[ericj]

Quote:

Originally Posted by Dave_M ok, it looks like alot of people feel strongly about this so how about if i put next to the cost of the amps, the typical cost to build a complete amp? That includes everything. I can put it next to it in brackets with a note saying somthing like "price in brackets indicates typical cost for a complete amp". Can anyone think of a neater way of doing it let me know. If you think its a good idea then list some costs for the amps and i will fill them in.

Maybe have a collumn for "missing from the BOM" for critical parts that aren't listed (pot, etc) and additional collumns for likely enclosure costs and likely power supply costs.

And maybe you should specify that these are the minimum costs if you go with the BOM, available kits, and popular enclosure methods.

I think it's important for some people to realize that while a dynahi maybe really does need the dynahi power supply - or at least benefits from it greatly - some amps don't need you to be particularly fancy.

The M3 is designed to be very forgiving of the power supply, so a TREAD and a thrift-store wall-wart is just peachy.

I think i spent $5 on my M3's power supply, because i don't need a printed board to build a power supply and my lm317T's cost me 30 cents each (real National parts, too).

The 30vdc 500ma brick cost me $2 at the thrift store - used to belong to an HP printer. it's held together with screws instead of glued or fused together, so i opened it up, replaced the filter cap, and put snubbers across the diodes.

Sure, the trafo in that brick is a laminated EI core and puts out plenty of EMI, but it's a few feet from anything that would care.

So, it would be reasonable to estimate that the minimum power supply cost for an M3 would probably be $17 for a TREAD kit and $10 for an elpac wart.

But many amps really should have more to the power supply than that.

(This is not to say that i don't believe in quality supplies - I have a nice 30vac 2a toroid that I'll be pairing with an lm338-K in a steps-alike configuration (mains filtering, etc) just for the hell of it. But the M3 sounds great with $5 worth of bodged & recycled regulated supply.)

 

[NeilR]

Nate,

I think everybody understands that this is only a starting point and not a final cost. If "reasonable" (not boutique) but good quality parts are selected and consistency and a well thought out methodology in parts quality selection is adhered to, it does have some meaning because it shows what a very basic sample will cost, before adding tax, tags, title, etc. In the case of a PPA, for example, it should not include a fancy battery/charging board because that falls under "Power Supplies", which is a separate item.

Also consider that, although Tangent designed the PPA and battery board as an integrated unit, I don't see a lot of them in the "Post your Build" threads. That should put some weight on the decision of what to do with the battery board. In this specific case, if it were my list, I would list that as a separate line item to resolve that problem, but that's just me.

Most of these standard BOMs include pretty similar parts quality, such as Pana FC or FM caps, or Nichicon UPW, which is the FC line. They are all very close in price. Unless you go beyond Vishay Dale resistors, using $0.10 vs $0.16 resistors is not going to change things very much, nor $0.30 vs $0.40 for caps.

Opamp choice is problematic here. One approach is to specify OPA2134, and leave it to the reader to see how much he wants to tack on. But that is not fair because an M3 deserves a better opamp, while a Cmoy should not be speced with an OPA627. Personally I think a Cmoy should be listed with OPA2134, and M3 should be listed with AD8610/8620 because that is the "base" spec for the respective amps and I think those Opamps are in line with the relative scope of those two projects. Same with the PPA; it was specced with AD8610 as the primary chip and the amp is a high end product deserving of that spec. Just examples to apply some sensibility. Something like a Pimeta is a more difficult choice because that amp falls in the middle of the extremes. Here Tangent specs OPA2132 as primary and AD843/AD8620 as alternates. Hmmm.... tough call here. I would go with the higher side but that's just me. Maybe the 8610/20 should be used just for consistency since one use of the list is "do I really want to go for an M3 or PPA or should I settle for a Pimeta (for cost reasons).

