGE crest factor, 20dB is not uncommon, so even turned well up that 50W amp was probably averaging under 1W with the very occasional large spike.
 
The way that the US consumer folks insist amps be specified does a poor job of dealing with this reality and actively encourages amps which clip just above their continuous rating, which is stupid design. I suspect your Marantz did not suffer from this, so that 50W was probably RMS Cont, but they may have gone for a design capable of much higher short term peak.

Lots of volts from the transformer, with a limited thermal envelope is probably the way to go for HIFI, and use a simple transformer/rectifier cap sort of supply, it will deal with short term overload just fine where a switcher will often go into hickup mode.

I second the comment about the speakers generally being the weak link (And the only reason for that "generally", is that sometimes the room is an even bigger issue).

Regards, Dan.

[Beamin]

Nowadays, there are so many cheap analog input audio amplifier modules that there's little point building one from scratch. For a PSU, telecom "rectifiers" are very cheap on the surplus market.

I'm going for high current low distortion <0.04THD. So I don't want cheap or chinese that would totally defeat the purpose. Yes I could go on ebay and get some pile of shit for a few dollars but I'm trying to upgrade not do a major down grade.

[Benta]

Do you risk damaging the speakers of the amp if the amp is under powered?

Under-Powered Amps result in Square Wave Distortion, Especially at higher Volume levels.
Speaker Cones Can't respond fast enough to Square waves, so this tends to tear the Speaker coils off of the Coil form.
Eventually resulting in a Badly Damaged Speaker.

As to High Quality Amps, The System is only as good as the Worst Component in it.
And Your SPEAKERS are the WORST Part of any Audio System.

This is not the failure mode I've experienced. The woofers and midrange speakers are normally quite well protected from high frequencies by the crossover filter (low pass limiting).
What really happens is, that when the amp starts clipping, a lot of power is suddenly transferred to the tweeter range, which causes them to burn out.

[NiHaoMike]

I'm going for high current low distortion <0.04THD. So I don't want cheap or chinese that would totally defeat the purpose. Yes I could go on ebay and get some pile of shit for a few dollars but I'm trying to upgrade not do a major down grade.

There are modules that use TI chips, the leader in the market.
Here's one example:
https://smile.amazon.com/Sure-Electronics-AA-AB32191-TAS5630-Amplifier/dp/B007ALBNP8/

[CJay]

Simple answer, no.

Old amps were as good as the technology they used/was available, that's not to say they weren't great in their day and a high end 70s amp would be passably OK now but they're nowhere near as good in terms of fidelity as a modern device that's targetted at the same market position, the technology has improved and you can get, in comparison with the old stuff, ridiculously good gear for peanuts these days.

 I spent a good few years repairing amps for HiFi and PA, the first MOSFET amps I heard were a revelation, at first I thought 'thin bass, lacking 'boom'' but it wasn't that at all, the 'boom' was a product of the old amps just not coping with high bass levels and making it 'smeared' and 'muddy'

What I was hearing from the MOSFET amps was actually a lot closer to the actual performance on stage or recording.

Note, the amps used on stage were and still are as much a part of the instrument as the strungs on the guitar being played through it so if we're talking amps for performance then you may be better making a faithful copy using the original parts, for recorded music, don't bother.

[Benta]

Note, the amps used on stage were and still are as much a part of the instrument as the strungs on the guitar being played through it so if we're talking amps for performance then you may be better making a faithful copy using the original parts, for recorded music, don't bother.

CJay, you're speaking to my heart. As an old rock guitarist, pairing the instrument to the Marshall/Ampeg/Acoustic/HH was part of the art.

[ejeffrey]

Under-Powered Amps result in Square Wave Distortion, Especially at higher Volume levels.
Speaker Cones Can't respond fast enough to Square waves, so this tends to tear the Speaker coils off of the Coil form.
Eventually resulting in a Badly Damaged Speaker.

That isn't the problem.  The amplifier output goes to a crossover, which provides a low frequency signal to the woofer and midrange with no square edges.  The fast risetime goes to the tweeter, which has no problem with it.

The problem with square waves is that they put a *lot* more power into the high frequency ranges than normal audio and that just overloads the tweeters.  A speaker with 100 W "system" power handling probably has a tweeter that can only handle a couple watts.  This is normally no problem since few musical instruments put out a lot of power in the upper ranges.

Playing a full power 10 kHz sine wave at max volume will destroy your tweeters too.

[chemelec]

I Use bi-Amplification, So the Square waves can go to the Woofer.
Electronic Crossover for the Tweeter.

[David Hess]

No, just using better parts except if they started out inferior or defective will not significantly improve the performance.

Further, layout is critical to avoid ground loops which is not always reflected on schematics.

Check out Designing Audio Power Amplifiers by Bob Cordell and Audio Power Amplifier Design Handbook by Douglas Self.

[CatalinaWOW]

Define better. 

