Symbol for MOSFET

[Medo]

Like everybody I give my components some generic designator. All diodes start with D (e.g. D1, D2...), all capacitors start with C, all transistors start with Q and so on...

I have issue with MOSFETs. Which designator should I use for it?
Usually I just call it MOSFET1 but I am wondering whether there is some standardized one-or-two-letter symbol for it...

[Simon]

normally "Q" is used

[Medo]

I usually use Q for transistors. I feel uneasy using it for MOSFETs.

[Alex]

Electronics in a very well regulated industry, you should be surprised if there is not a standard for something widely used.

Component 'classes' (eg. C, D, R, L etc) for schematic diagrams are covered by IEEE 315A and IEEE 315A (section 22).

http://standards.ieee.org/findstds/standard/315-1975.html
http://www.docstoc.com/docs/17118743/IEEE-315--1993-_Graphic-Symbols-for-Electrical-and-Electronics-Diagrams_

(page 217 in the second link)

Q is used for transistors (amongst other things).

A MOSEFT is a transistor too, for some reason we associate BJTs with the term though.

If someone knows where standards (IPC, IEEE, IEC) can be downloaded, please let us know! :-)

[cyberfish]

metal-oxide semiconductor field-effect TRANSISTORS

[MrPlacid]

Q_M works for me.

[allanw]

Some people use M's.

[Alex]

Some people use M's.

Non-standard.

Q_M works for me.

Which schematic editor do you use?

[MrPlacid]

Which schematic editor do you use?

MSPaint + infraview(for flipping, rotating), everything else is too complex for me One day, I swear I'll learn to use a real one.

[Alex]

If you can do a schematic in MSPaint you will find a proper schematic editor much easier!

I was asking because EAGLE does not have a facility to add a subscript designator on the class letter. So you can have QM but not Q_M.

Not sure how this is handled in say Altium Designer, if at all.

[Medo]

Q it is.
Thanks to everybody.

[MrPlacid]

I was asking because EAGLE does not have a facility to add a subscript designator on the class letter. So you can have QM but not Q_M.

I don't have much knowledge in those editor, but don't they allow you to add labels? If they do, can't you drag an "M" and place it slightly lower than the "Q" to get Q_M?

[Alex]

You can specify a class designator eg Q when making the library and then every time you place a component a number increments that is added to the class designator. But all of this is using the same character size and you cannot split the two. You can manually add an 'M' if you treat it as a text comment but that just takes too much work. You can see that its a mosfet from the actual schematic.

[Zero999]

LTSpice uses M for MOSFETs which is probably not right but I don't bother changing it.

[Scutarius]

Does anybody has a copy of the  IEEE 315A?

[skipjackrc4]

I usually use Q for transistors. I feel uneasy using it for MOSFETs.

MOSFET's are transistors, so...

[Kremmen]

I have the IEEE 315-1993 "Graphic symbols for Electrical and Electronic Diagrams (Including Reference Designation Letters)". No A suffix in the doc though.

It has a comparison list documenting the differences between ANSI/IEEE and IEC component lettering and transistor is Q for ANSI/IEEE and V for IEC and i have seen both in diagrams. Definitely neither separated a FET from a BJT on that level. It is obvious from the main text detailing graphic elements for FETs and BJTs that they are considered variations of one component - the transistor.
So do like the pros do - use Q for both (or V if IEC floats your boat).

M is reserved for many kinds of devices, none of which are even close to anything like a transistor.

[codeboy2k]

Some of the older schematics you come across use TR for BJTs  , i.e. TR1, TR2, etc... mostly TR101, TR102 though... like that, and M for MOSFETS...

I never liked it.  I always use Q.   to me, a BJT, UJT, FET, MOSFET, or IGBT is always a Q...

however, SCR, TRIACS, DIACS, SCS, GTO, ETO, IGCT, BCT, LTT, MCT, yadda, yadda, yadda.... everyone is confused on how to label those.

[armandas]

Here's a useful table from Tom Hausherr's blog.

[Kremmen]

Well, with no disrespect to Hausherr & al, not everyone is confused as to the lettering of components. There is little benefit in specifying the component too minutely using the ID symbol and for that  reason the standards use component classes, not individual component types in the designations. Thus all of the 3 and 4 letter comps in you post are simply Q or V, depending on whose list you prefer to use; ANSI/IEEE or IEC.

For reference, here is the comparison, attached as a PDF.

[codeboy2k]

...not everyone is confused as to the lettering of components. [...] all of the 3 and 4 letter comps in you post are simply Q or V, depending on whose list you prefer to use; ANSI/IEEE or IEC.

For reference, here is the comparison, attached as a PDF.

Thanks for the reference. Yes, Q or V. But if you look back at most publications from the 60's and 70's and 80s and even some more recent schematics, you'll find that those types of thyristors are labeled simply as they are.  SCS1, SCR1, DIAC1, MCT1, etc.. that's what I meant by everyone is confused.

