ESD: Truths, myths, and flat out lies

His understanding is incorrect. A resistor limits the amount of current that can flow by providing resistance to that current. The higher the resistance, the less current flows. If you know the voltage and the resistance, you can predict how much current will flow.

If the resistor is not there, there's zero resistance; zero ohms. That means there's no resistance to the current, and as much flows as the source can provide. If that current is going through someone's body into the wrist strap, it is bad news.

If the resistor is there, and presents a resistance that's low enough to bleed off static charge, but high enough to prevent shock from common 110 or 240 voltage sources, then we win.

Resistors are not semiconductors, nor are ESD mats.

A resistor doesn't really limit current, because if you increase the voltage, the current will increase.. so the current isn't really limited. It only limits the current at a given voltage is all i'm trying to say.. if you really want to limit a current you have to design a current limiting circuit that will not allow current past some given value to flow.. (i.e. the output current of a voltage regulator can be limited to 1A for example, if the load wants to sink say 10A, the regulator will only give it 1A because it is limited to 1A).. this requires the use of resistors AND transistors

Resistors can be semiconductors.. on chip resistors can be made using P-type-doped silicon.. however. it's more popular to create an on chip resistor using a transistor biased in the triode region.. even better is to use switch capacitors to mimic resistors in transconductor circuits.

but the bottom line is some resistors are indeed made out of semiconductor materials
 
Resistors can be semiconductors.. on chip resistors can be made using P-type-doped silicon..
This doesn't make them a semiconductor. (Or, makes them not a resistor, depending on the characterisitcs of the resulting device.) If the device doesn't have a constant relationship between I and E, it's not a resistor. Do a lab and prove it to yourself, if that'll help you come to the correct understanding.

A resistor doesn't really limit current, because if you increase the voltage, the current will increase..
Of course it does. At a given voltage, the current through a resistor won't be higher than E/R, even if the supply is capable of providing more current. Without the resistor there, the current that flows is as much as the source can supply.
 
the point of grounding a case is so that the case is always at a known potential... if it weren't grounded.. then the case is free to pick up charge relative to ground, and discharge somewhere (possibly causing damage)

grounding doesn't gaurantee ESD protection though.. but i'm sure most of you know that.. if you pick up a chage relative to ground.. and you touch your grounded case. there will still be discharge between you and the case... just hopfully not as much

Exactly why the person operating on the computer should be grounded to the mat. This further reduces the odds of holding a charge.

This is also why most "portable field kits" have THREE cables.

1)ground (wall to mat)

2) Device (Mat to chassis, sometimes integrated to the ground cord)

3) Wrist (mat to wrist, sometimes also integrated into the ground cord)

These three cables when used in unison, make it very hard to hold a charge....at that point, your only threat is if you are wearing wool or a sweater....neither belongs near a computer.

The resistor is usually small enough that a large voltage will burn it up....but to count on that is foolish. All it's there for is to reduce risk of electrocution (through typical mains voltage).

If you ever worked on a electronic production floor (like the one I'm typing this from)....you will even have your shoes grounded to the floor (heel straps or specially designed shoes...I use straps, I found they are more consistent). Problem is, when you have to hi-pot a unit (hi-potentcy voltage test....2.1KV for 2 seconds, slow ramp)....you have a VERY high risk of electrocution....so what do we do? We UNGROUND ourselves! We do this through use of a very thick rubber mat, or by simply flipping our hellstraps off the hell of our shoes.

Yes, it's a bit amusing, but it's needed....I've gotten zapped by a hi-pot before (not strap related, i was just having a stupid moment), it's a pain not worth experiencing, ever.

Yes, the heelstraps have resistors....it's just when dealing with that kind of voltage, it's best not to rely on an item made by the lowest bidder....protect yourself. Ungrounding of the operator during hi-pot is an industry standard for safety reasons.
 
Yes, the heelstraps have resistors....it's just when dealing with that kind of voltage, it's best not to rely on an item made by the lowest bidder....protect yourself. Ungrounding of the operator during hi-pot is an industry standard for safety reasons.

Policy where I work is not to wear ESD protection when working on anything that is powered. Most of our equipment is 28VDC and 115VAC 400Hz. And believe it or not kids, 28VDC bites if you touch it and are well enough grounded.
 