In the 70s I had a similar idea and had a complete schematic and parts list for the best amplifier out there (in the ever so humble opinion of a newly graduated engineer).  Went into the local electronics parts haven to buy the parts and found a stack of name brand amps surplussed and selling for less than part of the parts I went in to buy.  I bought that, skipped building the amp and have never looked back.

Were the specs on the amp I bought as good as what I was going to build?  On paper, not quite.  Who knows what the as built performance of my amp would have been. 

Did I learn as much by going this route?  No.  Building my own would have been better for that.  But not doing that amp let me do other things and learn things I wouldn't have learned building the amp.

Was the packaged product more pleasing than my homebrew thing was likely to be?  Absolutely.  Even today, with four decades of practice and experience my housings would barely match up to the commercial products, and then there wasn't even a comparison.

Can you trade labor in matching components for better performance?  Sometimes, but it depends on the design.  In most commercial designs a lot of effort has gone into making the design relatively insensitive to things which are costly in production, like matching.  So the design you copy is not likely to benefit.  Could another design approach that accepted the penalty of matching perform better?  Sure, that insensitivity to product variability usually comes at the expense of something else.  Maybe even something you care about.  But you won't find that design by looking at high volume products.

So the answer depends on you and what you mean by the question. 

[Berni]

Under-Powered Amps result in Square Wave Distortion, Especially at higher Volume levels.
Speaker Cones Can't respond fast enough to Square waves, so this tends to tear the Speaker coils off of the Coil form.
Eventually resulting in a Badly Damaged Speaker.

That isn't the problem.  The amplifier output goes to a crossover, which provides a low frequency signal to the woofer and midrange with no square edges.  The fast risetime goes to the tweeter, which has no problem with it.

The problem with square waves is that they put a *lot* more power into the high frequency ranges than normal audio and that just overloads the tweeters.  A speaker with 100 W "system" power handling probably has a tweeter that can only handle a couple watts.  This is normally no problem since few musical instruments put out a lot of power in the upper ranges.

Playing a full power 10 kHz sine wave at max volume will destroy your tweeters too.

Yep you can damage tweeters in such a way.

But its very unlikely to happen in a real use scenario. You can very quickly notice low frequency tones clipping even a little bit as it sounds absolutely awful. Also due to standardized line levels its very likely that the audio source is not even capable of putting a large enough signal trough the amplifier to cause such severe clipping.

Additionally normal music has quite low "power content". Its often the tiny voice coil in tweeters that gets damaged and that takes a sustained power trough it so it doesn't have a chance to cool down. This is the reason why i usually use a bass test song playing in to a pile of ceramic resistors to test amplifiers. Those tracks are usually almost entirely made up from full scale tones(sometimes even square wave) to push as much power as possible trough the amplifier in order to see if the cooling and power supply is up to the job. The difference can be really seen on my ceramic resistor dummy load, putting 100W of music into it gets it barely warm while 100W worth of bass test gets the resistors to water boiling temperatures.

So yeah you can damage speakers if you put the right abusive sort of track trough them(Even if both the amp and speakers are matched in power), but its highly unlikely to happen with any sort of normal music. (Unless perhaps you regard dubstep as normal music and like to listen it at max volume)

[Beamin]

Under-Powered Amps result in Square Wave Distortion, Especially at higher Volume levels.
Speaker Cones Can't respond fast enough to Square waves, so this tends to tear the Speaker coils off of the Coil form.
Eventually resulting in a Badly Damaged Speaker.

That isn't the problem.  The amplifier output goes to a crossover, which provides a low frequency signal to the woofer and midrange with no square edges.  The fast risetime goes to the tweeter, which has no problem with it.

The problem with square waves is that they put a *lot* more power into the high frequency ranges than normal audio and that just overloads the tweeters.  A speaker with 100 W "system" power handling probably has a tweeter that can only handle a couple watts.  This is normally no problem since few musical instruments put out a lot of power in the upper ranges.

Playing a full power 10 kHz sine wave at max volume will destroy your tweeters too.

Yep you can damage tweeters in such a way.

But its very unlikely to happen in a real use scenario. You can very quickly notice low frequency tones clipping even a little bit as it sounds absolutely awful. Also due to standardized line levels its very likely that the audio source is not even capable of putting a large enough signal trough the amplifier to cause such severe clipping.

Additionally normal music has quite low "power content". Its often the tiny voice coil in tweeters that gets damaged and that takes a sustained power trough it so it doesn't have a chance to cool down. This is the reason why i usually use a bass test song playing in to a pile of ceramic resistors to test amplifiers. Those tracks are usually almost entirely made up from full scale tones(sometimes even square wave) to push as much power as possible trough the amplifier in order to see if the cooling and power supply is up to the job. The difference can be really seen on my ceramic resistor dummy load, putting 100W of music into it gets it barely warm while 100W worth of bass test gets the resistors to water boiling temperatures.