In terms of casing, the case should be commmensurate with the primary intent of the amp. The mint was primarily intended for an Altoids tin and I think it would be proper to spec that case. For the Pimeta, I think it should be speced with a Serpac H65 or H67 because that board was cut to the precise dimensions of that case, as was the speced primary parts such as PSU caps. I know I will get an argument here because a lot of people built "desktop" or "transportable" versions in Hammond cases. However, the Pimeta's reason for being was to pack as much functionality and quality into a case that would fit in a pocket, to the extent that it was notched specifically for the Serpacs. I think it was Tangent's intent that a PPA be built for a more desktop and less portable role. Tangent's "amp line", if you will, is a progression of portability compromises, going from Mint to Pimeta to PPA. I think he did a great job delineating his lineup and I think that should be respected in the speced cases. In any case, (pun intended), the specced case should be mentioned in the list because cost more or less follows case choice to a great extent. The important thing is that if it fits in a Hammond, it should probably it should be specced with a Hammond, just to keep things consistent (Pimeta, etc., nowtwithstanding due to intent).

People are going to build Mints in Hammond cases and they may even send their panels off to FPE for laser engraving but that is their choice to go above and beyond the basic spec and that should not be the concern of this "base sticker price" list. It is inherent in DIY that few build exactly to spec and therefore few will build for the "sticker price". What the list does do is to give an overview of the relative cost, using consistent parts quality commensurate with the scope of the project, of different amps. It is a starting point for further research. Nothing more.

As far as jacks, pots, and other "casing parts", as I call them, some standard to follow needs to be determined if the list is to make sense. For example, if you specced $30 a pop high end Cardas or WBT jacks, that would be inappropriate for the intent of the list (IMO of course). I don't think $0.88 mouser RCA jacks should be speced because that part is not commensurate with the rest of the parts lists, such as Panasonic FC/FM caps and decent opamps. I used the cheap Cardas jacks for my Millet ($6.90 a pair) and I think that is a reasonable high end for a standard. Not because I happened to source that part, just that $7 for a pair of jacks gets you a good jack, much better than an $0.88 jack, and to the point beyond which the price/performance benefit diminishes greatly. Reasonable people can quibble if $4 or $7 is the right price for a pair of jacks, but the difference is relatively small as long as you stay in that range. Here I am just talking out loud and giving Dave M some ideas on how to approach this. The important thing is that, for example, the PPA and M3 should be specced with the same jacks because the amps are in the same league. A Cmoy should probably not be specced with the $3.69 Digikey/CUI minijacks. That is overkill for the spec. The $1.25 switchcrafts are probably more in line with the intent of that amp; they are good jacks and in line with a basic $25-$40 project.

I bet your $150 Millet was very close to spec. As to your $400 Millet, I want to see pictures and a BOM because I don't think I can figure out how to pack $400 worth of stuff on that board

Obviously, though, your gold plated Millet should not be the model for these estimates (nor even enter into the discussion) and anyone who puts that much into a Millet presumably knows what they are doing and is well beyond the simplistic purpose of the estimates.

P.S. Tangent's part lists DO make specific recomendations on cases, jacks, etc. He also includes alternates, in some cases because different people will use different jack types (mini vs 1/4" phone and RCA for example). However, he does stay within a general price point, speccing good quality, but not "boutiquey" parts. I don't think he lists $30 jacks as alternates and certainly not as primaries. He may list a $2 jack and a $4-5 jack or whatever, but I think he follows a very consistent price/performance strategy. He never specs (pricewise) beyond, say, a Hammond 1455 series case and I think that is a good reference point here. If you put all of Tangent's primary suggested parts lists together, I think you would have a very consistent and comparable set of costs.

I am sure that at this point, Dave M wishes he had never thought of this, but such is the nature of this sort of thing

A suggestion to Dave: If you try to "job out" some of the detail work here, be sure that you set the ground rules before you do so. Otherwise you will get chaotic results....

 

[SiBurning]

typical cmoy... $50
getting it right the first time... priceless
<standard_disclaimer>tools additional, cost overruns to be expected subject to 4% annual cola and congressional pork reapproval</standard_disclaimer>

 

[n_maher]

I'm going to try to keep this short as I've said most of what I have to say about this topic... All that I will add in the general sense is that when I do a BOM it's only for a new project where I feel it might be important for potential builders to decide if they want to persue the project. 99% of the information we're discussing in this thread is already here in other posts, you just have to use the now great functioning (thanks jude

) search button!