This doesn't make them a semiconductor. (Or, makes them not a resistor, depending on the characterisitcs of the resulting device.) If the device doesn't have a constant relationship between I and E, it's not a resistor. Do a lab and prove it to yourself, if that'll help you come to the correct understanding.

Of course it does. At a given voltage, the current through a resistor won't be higher than E/R, even if the supply is capable of providing more current. Without the resistor there, the current that flows is as much as the source can supply.

:eek:

the I vs V curve of the device has nothing to do with the classification of the device as a semiconductor... that's rediculous

and when you say E (electric field) you mean V (voltage) right?? there is a difference

semiconductor elements are on columns 3, 4, 5 (i think there is even one in column 6)on the periodic table of the elements (just to name a few, they are Si, Ge,B, Te, As, Sb).. anything made with these materials can be called a semiconductor device.. this includes resistors, transistors, and capacitors even (on chip inductors are made using metal layers in an IC so those definatly wouldn't be considered a semiconductor device)... seriously.. why are you arguing this? take some silicone, dope it with Boron(sp? )and now you have a P-type resistor which is made out of a semiconductor (2 semi conductors actually the SI substrate and the Boron dopant).. thus.. it is a semiconductor device.. anything on column 3 likes to accept an electron (thus is called an acceptor), anything in column 5 likes to give an electron (thus called a donor).. this is under-graduate solid state device physics.. trust me.. i know my solid state physics well, and i understand device modeling very well too..

a resistor doesn't exactly limit current becasue if the voltage changes, the current changes.. that isn't limiting at all.. a resistor only RESTRICTS the flow of current.. it does not limit.. i think you're just confused with teh term "current limiting resistor" for diodes and LEDs... again it's just usage of words really.. the "current limiting resistor" for an LED really only RESTRICTS curent flow, because if the input voltage increase.. teh currents will increase as well.. so.. where exactly is the limiting part??

on the other hand you can design a class AB amplifier that actually limits the output current to say, .5A. so no matter how small the load and how high the output voltage (conditions of high current sourcing).. you would never be able to supply more than .5A from the device.. that is the real definition of current limiting.. but people sometimes use it to describe things where current really isn't being limited.. just restricted (like in LED circuits)... voltage regulators have current limiting circuit in them as well.. so they don't supply large amounts of current in case of a short or something else.. well it depends on the regulator of course, cheap one's would fry right up, even well designed one's may fry too.

seriously.. please do not try to argue other wise.. when it comes to device physics and semiconductors.. i know my stuff

a resistors is just a linear device (thus the constant I vs V relation).. a transistor is a non linear device.. that distinction doesn't make one a semiconductor and another not a semiconductor.. lol where do you guys get this shit? it's laughable

the material the device is made out of is what defines if it's a semiconductor device or not..

any arguement against what i just stated is just arguing the definition of words really (like what exactly you mean by "limit").. and not the underlying electronics concept

by the way.. where in that link you posted does it say "a linear device such as a resistor is not a semiconductor".. lol

the question of whether or not something is a semiconductor, isn't so much an electronics question.. it's really a general chemistry question.. i mean shit, just look a the periodic table.. the semi conductors lay inbetween the metals (conductors), and non conductive materials (like chlorine).. thus.. they are semi.. they're somewhere in the middle.. which made it really nice because you can manipulate their electric behavior since they're not really a conductor, and not really an insulator

by the way.. i did that little lab you posted about 5 years ago in circuits 1 lab ;)
 
anything made with these materials can be called a semiconductor device..
I think a better critiera for classifying semiconductor devices includes identifying devices with a junction. Or devices that use charge carriers (that is, electron holes) for conduction.

Sure; you can fabricate something like a resistor out of semiconductor material. Your p-type resistor doesn't have the same electrical characteristics as a discrete carbon-film resistor, though; particularly for negative voltage. For an application where the voltage involved is limited to a narrow range, that might not matter. For the general classification of the device, it certainly does.

Are there any vendors who ship p-type resistors in the bleed line of their static grounding straps, by the way?

RancidWannaRiot said:
that isn't limiting at all.. a resistor only RESTRICTS the flow of current..
Isn't a restriction a limitation? Sure, the voltage can change -- and the flowing current would be higher with higher voltage if the resistor wasn't there, or was of a lower value. The current is limited to E/R. Is there some way it can be higher than E/R? If not, then I think you'll agree that it's got an upper limit.

take some silicone, dope it with Boron(sp? )
I should start with Silicone? Are you sure?
 
