SMD alternative to TIP31 and TIP32

[v8dave]

With the massive number of transistors on the market, I am looking for anyone who has done this and what options they chose.

I have a design that drives servo valves at 5V and max 500mA current.

The current design us all through-hole but I need to rework this to SMD to get it smaller. This uses the standard TO220 TIP31 and TIP32 and I'd like to use an SMD transistor instead. As these devices don't have any direct replacement for SMD use, has anyone got any experience of swapping these out for SMD parts?

I have a working simulation in Tina and this works with real hardware too so I can easily plug in the values from the SMD parts and make sure that it works, simulation wise at least.

[graybeard]

I have not looked but I would be shocked if there were not SMD equivalents.  Have you tried looking?

[jhpadjustable]

5V at 500mA is an easy spec to meet, and a BJT switch doesn't need very tight matching when substituting. Parametric selection guides on the manufacturer or distributor websites are essential resources for finding what's out there. On the other hand, maunfacturer-specific simulation/design tools tend to be stifling unless verifying one of their odd parts for which no generic model is available.

It's easy to find SOT-89 packaged BJTs from Nexperia, Infineon and other major manufacturers, also second-sourced by various Chinese semiconductor startups. For example, one recent addition to my home lab stock, BCX56:

• 1 A average collector current, 1.5A peak
• 80V Vceo
• 0.5W dissipation on a conservative footprint or 1.3W with more copper on the collector pad
• hFE minimum 25 at your chosen 500mA

BCX53 is its PNP complement with very similar specs. In any case, you should consult the full datasheets and specifically the safe-operating-area graph to make sure the collector currents and on-times in your application remain within limits.

You can get further size reduction if you can redesign for MOSFETs. There are hundreds if not thousands of part numbers in SOT-23 packages with multi-ampere drain-source ratings, 3.3V-friendly threshold voltages, and low on-resistance. AO3400 and AO3401 are my go-to components for switching an ampere or two at low voltages.

[george.b]

MJD31 and 32 are probably what you're looking for.

[mikerj]

The original transistors seem a little over-specified for the application, unless the valves are being switched rapidly and continuously.  There are plenty of high current bipolar switching transistors available in SOT-23 packages with gains much higher than the ancient TIP31/32 and lower saturation voltages.  e.g. On Semiconductor have a decent range, as do Diodes Inc.

[Yansi]

Why do you use linear bipolar transistors for 5V SWITCHING application?  Does not make much sense.

Use mosfets instead and think about protection circuitry, so that a random short on the output does not blow your driver circuit.

There is a metric shit ton of way more suitable N-MOS transistors for 5V 500mA switching.

[mikerj]

Why do you use linear bipolar transistors for 5V SWITCHING application?

As opposed to the digital bipolar transistors, or the digital MOSFETs?  Bipolar transistors are still widely used for switching, why do you believe this makes no sense?

[T3sl4co1l]

TIP31/2 perform worse than pretty much anything else on the market.  I wonder if they're even cheaper, all things considered?

The standard solution these days -- for CCFL backlights -- is a low-Vce(sat) transistor, probably DPAK or smaller, which has much higher hFE, lower Vce(sat) as the name suggests, and faster switching than a crusty old TIP31.

Likely with circuit changes, a MOSFET or integrated regulator is preferred.

There are switching regulators from SOT-23-6 to CSP size which are easy to use, and handily cover the ratings OP needs.  Highly recommended.

Tim

[Yansi]

Why do you use linear bipolar transistors for 5V SWITCHING application?

As opposed to the digital bipolar transistors, or the digital MOSFETs?  Bipolar transistors are still widely used for switching, why do you believe this makes no sense?

Why the hell would one even consider a poor linear BJT for a LOW VOTLAGE SOLENOID SWITCHING application in almost 2020, where you leave half (I am exaggerating) the load voltage across the saturated BJT?

There are tons of even SOT23 sized 3V3 level compatible mosfet switches, that would be sufficient for switching 5V 500mA solenoids.

Tell me, where are BJTs used for switching these days?  In a four decades old legacy designs, or in twenty or more years old dumb power ICs, that likely should not be even considered to new designs?

(I am not talking about offline SMPS, keep the topic around low voltage load switching please).

[floobydust]

Going from through-hole to SMT, the major design consideration is heat and temperature rise for every part, from transistors to resistors. It's extra math to use smaller parts.