Quote:

Originally Posted by NeilR I think everybody understands that this is only a starting point and not a final cost.

Again, I think you'll be shocked at the assumptions that some people will make regardless of what sort of disclaimer you put on the information. How do I know? Let's say I've fielded my fair share of PMs on the topic, and I reply to them all by the way

.

Quote:

Opamp choice is problematic here.

Exactly one of my points, there are crucial parts to each amp that can vary the cost significantly.

Quote:

Personally I think a Cmoy should be listed with OPA2134...

Exactly my point again, this is "your" idea of what each amp should use. As you noted the further you got into your examples it gets more and more muddy.

Quote:

In terms of casing, the case should be commmensurate with the primary intent of the amp.

Who am I to tell someone else what the primary intent of a DIY design is?


Quote:

I bet your $150 Millet was very close to spec. As to your $400 Millet, I want to see pictures and a BOM because I don't think I can figure out how to pack $400 worth of stuff on that board

What spec??? And the $400 Millett is certainly more than just the cost of populating the PCB, but price out a few Black Gate caps and you could probably hit $400 with just the board parts. Add a stepped attenuator and you'd hit the mark in no time flat, all without any gold plating

. Strictly speaking the $400 Millett isn't done yet so here's a $300 Millett - soon to get a price bump due to "new" buffers...

Quote:

P.S. Tangent's part lists DO make specific recomendations on cases, jacks, etc.

Yup, he does a great job of giving builders a baseline of suggested parts that they will need to complete a project. But what he clearly does not try to do is give someone an idea of how much it will cost.

And just one more time for those who've missed the first 2 times, I'm not trying to say that giving people a general idea of the cost of a project is a bad idea, but rather that it's liable to cause a giant headache for someone and if people would just take advantage of all of the knowledge that's already on this site might not really be necessary. Some of the projects being discussed here (M³, PPA, Dynahi) shouldn't be undertaken by someone who can't figure out how much it's going to cost.

Ok, enough of me for one night time to kick back with my $300+ PPA V2 (that still doesn't have a case

) and enjoy some tunes on my Senn 600s...

Nate

 

[Tril]

The price of the Dynahi psu for 140 is without the transformer. You should add the price of a typical transformer to it.

 

[jerb]

what about minimum part costs? sometimes you need to buy 5 or 10 of something to make a minimum order, should this be reflected in the cost?

are you going to match the ressistors? then you need to buy more than the two the Board needs

 

[kdaq]

Quote:

Originally Posted by n_maher Ok, enough of me for one night time to kick back with my $300+ PPA V2 (that still doesn't have a case )...

Hehe, still uncased? Eh, don't feel bad, I'm groovin on a half-cased Dynalo myself.

This thread raises a cool idea. Personally, I like it. I certainly understand some of the concerns that have been raised about interpreting prices in the correct light - that's the key, really. Emphasis can be placed on this point; that this is just a rough metric, and 'real' build costs will vary wildly.

 

[NeilR]

Nate,

I have seen comments on this forum to suggest that no one should be pointed to an answer or given an answer that has been answered here in the past five years, they should spend hours searching for it. I am not saying you have ever done that, I am only suggesting that this attitude is somewhat cultural here.

The M3 development and build threads total 85 pages.

Your Millett Hybrid Construction Forum currently has 831 posts in 84 topics. I recently posted something about the pad sizes for the i/o connectors and you responded to me that it had been previously discussed. Sure was, somewhere in those 831 posts I did find an offhanded comment about pad sizes for those connectors. I am not criticizing you or your work on diyforums; I am only pointing out an example of how difficult it can be to locate important data that is strewn through months or years of threads on these fora.