I think a better critiera for classifying semiconductor devices includes identifying devices with a junction. Or devices that use charge carriers (that is, electron holes) for conduction.

Sure; you can fabricate something like a resistor out of semiconductor material. Your p-type resistor doesn't have the same electrical characteristics as a discrete carbon-film resistor, though; particularly for negative voltage. For an application where the voltage involved is limited to a narrow range, that might not matter. For the general classification of the device, it certainly does.

Are there any vendors who ship p-type resistors in the bleed line of their static grounding straps, by the way?

Isn't a restriction a limitation? Sure, the voltage can change -- and the flowing current would be higher with higher voltage if the resistor wasn't there, or was of a lower value. The current is limited to E/R. Is there some way it can be higher than E/R? If not, then I think you'll agree that it's got an upper limit.

I should start with Silicone? Are you sure?

1) you're right, most resistors are not made out of semiconductor material.. i stated that like two posts ago.. BUT.. they can be made out of semiconductors materials.. and they are.. just not often. the fact there there is any sort of junction doesn't determin if it's a semiconductor device really.. again.. this is just arguement of definitions.. a P-type resistor would work just fine if negative votlage is applied across it.. what are you talking about? negative.. positive it doesn't matter as long as there is a voltage difference across it.. the sign of the voltage just tells you in which direction current is flowing (holes or electrons).. the definition that is used in industry for a semiconductor device.. is anything that is made of a semiconductor... for example.. an on chip semiconductor capacitor can be made using a double poly layer process.. you can search IEEE papers on semiconductor capacitors, and resistors.. semiconductor resistors use to be used for psuedo-Nmos logic as well

hell i've layed out a resistor made of P-type material in some of my IC design projects.. you can amke N-type mater as well of course.. but they tend to have less resistance (about half the resistance of the same sized P-type resistor)

2) i wasn't talking about wrsit straps. all i wa saying is that resistors can be semiconductor devices.. that's all.. I'm sure wrist strap resistors do not have a semiconductor resistor

3)silicone slilicon.. whatever it's called a typo.. i just typed this up real fast... fify ou just feel the need that you had respond..

3)
 
so.. the airplane itself is isolated from everything else.. and the electronics are referenced to the airplanes chassis.. so there is not earth grounding.. but there is grounding... if the chasis picks upa charge relative to earth ground.. it will not affect the electronics.. becasue the electronics are referenced to tha chasis... basically the electronics will not see the extra charge on the chasis... there is the possibility of things like ground loops and stuff.. where there is a bit of difference in the chasis.. but that's a howle other story.. and if something is designed well.. it's not an issue

Question: can this be compared with a PC by replacing "airplane" with "PC" and "chassis" with "case"?

I don't know what you do or don't know. Your original question was about the odds of damage occurring, which are very hard to set.
Well it wasn't really meant to be an exact question. I'm not interested in the statistics, I just want to know if it can cause damage like the way touching a chip can cause damage.

Earlier in this thread, I linked to the ESDA. On their site, the Basics of Electrostatic Discharge article provides an approachable overview about ESD management. It's aimed at manufacturers, so a few sections won't apply to computer hobbyists.
Thanks for posting that again! I read a couple of those PDFs, very clearly written and informative! I'm starting to understand this better and better.

Okay, some of you recommend touching a PC case (earth ground not even necessary) to achieve equal potentials before working inside. Then what is the difference with putting a PC on carpet (and leaving it there)?

Also, a wrist strap connected to a PC case slowly brings potentials to the same level, I imagine touching the case can produce sparks.. Isn't this a bit unsafe then?

And another thing..how would you guys install a new graphics card, or even motherboard? Don't you have to bring the case and component to the same potential slowly, before putting the component in?

For an analogy, think of a bullet. Fire one bullet from a gun and it can penetrate concrete, and perhaps even steel. No matter how many you throw by hand, though, you're not likely to break through a concrete wall in any significant amount of time.
I don't know.. Have you ever seen Shaolin monks throw a needle through a glass window? :D
 
I don't know.. Have you ever seen Shaolin monks throw a needle through a glass window? :D

That would be awesome to see. :p

Question: can this be compared with a PC by replacing "airplane" with "PC" and "chassis" with "case"?

