"Build quality". What do we really mean by this?

[knobtwiddler]

I wouldn't consider ease of access or ease of service as indicative of good design or build quality.

What if something contains parts that have an unavoidable shelf-life? There is an Italian sports car that needs to have the engine removed by hoist to change the cam belt. You wouldn't consider that 'bad design'?

You can't expect potentiometers in pro equipment to last forever, if they're used 24/7. You can swap them out in minutes with some gear, whereas other items might need a ground-up disassembly to change them. In the latter case, you'll get a whopping bill from the repair tech. Is that 'good design'?

(I am playing Devil's Advocate a little - I can think of well designed items that are laborious to service and need specific tools. I suspect they are intentionally so, in order that amateur electricians don't find access too easily).

[Smithdoor]

I am in America QUALITY BUILT has no legal teeth. So any can say the statement. The buyer is thinking it is built better. If have Harbor Freight everything thing is quality built there very few things I will buy most of tools from Harbor Freight most is just draw fillers .

It is like word ON SALE legal terms you selling something. Buyers thinking your getting a lower price.

Dave

[ToneArm]

This thread has had me thinking a lot since it popped up yesterday. Just like everyone, I guess, I always want my work to be of good quality with whatever I'm making, repairing or restoring. I'm puzzling over what it means now and how appreciating how subjective it is.

Build quality though.. Take two identical sets of parts to build the same amplifier (for example) but built by two different people. Any difference between the two must be build quality. Even if both kits work and even test the same, build quality can still be different. Certainly visually, but structural differences also might cause early failure, bad soldering, poor earthing, circuit cleaned of flux, etc.

I'm less certain about choice of parts. The urge to seek better parts is hard to resist, but unfortunately skill is needed, to know when a component is a limiting factor and selecting a suitable alternative. With better parts maybe one kit is performing better than the other, but I'm not sure that is build quality. The kit must have meaningfully better performance, or be more rugged, longer life etc for build quality to be improved. Otherwise it could be that getting the same result with less expensive components is a sign of better quality. Like the Japanese car analogy.

My first audio kit was a Bugle phono amp from Jim Hagerman, it worked great and I went on to build a valve one, the Cornet. Jim supported his customers on a forum and it must have been so dis-spiriting for him. Thread after thread from builders with problems, usually hum and noise, and almost every time the builders had gone off piste, trying to make their creation look like high end hifi, rather than the quirky (but proven) layout from the designer himself. Under the bonnet there'd be massive capacitors flapping around, cable-tied on to power wires etc. I did similar things myself, it was an expensive way to learn electronics! The perception of the builder was of higher build quality, Jim's perception probably the opposite!

Enough from me, but thanks for setting out a thoughtful issue I have enjoyed having it on my mind!

[Rich Woods]

Originally Posted by Jez1234

I'm always suspicious when I see "ISO9001" and other QA certification being bandied about as it is a paper pushers regime not engineers. IE it is intended to impress and satisfy often non/semi technical people, legal beagles, insurance/risk people and QA people. It's "something been seen to be done" which generates a paper trial big enough to cover the arses of anyone in an organisation who may have to "prove" that it wasn't their fault or the companies fault when something goes wrong! "Look we have all the certification to prove the quality of our product and that everything possible was done to ensure it".... but in practice it made no difference to the fact that someone made a very bad soldered joint to the emitter of Q231 and the unit failed after 7 months causing a production line to be halted for 2 days!! Ah well it keeps some employed I guess.

Having had the misfortune mid career to move into Quality Assurance Management for 14 years, from the mid 80’s perhaps I can add some insight. If (and it’s a big if) systems are allowed to be developed and / or improved to ensure “quality” is built in to a product and the level of failure and rejects are as a result reduced then it’s a very useful tool indeed. Unfortunately in my experience there are precious few companies who are sensible enough in their outlook to do this. The vast majority see ISO9000 as just another marketing tool and want to do the bare minimum to get approval. Of the multiple companies I worked on QA for, in reality only one of them was interested in doing it properly and they were a small engineering firm who had found a good business plan & had rapidly expanded into a large new production unit and understood that uncontrolled specifications and word of mouth “back of a fag packet” would no longer cut it. Even then it was difficult to promote what was often viewed as unnecessary bureaucracy. In the end after a difficult and turbulent few years the company gained ISO9002 with LRQA who certainly then were the most thoroughly assessors. After letting things bed in for a bit I left to do it all again in a different company. I am still proud of what I achieved and the company is still doing well. However, after doing QA & Business Improvement for 14 years in different firms & sectors I became totally disillusioned for the reasons above and took a big pay cut & went back to being an engineer with no regrets whatsoever.

