Creative ElectronGET A DEMO

X-ray News

Fireside Chat with the Xperts: Stump the Chumps, Round 2

It was a scintillating conversation as Drs. Thomas and Cardoso tackled more viewers’ questions about X-ray inspection systems.  There are no simple answers from these two, and that’s what makes the presentation so valuable.  If you want to get into the weeds as to how these machines work, you’ll really appreciate this episode of Fireside Chat with the Xperts.

Is higher resolution always better?  You’ll know after viewing, and as a bonus you’ll gain terrific insight  into the relationship between pixel size and resolution.  How is X-ray system resolution determined?  Watch, and you’ll understand it in a way that could actually be valuable to your decision making.  Keep those synapses firing: Register for upcoming Fireside Chats with the Xperts and view our archives here.

 

Transcript:

David Kruidhof:

All right. Well, it’s time to start another Fireside Chat with the Experts, and do another one of these stump the chumps. Let’s see if I can ask a question that might be a little bit harder this week for Dr. Bill and Dr. Glen we have again with us this week.

David Kruidhof:

A couple of questions, some follow-up from last week. So I’ll go ahead and start with those. But we’ll just see how many we can get through again this week just like we did last week. Sounds good?

Dr. Bill Cardoso:

Sounds like a plan.

Dr. Glen Thomas:

Perfect.

David Kruidhof:

All right. So the first question is can you have a detector that has too high of a resolution? Is a higher resolution detector always better?

Dr. Glen Thomas:

Not necessarily. It depends on the application. It depends on the power of the X-ray system. It depends on the sample. In some applications, too much resolution is actually a hindrance. If you have a standard you’ve been using like a medical device manufacturer or maybe in the military, you have a standard that you’ve been using for 20 years, your manufacturing process is more or less written in stone, and all of a sudden you produce an X-ray system that has too much resolution.

Dr. Glen Thomas:

You’re going to start seeing things that are insignificant, but concerning to certain people, maybe you’re the end user or…. And in a lot of cases, I’ve had some medical applications where I’ve supplied high resolution detectors and the doctors were confused. They actually rejected the product because it was too high of a resolution. There wasn’t what the radiologists and the doctors were used to seeing. So that’s the practical part as far as from a customer point of view.

Dr. Glen Thomas:

You have other concerns as far as detector size and detector gray scale, as well as the size of the pixels, and the ability to significantly light up or excite the scintillator and the diodes, photo diodes, and flat-panel detectors with an adequate amount of light. In our case, invisible photons. In some cases, you’ll need a much higher power.

Dr. Glen Thomas:

Then we get into the resolution side of the equation. Do you really need that much resolution to see the actual product? As far as a lot of people will get pixel size and pixels and image size confused with pixel size and resolution as well. A larger detector may be 1.3 megapixel, the same size detector in a 4×6 or a 3×4 or a 2×2 will have the same amount of pixels. So just because you’re getting a larger detector doesn’t always equate to a larger amount of pixels or a smaller, higher resolution system. Anything to add to that, Bill?

Dr. Bill Cardoso:

Yeah. I was chuckling earlier because I remember that, I don’t know if you remember the casting customer we had. They had a really old film system. You could barely see voids and issues and cracks in those castings.

Dr. Bill Cardoso:

Then we upgraded them with a digital system. All of a sudden, all their samples failed, remember? They were like, “There’s something wrong with your X-ray machine. Now all our samples are failing.” Like, “Oh yeah, now you can see what you’re making.”

Dr. Bill Cardoso:

But you’re right. there’s a lot of… Resolution is a tough one because it’s one of those things where you don’t know what you ask for. Sometimes you end up seeing information that’s really not valuable, and information that you don’t use is noise. So you end up with a lot of noise in your information that don’t really add anything to what you’re trying to do.

Dr. Bill Cardoso:

But the other aspect you touched on and I wanted to explore a little bit more is the relationship between resolution and pixel size and spot size. That’s one of the things that people, it’s blended all together and people don’t really understand. They look at, for example, a flat-panel with 100 micron pixel and they say, “Oh, 100 micron pixel. That means that the best I’ll be able to see from my sample is a hundred microns.” Which a hundred microns is four mil, which is a big feature if you’re talking about the microelectronics and some of these other applications, then a micron is a big feature.

Dr. Bill Cardoso:

What has to be taken consideration is magnification. So if you have a feature that hundred micron pixel is going to reflect one pixel if your sample is right on top of the sensor. Meaning that there’s one to one magnification.

