Amp design logic

Depends- does it have mayonnaise?

So no sandwich?

some Chinese component weirdness can trickle in

'Trickle' is really an understatement. I can give you plenty of other examples if you like. Just the tube sockets alone are example enough! Filter caps have been problematic as well as semiconductors, in fact in the semiconductor realm its become quite risky to buy parts off of ebay and the like because there are so many knock-offs now that won't meet spec.

I remember when US made Fender guitars SUCKED...from the late 60's until the late 70's...you had to look at 10 Telecasters until you found one put together correctly. Now they're mostly fine, even when made in Mexico or wherever. I understand that some Chinese component weirdness can trickle in, but I have some Chinese built gear that is amazing...a Kavent preamp, Jolida amp, and an original Reverend Goblin guitar amp (now made in the US by somebody who bought the designs) are 3 examples of "imports done right", likely due to a "lookout" posted at the gate (send him a sandwich).

Ralph the semiconductor industry is seeing a lot of rebranding of parts and miss markings as well;my customers are seeing more since they are now buying COTS (commercial off the shelf) components and are not doing incoming inspection with the exception of my military customers.It was a pretty good idea that you kind of did a DPA (destuctive parts analysis) of your own and saw it for your self;probally would have been used in a power supply application and resulted in early component failure and caused major/minor damage depending on where it was used.
Some have even reported blank die installed in chips and other components as well.

Wolf_garcia, 'capable', true enough. But if they do is another matter. We build a guitar amp for a company here in town. They got in some parts from China for testing.

One of the parts was a filter capacitor for the power supply, rated 47uf at 450V. It looked a little peculiar- most tubular electrolytic capacitors have a metal end of the part and an insulated end, with a crimp around the body of the insulated end. This one was crimped at both ends and was insulated at both ends. It was branded Nichicon, a respected Japanese brand.

We took it apart as it seemed odd. Turned out it was a shell, a ruse: inside we found a different part, marked 22uf and rated 400 volts (with a name on it I could not pronounce)! Imagine if we had used that in a 450V application...

So yes, the capability is certainly there, but the intention is another matter. And even if there is the intention, such a manufacturer has to be on the lookout for parts like this.

Any factory in any country is capable of building things well with high part quality. The key is establishing quality control and avoiding slave labor issues by paying fair wages...Jolida apparently does that and others might also but you don't hear much about it.

River251, here are some other things I did not see above.

A $1000 amp is likely mass-produced where is $50,000 likely is not. So any vendor offers price breaks on parts- the more you buy the cheaper they are. If you have an assembly line, which is practical for less expensive amplifiers made by larger companies, you can further drive the price down.

There are other factors- parts being one. However there is also the labor. Some $50,000 amps may be entirely handwired (our MA-2 is exactly that but is only $42,000/pr :) ). Handwiring is often more labor intensive than circuit boards. But it gives the designer the ability to control stray capacitance better than a circuit board, due to the dielectrics involved (fiberglass, which is a poor dielectric, as opposed to polyethylene, which is pretty good, for example).

In addition there is the overhead of the facility, the overhead of legal fees (patents, lawsuits, legal advice). There is also the formula for markup, which I think is where many audiophiles might feel that they are being taken for a ride. Some companies mark up by a factor of 5 and others even 10x. FWIW we have a markup of 2 times. This allows us to replace the sold unit and build another after that. IMO that is enough to make an honest profit, OTOH those that know me well know that I got in this business because I enjoy it rather than to make a fast buck. I should also point out that high end audio is a very finicky market that is also very competitive. Thus the saying about 'if you want to make a small fortune in this business, start out with a large one' applies.

Lately, Chinese parts and assembly has become an issue even in high end audio. Many companies that don't want to admit it have their gear made in China for a lot less. This puts a squeeze on those who are still holding out by using better, non-Chinese parts that should be more reliable (and also more expensive), not to mention the labor issue.

