SMD equivalents exist for pretty much everything on that board except the screw terminal connectors and perhaps the fuse carrier. SMD ceramic caps are commonplace, and small electrolytics are available in SMD designs. Have you looked at any of the major component suppliers?

Yes, I looked for some SMD versions of that components. In particular, I am not sure about SMD trimmer potentiometers: although they are not intended to use it very often, they are extremelly small size 5x5mm, and seems it will be very difficult to tune/adjust it with screwdriver. What's power rating of a resistors can be used? Power supply is +12V, each channel support up to 2A per light driver circuit.
Also, not sure about using SMD Power MOSFET 27N3LH5: there are used TO-220(DIP) package, and SMD version is DPAK: the difference is that DPAK have Total dissipation(P tot) 30Wt whereas TO-220 have 45W.

A better question might be: Why exactly do you want to change the technology?

There are several points you should consider:

How many do you build each order cycle?
When you have SMD boards produced you have to build at least 100 or the assembler will screw you mightily.
With thru hole there is no penalty for having just a few assembled.

Parts usually cost a little more in SMD form. Often resistors cost less but everything else more. Those
aluminum caps will cost probably 2X more.

Having ANY thru hole parts on an SMD board requires the assembler to run the board thru both of the
technology assembly paths. This ALWAYS adds considerably to the cost of assembly.

Having to heat-sink parts on an SMD board adds a lot of complexity and over a certain power dissipation is
not applicable at all. This is why those large, leaded TO-220 parts are still popular.(more thru holes retained)

Lastly, if you are going to the expense of all new parts and laying out a new board you should dispense with the
archaic pots due to their cost, the labor they require, and the early failures they represent by updating the design.

Keith Cress
kcress - http://www.flaminsystems.com

Its for myself, not for mass production, though. SMD PCB is much smaller and looks better. I can stay with through-hole design though, just need make new PCB a little smaller.
Design it not mine. Which terminal block model is usually used in such design for wire-to-board wiring? (commom Pitch, mm).

No standard on spacing.

Put those resistors on end and save at least an inch.

Keith Cress
kcress - http://www.flaminsystems.com

I like Smoked's pragmatic approach. If a smaller board is all you need, then do what Keith says.

You can also rise the power transistors with their heat sinks and perhaps use a quad opamp instead of four separate ones. There's also a lot of naked board area that can be reduced.

One off, you say? SMD? You work at Fort Knox?

Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

Quote (itsmoked)

No standard on spacing.
Put those resistors on end and save at least an inch.

Which resistors do you mean?
Its Terminal Blocks, there is pin spacing (pitch).

He probably means ALL the resistors; if they were all laid out in a single row, you'd get a considerable amount of board size reduction

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

All the resistors! It's a standard practice to save board space by putting all thru-hole resistors on end.

Yes, on terminal blocks there is no standard pitch. There are:
0.098
0.100
0.138
0.146
0.150
0.156
0.157
0.177
0.180
0.197
0.200
0.250
0.276"
and the list goes on..

Do keep in mind that as the pitch gets smaller they can get harder
to use and may need special small screwdrivers. They also get weaker so
wire strains can damage them if the wires gets yanked.

Knowing now that you're only doing one board which means you'll probably
be assembling it yourself I retract most of my concerns about mixed SMD
and thru hole since production aspects are null-and-void.

1) Go ahead and use SMD resistors everywhere. (Make sure you consider the power
dissipation of the smaller SMD resistors)

2) Use SMD caps for all the little decoupling caps.

3) Use SMD LEDs.

4) Change all the ICs to SMD versions.

5) If wire size and loading will allow maybe drop the terminal block size one major pitch.

6) Consider replacing the electrolytic caps with surface mount ceramics - NOT surface mount
electrolytics. If the ceramics won't make the reach do NOT switch to SMD electrolytics,
stick with the thru-hole ones.

7) The TO92 transistor can be switched to SMD if you can deal with the power dissipation question.

8) Keep the same pots and keep the big power parts and their heat sinks, also keep that fuse style.

Have fun.