Dr. Bill Cardoso:

As soon as you move the sample closer to the source, away from the sensor, now that 100 micron feature on your sample can light two pixels on the sensor. If you move closer to the source, now that 100 micron area on your sample can light four, eight, 16, 32 pixels on the sample, on your sensor. So that’s how we can get sub-micron resolution with a hundred micron pixels.

Dr. Bill Cardoso:

That’s something that sometimes gets confused out there. X-Ray companies don’t make it easier. I think sometimes they like the confusion so people just get “Okay, I’ll just buy it.” That’s what we try to do. We try to explain the science and the technology behind it so people can make decisions with their eyes open.

Dr. Bill Cardoso:

Then the other thing that gets confused all the time is, “Oh, I want the smaller pixel possible because that’s going to give me the best image.” Well, there are also trade-offs there. If you have a really, really tiny pixel, a really tiny pixel, remember that the amount of photons per square centimeter hitting your detector is not a function of the size of the pixel. It’s a function of the source that’s shining that sensor.

Dr. Bill Cardoso:

So if you have a hundred micron by a hundred micron pixel, you might have a flux of X-ray photons hitting that area. Now, if you break that in half, so 50 micron by 50 micron pixels. So you have four pixels instead of one big pixel, the flux in each one of those is going to be proportionally reduced.

Dr. Bill Cardoso:

That means that if you want the pixel to be as bright, you have to increase the flux of X-ray photons, which means you have to increase the power on your X-ray source.

Dr. Bill Cardoso:

Or, like everything in life, you increase the time that you’re going to be collecting those photos. Which in X-ray lingo is increasing the exposure of your sensor. What does it happen? Your image is going to get slower.

Dr. Bill Cardoso:

So you can’t have it all. I had an old professor that called nature.. The short blanket. You cover your feet, you expose your face. You cover your face, you expose your feet. You can’t have it all. You’re going to have to compromise.

Dr. Bill Cardoso:

So smaller pixels, you can get higher resolution, but your image is going to get darker. So I’d increase the power of the X-Ray source, which is going to, like we talked last week, increasing the power of the X-ray source means that you’re going to get more thermal noise, which is going to increase the size of the target, which is going to reduce the resolution, which is what you don’t want. Or you increase the exposure on the sensor, which makes everything slower which no one likes.

Dr. Bill Cardoso:

Some applications that do require high resolution no matter what, medical applications, some casting application, some industrial applications, they just live with a very long exposure time. You get to high resolution, but it takes 30 seconds. It takes a minute to collect an image. That’s why, because tiny pixels and you just keep collecting those photons and you collect, collect, collect, and to the point where you think you have a good image.

Dr. Bill Cardoso:

There’s also some considerations from the speed of the sensor. As you start shrinking the size of the pixel, how much of the electronics you’re going to keep inside each pixel? Because if you shrink enough, the electronics won’t fit inside the pixel itself, which means you have to bring, for example, the A-to-D converter, analog-to-digital converter, will no longer fit inside the pixel size itself. Now you have to bus the analog signal down, have the ADC outside, in the sensor. Then that means that you now have to have a pipeline of analog signals. And I think that’s more complicated. You have some more noise associated.

Dr. Bill Cardoso:

So there’s a lot of trade-offs that come into consideration when talking about resolution, speed, and the power of your X-ray source. So at this point, I don’t remember where the question was anymore, but I think I answered it at some point there.

Dr. Glen Thomas:

But another consideration that rarely we hear from an end user, they’ll look for pixel size, they’ll look for detector size, but one of the deciding factors on resolution as well is the scintillating material and the scintillating screen.

Dr. Bill Cardoso:

Yeah.

Dr. Glen Thomas:

So that will affect your overall gray scale values and it affects the perceived resolution. So it’s a complicated number. What we’ll do is we’ll query the customer and try to find the right combination. It’s not necessarily a super mega pixel. It could be a fairly benign number of pixels, but with a really fast or small grain scintillator.

Dr. Glen Thomas:

For people that don’t understand, the scintillator is essentially required in all X-ray applications because we have to convert invisible photons, invisible light, for lack of a better term there, to visible light or electrons.

Dr. Glen Thomas:

And actually on the scintillating screen is the ability to convert that. Essentially we excite a phosphorescent type material. Depending on the size of the grains of that material, we can change the speeds of the scintillator.