Well. My question about amp design turned into a huge lesson about the hifi business, and I wish I'd posed the question a little differently. I'm certainly in over my head in several ways. There is clearly some heavyweight experience and wisdom in these forums. Thank you for sharing your thoughts.

All I can say is, when I turn out the lights, kick back with my favorite cabernet, and just let the vacuum tube warmth of Diana Krall's impossible voice diminish my daily cares, I'm glad there are people who dedicate themselves to bringing this gear to the world.

Jim

Mechans, there is a difference between a modification and a straight parts swap as the OP is suggesting. Taking a $1000 amp and just putting premium parts in it. I have heard several modifications, mostly to digital gear, but also with amps. Some make a big improvement, some, not so much so. However, a modification implys a circuit (re)design, not simply a parts swap.

River251, With all due respect, based upon another recent thread you started, I think, at least for the present time, you might be in way over your head. Just think of the appropriate test equipment you'll need. It's not as simple as plug and play. You might want to begin with a simple starters kit, before attempting to supplant $50K amplifiers, regardless of their markup.

River251, perhaps you might want to consider cutting your teeth on projects like these:
http://www.diyaudio.com/forums/pass-labs/

You can't turn a Volkswagon into Bugatti, but you can alter/modify cars to excel at specific purposes. I think it is obvious that cars used in many races are based upon street models but have many parts changed/exchanged to upgrade performance.
Their is a thriving industry in this hobby living off the same idea. I can hardly think of any component that hasn't been or can't be modified. I myself can do little beyond rolling tubes, but I do that with some zeal, and I have employed others to mod a few pieces.
Some people do claim that modified components make their gear the equivalent of something much more costly. In any event it is abundantly clear that their is a strong belief in the value of modification.

As Minor1 has stated, there are many who have failed in this business. Many times that is due to cutting their margins too thin, or under-estimating their "true" costs. Trying to make something that is "too good of a deal" is a good way to go out of business. Great posts by Al and Byron too.

A manufacturer has to realize when a certain item is "too good" of a value and is dragging him down. I know of a manufacturer who is still in business, who stopped manufacturing a $6500 integrated amp because "he couldn't make any money on it". Knowing when to show 'em and knowing when to fold 'em is critical for staying alive in this, or any other business.

Cheers,
John

Sometimes, (not all)you actually get what you pay for. In terms of amplifiers, tube or solid state, it depends on;
1) time and salaries associated with design, test, manufacturer, R&D (this is huge), marketing, sales, repairs, etc. These all have to factored.
2) parts. if you require new parts or chasis designed and built that are not off the shelf, then you will pay a lot more money. Also, if you require a small amount of parts instead of a lot of parts made, then the manufacturers will charge more.
3) Transformers, yes this is parts, but transformers are the major factor in parts costs. These are not cheap and are made to your specs by transformer manufacturers in small quantities. hence high prices.
4) This isn't the car industry. If you are a small company, you have salaries and a price points you are designing to. You also have limited manufacturering capability.
5) repairs, service and admin support.
6) are the parts matched, and tested and the ones that didn't fall within the design criteria rejected? Well, these didn't come free. If you really are a stickler for matching parts and having parts within a certain criteria, well you have to buy hundreds or thousands of parts just to find a few that fall within the specs. you paid for those rejected parts and have to recover the cost.
7) design, manufacture timeline for a given product before new and better comes out. Say, three years. you have to recover all of your costs for the above, plus profit, misc. expenses, legal costs, human resources, (this is a business remember) and the costs for R&D for the new upgraded amp.

Several classes are taught regarding this type of thing. This is not straightforward and easy. It definitely isn't just the cost of simple parts. Go out and try to make an amp in the $5000.00 or $50,000 price range (sales price) and see what it takes.