Perhaps they are over it now, and it's not as bad as it once was.  And, thanks to your reference, even I know to call them all a 'Q'

[c4757p]

To what extent do those standards even get used? I'm pretty sure the only time I've seen a dedicated letter for "amplifier" was in LTC datasheets... otherwise they're just the usual U. E is specified by both for "miscellaneous", but that's almost universally U, unless I misunderstand what "miscellaneous" is. I don't recall ever seeing X for a socket (both agree on that), either A or BT for a transducer, or a dedicated designator for an isolator (interestingly, ANSI uses the same as for a termination - AT).

[Kremmen]

[...]
But if you look back at most publications from the 60's and 70's and 80s and even some more recent schematics, you'll find that those types of thyristors are labeled simply as they are.  SCS1, SCR1, DIAC1, MCT1, etc.. that's what I meant by everyone is confused.
[...]

I read your post carelessly Yes they used to just come up with a designation that appeared to make sense according to some logic. No prob, but one should make an effort to use standards if such exist and there is no penalty in doing so. But i am sure we agree on that point

[Kremmen]

To what extent do those standards even get used? I'm pretty sure the only time I've seen a dedicated letter for "amplifier" was in LTC datasheets... otherwise they're just the usual U. E is specified by both for "miscellaneous", but that's almost universally U, unless I misunderstand what "miscellaneous" is. I don't recall ever seeing X for a socket (both agree on that), either A or BT for a transducer, or a dedicated designator for an isolator (interestingly, ANSI uses the same as for a termination - AT).

They are certainly not mandatory - at least i think they aren't though i can't recall what EU Directives might say about it. But since you are not guaranteed to have a symbol for every conceivable gizmo, some leeway is needed. Standards are of course just someone's idea how things could be done in a uniform or "good" way, not a physical law or something. So you are free to ignore standards unless some outside agent (such as Customer or an Approvals body) makes demands.

At least for me the designator X is commonplace. In European (well, Nordic) schematics X is universally used to designate a pluggable connector of some kind, and i always mark terminal blocks using X.

[c4757p]

X is common here for crystals. It's a somewhat common American abbreviation style to replace a syllable containing K and S sounds - CryStal - with an X. (In the math and CS world, I've even seen "fxn" for "funCTion".) I haven't seen it for connectors, though. I've seen J and K, and the occasional misused P (should be for a Plug).

Personally I'd prefer to use a simpler system. There's no reason everything needs to have its own narrow little category. I use:

Code: [Select]

BT    Battery
C     Capacitor
CV    Capacitor - variable
D     Diode (forward)
DS    Display/indicator
DZ    Diode (Zener)
F     Fuse
J     Jack / female connector
JP    Jumper (configuration jumper, not jumper wire)
K     Relay
L     Inductor
P     Plug
Q     Transistor/other discrete semiconductor
R     Resistor
RT    Resistor - thermistor
RV    Resistor - variable
SW    Switch
T     Transformer
TP    Test point
U     Integrated circuit
V     Valve/tube

May have missed a couple. I make subcategories in cases where a component is significantly different in operation or use (Zener diode, variable passives, etc) but otherwise don't need a million different names. Anything that fits into multiple categories takes the more descriptive (a Neon lamp or VFD is DS, not V, for instance, just like an LED is DS rather than D). For more obscure things, I figure using an "official" single/double-letter designator is more likely to cause confusion when people aren't familiar with it. I can just use something like "SP" for speaker, "MIC" for microphone, "ANT" for antenna, etc.

[Kremmen]

I tend to use a similar logic - possibly even fewer designator classes. But i pay careful attention to use the proper graphic symbol if one exists. Thus i don't discriminate a bipolar, Schottky or zener diode, they are all D and so on. Likewise transistors and other similar are all Q's. If you really feel the need, IEEE 315 lists 13 separate diode symbols and if that is not enough you can use the construction rules to create even more, should anyone actually manufacture such devices. And check your medication while you are at it.

One rule of thumb i try to go by is to avoid using the same initial letter to denote completely different classes of device that are likely to appear in one schematic. I think IEC class system tries to work like this as well. Can't always be bothered but trying to, anyway.
Xtals seem to be Ztals here though they should really be Gstals i guess...

[c4757p]

I'm absolutely meticulous about schematic symbols. JFETs have to indicate whether the FET is symmetric (not that I use those much!), MOSFETs with extra diodes (like the ones with G-S Zeners) have to show them, etc.

I do discriminate Zener diodes because it's useful in troubleshooting. If a diode on the PCB says "DZ" instead of "D", you know that just probing V_F with a multimeter isn't enough to make sure it's working. I do RV and CV because it gives a quick indication of trim points and the like, either looking through a parts listing or just doing a Ctrl-F on the schematic. RT for Thermistor is more or less out of habit - now I'm considering editing that to R in my library. They're not fundamentally different and it's not really useful to have the indication.

[codeboy2k]

I'm absolutely meticulous about schematic symbols. JFETs have to indicate whether the FET is symmetric (not that I use those much!), MOSFETs with extra diodes (like the ones with G-S Zeners) have to show them, etc.