Dr. Glen Thomas:

Back in the old days, when we were doing film, we would use different cassettes with the film placed inside. You would use phosphorous screens of different speeds and grain size to increase the light being exposed to the film.

Dr. Glen Thomas:

So instead of taking a three-minute exposure, you can take a one-minute exposure because the… And in some cases, in film applications, you would actually use lead as a scintillating screen if you had enough power in your X-ray source.

Dr. Glen Thomas:

So scintillators are a huge portion of that resolution that most people don’t realize or even understand. So that’s where our expertise comes in. When we say this is the best detector for your application based on many values.

Dr. Bill Cardoso:

Yeah. It’s actually a pretty good point that we don’t talk very often. That screen, that scintillator layer that sits on top of the sensor, that defines how many visible photons are going to be generated by the screen that’s going to shine the CMOS, or whatever technology you have for your sensor as a function of one X-ray photon that hits it.

Dr. Bill Cardoso:

So let’s say a low efficiency scintillator for each X-ray photon emits one visible photon, for a high efficiency scintillator or for one X-ray photon emits 10 visible photons.

Dr. Bill Cardoso:

So that means is that with a sensitive screen, with a very sensitive scintillator, you don’t need as much radiation to produce the same image. So you can run the X-ray source at a lower power. That also means that you’re going to be able to see features on the sample that you might not be able to see with a lower sensitive scintillator because those few photons that actually show up on the sensor have to go through the source to your sample are not able to excite enough to produce any image.

Dr. Bill Cardoso:

So there are, nowadays, the two main scintillators that are used, Gadox and Cesium Iodide. Cesium Iodide is a higher sensitivity scintillator. Gadox is a little bit lower.

Dr. Bill Cardoso:

For a lot of applications, it doesn’t matter which one you use. You’re not looking for that extra level of gray scale to make a decision in your application. But for some applications, that’s what’s going to define success or failure.

Dr. Bill Cardoso:

That’s why you got to talk to people like us to figure it out, because you don’t want to overpay when you don’t have to. At least I don’t like to overpay when I don’t have to. It’s the thing I have.

David Kruidhof:

I think a lot of people have that one.

Dr. Bill Cardoso:

Yeah. I think it’s kind of a universal thing.

Dr. Glen Thomas:

People are funny that way.

Dr. Bill Cardoso:

In a way.

David Kruidhof:

Yeah. That brings us to another… I know it’s a long answer to a question as what’s the resolution of my X-ray system? How do you determine the resolution of the system as a whole? Are we talking about focal spot size? Are we talking about pixel pitch resolution of a detector? A lot of times we’ll see in spec sheets on systems saying the resolution of the system is this. Give a black and white number. It’s very clear. How is that determined and what’s your feeling on that?

Dr. Glen Thomas:

It’s never a black and white number. You can create a number, but it’s never black and white. As we were talking about detector resolution, we have X-ray tube resolution. We have magnification. We have how close the sample is to the X-ray source. A smaller focal spot will enable you to get more magnification by getting closer to the source.

Dr. Glen Thomas:

We do numbers. We do have image quality indicators of various types. With that, the same as with the detector, you have two different types of resolution there as well. You have contrast resolution and you have the ability to have geometrical resolution, a physical resolution of the X-ray tube, a set point, or a line in the sand that says, “This is your minimum resolution.”

Dr. Glen Thomas:

Then we have system magnification, which increases the resolution or perceived resolution. So if you think of it as a microscope, just because you look through a microscope, you haven’t really increased the resolution of your eyes. All you’ve done is magnified what you’re looking at.

Dr. Glen Thomas:

So that’s the same concept with system magnification. The system magnification includes the size of the monitor. If we have a 60-inch monitor versus a 10-inch monitor, you’re going to get a lot bigger image to view. Your defects or your artifacts or your landmarks are going to be huge compared to a smaller monitor.

Dr. Glen Thomas:

So there are a couple of different versions of resolution and how you achieve that. Just know that all manufacturers as a defacto will give you a resolution number based on the maximum magnification of the machine.

Dr. Glen Thomas:

If we use the point of contact resolution numbers, it would be just like looking at a small item with your eyes. You wouldn’t actually be able to pick out details. So that number would be quite small, somewhere four, five, six line pairs per millimeter, which is one of the indicators that we use to judge the overall resolution and magnification of an X-ray system.