Call Motorola, or tube manufacturers, for proprietary parts, small quantity and see how much they will charge you. Simple cheap sheet metal casings, aren't happening on quality amps. You need heavy metal, to help eliminate unwanted vibration. This does cost. Transformers? hah! this is stupidly expensive. Threshold like heat sinks for your class A ss amp? Get ready for sticker shock. matched devices? well get ready to buy thousands just to find a few that match. How much time did you take to design, construct and test? and how many people were part of this effort? You have to pay them and they aren't cheap. training, traveling, Human Resources, etc. Wow! And! one stupid review from some person that is full of himself/herself and you just lost a ton.

Also, where did you get the circuit design from? Osmosis? you had to go to school, and take years of R&D. Same for circuit board layout design, chassis design and construction, wire your own transformers? good luck. Make your own capacitors, have fun. Can't find the parts quality you need off the rack? Well then you have to do it yourself or pay someone to do it. these all mean up front the costs and you have to recover it from this amp sale. How many did you plan on manufacturering and selling? Take the total costs to design, construct, sell, HR, all of the above, including salaries, and factor that into the total number you plan on making. This determines your price.

So, yes, sometimes you actually do get what you pay for.

This is why many fail in this business. And also why many don't submit their products to magazines for reviews. Not that they don't make quality equipment, but because like wine, all you need is one stupid review, one reviewer that had a bad day and you are out of business.

enjow

04-23-12: River251
I was really hoping for a discussion of the components in a $1000 amp vs those in a $50,000 amp, and how much they, vs the designer's intellectual property, contribute to the quality (and cost) difference. In other words, can I take a standard amp design, use the best available parts, and get the sound of the $50K amp. If I can't even get close, the sound quality must come from the circuit design.

"Best" in this context would have to refer to using the same kinds of parts, but in versions that have tighter tolerances and/or more idealized behavior. My expectation is that Doug's answer is correct -- you can't make a VW into a Veyron by using upgraded versions of the same kinds of parts. You have to use different kinds of parts, that are not direct substitutes. That in turn will necessitate differences in circuit design, mechanical design, thermal management, and overall architecture.

Also, I would point out that "better" parts, however that may be defined, will not necessarily perform better if arbitrarily substituted into a given design, and may in fact perform worse, depending on how they interact with the surrounding circuit. For example, substituting a faster transistor or integrated circuit for a slower one can result in all kinds of problems, depending on the function the device performs. In digital circuitry that would include electrical noise problems such as signal crosstalk and "ground bounce," and increased sensitivity to impedance mismatches. In analog circuitry it would include possible increases in rfi susceptibility, intermodulation distortion, frequency response peaking, ultrasonic ringing, or even potentially damaging oscillations.

And I agree with everything in Bryon's characteristically well thought out post.

Regards,
-- Al

Here are the questions...

1. Between a 1K amp and a 50K amp, what accounts for the additional cost to the MANUFACTURER?

2. Between a 1K amp and a 50K amp, what accounts for the additional cost to the CONSUMER?

3. What is more important to the sound quality of an amp: its DESIGN or its IMPLEMENTATION (i.e. internal components)?

Here are my suspicions...

1. It depends on the amp.
2. It depends on the amp.
3. It depends on the amp.

RE: 1. Every manufacturer has costs, but not every manufacturer has the same distribution of costs. To state the obvious, there is considerable variation from company to company in the number of employees, the extent of marketing, the R&D man hours for any particular model, part costs, the costs associated with a particular country (labor, building, shipping from factory to market, monetary exchange rates, etc.). My suspicion is that there is significant similarity in the distribution of costs among manufacturers of $1K amps, but considerably less similarity in the distribution of costs among manufacturers of $50K amps.

RE: 2. Cost to consumer = answer to #1 + profit. With respect to profit, again I believe that there is considerably more similarity among manufacturers of $1K amps than manufacturers of $50K amps.

A consequence of #1 and #2 seems to be that there is considerably more variation in the PRICE/VALUE ratio for $50K amps than for $1K amps, which seems consistent with audiophiles' experiences.

