Why no SMT?
- Thread starter Pbassred
- Start date
Help Support GroupDIY Audio Forum:
BradAvenson
Well-known member
If they have the assembly equipment in-house, then it makes sense to use what they are equipped to manufacture. On the other hand, if you are outsourcing the asssembly work, then it can save you money to use whatever technology is cheapest for the pcb assembly shop.
Also, SMT makes it hard for Yamaha's service partners, i.e. local music shops, to repair a defective board. A guy who needs his bass for an upcoming gig will be seriously pissed off if his dealer has to order a replacement board from Yamaha instead of replacing a few parts.
Svart
Well-known member
Why would it be harder? Unless the problem is a BGA part, SMD is easier, cheaper and ultimately more reliable..
[quote author="Svart"]SMD is easier, cheaper and ultimately more reliable..[/quote]
Cheaper? Sure.
Easier? Please... :roll:
More reliable? Definitely not. I've had waaay more SM **** break than through-hole - both electrically and mechanically - from complex stuff like audio gear to simpler ones like TV remotes and lighted keychains (not to mention computer parts, which I'm not counting because there's usually no comparable through-hole alternative).
Peace,
Al.
Cheaper? Sure.
Easier? Please... :roll:
More reliable? Definitely not. I've had waaay more SM **** break than through-hole - both electrically and mechanically - from complex stuff like audio gear to simpler ones like TV remotes and lighted keychains (not to mention computer parts, which I'm not counting because there's usually no comparable through-hole alternative).
Peace,
Al.
Exactly. You need a good eye and a steady hand for SMT work. Plus, you need a whole new stock of SMD parts, including various form factors for what is essentially the same part. That's pretty expensive and requires extra space. A small music store won't do that.
Svart
Well-known member
Having been a manager for an electronics dept which included daily repairs of both SMD and through hole stuff, way more through-hole stuff was junked because of the damn headache of trying to get the parts off of the boards without tearing the holes and traces up. this is especially true of parts that have many connections to a ground plane.
There are tricks to SMD that make it almost absurdly easier to work with than through hole.
Not to be arguementative Al, but the things you mentioned that break a lot are almost certainly chinese made and mass produced. They cheapen on the PCBs and try to get away with the thinnest PCB possible to reduce cost. In a device such as a remote or keychain, that PCB will be subject to flex from the buttons being pushed to the device being sat on when it slips under a cushion.
So it's not a matter of SMD vs through hole, it's a matter of the cheapest you can go and still have the device work through the QC test.
As far as the electrical portions of SM parts breaking easier, I doubt it. Most of these parts are identical inside to their TH counterparts.
different form factors for the same part? Of course you will see that. that is the same with TH parts as well, but it is also very rare just the same. Products improve and they are made smaller over time. The manufacturers bring out several footprints of the same part for the same reason that car manufacturers make several different models of cars, some designers prefer a certain choice over others.
That's the same reason that the IC is now used.. If everyone were as resistant to modernization as you guys we would still be using punch cards in our computers! :green: :green: :green: :green:
All kidding aside, I have seen problems with both sides and the problems can usually be traced to human error, not necessarily the fault of one topology or another. The world is going SMD no matter what.
:guinness: :guinness:
There are tricks to SMD that make it almost absurdly easier to work with than through hole.
Not to be arguementative Al, but the things you mentioned that break a lot are almost certainly chinese made and mass produced. They cheapen on the PCBs and try to get away with the thinnest PCB possible to reduce cost. In a device such as a remote or keychain, that PCB will be subject to flex from the buttons being pushed to the device being sat on when it slips under a cushion.
So it's not a matter of SMD vs through hole, it's a matter of the cheapest you can go and still have the device work through the QC test.
As far as the electrical portions of SM parts breaking easier, I doubt it. Most of these parts are identical inside to their TH counterparts.
different form factors for the same part? Of course you will see that. that is the same with TH parts as well, but it is also very rare just the same. Products improve and they are made smaller over time. The manufacturers bring out several footprints of the same part for the same reason that car manufacturers make several different models of cars, some designers prefer a certain choice over others.