Dr. Glen Thomas:

So again, like Bill said, it’s a short blanket. There’s compromises. X-ray in general is as general rule, building an X-ray system is a compromise. Everything you do is a compromise. Your magnifications a compromise, your resolutions a compromise, the amount of power that you’re producing in the X-ray tube is a compromise. Everything is a compromise. So you actually have to pick the best of all of the means available.

Dr. Glen Thomas:

One thing since we’re talking about maximum magnification gives you the maximum perceived resolution. A short source to image distance, like a benchtop X-ray system, it’s going to be limited on your geometrical magnification. So we may be depending on some system magnification in there to actually give you, say, 50, 60 line pairs per millimeter, which is what we would consider a high resolution X-ray system. Back in the day 30, 40 line pairs were pretty decent.

Dr. Glen Thomas:

One of the problems with the line pairs per millimeter is they were designed back in the day, back in the fifties or early sixties, and they only go up to 20 line pairs per millimeter. So that’s again, another compromise.

Dr. Glen Thomas:

So to eliminate that, we do some tricks with the line pair gauge to actually reduce those distances between the lines to give us the ability to measure up to 40 and 60 line pairs per millimeter. So resolution and magnification are key in, especially in microelectronics.

Dr. Glen Thomas:

In resolution-wise, if you’re looking at larger focal spots, you can increase the perceived resolution by having higher magnification, but you can also open up your source to image distance.

Dr. Glen Thomas:

Medical X-ray systems use a 3000 micron focal spot, which is huge, three millimeters. They do that because they have to get a lot of power and they have to do it very quickly in between breaths for the most part.

Dr. Glen Thomas:

So what they’ll do is the standard distance or source to image distance on a medical X-ray system is 36 inches. They can go up to 48 inches.

Dr. Glen Thomas:

In industrial applications, they may run a source to image distance out to 96 inches or so. The reason for that is they’ll get the best resolution out of the X-ray tube and out of the detector by opening up that source to image distance, and essentially, just using the sweet spot right in the center of that X-ray beam, which is essentially what we do with a microfocus X-ray source. In the five micron range with high magnification, we’re using the perfectly aligned photons. We’re not dealing with any of the penumbra or scatter or some of the other things that would affect resolution and affect that ability to measure that number.

Dr. Bill Cardoso:

Yeah. That’s the other thing that when you look at those, like what David was saying, the black and white. The new Ford Mustang GT 500 claims 0 to 60 miles an hour in 3.3 seconds. Now, do you think if I buy one, I’m going to get 3.3 seconds if I try? Yeah, right? That’s a special track with a special driver with the special tires. That’s what X-ray companies do like everyone does. You do your special, the best with experts running and then, yeah. I’m not lying. If you do all that, you can get those results. But the point number one is, do you need a GT 500 to go to Costco on a weekend? You don’t.

Dr. Bill Cardoso:

At the end of the day, and that’s something that we’ve been doing since we started the company is we’ve got to do what’s right for your sample in your company. You don’t have to spend $70,000 on a car to carry groceries, that actually, the trunk is like this big and end up overpaying for something you don’t need. You end up not getting things that you actually need.

Dr. Bill Cardoso:

So looking at specs and just doing a spreadsheet type of comparison of machines is very dangerous because you might end up cornering yourself on a hole and buying the wrong stuff. That happens all the time. It happens all the time.

Dr. Bill Cardoso:

Customers, if you’re not careful, you can just buy in some company’s marketing BS and like, “Oh yeah, I really need that. Oh, that’s the greatest new thing.” They wind up overspending for something you’re not going to use, and the stuff that you really needed, gets neglected.

Dr. Bill Cardoso:

So we set off upfront talking about resolution because resolution, like Glen said, line per gauge. You know what a line per gauge is? It’s a really thin film with gold-plated lines. For an X-ray system, it’s the best case scenario to produce any image. A really dense line, gold-plated, very heavy metal against a really thin background. How often do we have an application or customer with that? It doesn’t exist.

Dr. Bill Cardoso:

So what happens, as soon as you have mass and you have to go through different metals and different epoxies, et cetera, well you end up having to crank up the power on your tube and your resolution lowers. That’s just the physics. There’s nothing we can do about it.

Dr. Bill Cardoso:

So you think you’re getting that resolution. You’re not going to get it because that was done, like the GT 500, in a special track with a special driver with the special tires, the perfect condition, sunshine, perfect humidity on the track and you get that number. Yes you do. But does it matter to you?