RE: 3. There are several people on this thread who have far more informed opinions than I do, but from what I can tell, the relative importance of design vs. implementation (i.e. internal components) to sound quality varies from amp to amp, and probably from amp type to type. And I suspect that the relative importance of design vs. implementation varies among sections/stages OF THE AMP ITSELF.

I also suspect that the attribution of sound quality to design vs. implementation is highly speculative, except for manufacturers who have personally built many similar versions of the same amp over a number of years (e.g., Nelson Pass, Ralph Karsten), and so are in a position to have heard (a) similar designs with varying implementations, and (b) varying designs with similar implementations.

And finally, there is a level of analysis at which the distinction between design and implementation breaks down, but that is a bit of a philosophical quagmire.

Bryon

Can I take a VW Beetle and use superior parts and get a Bugatti Veyron? Because, if it's that easy, I think I'll do it! :)

Thanks for all the enlightenment everyone. I was really hoping for a discussion of the components in a $1000 amp vs those in a $50,000 amp, and how much they, vs the designer's intellectual property, contribute to the quality (and cost) difference. In other words, can I take a standard amp design, use the best available parts, and get the sound of the $50K amp. If I can't even get close, the sound quality must come from the circuit design.

But I see the way I asked it, isn't exactly like that.

So, besides proprietary circuit and transformer designs, how much of the quality difference between $1000 and $50000 amps can be made up by putting the very best components in?

All the other discussion is pretty enlightening too, thanks.

Jim

Stanwal, my friend was planning to do the same with amplifier he designed. Unfortunately this amplifier had tendency to oscillate in certain setups. I told him that it would be financial suicide if his amp will start damaging customers' speakers. He couldn't afford to buy many common speakers or spend a lot of time testing. Fortunately he gave up this idea. It is very difficult and costly to start such business and get any recognition.

Kijanki, I not only admit it but I second it :-)

Great line! Best regards,
-- Al

A friend makes tube preamps by hand; when I bought one in the late 80s the price was around $1000. He investigated producing them commercially but found that setting up an actual production line would result in them having to retail at $3000 to $4000. One at a time he could hand built them in his basement and check everything himself; but having a factory, reps, dealers , advertising etc would require a much higher price. And this was for a produce that had already been developed; no research and development costs. Granted that some seem overpriced; quality seldom comes cheap for any type of product.

Al, you're right of course, but you have to admit that my animal joke fits to expensive cable trade.

Many factors presumably contribute to that $50K sales price, many of them having been mentioned in the posts above. As someone experienced in electronic design (not for audio), one factor I would particularly emphasize is amortization of design and development costs.

Designing a sophisticated high performance electronic product involves many months, perhaps even years, of effort, involving multiple technical disciplines. For starters: Electrical design, including analog, digital, and power supply design, which are essentially different disciplines; mechanical design; thermal design; and in many audio products these days also software or firmware design. Significant payments to subcontractors and suppliers are also involved, not just for parts and materials but also for work by the subcontractors to design and develop custom integrated circuits that may be required, and to perform the non-recurring efforts that are required to prepare for production of major subassemblies such as the chassis, transformers, etc. Specialized test equipment and laboratory facilities will also be needed to support the design and development process.

All of those costs, and many others that have been mentioned, will have to be amortized across a total lifetime production quantity that is perhaps measured in dozens. Frankly, considering that and considering the limited market for high end equipment and the vast number of products that are available to choose from in any given category, what surprises me is not the amps that sell for $50K, but the ones that manage to sell for only $4K.

Now if you were asking about cables, that would be a different story IMO :-)

Regards,
-- Al

When I see people buying and selling $50k gear it reminds me this guy who successfully sold dog for $50k by trading it for two cats $25k each.

What's the difference between a $2.4 million Bugatti Veyron and a $12,000 Chevrolet Aveo? They both get you from point A to point B.

