What Makes Galvo's Faster?

[James Lehman]

Ah, okay, Just like the steps in a stepper motor. Obviously a stepper motor makes much larger steps and is not nearly as quick.

Stepper motors have very complex magnetic structures with many poles. They are designed to step from one pole alignment between the rotor and stator to the next, (or half or quarter step). None the less the idea is to be able to repeatedly move an exact portion of the circle CW or CCW and stay there in a single action; as orchestrated by a stepper motor controler.

Scanning galvanometers have a much simpler magnetic structure, much like a needle movement in your analog meter (which is a kind of galvo!). They have no tendency to locate and stick to any pre-arranged system of poles. They are truly analog.

James.

[300EVIL]

Stepper motors have very complex magnetic structures with many poles. They are designed to step from one pole alignment between the rotor and stator to the next, (or half or quarter step). None the less the idea is to be able to repeatedly move an exact portion of the circle forward or backward and stay there in a single action. Scanning galvanometers have a much simpler magnetic structure, much like a needle movement in you analog meter. They have no tendency to locate and stick to any pre-arranged system of poles. They are truly analog.

James.

Yes, I completely understand that. I was just making the comparison of degree points to radians. Probably a bad one.

[Pangolin]

I really think their should be some more accurate standards for rating galvo/amp kits. It should be something like a perfect test pattern @ 30KPPS @ X Scan Angle for 1 hour @ X maximum galvo temperature and X maximum amp temperature. I don't know... Just a thought.

Yes, it can be confusing. For scanner experts, a handful of patterns is all that is needed. I think the biggest problem is lack of knowledge on the part of scanner manufacturers as well as users. While I was in China, I visited a number of scanner manufacturers. Tried to give all of them good advice. All appreciated the advice, but a few hours spent here and there is really not enough. It would take days...

So, in summarizing your detailed response, It sounds like the quality and heat dissapation of the windings, machining tolerances of the galvo bodies and rotational mass seem to be the biggest factors in a galvos rating. Is that correct?

Yes, although "rotational mass" (actually, it's mass at a radius, which is called moment of inertia) is not much of a factor in the DIFFERENCE between the kinds of scanners we would be naturally talking about, since everyone is making copies of the original 6800, and thus, all have roughly the same "rotational mass". The mirror sizes may be different, but don't be fooled. The mirror has so little inertia when compared to the rotor, that it's like a jockey on the back of a race horse.

Obviously the amplifier design has a lot to do with performance as well.?

LOL! Well, they are all supposed to have the same design. Or, let me state it another way. They all kinda start out with the same design -- that of the original Cambridge model CB6580. But then some people change components, I would say, so that they can feel good about themselves and pretend that it's not a copy of the CB6580. But any change to that design will lead to sub-optimal performance. (Consider an automobile -- four wheels is the optimum number. Reducing or increasing the number of wheels will lead to sub-optimal performance.) So the question becomes -- exactly how much did they deviate from the CB6580 design, and exactly what quality level parts do they use?

The reason for my curiosity is that the galvo's I have had have all appeared pretty much the same with very large performance jumps.

Then you got lucky . As people have sent me their amps to tune and modify, I see vast differences in large-signal performance. Also, if the amp is over-tuned (high frequency damping way too far up), the large signal performance will really suffer. This was the case with MechEng3's amps before I retuned them. I am sure he will give a few words. But there were countless others as well...

I also understand the "bin-out" method. This is done in a number of industries. While I don't agree with it's practice, I could definitely see where faulty quality could be sold as a lesser device.

Bin-out is a fact of life, so this, in and of itself, is not that surprising. For example, all TL084 start out the same. The only difference between an TL084A and TL084B is due to bin-out. The best parts come from the center of the die, but this isn't related to scanners.

Basically, if you machine things in large lots, there could be a range of different parts you wind up with in terms of tolerance. Controlling these tolerances costs money, so if you want to make things cheep, one way is to not control tolerances all that tightly. Note that not all G-120 are exactly cream of the crop .