Dr. Bill Cardoso:

So that’s what the conversation about resolution, that’s where we have to get through is what really matters, so you don’t try to buy, spend tons of money on a nano-focus source that you know you never going to get nano-focus resolution. First, you never needed it. And second, because the power out of the source, by the time it’s out, it’s not in a focus anymore.

Dr. Bill Cardoso:

So the bottom line is resolution on those data sheets are based on ideal conditions and should be taken with not a grain of salt but a bucket of salt.

Dr. Glen Thomas:

Right. The perfect storm for destroying your resolution is a aluminum wire or a bond wire under a stainless steel heatsink.

Dr. Bill Cardoso:

There you go. Yeah.

Dr. Glen Thomas:

Yeah. It doesn’t matter what your resolution is. You’re not going to be able to image it. The basic reason is when you shoot through that stainless steel, it’s going to cause all kinds of X-ray scatter and it’s going to essentially deflect those photons, and you’re not going to get a pristine image no matter what you do.

Dr. Bill Cardoso:

Even we talk about the simple PCB. A simple PCB nowadays, even the simplest ones, have multiple layers. We can have a power layer, or ground layer made of copper. As soon as you start putting those metals, you have scatter.

Dr. Glen Thomas:

Right.

Dr. Bill Cardoso:

There goes your resolution. So if you’re going through a standard PCB, you’re going to have… Again, unless you’re looking at line per gauge, everything else is going to get worse.

Dr. Glen Thomas:

Right. Right.

Dr. Bill Cardoso:

So beware people.

David Kruidhof:

Practically, the best thing to do is basically take a look at what you want to look at, right?

Dr. Bill Cardoso:

Yep. Best thing.

David Kruidhof:

If you inspect cell phones, then let’s X-ray the cell phone and see can you see what you need to see.

Dr. Bill Cardoso:

Exactly.

Dr. Glen Thomas:

Well, that’s the bottom line. Then that’s the most beneficial question you can ask yourself if you’re buying an X-ray system. Can I see what I need to see? Specs don’t really matter. It doesn’t matter if I have the highest resolution X-ray tube or even the highest resolution detector. Can I see what I need to see? That’s the simple question. Then everything else is based off that. All the specifications would be based off that simple answer of yes or no.

David Kruidhof:

No matter if you’re using algorithms to analyze it or you’re using our AI to analyze it for you. Can that see what you need to see?

Dr. Glen Thomas:

Exactly.

David Kruidhof:

Maybe it doesn’t look pretty to you in our eye, but the computer is like, “Oh yeah, that’s good. That’s bad. That’s obvious to me.” That’s enough, right?

Dr. Bill Cardoso:

Or even simpler stuff. We’ve had customers who bought machines and they realized that the samples don’t fit. Like, “Well.”

David Kruidhof:

Yeah, that’s true.

Dr. Bill Cardoso:

Or if you’re expecting a 48-inch board and you buy a 50-inch machine, there’s a chance that… “Well, but it fits.” “Yeah. It fits, but it doesn’t move. It’s just stuck in there. And once you put it in, it doesn’t come out.”

Dr. Bill Cardoso:

I think the danger in this, and we’ve seen this from time to time, some companies try to focus the customer on one specific aspect of the machine, resolution and… They corner you in that aspect until that’s the most important thing. Don’t think of anything else.

Dr. Bill Cardoso:

When you buy capital equipment, you buy a machine to inspect a wide range of product that most of our customers do, you’ve got to be more holistic. Yes, some of these things are very important, but the overall package, if you can’t move your boards in and out of your samples, and samples in and out of the machine as fast as you want or need, integration of the machine to your ERP or MRP. All these different things that play a role. Support right afterwards, after you buy a machine, things break, things change in your manufacturing, in your processing.

Dr. Bill Cardoso:

So are you going to have someone who can help? So it’s a much broader choice that you have to make. And if you have, it’s interesting. So if one company is focusing you in one aspect only, be suspicious. There are probably something else missing they don’t want to tell you.

Dr. Glen Thomas:

Right. Because keep in mind, every thing with X-ray’s a compromise.

Dr. Bill Cardoso:

Yeah.

Dr. Glen Thomas:

Right.

David Kruidhof:

Well, thank you, gentlemen. Appreciate the time.