That's the same reason that the IC is now used.. If everyone were as resistant to modernization as you guys we would still be using punch cards in our computers! :green: :green: :green: :green:
All kidding aside, I have seen problems with both sides and the problems can usually be traced to human error, not necessarily the fault of one topology or another. The world is going SMD no matter what.
:guinness: :guinness:
NewYorkDave
Well-known member
Spoken like someone who's never worked in the repair business. :wink:
Svart
Well-known member
"Spoken like someone who's never worked in the repair business. Wink
Well my question was more of a rhetorical one.. but since my background is being questioned:
I have 12 years in the repair business, 6 of those have been in R&D which includes testing and debugging both the design and the hardware itself while 3 of those years were as a manager of 2 repair technicians. I have 2 commercially sold designs with PCB layouts that I have done for a previous employer. Those 6 years of R&D have been almost entirely in SMD and a lot of that time has been spent phasing out through-hole parts and bringing in their SMD replacements. Both the quality and the reliability of SMD has been refined and most of the problems I have seen have been the result of humans not following specific guidlines for PCB layout or for the poor design of a product.
I may not be anywhere near a lot of the fellow members here in terms of experience and education on many issues but I can say that it seems that I have more experience in the SMD areas than most here and from that I am honestly trying to fight this "SMD is the devil" attitude because there is no basis for a lot of the statements that I have read here and elsewhere concerning this SMD vs. TH fight. I do, however, understand that people are put off by the unknown and generally resist ideas that are foreign.
Well my question was more of a rhetorical one.. but since my background is being questioned:
I have 12 years in the repair business, 6 of those have been in R&D which includes testing and debugging both the design and the hardware itself while 3 of those years were as a manager of 2 repair technicians. I have 2 commercially sold designs with PCB layouts that I have done for a previous employer. Those 6 years of R&D have been almost entirely in SMD and a lot of that time has been spent phasing out through-hole parts and bringing in their SMD replacements. Both the quality and the reliability of SMD has been refined and most of the problems I have seen have been the result of humans not following specific guidlines for PCB layout or for the poor design of a product.
I may not be anywhere near a lot of the fellow members here in terms of experience and education on many issues but I can say that it seems that I have more experience in the SMD areas than most here and from that I am honestly trying to fight this "SMD is the devil" attitude because there is no basis for a lot of the statements that I have read here and elsewhere concerning this SMD vs. TH fight. I do, however, understand that people are put off by the unknown and generally resist ideas that are foreign.
AnalogPackrat
Well-known member
[quote author="Svart"]Having been a manager for an electronics dept which included daily repairs of both SMD and through hole stuff, way more through-hole stuff was junked because of the damn headache of trying to get the parts off of the boards without tearing the holes and traces up. this is especially true of parts that have many connections to a ground plane.[/quote]
There are tricks to remove such parts. :razz:
There you go.
Thru hole parts are in much larger and more mechanically rugged packages. They are less prone to thermal damage during soldering because of the increased thermal mass of the package and the ability to heatsink via pins while soldering. Mechanical and thermal reliability is lower for SMD. Small boards with nine layers and buried vias are just asking for failures due to mechanical stress.
Cheaper, cheaper, and cheaper (and smaller). If it breaks, buy another one. Keeps them in business. It's not BETTER.
Are you calling us old! Watchit!
And crappiness will increase as a result. Take my cell phone...I recently replaced a 6 year old Nokia 5165 with a flip top Nokia 6061 GSM phone (no camera, no Bluetooth, just a fooking phone). Voice quality of this spanking new unit sucks. And yes, I have disabled the "automatic gain" audio function which REALLY sucked. This is progress?
Luddites Unite!
A P
There are tricks to remove such parts. :razz:
There you go.
Thru hole parts are in much larger and more mechanically rugged packages. They are less prone to thermal damage during soldering because of the increased thermal mass of the package and the ability to heatsink via pins while soldering. Mechanical and thermal reliability is lower for SMD. Small boards with nine layers and buried vias are just asking for failures due to mechanical stress.