Consider also making changes to the design to make it more efficient. We can help here if you share the circuit portions.
MOSFET's are superior to BJTs for switching motors, but fragile with overvoltages and overcurrent. A MOSFET's low on-resistance means much less heat so smaller packaged parts can do a lot.

TIP29/30/31/32 is dinosaur technology from the early 1970's, I tossed them all in the garbage they are the worst for hFE, speed, beta droop.
But the cheapest in town for what they do.

[mikerj]

Why do you use linear bipolar transistors for 5V SWITCHING application?

As opposed to the digital bipolar transistors, or the digital MOSFETs?  Bipolar transistors are still widely used for switching, why do you believe this makes no sense?

Why the hell would one even consider a poor linear BJT for a LOW VOTLAGE SOLENOID SWITCHING application in almost 2020, where you leave half (I am exaggerating) the load voltage across the saturated BJT?

Why do you keep emphasising linear BJT? Exactly what other kind of BJTs do you think there are?

Tell me, where are BJTs used for switching these days?  In a four decades old legacy designs, or in twenty or more years old dumb power ICs, that likely should not be even considered to new designs?

Obviously all these semiconductors companies offer a wide range of bipolar switching transistors purely for their own amusement, there can be no other explanation can there?

[Yansi]

TIP31 is not a switching transistor, it is a (poor) linear amplifier.  Quite unsuitable for switching applications.  (Even though it may be fast enough for a solenoid)

There are different types (whole product lines) of BJTs intended for switching applications, optimized for much lower Vcesat and faster switching times.

What I want to say is that I would look for a very different BJT (than TIP31), if that much you (or anybody else) needs a BJT SWITCHING application.

[T3sl4co1l]

I mean, they certainly found them suitable back in the day.  Nothing wrong with that.

But arguing semantics is rather useless without any further input from the OP.

Tim

[langwadt]

With the massive number of transistors on the market, I am looking for anyone who has done this and what options they chose.

I have a design that drives servo valves at 5V and max 500mA current.

The current design us all through-hole but I need to rework this to SMD to get it smaller. This uses the standard TO220 TIP31 and TIP32 and I'd like to use an SMD transistor instead. As these devices don't have any direct replacement for SMD use, has anyone got any experience of swapping these out for SMD parts?

I have a working simulation in Tina and this works with real hardware too so I can easily plug in the values from the SMD parts and make sure that it works, simulation wise at least.

servo and npn+pnp does that mean linear control and not switching?

[mariush]

I wonder if it would be possible to use darlington arrays like ULN2003A : http://www.ti.com/lit/ds/symlink/ulq2003a.pdf

They're 500mA per "channel" but you can parallel multiple inputs to get the higher current, so you could do for example 2 x 3 channel and leave the 7th unused.

[Yansi]

Beware of the maximum combined current with these arrays.  Loading all 7 ( 8 ) channels fully at once is not permissible.

But there are other types of arrays, even protected ones (both OCP and OTP), that would be way more suitable for any interfacing cards. Downside of these is usually quite higher price, but they are VERY robust. (typical quad and octal switches for PLC application).

I think a very interesting one for the OP might be for example Rohm BD8LB600FS, octal protected low side switch.  (or any variation on this chip, I think I have seen one even cheaper, about $2 in small quantities - which is a very good price, considering this is a rugged switch, not just a bunch of unprotected transistors)

[rsjsouza]

I am not sure if I am missing something, but a cursory glance at a SOT23 device (Onsemi FSB560A) shows 2A of maximum IC and 300mV of VceSat - however, only rated at 500mW of maximum Pd. The ancient TIP31 respectively handles 3A, 1.2V and hefty 40W on its beefy TO220 package. I can't help but wonder that the old device would be much more reliable than a puny modern one.

[T3sl4co1l]

SOT89s go up to 5A and more (and a watt or two).  Perfectly reliable if you're not a dick to them.

Tim

[floobydust]

Big ugly through-hole is tough. TIP-31 datasheet is 2W no extra heatsink and 40W on a heatsink. If the servo motor can jam or get overloaded, a SOT-23 BJT would exit in a puff of smoke. MOSFET's look much stronger, or Zetex BJT's which are great for low VCE(sat).

I wonder what OP is using for overcurrent protection, aside from the TIP-31 heating up.