[G6Tanuki]

TBH I rather liked the formalisation of 'quality methods' and embedding them in businesses, whether they were of the ISO version or the likes of PRINCE. Getting proper appreciation of quality- and process-management made eminent sense to me.

[The thing I liked about PRINCE was that it was a menu-board/toolkit: you could take the bits of it that were appropriate for your business needs and ignore the bits that were irrelevant]..

Whaty always annoyed me was the reluctance of suppliers/contractors to take the client-feedback implicit in such methods and apply it to their business.

An example: one of my competitors designed a RF power-amplifier, which used a push-pull-parallel quad of rather-expensively-matched-by-TRW bipolar transistors. The prototypes worked brilliantly - but then they subbed the manufacture of a few hundred of these to a contractor, all of which came back after a few weeks/months with one or more od the transistors having failed.

Because they were a matched-quad, one failed transistor meant that all four had to be replaced. That's serious $$$$

And still they kept coming back. The company was going under with the penalty-cost of this.

Spotting a business-opportunity, I got access to the contractor [a couple of cases of single-malt were involved].

Seems that though the playbook for assembly said that the TRW transistors should be fitted to the heatsinks and the clamping-nut [they were 'capstan' type transistors] torqued down to the specified value, and only then should the four 'fins' be soldered to the PCB, the production-types were soldering the transistors' fins to the PCB, then feeding the stud through the heatsink and tightening down the clamping-nut 'until it felt about right' without using a torque-driver.

Results: the stud was being overstretched, and there was also a torque being applied to the transistors' ceramic encapsulation which caused it to snap off, potentially revealing the Beryllium Oxide inside!

You can't assess this sort of 'build quality' thing from 'visuals'.

[For this bit of analysis I got paid rather well, and also we got to take-over the company whose build-processes were found lacking]


For stuff I designed/specified, I liked using 'esoteric' fasteners [Torx, or a trilobular socket-headed fastener whose socket had a constant-diameter, like a 50p piece, so you couldn't get them out by jamming a screwdriver into the socket. I forget the name of these] We also would wax-fill the recessed sockets of fasteners in order to determine whether there had been 'unauthorised' fiddling.


You can't honestly send something out with a warranty after _you_ have rebuilt it, if someone else unauthorised has been in there before you.

[Bazz4CQJ]

Originally Posted by G6Tanuki

I wouldn't consider ease of access or ease of service as indicative of good design or build quality.

the Japanese instead produced sets that didn't need service or repair so accessibility of internal components was largely irrelevant to their design process.

Yes, that is quite right; within their expected working lifetimes, access for repair was not required.

But, now we have Japanese equipment which is well in to the vintage classification, lack of access to enable repair work can be a real problem. I have a Sanyo LW/MW/FM set, made around 1980, which was a wonderful radio, but it now needs work on it. An initial assessment suggests it will require more time, effort and shear dedication than I can generate . The cord-driven tuning scale obscures a large area of the rear of the PCB; I think it would be very hard to get it back together and working properly.

B

[Michael Maurice]

If you look at the build quality of a Revox and compare it to the Japanese tape recorders of the same era, you could say chalk and cheese. The Revox is a much better, easier to service than any of the Jap machines

[knobtwiddler]

And what about bodge wires and the occasional component wired P-P in a 'dead bug' style. Would that count against build quality, or is that something completely different?


I think that Tonearm makes some good points. If you look at diyaudio.com for example, there are threads where many members have made the same project, from the same PCB etc. I'd take amp made by someone knowledgeable on a low budget, over one made by someone with budget and ignorance.

[kalee20]

Originally Posted by knobtwiddler

And what about bodge wires and the occasional component wired P-P in a 'dead bug' style. Would that count against build quality, or is that something completely different?

It would count against design quality.

But, if well done - component secured neatly with a blob of suitable adhesive, wires (anti shrink-back insulation) joined to component legs with workmanlike joints, flux cleaned off, then built quality could still be high.