I am trying to say that, even after zeroing in on a specific project, it is very easy to miss important information. If one were to attempt to consider a pile of projects in order to select one, the amount of reading necessary is prohibitive for anyone with a day job or anything else occupying their life.

I KNOW that many people must come in here looking to get started in DIY and throw up their hands and either walk away completely or find some other venue to get a start. For a newcomer it can be very intimidating. You started a year ago or so, and just in the past year there are a number of new projects. Each year there will be more and more, resulting in more and more chaos and confusion for newcomers.

I truly believe that anything that creates a little order out of this chaos will do more good than harm. Yes, there will be PMs mailed about from newcomers who discover mission creep or who take the estimates to mean final invoice price despite all the disclaimers. But if the attitude is "publish nothing that might generate a complaint" then taken to it's logical conclusion, nothing will ever be published because otherwise someone may misinterpret it.

No, Tangent does not publish a project cost per se, he does everything but that. The only thing missing is the price for the specific part numbers and quantities mentioned. I don't think I took any liberties in divining his intent for the various projects; in most cases he puts it in black and white. If you lay a Pimeta board in a Serpac case, the intent is pretty obvious even if he didn't say it in writing, which he basically did anyway.

Nate: Do you think $77 is a fair estimate for a Pimeta?

 

[kdaq]

Quote:

Originally Posted by NeilR I truly believe that anything that creates a little order out of this chaos will do more good than harm. ... But if the attitude is "publish nothing that might generate a complaint" then taken to it's logical conclusion, nothing will ever be published because otherwise someone may misinterpret it.

There's a line between science and art. Some would call it a fine line.

On the surface, it looks like naming a typical DIY amp build price is a science, and a straight-forward BOM-reading one at that. Due to sheer amount of variables (not least of which being differing conceptions of what 'typical' means), though, I say it's closer to an art.

As I stated earlier, I believe this exercise is useful. But, 'order coming from chaos' can only come if everyone is interpreting this information the same way. If not, then things might end up more disorganized then when they started out.

Not that anyone needs to speak for Nate, but I think he's actually trying to help organize, by speaking from his experiences. Which, incidently, is a good thing to do when dealing with things that are hard to predict and quantify (like social tendencies).

 

[NeilR]

Quote:

Originally Posted by jerb what about minimum part costs? sometimes you need to buy 5 or 10 of something to make a minimum order, should this be reflected in the cost? are you going to match the ressistors? then you need to buy more than the two the Board needs

OK... who among us is so anal that they actually buy FIVE of each resistor to match a pair????

hehe.... I always do that

Seriously, though, you have to keep in mind that...

a) The cost will be somewhat theoretical. If exact precision is required, then this is a hopeless exercise.

b) You do not HAVE to match resistors, just like you do not HAVE to buy Cerafines to get a good working project. As an aside.... I would defy anyone to pick, in a blind test, between two identical amps, the difference being only resistor matching. Does anyone over buy and match capacitors? I have never seen that discussed here. Why not? The tolerances are much greater....

c) I have found that minimums might make two $0.16 ceramic caps cost me $1.60 because I have to buy 10 instead of the two I need. I don't see it that often that it significantly affects the project cost.

d) Most importantly, your issues will, on average, affect each project equally in proportion to it's complexity and by extension, it's "base cost". It might add 5% to the cost of the project but that is ok because it should affect each project equally.

A bigger problem for me has been spec'd part numbers that currently have a minimum of 100 $5.00 parts, making it impossible to buy the part. Presumably the minimums were raised subsequent to the publishing of the parts list. I have a done a lot of part chasing because of that.

Don't lose sight that the primary objective is to place the various projects at their relative cost levels. It is a rough estimate, not an offering price or a kit price. The absolute, actual precise sticker cost is beyond the scope of what you are trying to do.