I guess it's just a good thing that you are not asking about the REALLY expensive amplifiers. Sheesh....

It would take me years of study to get to the point of building my own tube amp. That takes time and time and expertise is what you also pay for. Time is expensive and means a great deal to us humans. It has been said that to become an expert in any field takes at least 10,000 hours of study and time. 10,000 hours is a lot of money :-)

My tube preamp amp sells for $4000 and the parts including the case, tubes etc. cost far more than $400. Heck two capacitors in the unit cost $400 total!

I am also paying for the mind, ear and experience of the amp's designer. I love the sound of my preamp and it sounds far different and better to my ears then all other preamps I have heard. Are we wanting and willing to pay for this musical experience which is unique compared to other offerings?Yes, for me!

In addition to the parts, time, expertise, unique design and sound you are also paying for these real business costs;

- marketing
- building, tools, employees, benefits, insurance, all sorts of costs from utilities to toilet paper in the bathroom
- costs to travel, attend, set-up and work all the industry shows and events

What about the need to price with a sales profit. Yes profit. The builder needs profit as well as the retailer be it internet based or brick and mortar.

How many $4000 amps can really be sold today? This is not high volume business, so since the business model is not based on volume it must be based on high margin return. That is the business reality. If the builder is only going to sell 100 units, then they better make good money on each unit to pay all the costs listed above and make a living.

Seeing amps that sell for $4000 - $15,000 is certainly no surprise to me. It seems like a must based on the business model. Just attending the shows must carry an annual budget exceeding $50,000 for many.

Your question has many moving parts that must be considered to give this topic a full and realistic viewing.

As for SOTA attempts that sell for $50,000, well I can assure you that you won't ever be able to build these. The cost to market these and get to the shows required is very high indeed. If only a few dozen are sold, then the margins must be very, very high.

IMO, part of this pricing is hype. Boutique parts can be very pricey, especially if machining costs do not benefit from the economy of scale. Unless the manufacturer is copying a time proven circuit and modifying it, there could be significant time spent engineering the circuit, plus there are all kinds of esoteric build "tricks" that experienced designers/builders know and frequently don't share. Things like: wire routing,component layout, capacitive coupling, and stuff that's just plain weird like eddy currents and transformer core-saturation.
It may be a case of the 80/20 rule. You might be able to get 80% of the way there for 20% of the price, but the final 20% is REALLY gonna eat up some time and money.

"Excitable boy they all said"....all apologies to Warren Z!!!!

Also remember that generally the dealer makes more on the product then the manufacturer does. Yes the dealer has overhead, but the manufacturer has much more overhead. That is why some manufacturers tend to sell direct. By selling direct, it cuts out the dealer, and the mark up. It can be an inconvenience if the manufacturer doesn't allow you to return it and get you money back if you don't like it however. It also becomes an issue if you like to see it and touch the product before you buy it. That being said, some companies find dealing direct with the consumer a better, more cost effective approach.

The price paradox: Expensive gear adds to its own "excitement value" when you wait and buy it later (used) at a much cheaper price. It's also conversely true (for me anyway) that finding an inexpensive new thing that sounds amazing (good cheap DAC or a bargain tube amp) is also very cool. Maybe I'm easily excited.

Thank you all for the great discourse.

I don't know that starting an amp from scratch is better than getting something like a CJ MV-55 and trying to make it sound like the way high end.

So the parts (I am guessing) are 10% and a retailer marks up double in most cases. How much is the manufacturers and distributors profit? Isn't overhead part of construction costs?

Look at car prices; a Kia, Mercedes and Rolls Royce all have 4 wheels, a seat and will get you there but the price differential is enormous. Remember that construction cost usually equals 10% of selling price; the rest is overhead and profit. If this seems excessive look at the mortality among high end companies. The best product in the world requires distribution and advertising; things do not sell themselves. Also when you build your own gear your time is free; don't forget to factor that in.