Cheaper, cheaper, and cheaper (and smaller). If it breaks, buy another one. Keeps them in business. It's not BETTER.
Are you calling us old! Watchit!
And crappiness will increase as a result. Take my cell phone...I recently replaced a 6 year old Nokia 5165 with a flip top Nokia 6061 GSM phone (no camera, no Bluetooth, just a fooking phone). Voice quality of this spanking new unit sucks. And yes, I have disabled the "automatic gain" audio function which REALLY sucked. This is progress?
Luddites Unite!
A P
NewYorkDave
Well-known member
I'm not talking about manufacturer's in-house rework. I'm talking about being a technician out in the field, servicing a variety of products from a number of manufacturers, and having to maintain a stock of parts and tools to take care of whatever comes down the pike. If you're working in a lab that's stocked with exactly the tools and parts you need, then yes--SMD is no harder to deal with than TH. But to the tech in the field who has to get a TV remote back on the air in the middle of a primetime newscast, or fix an audio amp or mixer that's gone down an hour before a gig, it sucks balls.
I had this argument with a manufacturing engineer from a major audio manufacturer. After going around for a while, he finally confided to me that he agreed that SMD sucked from a serviceability standpoint, but the mandate from above was for higher volume and profits, so SMD was the way to go. I believe that particular production line has now moved to China, anyway.
I had this argument with a manufacturing engineer from a major audio manufacturer. After going around for a while, he finally confided to me that he agreed that SMD sucked from a serviceability standpoint, but the mandate from above was for higher volume and profits, so SMD was the way to go. I believe that particular production line has now moved to China, anyway.
AnalogPackrat
Well-known member
[quote author="NewYorkDave"]Iit sucks[/quote] hairy moose
@Svart
I don't question your experience, and I'm not against SMD per se, but from a practical standpoint SMD is difficult for small repair shops and DIY.
You do need different size parts for exactly the same function, and you know it. In TH a slightly smaller or larger resistor or cap is no big deal. You can make a 5mm pitch part fit into 2.5mm pitch holes and vice versa. In SMT you have to have a part that fits exactly. Also, it's often difficult to see what a particular part really is or what value it has. Or take the venerable NE5532 there is a 16 pin version and an 8 pin version. No such problem for Non-SMD.
I don't question your experience, and I'm not against SMD per se, but from a practical standpoint SMD is difficult for small repair shops and DIY.
You do need different size parts for exactly the same function, and you know it. In TH a slightly smaller or larger resistor or cap is no big deal. You can make a 5mm pitch part fit into 2.5mm pitch holes and vice versa. In SMT you have to have a part that fits exactly. Also, it's often difficult to see what a particular part really is or what value it has. Or take the venerable NE5532 there is a 16 pin version and an 8 pin version. No such problem for Non-SMD.
Svart
Well-known member
Increased mass of PLASTIC? doubt it. the amount of cast plastic will not help you here. the legs are longer internal to the body of the part but that won't matter, their mass is so small that the 50ms of extra time it takes to heat them to the iron's temp is of no use.
Sure, cut the legs from the body, then heat and pull each leg from the hole, then solderwick the holes you can and use a solder sucker for the harder ones. total time=3 minutes for a dip14. total time for removal of soic14 with a standard iron=3 seconds.
what are those shiny doodads sticking out of the SMD packages.. those are pins too! they are made of the same metals and will conduct the same heat. Trust me, i've had to study this.
sure, if they were designed incorrectly and/or attached to the chassis of the device incorrectly. Bend a 2 layer TH board enough and you will get the same problems, generally opens near through hole pads where the thin trace meets the pad itself.
Smaller is better, your girl has been lying to you.
@NYD:
I did not mention the frequent trips to end user installations in the field that I had started to do with some regularity. This was not my official job so i did not mention it but I got stuck with it more and more so I will mention it now. It got to the point where I traveled about 3 times a month to various places around the country and did SMD work with a cheap weller iron.