 

[NeilR]

Quote:

Originally Posted by kdaq There's a line between science and art. Some would call it a fine line. On the surface, it looks like naming a typical DIY amp build price is a science, and a straight-forward BOM-reading one at that. Due to sheer amount of variables (not least of which being differing conceptions of what 'typical' means), though, I say it's closer to an art. As I stated earlier, I believe this exercise is useful. But, 'order coming from chaos' can only come if everyone is interpreting this information the same way. If not, then things might end up more disorganized then when they started out. Not that anyone needs to speak for Nate, but I think he's actually trying to help organize, by speaking from his experiences. Which, incidently, is a good thing to do when dealing with things that are hard to predict and quantify (like social tendencies).

You're right, there is more art than science here. I think the designor's BOM should be given the most weight because it may be a more objective approach. There is no "ultimate truth" here. I say this after Nate called me on my RCA jacks a few posts above. The problem I have with blindly using the BOM is that one designor might spec a $1 part and another spec a $5 part for what is essentially the same level and otherwise quality of design. Pick your poison.

You can lead a horse to water.... the important thing is that the various costs are based on a single consistent set of criteria, whatever that criteria is, and the criteria is published with the list. In my RCA jack example, it might not matter much if an $0.88 jack is spec'd or a $3.50 jack as long as the BOMs are consistent in specing the same jack.

The alternative, as Nate suggested above, is to search the forums for cost estimates. The problem with that is that if you search 'PIMETA PROJECT COST' you will get 34 threads and probably 15 different opinions on the cost of a Pimeta by 15 different people. You will also get 19 discussions of everything other than Pimeta costs but that's a problem with generic searches

Few, if any of those 15 widely varying opinions will state the basis for their estimate. I think in practice, there are more individual opinions on the "base cost" among all the members here (hence much confusion) than interpretations of a single consistently applied set of rules.

 

[blip]

What if we standardized on one supplier... say DigiKey or Mouser... On each BoM we could try to get all the parts from that source thus standardizing the majority of common parts (resistors, caps, jacks etc.).

Of course, for many projects we would need to order parts from other locations but at least by standardizing what we can it would even the playing field a bit.

 

[Francis_Vaughan]

Can I suggest a totally different approach?

Everyone who has built any of the listed amps should be encouraged to submit an entry to a poll that builds a set of statistics for the price of the amp.

The builder needs to provide the year it was built - so that some basic price normalisation can be done, and should also rate their build on a scale that goes something like this.

1. A real cheapskate build. Scrounged parts, second-hand, nasty case, etc etc.
2. A pretty cheap build. Generic quality parts, but all new.
3. An average build. Used better parts where needed, but I was watching the cost.
4. A good build. I used good parts, and spent maybe a little more than I might have.
5. A premium build. I used premium components, worked really hard on a nice case, knob, connectors. It is a build I am proud of. I was happy to spend the extra money for the result.
6. Seriously premium build. I went for the best I could reasonably source everywhere. The money was not a concern.
7. Utterly silly build. I simply went nuts and my wallet still hurts.

*. The designer's BOM and recommended parts. I didn't vary much from this, and if I did it was only to source a compatible rather than superior item.

The poll then lists the normalised average cost in each category.

The poll also invites the builder to place catagory * on the 1-7 scale. In other words, to rate the designers BOM selection.

 

[n_maher]

Quote:

Originally Posted by NeilR Nate: Do you think $77 is a fair estimate for a Pimeta?

For a single buffer per channel, panasonic pot, run-of-the-mill components, yeah that's about right. But I'd round it to $80

. Let me know if you want the mother of all PIMETA spreadsheets, I didn't design it, I just have a copy.

And if there are folks try to compile this list of amp prices and have any questions about how much I've spent building my various amps I'm happy to try to answer, just please realize that sometimes I try to forget just how much I've spent...

Nate

 

[NeilR]

Nate,

Are you talking about "Sanaka's Pimeta Parts List"? If so I have that. that was a great roadmap to make sure my $80 Pimeta cost me at least $225

Combining this with Francis's comments, it would be interesting to see his suggested survey. Based on what I see posted here, there are more $250 Pimeta's than $80 Pimeta's. Has anyone actually built an $80 Pimeta? His survey would probably have the unintended result of convincing people they really had to spend $250 just to keep up appearances with the neigbors