Rule of thumb for replacing power transistors with MOSFets?

I'm updating some game board designs and was curious if there is any sort of printed guideline for replacing silicon driver transistors with MOS? Both NPN and PNP.

For example TIP122 Darlington transistors driven by another pre-driver, that is the TTL interface. I've seen many a substitute, but would like to have the theory explained so I can make better use of it and understand exactly what I am doing.

A good reference book you can recommend perhaps? Books and I usually get along very well.

Too many projects and not enough to hunt down all the solutions!

Thanks, and sorry for the on-topic post...

John :-#)#

Reply to
John Robertson
Loading thread data ...

TTL can drive low-threshold, aka logic level, fets. There might be a base voltage divider that you can booger to get maximum voltage swing.

I like FDV301 for low current, 0.2 amp loads. STD16NF06 for heftier stuff; it will easily sink 15 amps with +5 on the gate.

Just don't let that happen again.

Reply to
John Larkin

Digikey, Mouser, and Newark are all months away from having stock on the STD16NF06...could use the STD25N10F7 I suppose - a bit of overkill.

Oh, well, I'll just look for TTL compatible I guess - I just thought there might be some theory behind it that I could brush up on. Not that I ever really understood transistors, too much magic in those parameters!

John :-#(#

Reply to
John Robertson

There are lots of "logic level" low-thresold p and n channel fets around. I just named a couple that we use.

If you post a link to your schematic(s) we could check the circuits for drive compatibility.

Reply to
John Larkin

Hey, your kettle's black.

Just deleted at least ten of your most recent posts as being OT.

RL

Reply to
legg

You have something personal against John?

Reply to
Bob Engelhardt

Never miss the opportunity to redirect a technical thread to an insult fest.

Reply to
John Larkin

All FETS have 50% threshold tolerance on gate voltage and those std types with Vt= 2 to 4V need at least 250% times the actual threshold to make a good switch.

Although BJT transistors have a vast range in hFE, however when used as a switch, it is the actual Vbe that controls the current. Normally they are typically rated as switches with base current = 10% of Ic unless they are super beta types hFE> 300.

The critical power dissipation in switches is the ON resistance, such as Vce(sat)/Ic=Rce and RdsOn @ Vgs.

But the secret in fast pinball bumpers is the low T=L/R. That means lower inductance and higher resistance but hotter or less efficient but faster. This is important for spring back solenoid bumpers and flippers. Dry contact switches produce very high back EMF so a clamp diode is needed but this slows the L/R=T time even with a stiff spring. So instead of a diode clamp by adding a Zener & diode in series makes it faster and limits the maximum voltage for the switch for safety.

As far as substitutions, one needs to know the load, current, voltage and function. Otherwise, why would they stock seventy-five thousand different FETs.

It's a matter of the heatsink method, size, cost, drive voltage, load current, availability, and compatibility. Copying newer designs can help.

Reply to
Anthony Stewart

Can you say what kind of loads they are switching and at what rates?

piglet

Reply to
piglet

Given sufficient voltage in the gate drive, just make sure RdsxIds is less than the old Vcesat that was present.

I used to stick fets in older bipolar SMPS and could usually increase unit rating till something else in the power train started to sweat.

Often a magnetic part.

RL

Reply to
legg

FET is a fetish to whine about grammarschoolmarm rules. A 2N7002 doesn't care what you call it.

What's the last thing you designed with mosfets?

Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.