Parts-Sourcing / BOM Price Estimating

[WyndBalduram]

So, sourcing parts has been a bit of a pain in the neck for me.  I like to know about how much I should expect to spend on a project before I start ordering components.  So far, the only way I have thought to do this is to manually look up each individual component and compile it in an excel spreadsheet.  This takes awhile, even though I can save the spreadsheet for later so I know how much a particular project will cost.

What are some of everyone's different methods for getting price estimates on their builds?  I am wondering if there might be something easier / quicker or if what I'm doing is a waste of time.

[R.G.]

There is no substitute for listing all the parts and looking up the prices.

A good thing to remember is that compared to boxes, jacks, switches, and knobs, electronics can be estimated at some low amount for the whole electronics parts kit. The money is all in the mechanics and controls, to a first approximation unless you're using some expensive chip inside the box. If the purely electronic parts were free, it wouldn't get much cheaper.

And there's a way to speed things up. Look up the price for a 1/4W resistor, ideally in packs of 100 to 200. Buy the pack, and put in the $0.02 to $0.04 for each resistor you use. Caps are about $0.10 to $0.40 each, with 1uF to 22uF electros at the $0.10 end and big electros and large film caps at the other end. Bipolar transistors are $0.08 each, JFETs are $0.25 each, CMOS ICs are $0.50 each, dual opamps $0.30 each. If you simply use this estimate, and total up the enclosure and controls, you'll be very close.

If the number of resistors or the price per each makes a difference, you can't afford it on a one-box basis. If you're totalling BOMs for manufacturing, you need to go do the work to get familiar with the business side. Forums are a poor place to get business advice.

[tubegeek]

Just to add one bit of advice: you can get assortments of, for example, resistors containing 20 pieces of every useful resistance value, for dirt cheap - eBay is one source. Deal Xtreme is another. Then when you do your next project, you'll have the resistors on hand - nice convenience - and dirt cheap - nice price.

Stocking up when you see good deals can really make things easier in the long run, you can get more done when inspiration and opportunity strikes, and never rarely have to go to Rat Shack for that one missing part, as an added bonus.

[PRR]

Resistors are 12 cents each (25c in bulk because you will be 85 years old before you use up all those 39K and 75K resistors you bought for 2c each in strips).

Generic transistors 30 cents. Capacitors 50 cents. (Maybe 20c to $1 depending on size and source, but it averages out.) Most chips 50c to $1, some MUCH higher. Pots are a buck, $3 with nice knobs. Jacks are 75c (or a baggie of 20c cheapies that don't last).

There's sometimes some part which is MUCH more. The BBD delay chips, the 3080 OTA, Vactrols, magic-eye tubes, cold-war transistors. Perhaps the key to budgeting is to know what parts are "common and cheap", and what parts are real money-sinks.

As R.G. says, the box can cost a lot more than everything else.

At retail, assume $10 per knob or jack, *times* a multiplier for the brand-name. JBL's fixed-install 6-in mixer has 20 knobs, 10 jacks, so $300, but the JBL badge is worth 2X so $600. Berenger has worked to make their badge worth 0.25X, so their 6-in is $75. A pedal with 2 jacks and 3 knobs is worth $50 ($29-$500). The all-up parts-cost, no pay for labor, is half of that (Banjo World marks-up at least 30%, and the rest is labor and shipping).

[R.G.]

The funny thing about accounting is that how you consider the cost of parts does depend on when you use them.

Let's say you need some 100K resistors for a project. You have to ask yourself - will I ever build anything else? If the answer is no, buy them one at a time. If the answer is yes, make the list and buy them in 200 packets.

Mouser sells a 100K 1/4W 5% carbon film resistor for $0.10. It sells packets of 200 for $1.90. So if you buy 20 100Ks, you would have paid more than for a 200 pack. So in 200 packs, the first one costs $1.90. two cost 0.95 each. Three are $0.63 each. Four are $0.47 each. Nineteen are $0.10 each. And after 19, the cost is ... zero. 

Buying in 200-each packets constitutes a bet that you'll use more than 19 in a short enough time to care. For 1K, 10K, 100K and 1M, this is almost a sure thing if you build more than one pedal. Odd values are taking a different chance.

The metric I use is that if I use it once, I'll probably use it again. So when I run into a new resistor value, I buy 200. The 39Ks and 75Ks last forever for the first $1.90, and the others are $0.019 each. Resistor values I have never used never get bought. I've refilled 1K/10K/100K/1M/2.2K/3.3K/4.7K, all the usual suspects multiple times, and my resistor box now contains at least a few of every value I've used over the last 20 years. So in most cases, building a new pedal costs *zero* for the resistors used because I've already bought them at $0.019 each and there is no additional money outlay.

