TO-92 vs sot-23 transistors for power

Blue boards work much better than green boards.

Are the outputs bips or mosfets? Or something exotic like GaN?

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

[ snip ]

I mentioned before my worry about replacing a set of TO-92 transistors with sot-23 parts. Not only do the sot-23 parts usually have a lower Pdiss rating, they dissipate their heat into the PCB (unlike TO-92 into air**), so a set of them in close proximity would raise the PCB ambient, further reducing their real power-dissipation handling capability.

Specifically I've been using Zetex' ZTX450 and '550 E-line complementary transistors,

Their corresponding sot-23 types might be FMMT619, '624, or '722, all rated at 625mW.

The datasheets have a note: For a device surface mounted on 25mm X 25mm FR4 PCB with high coverage of single sided 1 oz copper.

Can anyone give me some guidance on further derating for closely-spaced sot-23 parts?

BTW, going to packages like sot-223 could mean an even larger PCB footprint.

** I know that TO-92 transistors dissipate some heat into the PCB, so I'm limiting their dissipation to under 500mW.

--
 Thanks, 
    - Win

Really?

They're low-capacitance video transistors, etc.

--
 Thanks, 
    - Win

I spy a bunch of carbon comp resistors. (Can you say why?)

I assume the amp is class A?

George H.

Never did it... my first pass guess would be to take the board area you have divide that by the number of sot-23's and get an area per device... then scale the power down by the area ratio.

George H.

I don't think the single fact of using SOT-23 devices crowded together means you will be able to dissipate less heat. When coupling into the air you need to consider how the air will be cooled. When coupling into the PCB the construction of the PCB is important as well as how the PCB will be cooled. Thermal impacts are a system design issue, not just a transistor package selection issue.

Is this in an enclosure? Is the PCB cooled by the air alone or will there be significant conduction away from the PCB? Is this a single sided PCB, two sided, multi-layer? Can you add power planes to the board to conduct heat away from the hot parts? The board layout I see looks like it can easily support heat conduction away from SOT-23s depending on what the rest of the parts are doing.

--

Rick

On a sunny day (27 Jan 2016 07:05:21 -0800) it happened Winfield Hill wrote in :

It is full of holes!!!

No SMDs?

Yes, 20 big holes for the a screwdriver for the TO-220F screws, and more big holes for air flow for some of the power resistors.

There are some on the bottom.

--
 Thanks, 
    - Win

Low inductance and transient power-handling.

Yes. Here's a schematic of a slowish version.

--
 Thanks, 
    - Win

What I've done with SO8 packages I didn't want spaced was use thermal insul ating pads and heatsinks. Aavid makes these BGA heatsinks with springs to p ush down on the board. You then need to use the power drops of each transis tor and the junction to package thermal resistances to work out hot the jun ction is going to be. You can also mount heatsinks on the copper clad under neath the board if you got the space.

The Zetex E-Line parts did not use conventional epoxy for their encapsulant, and derated based on 200degC junction temperature to achieve the 1W rating in free air. Rthjc 175degC/W is achieved through actual package material superior thermal conductivity. These parts can be profitably heatsunk.

This is one reason why the generic substitutes in conventional epoxy have a lower free air listing, as derated based on 150degC Tj max at

200degC/W or ~625mW.

The SOT23 or SOT223 bodies are similarly restricted to the same

150degC Tj max by the epoxy encapsulant.

In an unimproved redesign, your first questions should be what junction temperatures are actually present in the current product, and what power dissipation and ambient temperatures are currently handled.

A reduced Tj max means more area will be required for the same reliability margins, regardless of package size changes.

RL

Wait, that was supposed to be another thread!

Do you have a ground plane? That, and power pours, spreads heat pretty well laterally. Vias from parts to power pours can radically reduce theta.

A SOT89 takes about as much board area as a TO92 but can dissipate more power, a couple watts, and is easier for assembly.

1 oz copper (if you can really get it) is about 70 K/w per square, which can be used to very roughly eyeball heat spreading. We played with Sonnet Lite for a while and it looks like it can do 2.5D modeling of the electrical or thermal resistance of arbitrary shapes, but I'm doubtful that it's worth the effort, with Dremels so common.

TO92 types are good if you plan to have air flow (a fan) over the surface of the board. Not so good if you want the heat to flow through the leads into the board. (ROT: 150,000 K/w per ohm of copper)

I sometimes add a gap-pad between the PCB and a baseplate or box. Heat from the surfmount parts goes into the board, gets spread by the planes, and is conducted by the pad to the base.

Aluminum or brass spacers can move heat to the baseplate, too. Or blocks of aluminum.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

It looks like it might be difficult to align those bottomside TO220s and tighten the screws, especially if there are insulators and grease involved.

We had a similar problem recently, and decided to use lots of D2PAKS on top, or cut rectangular holes in the board for TO247s so they could be easily screwed or clamped to the cold plate and "surface mount" (lap solder the leads) to the board. That job is still unresolved.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

What's the circle with JL inside?

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

A John Larkin copyright notice? Dunno, where is it?

--
 Thanks, 
    - Win

There are twelve 10-ohm 1/4 W carbon comps, for power handling, as mentioned. But the standout 27-ohm 5% resistors are actually TE Connectivity ROX3S 3-watt metal oxide film types, with very low self inductance.

--
 Thanks, 
    - Win

I see it to the right of your Dropbox schematic. Maybe Dropbox is goading me to add a comment.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

On a sunny day (27 Jan 2016 08:19:21 -0800) it happened Winfield Hill wrote in :

Strange, seems to be 2 completely identical amplifiers driven by 1 opamp, but one negative and the other positive, each a carbon copy of half the old audio amp.

Could that not be done like the old audio amp a lot simpler? I remember doing a spice simulation for something like that to 1 MHz or was it more, just like an audio amp.

Or is there a reason for all the complicatiatitiated stuff?

Look at the "Thermal characteristics" table on page 2. The thermal resistance, junction to ambient, is 200 C/W with that inch-square copper, while the thermal resistance, junction to leads, is 194 C/W.

So the bulk of the thermal resistance is junction to lead, with a tiny bit left over for the actual pad.

Making the bone-headed -- and probably erroneous -- assumption that the copper pour is all at one temperature, ten devices on that same pad would have an aggregate thermal conductivity of 25.4 degrees C/W, with 19.4 C/W of that from chip to copper, and the same ol' 6 C/W from copper to ambient.

I'm sure that there are people out there who are much better at this stuff than I am, but I would consider making the board with components only on the top side, putting lots of thermal vias to big copper pours on the back side, and gluing it to a heat sink (or heat spreader) with thermally conductive glue.

You don't have to fly in the dark on this -- I know there's white papers out there that detail the methods to calculate the thermal resistance of vias, and -- if you don't want to use FEM calculators -- estimate the thermal resistance to ambient of copper pours to ambient.

If your collectors have RF on them then you'll have all sorts of interesting capacitance-to-ground issues to deal with, though. I'm sure there's ways around that problem, too, possibly involving alumina or beryllium oxide.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com