Sure sounds that way, with one minor note. A PC Case is connected to Earth Ground as long as it is plugged in and you have not bypassed the third prong of the cable with an adapter or other such device.


Thanks for posting that again! I read a couple of those PDFs, very clearly written and informative! I'm starting to understand this better and better.

Okay, some of you recommend touching a PC case (earth ground not even necessary) to achieve equal potentials before working inside. Then what is the difference with putting a PC on carpet (and leaving it there)?

Also, a wrist strap connected to a PC case slowly brings potentials to the same level, I imagine touching the case can produce sparks.. Isn't this a bit unsafe then?

And another thing..how would you guys install a new graphics card, or even motherboard? Don't you have to bring the case and component to the same potential slowly, before putting the component in?

Due to the nature of most carpet material, other objects that come in contact with carpet have a very high chance of generating a charge. Therefore, while your case might be at the same charge level as the carpet, the carpet is causing surrounding objects to reach higher or lower charge levels. If you work with your case while it's on carpet, you simply need to be more diligent in ensuring that you ground yourself each time you touch the internals of your computer. This dissipates the charge you may have built up from contacting the carpet.

Touching the case with a high charge in your body can indeed cause a spark, however, if the case has a PSU in it and the PSU is connected to a properly grounded wall outlet, then any charge you direct to the case should be directed to the ground rather than to your components. If the case is not connected to a ground of any sort, and is isolated, then I suppose it's feasible that if you had enough charge to generate a spark when you touched it, that it might travel through a circuit or component.

As far as installing hardware, If you are holding the mounting bracket of your video card, and you touch the case, then you have just brought the case, yourself, and the video card all to the same potential. When you say "Slowly" you might be thinking of over a period of a few minutes. As far as ESD is concerned, "slowly" could be a few milliseconds. I don't know of any real figures offhand, but I doubt it would take a wrist strap more than 2 full seconds to bring you to the same potential as the case.
 
Question: can this be compared with a PC by replacing "airplane" with "PC" and "chassis" with "case"?

yes, but.. the PC is sitting on the planet earth, so there is more danger to the electronics if the PC picks up a charge relative to earth ground.. because then there can be a discharge between teh PC and earth ground.

an airplane is thouasands of feet away from earth ground.. so it's less likely to discharge to earth ground... that's not to say an airplane is totally safe.. I think there is probably more of a danger to the airplane of ESD actually.. given that it could fly through clouds and stuff that have different charges.. the airplane will never discharge to earth ground though

there is more to it than that.. and the airplane has protection circuitry in place...

BTW i don't claim to be an expoert in ESD, and ESD protection
 
That would be awesome to see. :p
They showed it on NGC once. :)

Touching the case with a high charge in your body can indeed cause a spark, however, if the case has a PSU in it and the PSU is connected to a properly grounded wall outlet, then any charge you direct to the case should be directed to the ground rather than to your components. If the case is not connected to a ground of any sort, and is isolated, then I suppose it's feasible that if you had enough charge to generate a spark when you touched it, that it might travel through a circuit or component.
I felt a spark yesterday when I touched my main PC. I'm glad it's connected to earth. So if I understand things correctly, you can't walk across the room holding a PC or put it somewhere else, unless it's grounded? I must have damaged many, many things.

As far as installing hardware, If you are holding the mounting bracket of your video card, and you touch the case, then you have just brought the case, yourself, and the video card all to the same potential.
But the case has to be grounded when touched or otherwise damage may occur?

Another thing, is it safe to hold an ESDS device enclosed in a shielded bag when you're not grounded? I bought 2 SoundBlaster Live! retail cards that were on sale recently. I opened both boxes to examine the contents and took out the cards (which were enclosed by their shielded bag) by holding the mounting brackets and ESD bag ends only. I did not go into the bag itself.

I'm a bit concerned I didn't use a wrist strap, but then I'm thinking that even if I had used one, there could still have been an ESD if the bag wasn't at the same potential as the ground, right?

I noticed Creative didn't seal the bags with tape, but they were folded. I tried to keep the bag folded together as much as possible. Both bags were also punctured a bit by the cards at the same spot. One worse than the other. Does this damage a bag's shielding ability?

Yes, I'm paranoid. :p
 
I've left an older motherboard on a carpetted surface for well over a year and a half.

