Surface mount components for analogue audio... Yes or no???

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi.
I'm just wondering what people's thoughts are on using surface mount components for analogue preamps. Mounting them isn't an issue in this case, and because of the physically smaller footprint I'd prefer to use them (lots to pack in to the case!!!).

I'd like to go all out with surface mount - Op-amps, resistors, capacitors, etc.

What's people's thoughts on this? I don't want any loss in quality of output or componentry by using these.

Cheers!
 
bacon665 said:
If you canw ork them then go for it but keep in mind they cant take the abuse their larger kin can and that cooling can be an issue


Cool, thanks for that! Good point about the cooling. My prototypes will be all PDIP etc so I can see how hot things get then. I was thinking about cooling the other day, this project is going to have something like 150 op-amps in a 4U case!!! I'm thinking a proper 100 hour test may be necessary - I need these things to go the distance!!!
 
star882 said:
Almost every modern sound card uses surface mount components on both the digital and analog sides. In fact, some sound cards have no through hole components at all.


Thanks. I don't normally consider soundcards to be of great audio quality anyway, unless its a PCI M-Audio (or similar) analogue card. Even then, they're still pretty bottom range as far as the professional industry goes. Some of the better brands do PCI/PCI-E cards, but these are normally digitally based.

I'm more referring to using surface mount in professional analogue preamps - Rackmount or Home Theatre.

Thanks for your input though!
 
There is a great range of thin-film surface mount resistors available, which will give you much lower noise than thick-film. You also get shorter pathing and tighter bypassing capabilities (caps).

If you use ceramic caps, try hard to stick to C0G/NP0 rated caps (usually available up through .1uF), falling back to X7R for larger values if needed (up to >10uF).
 
BrianDonegan said:
There is a great range of thin-film surface mount resistors available, which will give you much lower noise than thick-film. You also get shorter pathing and tighter bypassing capabilities (caps).

If you use ceramic caps, try hard to stick to C0G/NP0 rated caps (usually available up through .1uF), falling back to X7R for larger values if needed (up to >10uF).


Great! That's the sort of information I love to hear!!! Thanks heaps!
 
I could work but....

I build alot of SMT boards at work. I am not real sure you are saving a lot of area. If you change a 1/4 W resistor for a 1206, I think it is a wash size wise. Power rating of SMT parts can become an issue. No .1 --> .47 ohm 5 W SMT emitter resistors that make sense to use and cost fairly cheap. Still will need TH parts for them, IMHO. SOT-23s for xstrs/fet - small signal stuff. Caps no problem but I hate hand soldering SMT Panasonic 'lytic caps. It's hard to get enough heat UNDER the cap to do an adequate job of attaching the caps. If you pay a board house to build the boards all the problems do go away. I use a Metcal iron with a variety of tips. My favorite tip is a .016 conical tip (sttc-145p) since we use mainly 0402 resistors and caps. Sometimes I have to solder 0201 sized parts. That's about my limit.
 
Re: I could work but....

Pilbromatic said:



Haha, yep, that's the plan! :D

Also some caps would be standard electrolytics. Need some 2,500uF or thereabouts rated at 60V.

one problem with small volume SMD is the setup charges. A solder stencil has to be made for screening on the solder paste. It's around 300-500USD where we go. Plus the charge to do the board(s). Those charges per piece add up. Programming the placement machine....getting parts on tape and reel/tubes. Cut tape won't do it - no leader on the tape for the machine. We build small volumes (10s to 100s) and the charges are not cheap. When we want perfect boards we pay to have them built.

I just finished building 2 boards with around 200 0603 parts and various SMD ICs plus a few SMPS modules. Took me 8 hours per board but I'm cheap compared to the setup fees for 2 boards. Just needed 2 boards and no more.
 
I like the approach QSC takes to SMD components in their PLX series amps - use where appropriate.
In general all low dissipation resistors and semis are SMD, and all things that get hot, and electrolytics, are through hole.
OTOH Crown seem to like having things like SMD emitter resistors in their big amps,
which leads to this sort of thing.....
An externally hosted image should be here but it was not working when we last tested it.

M
 
Pilbromatic said:
Hmm, cheers for that! I'll see how things go. I plan to sell my prototypes, so setup costs shouldn't be a major problem. Thanks for the heads up though! Also the main part of my prototypes will be PDIP components. Well, as much as possible anyway.

Talk to Newton at NJS Electronics in Waihi - he recently did an SMD run for me and the results were excellent.
He provided a pannelisation spec, which I passed on to the
Chinese boardhouse that made the PCBs.
Reasonable price too.
M
 
Re: Re: I could work but....

jcx said:
power dissapation has to be watched

for resistors going NiCr is relatively cheaper though

there is little choice in signal path C - PPS is likely the best smt film - also hand soldering smt flim can be difficult due to melting the cap


I haven't heard or seen any reports on PPS (polyphenylene sulfide?) "sound", dissipation etc. They were not (apparently) around when Bob Pease wrote his "soakage" article!

I've used SMT for the Linear Tech ultra low THD% Signal Generator, for a high speed comparator, for several preamps (one with CFB opamps) and have, like Peranders, built the Jung regulators in SMT. I use straight pin to apply solder paste, then fry the board in a toaster oven :bigeyes: I still don't have a properly working mcu controlled toaster with controlled heating/cooling.
 
I'd highly recommend SMT components for audio, especially small-signal and preamp gear. When it comes to high-density builds, like you described your project, there's little other choice.

As far as device ratings and dissipation, treat the components like everything else you design; select the correct part for the job with adequate overhead for dependability and longevity. Prototyping is fine, but touching parts to see if they're dissipating too much heat just won't cut it for little bitty 0602 resistors. If they feel too hot, it may be too late.

I've done a lot of work with SMT and BGA components, and for high-speed, low noise, high density, low inductance, and precision, it's really the only way to go. Most of my work was on PCs and servers, but some was on analog, and if it's good enough for Ghz speeds with low noise and pS timing, it should work just fine for very high quality audio.

Attached is a picture of a voltage driver and voltage switching board from a 4KW Mackie amp. Even though these are from a blown amp, the problem was just a rectifier (thanks eBay).

I'd go for it!

Paul
 

Attachments

  • pict0641.jpg
    pict0641.jpg
    89.1 KB · Views: 297
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.