Sure - some of them are going to cost me $0.30-0.40 each. But I make it up on the volume.    Then there's the time saved when I want to try something. Chances are I have the resistors I need already here.

I guess my advice is - if you're making one pedal, worry about the cost per part. As soon as you know you're doing more than one, start planning for the long haul.

[Ice-9]

If all you want to do is build one kit board then buy  a kit of parts, if you are getting into electronics and experimenting the you will not regret buying a couple of component draws and a full order of 50-100 each different caps and resistors. For less than half the price of one MXR pedal  (sub any other brand here)  you could have all the caps and resistors you need, just add specific IC's TR's to each of your builds.

The biggest expense in pedal building is the hardware like enclosures pots etc , on one build I make the knobs alone for one pedal cost £15 for 5 knobs and that is nearly more than the cost of  everything else.

[WyndBalduram]

R.G. (and everyone else), your advice is enlightening.  I did not realize that buying certain parts in bulk actually reduces the cost so dramatically.  I knew buying 10 or 20 at a time cut the cost a little, but I guess I didn't pay attention to how much the price fell closer to buying 100 or 200 at a time.  I have noticed though that hardware tends to cost the most, especially enclosures.  I honestly wonder why enclosures cost so much, but maybe my ignorance on how they are manufactured has me deceived.

I have so far built several different pedals for my own personal use - I like making them and have adopted the hobby.  I would like to design a few pedals with the intention of selling them after I get some more experience under my belt; though from what I understand, unique designs are few and far between.  I think I would like to focus on boxes with multiple effects in them, but that is probably for another topic.

As far as cost of parts is concerned, can it be safely assumed that the less a component costs - especially hardware, such as a jack or pot, for instance - the lower the quality of said component?  I have ordered most of my parts from Tayda, and all of those parts, except one of their 3pdt stomp switches, have worked without any problems.  Then again, I have not used / abused the pedals I have built with their parts long enough to know how long they will last or how much abuse they can endure.  That being said, their parts are inexpensive - but does that mean they are cheap (quality wise), and would it be smart to invest in their parts for larger scale operations, or to go with another company such as Mammoth, Mouser, Digikey, etc?

[PRR]

> ...cost of parts does depend on when....

Buying parts "for future need" is an Investment, same as a bank deposit or a share of Sony stock.

My 1970s stash was running out of 10K etc, so I got a big assortment for $15, perhaps a penny each. I actually used less than 50 in a few years, and then my life took a big shift. The box was left behind, and I've hardly missed it (been doing bigger electrics).

So it turned out I paid 30 cents each (but always had a value I needed, sometimes more important than a few pennies).

If I'd brought the box along but did 50 resistors a year, it lasts 30 years. If I put the $15 in an interest-bearing account (not in this recession but in normal economy) it would be worth $60 in 30 years. Or I could pull $0.75 interest each year and buy a dozen resistors. The investment in resistors is not a bad investment, *assuming* I keep-up needing a dozen resistors a year for 30 years. (And at a certain age, that gets less likely; also a question of if leaded-resistor technology will still be interesting in the year 2044.)

> the time saved when I want to try something.

Yes. If I have a 4.7K and a 7.5K, and decide I need to try a 6.8K, that's $5 gas to the Shack if they even have resistors. And I don't go to town for a week at a time. Or $5 shipping from someplace 3 to 13 days away. While waiting I will lose my train of thought, could even lose a client or sale. And if I have to pay $5 even a few times, it's worth having 1,423 excess resistors for my estate sale.

> parts in bulk actually reduces the cost so dramatically.

No"body" makes resistors. Un-touched by human hands. Mechanized jigs assemble the body and glop and leads and tapes them into 1000-part reels for mass-production mechanical stuffing. While the machines cost a lot, labor is nil, production is huge, so the per-each cost is tiny. When you buy a 1000-reel.

Now you want to buy just one? Steve or his staff have to take-down the reel from a shelf, count "one", snip-clip from the reel, check it off on the pick-list, make sure it gets in the box. There is FAR more paid labor in one resistor than a 1000-reel. Also in 30 years Steve will still have 456 of 6.8K un-sold. That's why penny resistors cost 12 cents each retail. (But now that resistors are reeled, staff can cut-off 10 or 100 for nearly the effort of one, and there won't be so many left-over down the years, so the price *does* drop rapidly with quantity.)