She still boots up nicely.

Moral of the story: Tom, if they were in the bag, you're allright.
 
Moral of the story: only idiots make global generalizations about limited personal experience.
 
Moral of the story: only idiots make global generalizations about limited personal experience.

:eek:

at what point is it not considered limited :eek: I'm sure i have more experience designing shit..

ESD protection circutry is pretty damn beefy these days.. as i said.. we should take basic precaution.. but let's not get our panties all worked up over it..

what kind of non limited experience do you have?

I've had conversasions about ESD with guys that have 30 -40 years of experience..

let's stop trying to have the last say (as i am right now lol.) or i should say.. make your last say meaningful
 
OK folks....enough cockswinging.

Back to topic, or I shall find you all and you all shall wake up with the a hard drive read/write head at the foot of your beds.

fo shizzle.

....seriously, just go read up on ESD practices and principles, you'll figure out what's best and what's not quickly that way. Just follow their guidelines, and your components will be happy, electroncs frolicking through the circuits, and being merry with each other. It's just so awesome.
 
1. ESD is a problem in manufacturing. If left uncontrolled it can cost manufacturing facilities millions. It can reduce reliability of the product (more warranty replacement) as well as causing the engineers to scratch their head when something just stops working and they can't put their finger on it.

2. I've done plenty of ESD testing and plenty of lightning testing (lightning is more fun). And guess what...it does cause damage. Pin injection testing is fun for helping parts let out their smoke.

However, the reason why most consumers don’t “appear” to have ESD problem is because, in our own homes, we are naturally “ESD smart” when working with electronics w/o even trying.

The biggest culprit of damage is when we build up a voltage differential compared to what we are working with. These are the big scary “V” numbers being thrown around. What causes the damage is when the item we touch can’t quickly get rid of that energy. All and ESD is just moving of energy. The problem is when we move that energy into a “small” area like ‘1’ pin or input instead of the board and it can’t spread it out.

However most times when we working with electronics we “touch” the case first or “touch” the edge of the board/component first. Since we touch something that is low impedance we quickly “share” the energy with the item touched and thus “prevent” a large voltage differential on something that is sensitive.

Remember, in order to create ESD damage you MUST move that energy in a differential fashion through something that is sensitive to it.

I’m going to use an analogy here…I just hope it works. Take yourself and you start walking and a certain pace (i.e. a build up of energy, like ESD). Now imagine if you walked into something like a mattress. It doesn't really hurt does it? This is basically no different than grabbing your case or the edge of a board. You quickly spread that energy out over the mattress. Now imagine if you did the same thing again with a knife pointed forward. Now you have the same energy but now the resultant "force" is applied to a very small point. This would be like touch an individual pin on a board or chip that you were holding. This is what causes the damage. As for the people that carry around RAM in their pocket...no different than carrying around the mattress with you after you walked into it. :D

Therefore all the ground strapping and mats and bags and stuff are good if you working in an environment where you can’t control the flow of energy consistently or working with millions of dollars in parts per day. However, for the average Joe in his house…if you stop moving and touch your case…you have covered a huge portion of what you need to worry about.

Easy rules of thumb:

1. Always touch your case before touching the electronics
2. Try to keep parts in ESD bags
3. Don’t touch parts by their pins (grab on edges)

If you do that….as a consumer, you have done enough IMO.

The next question I often get is when people say "Well, I zapped my TV when touching the power button but it still hasn't broke yet". The big reason for that is you haven't "zapped" the button. What happened is that the energy on your body conducted through the metal foils around the button. The energy never hit the circuit (well, at least not enough to damage it). The energy gets quickly absorbed and spread out in the foil. Even things like IPODs even have ESD dissipation circuits in them. The industry has chosen to protect you from yourself so they don't have fix the shit you broke. ;) Open up your digital camera sometime…see if you notice something.

In short the myth people are wrong and right…and the paranoid people are wrong and right.
 
2. I've done plenty of ESD testing and plenty of lightning testing (lightning is more fun). And guess what...it does cause damage. Pin injection testing is fun for helping parts let out their smoke.

No one has stated that ESD doesn't cause damage
 
No one has stated that ESD doesn't cause damage

Then what about the Nigerian businessman’s posts? I'm confused...I thought he called ESD a myth? I would also be willing to be that there are quite a few people that will not post for the fear of being "beat'n down" by those in the know that may also think it is a myth. I was trying to bring a tempered viewpoint to this thread...

