smt costings

Subscribe to Thread

Hi Guys, I have been given the dreadful task of looking ove... - Mar 30, 2015 by Daza

Seems like you've got it about right....though, it's a rathe... - Mar 30, 2015 by robgd3

Cheers Rob, I thought that I wasn't far from the true answe... - Mar 30, 2015 by Daza

I've always hated using machine specs....they're always best... - Mar 31, 2015 by robgd3

Hi Robgd3 for CEM work this task is a doddle especially whe... - Mar 31, 2015 by Daza

All you have to do is to go on the floor when the machines a... - Apr 01, 2015 by Evtimov

Daza

#73750

smt costings | 30 March, 2015

Hi Guys, I have been given the dreadful task of looking over our SMT costing to make us more competitive in the industry, I am suffering with the cost per comp, I am assuming that the equation: Loaded cost divided by placement rate should land me somewhere near where I need to be, I am looking for further suggestions that would tie up any doubt in my assumption....any other guidance would be gratefully received.

reply »

robgd3

#73751

smt costings | 30 March, 2015

Seems like you've got it about right....though, it's a rather complex equation. Loaded labor cost is only a part of it, if you're trying to estimate down to the last cent.

You have to account for energy utilization, space utilization in the factory, set-up/break-down times (vs. run time), average up-time, maintenance time, and historical down-time (lost production), and consumable (MRO) costs (nozzles, filters, etc).

That will get you to the closest true answer. Though, unless you're in a pretty high-run rate and high sales volume company, I think getting to that number can be a little ridiculous. It all depends on what your bean-counters want to know, and what is trying to be accomplished.

cheers, ..rob

reply »

Daza

#73752

smt costings | 30 March, 2015

Cheers Rob, I thought that I wasn't far from the true answer....I figured that if we were to use a "first piece assembly " quote to get a first off through smt production, This would include start to finish process....e.g Programming to first board built and inspected on a time spent basis, and then price as discussed above by working out a price per part scheme.. This is the first time I have been tasked with this and I pray it is the last... Unfortunately for me I am working with no SPC data for average placements and we all know that machine specs are synthetic, I am told to use machine spec with a deviation of 25% - 40%.....mmmmmmm

reply »

robgd3

#73756

smt costings | 31 March, 2015

I've always hated using machine specs....they're always best case scenarios, and nearly never capture actual through-put rates.

Personally, I don't like using first-piece rates, either...unless you're looking for worst case scenarios. Usually, the first article runs much longer than subsequent ones. It's a good data point, especially if you're in a high-mix/low-volume environment.

For time-studies/estimates, I've always tried to use a board from the middle of the run. That way, first piece issues aren't cluttering up the data, nor last piece issues (reel changes, etc). Then, of course, use the total time including set-up and break-down to capture actual utilization.

In the end, it comes down to how granular you want the number to be. A couple of life-times ago, in a very high-mix/low-volume shop, I had estimates down to half shifts (we had realized that nothing takes less than four hours, no matter what we did); so, all capacity planning and cost estimating was based on half-shift utilization. Worked great there, but, we hadn't gotten to the point of needing to care about hour by hour productivity of either our machines or our employees.

cheers, ..rob

reply »

Daza

#73757

smt costings | 31 March, 2015

Hi Robgd3 for CEM work this task is a doddle especially when I had line manager software at my disposal....This is like pulling hair, If I have worked out how quick my machine can place a certain comp and then match that to the BOM it should give me an idea of what time it takes to build one board....divide that one board time (complete from printer to aoi) to the amount of boards that sit on the line up until the oven and that should give me an actual board build time as a full process.... e.g how much time passes between each board coming out of the oven= build time per board..... am I going crazy or is the norm.....

reply »

Evtimov

#73758

smt costings | 1 April, 2015

All you have to do is to go on the floor when the machines are running and see yourself what the speed is. All machines show this data and even more. Most machines show some statistical data too - you can dig a little bit into that and you will get your accurate numbers. It probably can't be done from the office though.

reply »