I agree with this.. I want the symbol to be accurate. I always draw my MOSFETS with the body diode, its really there, and it helps to visualize the flow better. and definitely need the schematic to show the gs zeners.

I do discriminate Zener diodes because it's useful in troubleshooting. If a diode on the PCB says "DZ" instead of "D", you know that just probing V_F with a multimeter isn't enough to make sure it's working.

I'm like Kremmen, my diodes are all D's, however you make a good point.. I think it convinces me to also start using a different designator for zeners. I've actually seen schematics with ZD1, not DZ1, and I sorta like that better than DZ1

[c4757p]

I'm absolutely meticulous about schematic symbols. JFETs have to indicate whether the FET is symmetric (not that I use those much!), MOSFETs with extra diodes (like the ones with G-S Zeners) have to show them, etc.

I agree with this.. I want the symbol to be accurate. I always draw my MOSFETS with the body diode, its really there, and it helps to visualize the flow better. and definitely need the schematic to show the gs zeners.

Urgh... pet peeve alert!

I've actually seen schematics with ZD1, not DZ1, and I sorta like that better than DZ1

I'm OK with both. I tend to think of "encoded" things like that in a top-down sort of hierarchical notation, so I think "Diode - Zener".

[Dave]

I still can't figure out where they found "CR" for diodes. Anyone got any clues on this?

[c4757p]

IIRC "Crystal Rectifier"

[Dave]

Ah, that makes a whole lot of sense. Mystery solved.

[codeboy2k]

re: pet peeve.. yes, I know that's the body diode, and it's connected to the source internally during manufacturing, but  it's not as clear (to me anyways) how it affects the operation as when it's drawn explicitly across the source and drain like in the second picture.. I just like that one better.

But I agree with your peeve.  I will re-train myself 

[Kremmen]

[...]
I do discriminate Zener diodes because it's useful in troubleshooting. If a diode on the PCB says "DZ" instead of "D", you know that just probing V_F with a multimeter isn't enough to make sure it's working.

That would be one valid reason to use a different symbol.

I do RV and CV because it gives a quick indication of trim points and the like, either looking through a parts listing or just doing a Ctrl-F on the schematic. RT for Thermistor is more or less out of habit - now I'm considering editing that to R in my library. They're not fundamentally different and it's not really useful to have the indication.

You do have a point in the logic, but i doubt i am going to change the way i do it.
My reasons:
- There are standards that define the designators you should use. Observing the standards is voluntary but they are the result of some effort and should not be dismissed out of hand. Observing half a standard is the same as observing none of it, as far as i see things.
- Looking at someone else's schematic, you cannot always know what internal logic they have used to adjust the standard designators. So in the end it tends to add more noise than signal to everyoone else but the designer him/herself. A thing like 'DZ' instead of 'D' is easy but there are so many completely opaque cases where you just cannot know.
- The component designators are intended to classify the components themselves for clarity. Now different people have different values for 'clarity' and i only speak for myself. My logic is that you don't attach meanings to the component symbol other than those helping to classify the component itself. Not how it is used, not any ancillary info related to the physical placement or whatever secondary info. That introduces uncontrolled extra dimensions into the classification and causes the classification to lose its coherence.
- All classification info is contained in the class designator and component symbol. You don't repeat the info across those two. This is the reason why my diodes are all D and the graphical symbol distinguishes the further nature of the diode. I do understand the point of 'DZ' but would not use it myself.

I would prefer documents other than the parts list to specify things like adjustments and test points and similar. I guess if you have the trimmable components designated differently then you can pick them out easily from the parts list. That however doesn't tell you anything about how to actually perform the adjustment itself, only what components may be adjustable. If you don't want to leave the adjustment rules as an exercise for the reader then you must document the procedure. And if you do then do you really need the separate designators any more? This has been my logic, so far.

I don't have a big argument against your way of doing things, i just go by a somewhat different logic. Maybe it comes from how i have experienced and assimilated the logic of the big projects in my earlier jobs. I am thinking of say the electrical and electronics "schematic" of a paper machine. The complete set of schematics is several meters of large folders on a shelf. Practically every different kind of electronic or electromechanical component manufactured by man appears somewhere in the diagrams, or so it seems. I guess the principles i listed above were taken to their logical endpoint there, especially the part concerning documentation of testing and adjustment. Certainly, a lot of that was "standard" in the sense that it didn't vary between projects but documented it was, nevertheless.
I have found that the principle scales nicely downwards - you just need to keep things in their logical compartments and see that every compartment is done as required. Not a big deal and not that much extra work really. I have yet to encounter a software project that failed because the programmer coudn't type fast enough after spending time documenting the architecture of the solution, or an electronics project that failed because there was no time to write a necessary design document. On the other hand, i have seen projects fail because those things were not done.

OK i digressed quite far from the question of component designators. I just tried to make the point why i do things my way, and that there is a reason for it. Other people have different priorities and life manages to go on..

[IanB]

Urgh... pet peeve alert!

When I look at that I have the hardest time not translating it into British (or Australian):