So yeah you can damage speakers if you put the right abusive sort of track trough them(Even if both the amp and speakers are matched in power), but its highly unlikely to happen with any sort of normal music. (Unless perhaps you regard dubstep as normal music and like to listen it at max volume)

Yes it is.

No it's not.


Yahuh!

Notunh!


Can anyone back up what they are saying? I feel like I'm reading youtube comments where people know absolutely nothing make a statement as if it was fact then post some crazy idiot video on youtube and call it "research". My college education and fancy book learning trumps your research you did on youtube. It's like old wives tales that just because they heard it 100 times it must be true. The internet made the smart people smarter and the dumb people a lot dumber and far more dangerous. I saw a video where a guy drinks turpentine everyday to be "healthy". 

[BrianHG]

     The truth is if you have a quality design schematic from the 70s, and use parts from the 70s, you will make a good amp.
Modern parts, what you might think of as high end parts today, when it comes to the transistors, today, they have been optimized for switching performance, not necessarily linear drive.  This can alter the sound, quality and steady function of the amp which was designed with 70s transistors in mind.

     When it comes to resistors and caps, you have a much better selection for resistors.  With today's caps, if you need it, the higher quality ones will cost you money.  The life of these components will be better than the ones from the 70s if you pay for the good ones where needed.  Be careful of any cap which needs to handle high ripple currents in your design, the odd high end ones from the 70s will need high end ones from today as well.

     Make your PCB with 2oz of copper at least.  Old 70s design PCBs had no soldermask and usually flooded the traces with solder to thicken them up to handle higher current where needed if not everywhere.

Everything else is BS.   Good luck, rejuvenating old analog circuits from the 70s will be a fun experience and if you choose the right design, don't believe anything said above, and don't strive for that absolute minimal distortion, the design should sound perfectly fine.

As for transistors, motorolla/On semiconductor still makes a nice NPN/PNP matched pair designed for linear operation for servo motor and HIFI audio in a TO247 package, designed to maintain at least 100 watts continuous on a generous heatsink.  be careful as Chinese crappy knockoffs exist.

[BrianHG]

The modern Onsemi complementary 150watt pair:
NJW0281G (NPN)
NJW0302G (PNP)

Respectable 30Mhz bandwidth and wide safe operating area, plus, averaging 2.50$ each at digikey for originals.

NJW3281G and NJW1302G for 200 watt handling.

[Beamin]

     The truth is if you have a quality design schematic from the 70s, and use parts from the 70s, you will make a good amp.
Modern parts, what you might think of as high end parts today, when it comes to the transistors, today, they have been optimized for switching performance, not necessarily linear drive.  This can alter the sound, quality and steady function of the amp which was designed with 70s transistors in mind.

     When it comes to resistors and caps, you have a much better selection for resistors.  With today's caps, if you need it, the higher quality ones will cost you money.  The life of these components will be better than the ones from the 70s if you pay for the good ones where needed.  Be careful of any cap which needs to handle high ripple currents in your design, the odd high end ones from the 70s will need high end ones from today as well.

     Make your PCB with 2oz of copper at least.  Old 70s design PCBs had no soldermask and usually flooded the traces with solder to thicken them up to handle higher current where needed if not everywhere.

Everything else is BS.   Good luck, rejuvenating old analog circuits from the 70s will be a fun experience and if you choose the right design, don't believe anything said above, and don't strive for that absolute minimal distortion, the design should sound perfectly fine.

As for transistors, motorolla/On semiconductor still makes a nice NPN/PNP matched pair designed for linear operation for servo motor and HIFI audio in a TO247 package, designed to maintain at least 100 watts continuous on a generous heatsink.  be careful as Chinese crappy knockoffs exist.

This makes sense. No way am I going to buy some shitty Chinese chip amp and build around then hope it sounds good. That's like building a race car then buying Walmart brand chensing tires to save costs.

[BrianHG]

This makes sense. No way am I going to buy some shitty Chinese chip amp and build around then hope it sounds good. That's like building a race car then buying Walmart brand chensing tires to save costs.

Do not discount some of the Chinese class D amps.  The problem is that you have no control overs the choice of architecture or main IC & source of the components and the quality of PCB, PCB layout & design.  Just placing a few fancy quality caps here and there with the highest spec class D amp IC doesn't necessarily make the amp great.

Also, reading deeply into the top Texas Instruments class D amp IC, the 75-100 watts devices, powered to the max, if you look at the THD charts, the 100 watt setup has incredible performance to around 35 watts, then the THD begins nose dive right after that point up to the peak 100 watts.  With today's popular music recordings, this is fine for most junk.

A properly biased regular analog class B 100 watt amp will have have a fairly consistent THD way up to 80 watts, if really well designed, the THD will be clean up above 95 watts, true RMS level, not peak like the class D amp specs.