With Bill advising these guys in China and IF they take on board the hints and tips he gives them (no doubt free if I know Bill) we should see a slow improvement in their product.

Yes, and yes . Yes the advice is free, and yes, if they would actually take it, it would result in better products for everyone. But sometimes I have to tell them over and over again. And lets say that they only take 70% of my advice. So we have to go through around 5 iterations of me telling them what to do, before they get to the point of taking 99% of my advice...

Why do you think there is always the qualifying "at-so-many-degrees" when they rate scanners at points-per-second?

This is going to sound like an odd answer, but it's because of "power supply". If you take the ILDA test pattern, and start displaying it small, then slowly increase the size, what you will notice is that, at some point, the circle will stop growing (and probably get distorted). At that point, you have "run out of power supply" and reached the peak current available. The current limit may be dictated by several factors (including the power supply voltage itself). But the primary reason why we say "at so many degrees", is so that people don't expect to display the ILDA test pattern at full size.

Best regards,

William Benner

[survey]

It's allways a pleasure to read posts like this.
It remember's me high school times, and Bill you're a good teacher.
Your explanations allways hide something, like if you want us to think about what you're saying and put more questions.
At this moment I'm studying 6210H vs 6215H, finally I figured that

RMS Current: 4.1A, at T

case=50°C and Peak Current: 20 A, Max
on the 6215H well deserve the waste of 30 micro seconds in
Small Angle Step Response.

(Maybe that's why Bill recommend them).

But besides the galvo itself, I'm trying to figure if a true hi-fi mosfet amp (tired of iC power amp's) and real power supply with trafo, bridge rectifier, and large capacitor's (10000uf) would do a better job.

Even if I have to put that PS away from projector

Cheer's to all

Jose broco

[Pangolin]

Your explanations allways hide something

Hide something? My explanations are not meant to hide anything. If they do, then maybe I am not such a good teacher after all ...

At this moment I'm studying 6210H vs 6215H, finally I figured that

RMS Current: 4.1A, at T case=50°C and Peak Current: 20 A, Max

Unfortunately those specs won't mean much to most people. 20 Amps Max is what I call the "fuse current". If you put 20 amps into that scanner, the coil will blow like a fuse. In reality, because of the inductance of that scanner, there is no practical way that the current would ever reach 20 amps anyway. So this particular spec is pretty much meaningless...

As for 4.1 amps RMS, this is very meaningful, and very impressive. It means that you can put 4.1 amps into the scanner at DC (i.e. forever) as long as you are able to maintain the case of the scanner at 50 degrees C. The reason why this is meaningful and impressive is that, compared with a Cambridge 6800 (and remember, that's the scanner that most Chinese companies copy), the 6800 can only withstand an RMS current of around 1.5 amps with a case temperature of 50 degrees C. So the 6215 can withstand 2.7 times as much RMS current. The KT (torque constant) of a 6215 is also higher than a 6800 (or 6210), so not only can it withstand more RMS current, but also produces more torque for this current.

Basically, it can be said that the 6215 is four times as good as a 6800 at converting heat into motion. But unfortunately, due to "square law" phenomenon within the mathematics of motion, this does not give you four times the speed. It does however deliver twice the speed of a 6800 for a given angle of comparison, or twice the scan angle at the same speed.

There are many beneficial qualities of a 6215 which are too compicated to discuss in this kind of forum. Lets just say, I like it for a number of reasons. The price is generally only incrementally more expensive than other Cambridge scanners as well.

on the 6215H well deserve the waste of 30 micro seconds in Small Angle Step Response.

Waste of 30 microseconds? I am afraid I don't understand the reference...

With the 671xx high power amp, the 6215 is the fastest scanner that currently exists on earth. I don't think any time is being wasted...

Best regards,

William Benner

[Pitts]

Bill, as we say down here ....


Good on yer mate !!

Thanks for all your help here on the forum and with the Chinese guys, must be frustrating as hell at times.

Cheers

Ray
NZ