Keith Cress
kcress - http://www.flaminsystems.com

Every one of those ICs likely has a SMD version that is 1/4th the board space as its TH parent. Very little heatsinking going on with those, and SMD versions are often served just as well with a ground pad and lots of capper on an inner/outer layer. I could make that board 1/4th its current size and STILL keep your pots... though I would try my best to design them out. If it's a one-off, measuring for the proper resistor value and inserting that instead of a pot is better anyway.

Dan - Owner
http://www.Hi-TecDesigns.com

Personally I hate the pots with sunk adjustment screws. The adjustment tool slips off too easily. I've had very good luck with SMD pots with 'normal' adjustment screws, like Copal ST5 series. We use a Vishay adjustment tool and they work well.

Watch out for 'hidden' thru hole components; although probably not a problem as this is your design. In 1994 I inherited a radio design that had just changed from thru-hole to SMD. it had poor reception according to the spec and a capacitive load in the IF section that made no sense. It really needed an inductor for the proper impedance in the IF circuit, but we were told that 'inductors were too expensive' and 'the old design didn't need one'. I helped another engineer tweak the circuit for weeks to get marginal performance. About a year later while cleaning out an old lab I found a sample of the old radio. The thru hole capacitor had one leg wound into an air core inductor and the engineer who converted it to surface mount never caught it. Grrrr... (C) 2015 because I should write this up for EDN one day.

Z

@itsmoked Thanks for the helpful advices! Will try to follow it.

Those TO-92 transistors - 1 x FET J310 and 5 x 2N3904 is only available in TO-92 package.

Power MOSFET 27N3LH5 have DPAK SMD package, but with less Total power dissipation 30W, these Power MOSFETs drives the current (1W LED diodes connected to the output).
There is a Surface mount heat sinks available for D-PAK (TO-252) package semiconductors, but they are wider (0.75") than the existing heat sinks.

@MacGyverS2000 Yes, exactly I'm thinking to find a proper resistor value and replace trim-pots with single resistors. Trim-pots used to adjust center frequency.

Is there anything else wrong in mixed technology, when use SMD + some Thru-hole components in one PCB, except more complexity and thus extra cost?
In terms of reliability?

Depends on the environment. How much vibration and temperature are you expecting

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

Normal conditions, no anything special.

Hi Sanvik,
You said in an earlier post that you are 'making this for myself', well have you considered how you will assemble it?. Unless you are having a one-off p.c.b. made and assembled professionally (with all the usual methods such as glue spots, solder paste, hot air reflow etc.), have you actually tried hand-soldering surface mount components? Some of the ICs, and especially the single transistor packages are almost impossible to handle, and even with the smallest soldering iron can be a bit problematic to solder first time without getting solder bridges on such close lead spacings.

I suspect there won't be any issues with switching or mixing SMD into your single board.

Do keep BrianG's points in mind though all the parts you contemplate are on the easier end of
SMD soldering.

Both the FET and the 3904's have SMD equivalents.

Depending on the circuit design and accuracy needed, adjustment via 'resistor selection' could easily NOT work.

Keith Cress
kcress - http://www.flaminsystems.com

"Normal"

Seriously? Normal for me is MIL-STD-810. What is "normal" for you?

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

For low volume SMD and thru-hole can go on the same board with ease. At higher volumes they each need a different soldering process (hot air reflow for SMD and wave solder for thru-hole) so that drives up the time to build and thus the assembly cost.

Z

Quote (IRstuff)

"Normal"
Seriously? Normal for me is MIL-STD-810. What is "normal" for you?

Normal means "for indoor use". -10C to +60C / 14F to 140F Not in aerospace industry. :)

@BrianG For myself, yes. You are absolutely right, assembling SMD PCB professionally at home is very difficult, and also requires a very sharp vision, and solid hand-soldering skills. I didn't plan to assemble SMD boards myself. I have good contacts with small PCB assembler, and I'm expecting they can help me with that. If not, then I will go with through-hole PCB, or will looking for another assembler.

I find SMD equivalents for MOSFETs and transistors, so all can be done in SMD, except terminal connectors blocks. Two separate terminals on the left side will be moved to the other end of the board, so PCB will be twice less in width.
Can anyone advice about SMD Fuse 10A?

sanvik,
Go to digikey.com on the most basic search they yield 84 SMT/10A fuses..
BUT.. I'd suspect what you really want is a SMT fuse holder to accept the same/similar fuse.. again digikey and look for one that fits your needs..
Same with everything else.