[v8dave]

Sorry, all. I got the current wrong. It is 40mA max as the coils are 250 ohm and 2 of them wired in parallel. I was thinking of the solenoid valves which are 500mA. I should have looked at my circuit before I posted.

The circuit is shown here. The drive to the servos is +-5V with 0V obviously being off and +-5V fully on for each respective direction.

Input T4 is 0 to 10V analog with 5V being the midpoint from a micro controlled DAC.

T4A and T4B are the connections to the servo. As said, this has 2 coils, each of which is 250 ohms and they are wired in parallel. The feedback is used to give an indication on the control panel and if the servo is faulty, there is no feedback.

The design from the DAC side cannot be changed as the input from the micro already exists.

[T3sl4co1l]

Haha ew, double crossover distortion.  LM324 plus an unbiased (class C) emitter follower!

If the servo motor can jam or get overloaded, a SOT-23 BJT would exit in a puff of smoke.

Well obviously, you don't get much power through a SOT-23 regardless.  You wouldn't dissipate it in the first place -- use a class D driver.  At which point you can plop in a cheap 8-pin audio amp, with self contained oscillator and biasing, and you don't even need to add an RF filter if the motor leads are short and shielded.  Bridged amps are also abundant so you might use 5 or 12V instead of +/-15.

Could even just use a LM393 for osc/PWM and a gate driver (TC4428ish?) if you can be reasonably sure the load won't go short circuit or anything.  But any discrete solution will be hilariously larger than a self-contained amp, it really is a very good option here.

Tim

[rsjsouza]

It seems an overcomplicated alternative to an on/off operation. Do you have links or rough drawings to what you are saying?

(Of course, my interest is purely curiosity as I saw myself in the same corner in the past and opted for the well known to me TH option).

[T3sl4co1l]

This plus a gate driver gets you a solid amp or so of output drive:



Add on a feedback error amp (to match the above servo example, it'd be the 2 x 82.5R current feedback and U?C error amp) and you've got a class D amp with constant current output and no dead band (give or take if you replace the crusty old LM324 with a nicer e.g. TLV2372 ).

Something like TC4420 has a 6A peak output capacity, from a few ohms output resistance; it would be fine at 100s mA continuous duty, if maybe not an ampere or more.  It's actually rated for reactive current of similar magnitude (500mA without causing CMOS latchup).  Else you can add any inverter you like (last example I used, TPS28225 + SIS932EDN).

And as power goes up and dissipation rating goes down, you can consider other current sense or fault protection mechanisms.

Again, you can get various chips that basically integrate all this functionality into a little amp, all you need are a couple bias resistors (if that), a bypass, and some filtering (which you may be able to waive).

Even for something like a solenoid valve, this approach isn't completely off the wall.  Higher voltage drive can be used to accelerate response time; a constant-current driver should then be used to maintain holding current.  You can actually get higher performance at lower overall power consumption this way (depending on how close to minimum holding current you want to run at, and how often it's switched).  When this is done with switching sources, you can get even better improvement.  There are relay/solenoid drivers/controllers for this purpose.  (I don't know any part numbers offhand, though.)

Tim

[v8dave]

It seems an overcomplicated alternative to an on/off operation. Do you have links or rough drawings to what you are saying?

It is not ON/OFF as you have with solenoid valves, it is a servo so it is a linear output to drive the servo from closed to partially to fully open in both directions. +-5 with 0 for off. 2.5V would open the valve midway.

[v8dave]

This plus a gate driver gets you a solid amp or so of output drive:

Even for something like a solenoid valve, this approach isn't completely off the wall.  Higher voltage drive can be used to accelerate response time; a constant-current driver should then be used to maintain holding current.  You can actually get higher performance at lower overall power consumption this way (depending on how close to minimum holding current you want to run at, and how often it's switched).  When this is done with switching sources, you can get even better improvement.  There are relay/solenoid drivers/controllers for this purpose.  (I don't know any part numbers offhand, though.)

Tim

Thanks, Tim, but this is to drive a servo valve which can be positioned part open to control the flow rate. I have to get 16 of them onto a board that is VME form factor along with the feedback and the ADC/DAC etc.

As for solenoid, I've used PWM before to reduce the total current when the valves are held in the on state. I used 100% to pull the valve in and then reduce this to under 50% to the point where the valve stays open and this reduces the overall current. Ideal when you have 24 of them with about 10 possibly on at the same time.