Motors and MOSFETs

[audioguru]

Taken to extreme, if you always register all the worst case interpretations of all the worst case parameters simultaneously, you end up with a design that's too expensive to build, an aeroplane that's too heavy to take off, and a manufactured product that's not competitive in the marketplace!! If you don't cut costs, your competitors will!.

I always use minimum and maximum values in my designs so that they all work perfectly every time. I never waste time testing parts to find the best ones.
Transistors have wide spec's so my designs allow for the wide spec's.

For a product that costs $10.00 to build cheaply or costs $15.00 to build so that all of them work perfectly and the product is sold for $500.00, then my way works fine.

The high markup in price is because most of my designs had no competition.

 

[marcbarker]

Well good for you AG that's 'tolerancing' you're talking about, a skill that's part of the overall risk management, sometimes the idea is to net maximum yeild in production, .....and sometimes it's not always economic to do so (i.e. yeild on silicon wafer).

But the point is, voltages are coming down lower and lower all the time in circuits. Using "10 V" of gate drive because 'that's what the FET is characterised at in the datasheet' becomes a religion!

The same religion of those people who thought that charging AA NiCad's at 50.000 mA would increase their service life...

 

[audioguru]

I don't manufacture cheap Chinese products where only half of them work.
All my circuits work and continue to work.

All my Ni-Cads failed and were replaced by Ni-MH cells with a capacity 5 times as much. A huge improvement.

All my "10V" Mosfets have 10V on the gate. All my "logic level" mosfets have 5V on the gate. They all work perfectly.

 

[Hero999]

But the point is, voltages are coming down lower and lower all the time in circuits. Using "10 V" of gate drive because 'that's what the FET is characterised at in the datasheet' becomes a religion!

Like the always allowing 3V of drop-out for an LM317 (regardless of the output current) religion.

I always look at the graphs as well as the worst case when determining the gate drive or drop-out voltage.

 

[audioguru]

I don't believe the "typical" graphs.
What happens when the manufacturer has a bad day (a poor yield)? Then the spec's are poor and are not typical.
What if a major product manufacturer bought all the good ones and only the poor ones are left for you to buy?

 

[Hero999]

I don't believe the "typical" graphs.

But if you assume a drop-out of 3V you'll probably end up making the the product much more expensive than it needs to be, causing lower profit and more grief from the boss.

What happens when the manufacturer has a bad day (a poor yield)? Then the spec's are poor and are not typical.
What if a major product manufacturer bought all the good ones and only the poor ones are left for you to buy?

Yes I know, so what should you do?

You do what any decent engineer with common sense does and normalise the typical graphs to the worst case.

For example, you want to use an LM317 in an application that only requires 200mA at a minimum ambient temperature of 0°C.

From the graph on page 6 of the datasheet, the typical dropout voltage is 1.75V with 200mA drawn and an ambient temperature of 0°C.

https://www.electro-tech-online.com/custompdfs/2009/08/LM117.pdf

We know that the maximum drop-out voltage is 3V.
What's the maximum typical value on the graph? 2.5V

Therefore if we multiply any typical value from the graph by 3/2.5 = 1.2, we should get the maximum drop-out voltage.

1.75×1.2 = 2.1V

As long as the input is 2.1V greater than the output then there'll be no problems.

Using 3V might mean a larger filter capacitor, transformer and heat sink or using a LDO regulator, all of which will add cost to the product with no benefit.

Exactly the same common sense procedure can be applied to the gain of a transistor, MOSFET threashold, op-amps etc.

 

[indulis]

What happens when the manufacturer has a bad day (a poor yield)? Then the spec's are poor and are not typical.
What if a major product manufacturer bought all the good ones and only the poor ones are left for you to buy?

Every hear of things like "SPC", "6-sigma" or "Cpk" and what they mean for a manufacturers process???

 

[Hero999]

6σ is relatively new so I don't think people as old as audioguru will have learned it at college.

Yes, I vaguely remember touching on 6σ at college although we didn't study it in any great detail.

Statistics are not always helpful because there isn't always a Gaussian distribution centred around the typical values. For example I've measured 25 of a batch 5% tolerance resistors before and found the mean value was something like 1.03% below the rated value and no, this was not my meter's fault; it had been recently calibrated and had an accuracy of 0.1%. If you did a statistical calculation assuming a Gaussian distribution centred around the nominal value this would be nearly impossible.

So why did this happen? Could I have bought a duff batch?

No, none of the resistors were outside the tolerance band.

The answer is the consistency of the process must have been much better than 5%, probably closer to 2.5%. That way it doesn't matter if the process drifts slightly in either direction, it can be corrected every now and then. Heck I wouldn't even be surprised if there's some oscillation in the range of values produced.

marcbarker is probably right that most components are better than typical but it's always wise to at least assume slightly worse than typical values, especially when it's critical to the operation of the project.

How critical depends on the consequences of it going wrong.

If it's something non-critical such as battery life in a consumer product, then I'd use the typical values for say regulator drop-out because nothing bad is going to happen. However, if it was for a medical application were someone's life could be threatened if a certain component was off tolerance, I would go for worst case with an extra safety margin.

 

[marcbarker]

However, if it was for a medical application were someone's life could be threatened if a certain component was off tolerance, I would go for worst case with an extra safety margin.

Now that is a field of electronics that where you become more of a lawyer than an engineer, and you become the most pessimistic engineer, treating the slightest risk of 'mis-interpretation' as though it came directly from the devil!.

Tolerancing and WCA aside, if a datasheet had an error in it (i.e. a typo in a pulse rating making interpretation ambiguous), you still have to "go by the data sheet", no matter if you were smart enough to of noticed and got yourself past the 'heretic' stage and had obtained enough peer support to protect yourself with. But it's often by then circuit got too complicated anyway, and the detailing engineers have already started embroidering it. But only a foolish engineer wants to jump in too early and take this risk of being a scapegoat if something ever later went wrong* in the product. Nah not medical electronics me!

That's the trouble with Life Support Equipment, it kills the spirit of design.

(* not even related to your circuit!)

 

[Zaeed]

Ok after reading that i'm totally confused...

I'm replacing a key switch with a circuit controlled by a PIC.. To do this i'm using an IPS1031 Low Side MOSFET (https://www.electro-tech-online.com/custompdfs/2009/09/ips1031.pdf) to drive a 12V Relay..

If the PIC gives out a ~5v signal, what resistor values should I be using?

 

[MikeMl]

As long as your Pic puts out 5V, the FET will turn on. The series resistor called out on the sped sheet can be from 0.5K to 10K.