I disagree strongly with the engineer you talked to, SMD can be much more reliable than through hole if the design and layout are done well. The keywords being "done well", something that I have seen very little of in the mass manufacturing world.
@Rossi:
not necessarily. It is ideal to have the parts fit well but you can make 805, 1206 and 1210 parts all fit a 1206 footprint if you need to.
who's every actually used the 16 pin version?
[quote author="Svart"]
who's every actually used the 16 pin version?[/quote]
It's what I got in an electronics store when I ordered an NE5532 SMD. Of course I didn't notice until I was home. What I needed was the 8-pin version. Which was unavailable in any local store. It's quite a pain in the ass when you have a number of TH parts of what you need and then a unit requires an SMD part. Then you go buy it and still don't have a part that fits.
BTW there was a thread about excess noise in SMD resistors. In some applications size does make a difference. Also, I dislike the fact that there is very little room to work with. You can't just replace a ceramic cap with a polyester or polypropylene one. It won't fit. And I'm pretty sure some cheaper brands go to the very limits of a part's specs to make it fit (electrolytics voltage rating for instance). Or they take a dull multilayer ceramic cap instead of a nicer sounding polyester cap, just because it is smaller. So no love here. However, I would have liked SMT for a CMOS based DC converter I built for a smallish SD microphone.
who's every actually used the 16 pin version?[/quote]
It's what I got in an electronics store when I ordered an NE5532 SMD. Of course I didn't notice until I was home. What I needed was the 8-pin version. Which was unavailable in any local store. It's quite a pain in the ass when you have a number of TH parts of what you need and then a unit requires an SMD part. Then you go buy it and still don't have a part that fits.
BTW there was a thread about excess noise in SMD resistors. In some applications size does make a difference. Also, I dislike the fact that there is very little room to work with. You can't just replace a ceramic cap with a polyester or polypropylene one. It won't fit. And I'm pretty sure some cheaper brands go to the very limits of a part's specs to make it fit (electrolytics voltage rating for instance). Or they take a dull multilayer ceramic cap instead of a nicer sounding polyester cap, just because it is smaller. So no love here. However, I would have liked SMT for a CMOS based DC converter I built for a smallish SD microphone.
Svart
Well-known member
sure noise can be a problem but you can get the thinlayer metal film resistors for SMD work.
Leaded parts can most definately be placed on SMD pads, it just costs more as you have to do it by hand.
Leaded parts can most definately be placed on SMD pads, it just costs more as you have to do it by hand.
DIP packages are now getting harder to source from subcontractors as well...
There is a definite drive to move to SMD from the manufacturers point of view. I'm told that it's also easier for things like noise, capacitance and esd on the pins as well. (for processors etc).
In a mass production environment, the advantages of SMD far outweigh it's disadvantages. As much as I'd love to say the semiconductor companies support the DIY'er (we do) - what pays the bills are the companies buying thousands of parts at the time.
In the mean time, you could always get some of those "adaptors" that take SMD parts and make them through-hole.
There is a definite drive to move to SMD from the manufacturers point of view. I'm told that it's also easier for things like noise, capacitance and esd on the pins as well. (for processors etc).
In a mass production environment, the advantages of SMD far outweigh it's disadvantages. As much as I'd love to say the semiconductor companies support the DIY'er (we do) - what pays the bills are the companies buying thousands of parts at the time.
In the mean time, you could always get some of those "adaptors" that take SMD parts and make them through-hole.
AnalogPackrat
Well-known member
[quote author="Svart"]
Increased mass of PLASTIC? doubt it. the amount of cast plastic will not help you here. the legs are longer internal to the body of the part but that won't matter, their mass is so small that the 50ms of extra time it takes to heat them to the iron's temp is of no use.[/quote]
yes, the plastic, to a limited degree, and the much larger lead frame. Lets not forget the ceramic packages or the metal can TO parts, etc.