BTW, any reason for the hostility?
 
Yawn.... Some of you kids....

ESD can cause damage... OF COURSE. But to get on and talk about how much energy a part can transfer or withstand in the absolutes I've just read here is retarded.

Every component is rated on its sensitivity to discharges. FETs (transistors and gates) are the most sensitive and too bad for you carpet munching sock scoffling yahoos they are almost everywhere. Ram sticks are fairly sensitive, individual ram chips are not so much. Basic voltage regulators are not, good ones are. Drives (led, i/o, etc) are, diodes and resistors are not. Every data sheet for every part made has a rated ESD value and its conditions (Human Body Model for ex)


Most electronic systems that are expected by the user to come in contact with have TVS, diodes, resistors, decoupling capacitors, etc etc etc on the pins you may touch. Thats just good design and its really really common. All of my projects have at least tvs, diodes, ferrite beads or something on them b/c I don't ever want to here back from a customer, and neither does anyone else.

ESD isn't a magic evil you can denounce or build up as some god you can live in fear of. Its just a common fact of electronics. Don't touch pins with your fingers, try and keep a common gound with whatever your working on and relax. If a part was designed so poorly that you mess it up without any effort there is a really good chance it will destroy itself in its intended application anyway.
 
Then what about the Nigerian businessman’s posts? I'm confused...I thought he called ESD a myth? I would also be willing to be that there are quite a few people that will not post for the fear of being "beat'n down" by those in the know that may also think it is a myth. I was trying to bring a tempered viewpoint to this thread...

BTW, any reason for the hostility?

No hostility.. i just don't recall anyone saying ESD isn't real... then again.. it's been a while since i read all the posts
 
ESD isn't a magic evil you can denounce or build up as some god you can live in fear of. Its just a common fact of electronics. Don't touch pins with your fingers, try and keep a common gound with whatever your working on and relax. If a part was designed so poorly that you mess it up without any effort there is a really good chance it will destroy itself in its intended application anyway.


/thread
 
Mods should rename this thread "ESD: Truths, myths, flat out lies, and bullsh*t". If I were a person looking for accurate information regarding ESD I would become completely confused by the variety here.
 
If I were a person looking for accurate information regarding ESD I would become completely confused by the variety here.

Absolutely, especially the contradictions are confusing. It's still the best discussion on the topic I've come across. I've read it several times, hoping each time it would become clearer. :D But alas..

On the other hand, it IS a discussion and everyone is allowed to participate in it.
 
I still think my original post that started the whole thread is the most accurate out of them all.
 
It's got at least a few errors. About ESD foam being a semi-conductor, for example.
 
Right. But it's also not a semi-conductor. It shouldn't be hard for you to edit the post and get it corrected.
 
Right; it is a conductor, with a high resistance. It's actually called "conductive foam".

The ANSI/ESD S541 standard explains that there's a couple of characteristics that are interesting. One characteristic is that the material itself is inert; that it doesn't generate much of a static charge. Another characteristic is that the material helps dissipate a static charge, or shield electrical fields. A third is its resistance.

The standard explains how these materials are classified; "low charging" or "antistatic", "conductive" or "dissipative", "discharge shielding" or "electric field shielding".

Dissipative ESD foam has a high; usually in the millions of ohms per centimeter at the surface, when the foam isn't compressed.

ESD STM 11.12 and ESD STM 11.31 explain the specific standards for planar materials and loose-fill materials, respectively, but I don't think those standards are available for free.
 
So the foam does not conduct fully from one edge to the other edge, thus, semi-conducting electricity. Correct?
 
So the foam does not conduct fully from one edge to the other edge, thus, semi-conducting electricity. Correct?

Direct quote from www.whatis.com defining a semiconductor:

A semiconductor is a substance, usually a solid chemical element or compound, that can conduct electricity under some conditions but not others, making it a good medium for the control of electrical current. Its conductance varies depending on the current or voltage applied to a control electrode, or on the intensity of irradiation by infrared (IR), visible light, ultraviolet (UV), or X rays.

Because the ESD foam is a conductor at all times, it cannot be qualified as a semiconductor.
 
Back
Top