[Big Bad Bob]

On 05/12/11 20:26, flipper so wittily quipped:
On Tue, 10 May 2011 01:09:06 -0700, Big Bad Bob
wrote:

On 05/09/11 21:36, flipper so wittily quipped:
but at the end of the day
if there are 200 man hours in a chinese amp, ( 5 men working for a
week ), the cost of wages is $128, and if materials in chinese prices
are also $128, then total cost of production = $256,and if you double
that for the profit made my whoever owns the chinese company, maybe
the Chinese Communist Party, then the export price might be $512.

All ridiculous assumptions, not to mention the cost of doing business
is more than simply COGS.

my point

That reply was to Patrick, not you.

heh, sometimes hard for me to follow.

was that the design would determine how many man hours you
need. there are really a LOT of variables in the cost of manufacturing
something. mfg engineers get paid HUGE bucks to figure out inexpensive
ways of building something (without compromising quality even). But if
you can have it all done via automation, so much the better. Then the
assembly production costs can become "COGS" eh?

COGS depends a lot on how one does the accounting. Like, for one,
whether R&D is amortized into the product cost or overhead margin,
often a combination of both.

My point was similar to your "there are really a LOT of variables" and
that simply adding up some parts costs and assembly labor doesn't
begin to paint the picture.

yep. being in the manufacturing sector (at various levels) helps you
understand that.

For one, it leaves out the entire financial, legal, and regulatory
departments dedicated to sifting through the bazillion regulations
socialists love to impose on everything and, of course, the ever
changing 'screw you' tax laws.

don't forget potential patent infringement and international regulations
and requirements for exported products (like ROHS for one). There are a
number of businesses oriented around getting 'approval stickers' on your
product. Look on a laptop power supply some time and all of those logos
(UL, ROHS, etc.) involve some company testing your product or reviewing
the manufacturing process (or both). From what I can tell, Chinese
manufacturers seem to have streamlined this process somewhat, so when
you outsource your manufacturing to them you can set up a 'package deal'
pretty fast and get all of the logos with it for an up-front fee.

Overhead costs can easily be more than the product, especially in
niche markets.

Final assembly and test
would take very little time for a reasonably experienced tech.

And what are you going to do with the faulty ones? Scrap or repair?

scrap is probably faster, if the cost is low, but testing and rework is
more likely since you could swap out the faulty component. But what you
have to do is set the process up to be foolproof: Board(s) + tubes +
transformers + chassis + front panel assembly + case with associated
screws, washers, insulated binding posts, crimp connectors, and so on.
Each component sent to you would either be pre-tested or process
verified, like if you bought them from directly from hammond or similar,
so the likelihood of fallouts is pretty small. If you can't get it
below 1%, you're probably NOT going to do well, and that's easily
achievable nowadays. "Final test" would involve a warmup, a test
signal, and a dummy load, possibly run by computer (so the tech just
plugs it in). If the board is bad, you swap out the board, then tubes
can be individually tested or you could use a test point on the board
and a 'special test unit' (computer driven - or someone like me could
program a microcontroller to do it for ya). Anyway, that's how things
are usually done by companies that are profitable.

That's not a simple choice, nor is the choice of when to do testing,
or how long (aka burn in).

You could have 2 stations per employee for testing, and rotate the
devices in. That would give you at least 10 minutes per device to run a
series of standard tests on it, and you would randomly take 1 of 'xx'
units out for more stringent testing by a reasonably trained tech.

Those costs have to be rolled in too. And how many do you predict will
be returned? Damaged in shipment? Are you insured?

heh, yeah, these are all factors. Normally you would initially make
assumptions based on gross margin percentage along with expected sales,
and then you burden the costs with 'standard values' based on normal
business practices (this would be presented to the board of directors
and potential investors), then add in your fixed costs and so on. So
yeah, this would be part of the 'sell the business' end of it. After a
time you would adjust your costs based on actual rather than 'estimated'
values. 'One man band' shops could (instead) grow slowly, doing the
work yourself at first, then hiring people as your sales grow.

It's not easy starting a business these days.

yeah, no kidding. fortunately you can't eat an amplifier. and let's
not forget, Oregon is further left than California.