For through-hole mounting trimpots, Digikey offer 120 models for given criteria. What type is better suitable for application? does Tolerance % matter?

Given that you are tweaking the pots anyway, the absolute tolerance isn't particularly relevant, since I assume that you'd remove the pot, measure its value, and replace with a fixed resistor that's closest to that value.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

For THT cdesign, would be fine to use resistors with 1/4W power rating for all elements or it need be calculated?
If use SMD design, what is common SMD resistor size (Code) that can be used here? (Imperial/ Mectric)

Why would you assume 1/4W is acceptable in all locations? If this is your design, put in the appropriate power value for each component.

Dan - Owner
http://www.Hi-TecDesigns.com

0805 are common, rated 1/10 to 1/8 watt, and fairly easy to hand solder. 0603 get trickier to solder and the power rating goes down. 0402...

If you need higher power ratings then you'll need to go to larger sizes.

Z

Can be 27N3LH5 Power Mosfet TO-220 replaced with DPAK version without heat sinks?

?? You've done nothing to make the power dissipation lower? So, how are you going to deal with the 15W lower power rating on the DPAK?

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

Theoretically, max power is 2A x 12V = 24W per driver circuit: so I still have some power reserve?
Possibly, I can use copper area around DPAK Mosfet for heat dissipation?

The ST datasheet has numbers that don't make much sense; however, it's pretty clear that the 40ºC difference in junction to case temperature cannot be made up with external improvements. Therefore, if the ambient is going to get above 30ºC, you're going to have potential reliability problems with the DPAK.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

Should I consider another Mosfet with more power dissipation? Which model would be suitable?

The issue isn't the power dissipation spec, it's the thermal resistance of the dinky package you want to use.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies
[IMG http://tinyurl.com/7ofakss]
Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com: http://www.engineering.com/AskForum/aff/32.aspx

So I should use MOSFET with lower Thermal resistance junction-case R_thJC value, smaller the value it is better the heat radiation.
Would be these a good substitute: Infineon IPD90N03S4L-02 or Vishay SiS476DN ?

As I stated earlier you should stick with the existing thru-hole parts, not for the parts but for access
to the heatsinks those parts work with. I don't think that you're getting that the
whole game is surface area. Changing from a bunch of metal fingers exposed to the air down to no metal
fingers and just some flat board area is going to be a dissipative back-step. You can dissipate some
reasonable power that way as long as you trade up the area dedicated to it on the board, the opposite
of what I thought you were trying to do (smaller board). About the maximum area you can achieve with
standard circuit board materials is an area around the parts you contemplate of about two square inches
each on both sides of the board. That much space would probably account for about 1/2 of the existing
heatsinks.

You could make the switch to SMD and maybe a square inch per part if you had forced air blowing across
the heatsink pad area but that requires noise and a mechanical failure down the road.

Keith Cress
kcress - http://www.flaminsystems.com

You could go with a metal core board... but then you would be trading a higher cost for reduced space (not necessarily a bad thing, mind you, but that's a project-specific decision).

Dan - Owner
http://www.Hi-TecDesigns.com

It's possible that this design was created with a big reserve in mind, so using heat-sinks & fuse might be overkill for the purpose.
I saw similar device (SMD PCB), where output connected to IRF8313 Power Mosfet without heat-sinks. It max total current is 5A though.

Run the existing board at the highest power dissipation you contemplate for
about 1hr then use a temp-gun and see if it's an option.

Keith Cress
kcress - http://www.flaminsystems.com

Heatsinks for DPAK surface mount devices are available.

http://www.alliedelec.com/comair-rotron-8339trt000...

Of course they seem to use more area than the through hole equivalents with a heatsink.

The power dissipation in the transistor isn't the supply voltage times the current.

Quote:

Of course they seem to use more area than the through hole equivalents with a heatsink.

Yeah.. Exact reason I didn't suggest it. :)

Keith Cress
kcress - http://www.flaminsystems.com