> less a component costs - ...the lower the quality of said component?

Yes, of course; but sometimes no.

Supermarkets offer frozen pizza, peas, and drugs in both Fancy Brand and House Brand, and very often they are the SAME thing at different price. I used to buy a lot of Dristan, and I'm sure the generic came out of the same press. So sometimes the price depends on who is selling it.

Also sometimes there is a design revolution. Someone looks at the way they have always done it, flashes on a simpler or less-work way, and gets an equal/better product for less money. The early Model A had forged body brackets, the kids showed Henry that a stamped bracket was just as good and a lot cheaper to stamp-out. The SwitchCraft jack was a revolution on telephone operator jacks; not as good but a LOT cheaper and good-enough for many non-telco uses.

Resistors: there is NO serious market for bad resistors. Resistors are THE most common electronic part. They have to be cheap, but they HAVE to be reliable. If a penny resistor isn't what it says it is, the TV/phone/PC doesn't work when assembled, it has to be diverted to a Rejects room where a person with a brain and a meter has to find the problem, which will cost at least $10. Or 1,000 times the cost of the resistor. If a mass producer finds more than one bad resistor per truckload it makes sense to find a more reliable supplier even if the price is more.

That's for big buyers. Every once in a while *here* where people buy from odd suppliers we hear of a mis-marked or open resistor. Even if you assume some folks can't read a resistor and other folks can't solder, I can still picture a back-alley market for "sweepings", resistors which were kicked-out of the automated machinery. Thrifty merchants could re-pack those parts (with or without testing) and sell them to less fussy customers. It's a hard way to turn a buck, but there is motivation everywhere.

Now when it comes to "mechanicals".... IMHO a good jack, with a robust bushing, spring fingers of the correct dimensions with metal that stays springy most of a lifetime and doesn't tarnish much, should cost $20. But bad (cheap) drives-out good. SwitchCraft ain't what it used to be. Neutrix seems to have lost control of their trademark. Telco quality jacks are hard to find, and telco relays (40 years steady use and millions of operations) seem to have all been melted-down. I admit I buy $5 and $3 jacks, but I've also tossed-out a lot of those. But I _am_ fussy. I apprenticed in larger systems with hundreds and thousands of jacks. You can spend a lot of time debugging bad connections in jacks. A typical guitarist, even with large pedalboard, does not have so many jacks and may be a little more tolerant. But there ARE jacks too crappy to take seriously. Since the sellers never say "crap jack", some blind-buys and toss-outs are part of the process.

That's all for building one pedal at a time. Folks like Peavey have a whole Purchasing department which evaluates samples and maintains an Approved Parts list which designers must work from. Trying to get a special opamp into a design, when TL072 and 5534 are already Approved and regularly stocked, takes a major begging effort with Business justifications.

> go with another company such as

For *pedals*, don't overlook Small Bear. Tayda and Mammoth focus on low price, which may be fine or disappointing. DigiKey Mouser et al have *everything*, parts you don't know what they are, and they do NOT know what you are doing. Steve does, his customers tell him when a part is not satisfactory, and he services a few high-volume (relatively) clients who depend on getting the "right" parts at reasonable price.

[tubegeek]

(much rather-brilliant explication snipped)

For *pedals*, don't overlook Small Bear. Tayda and Mammoth focus on low price, which may be fine or disappointing. DigiKey Mouser et al have *everything*, parts you don't know what they are, and they do NOT know what you are doing. Steve does, his customers tell him when a part is not satisfactory, and he services a few high-volume (relatively) clients who depend on getting the "right" parts at reasonable price.

And I can also confirm that Steve is the kind of guy you'd enjoy sitting and having a pint with, no small concern. He's about half as nice as his wife - which is really saying something I wish I was 1/10th as nice!

[duck_arse]

I have so far built several different pedals for my own personal use - I like making them and have adopted the hobby.  I would like to design a few pedals with the intention of selling them after I get some more experience under my belt; though from what I understand, unique designs are few and far between.  I think I would like to focus on boxes with multiple effects in them, but that is probably for another topic.

if you're going to do this, you'll need to prototype. and that means a breadboard. and then you need parts to plug in, swap around, try out. so, now you needs a multi-drawer cabinet, w/ all those in-between resistors, 10 of each. (the 39K lasted me til 50.) and caps. and transistors. and so on. as you go, you just end up collecting parts, or making lots of ordering runs.

the more breadboarding you do, the less pots/jacks/knobs/enclosures you waste.