The price of anything is always what someone is willing to pay for it. A $50K amp that has never been sold is not worth $50K, it is really quite simple.
OTOH some of the "best" parts prices (called components by manufacturers BTW) are higher than you might imagine . Look up Deuland capacitors for instance and think of multiples of that cost.
Labor/benefits costs also seems to matter greatly. I needn't tell you that some populus countries that are making most of the consumer electronics now, have very low costs. In fact all overhead costs energy, space etc. vary widely by place of manufacture.
Many things we covet are kept intentionally scarce and never marked down these have a quote a pride of ownership cost assiciated (Think Patek Phillipe, Cartier, ST DuPont) etc.) Certainly this phenomena is part of this hobby.
Don't forget patented proprietary circuit designs. A lot of manufacturers do not release diagrams (schematics) and code parts with their own numbers, making duplcation and reverse engineering very difficult.
Well that is a start but when you finish your perfectly built DIY $50K, for only ?$1K?? let us all know if it sounds as good as the original.
But in the end only the demand will ultimately decide the price.

Take your own advice, just build your own. After you learn to solder, of course. I'm sure it will sound stellar. Hell, start your own business and grab a piece of the highend pie.

I am surprising myself by asking the same sort of question lately about vintage after dragging an HK Citation 16 out of the closet manufactured in around 1975. A google search tells me that some audiophiles buying vintage and upgrading or having upgrades are very happy with the result. I don't know that I'm completely sold but I'm sure I've been undervalueing vintage. We might not need all the new amps, let alone a $50,000 one.

(oops, hit the post button too soon). These people approach buying audio the way they may buy a new car for themselves - they may ask their friends what they like, take a peek on the internet to see what folks are saying about it and they may also pay attention to the advertising. Will I be buying a Mercedes or a Jag or a Rolls?? To a real degree it may come down to status perhaps more than anything else, but, for them, attempting to get inside the heads of manufacturers long enough to second guess them at every turn is perhaps an unassailable task, so they willingly trust makers to do as they will. But, since these people live in a world of advertising and word of mouth (and word of mouth based on advertising) and little else on the subject, I think it instructive to see that their willingness to rely on what makers say fits like a hand in a glove to how such items are priced. So what's the difference between a $50k amp and a $10k one? For the well-heeled non-audiophiles like them, one of them must be more special than the other. For us, usually about $40k... :o) It's interesting to me that many of us audiophiles also fall victim the further up they move on the industry food chain. Once you've spent that $50k on a component it's that much more difficult to proclaim it to be significantly no different than that $10k one. Of course some audiophiles in these pages have been there and done that and have said to have later come their senses, but makers know that there's always that next audiophile that may take the bait. But, these days are hard times for everybody and $50k amps are reportedly indeed a tough sell, even at the very few number of units it takes for it to be worthwhile.

You're right that the total components don't nearly add up to $50k. What you're basically paying for, apart from good design and execution, is for the most part the sheer mystique that such a product has - especially the mystique that becomes attached to it as soon as it's given a $50k price tag. You and I are clearly not the intended buyers for a product as this. We've spent at least some time getting to know audio gear to some degree in some technical way or another...that is, we've bothered to in effect actually glimpse the little man behind the curtain of this industry. And why did we go there in the first place? Because we understood that buying outright the kind and quality of gear that we knew would likely be necessary for us from the outset was going to be more expensive than we could afford...so instead of giving up on our dreams, we decided to figure out a way to do things as cheaply as possible ourselves and 'learn from the ground up' to enlighten ourselves as to what it would take to do more for less. We may have that kind of time, but there are others who are perhaps married to their careers or whatever who simply don't want to have to spend so much time, effort and energy just trying to figure out where to start. What they want is for someone that they feel has already taken the guesswork of it all out for them. These people may feel it a privilege of their level of income that they indeed don't have to bother with any of that. Their approach to audio

Very good question!