And to replace the part? Dip: 45 sec, SOIC--depends on how many times you drop the little bugger. I also don't need a mag-eye to solder TH.
I'm talking about clipping a copper gator clip to the lead between the part and the board while soldering. I commonly do this with polystyrene caps as they are prone to damage. Oh yeah, you can't get SMD polystyrene caps.
The point is that it is quite often sufficient (or even advantageous) to use simple TH construction. Opens near thru hole pads can be found and repaired, buried vias cannot.
No comment.
Great. But you cannot deny that TH parts are much more flexible than SMT. You just cannot try out all the different kinds (and sizes) of caps in that little SMT world. And FYI people have been surface mounting DIPs since the 70s (go read some of the original Bernie Hutchins' Electronotes articles on making PCBs with minimal or no drilling). You can dead bug parts, you can point-to-point stuff, you can hang stuff over the ground plane or fly leads directly to any part.
Can you even match SMD resistors? How the hell do you physically get your meter probes on there?
A P
Increased mass of PLASTIC? doubt it. the amount of cast plastic will not help you here. the legs are longer internal to the body of the part but that won't matter, their mass is so small that the 50ms of extra time it takes to heat them to the iron's temp is of no use.[/quote]
yes, the plastic, to a limited degree, and the much larger lead frame. Lets not forget the ceramic packages or the metal can TO parts, etc.
And to replace the part? Dip: 45 sec, SOIC--depends on how many times you drop the little bugger. I also don't need a mag-eye to solder TH.
I'm talking about clipping a copper gator clip to the lead between the part and the board while soldering. I commonly do this with polystyrene caps as they are prone to damage. Oh yeah, you can't get SMD polystyrene caps.
The point is that it is quite often sufficient (or even advantageous) to use simple TH construction. Opens near thru hole pads can be found and repaired, buried vias cannot.
No comment.
Great. But you cannot deny that TH parts are much more flexible than SMT. You just cannot try out all the different kinds (and sizes) of caps in that little SMT world. And FYI people have been surface mounting DIPs since the 70s (go read some of the original Bernie Hutchins' Electronotes articles on making PCBs with minimal or no drilling). You can dead bug parts, you can point-to-point stuff, you can hang stuff over the ground plane or fly leads directly to any part.
Can you even match SMD resistors? How the hell do you physically get your meter probes on there?
A P
Svart
Well-known member
ceramic=$$$$ metal can parts= hard to get.
again, how much larger are we talking? a few mm at best, not much in the terms of heatsinking. You might sink a few extra uW before the metal reaches it's soaking potential but again you are really gaining nothing.
TH, sure you can do that in 45 seconds. No arguements there. Soic and others, you place a dot of solder on ONE pad, then place the part over the dot and heat the leg, it will drop perfectly into place. proceed to use a high flux solder and run a bead up and down the opposite side of the IC. let cool then repeat on the other side. total time=~45 seconds as well much less once you do it a few times. try doing that quickly with TH and you'll get huge globs on the bottom of the holes that you will need to go back and clean.
Use a hot iron, not a weller crapstation. hot irons will melt solder fast and you can work much faster therefor minimizing heatsoak. Heatsoak is the killer of parts, not the temp of the iron's tip.
As far as Polystyrene is concerned, wait a few more months and no one will make them anymore. And yes there were some SMD polystyrene caps being made but they were almost stictly military. Other types of plastic caps are just as good and are now made in SMD. Again, this dips into the Audiophool territory and cannot really be validated or invalidated.
true. however I have never seen a buried via go bad without some outside force causing it. bending a board should not be allowed by design, and if the board can be bent then the supporting structure was poor. Again this is a human error, not the error of the topology.
again, WHY NOT? because it seems that it would be hard? I've been doing this on SMD for years and I have yet to find it any more challenging than TH parts. you use tweezers and a good iron tip. and it's the same. I've paralleled resistors, put two resistors in series like a tent on one pad, used leaded caps with SMD pads and a whole lot more. I've soldered wires to a tqfp44 before for testing and it's not hard.
Similar threads
