Analog modulation connections

[jmilan]

Im building a Green/Red laser proyector with analog modulation and I whant to make sure this is the right connections, this is going to be controlled using FB3

Green Laser
+ Lead to PIN 6 on DB25
- Lead to PIN 19 on DB25

Red Laser
+ Lead to PIN 5 on DB25
- Lead to PIN 18 on DB25

Do the laser power supplies need to be common grounded to PIN 25?

Thanks

[buffo]

I always thought that the negative modulation (blanking) leads from your lasers should ALL be connected to pin 25, not to the negative color signals. (That's the way mine is conencted, and several other PL members have their projectors wired this way as well.)

I assumed that those negative color signals (pins 18 and 19 in your example) are for inverted color signaling (Laser is OFF when the modulation line is positive), or for the rare case when you want to use differential signaling for color.

However, after reading some of Bill Benner's recent posts concerning pin 25, I'm starting to wonder if this is, in fact, incorrect. Bill has mentioned several times that pin 25 should only be connected to the projector power supply ground, and nothing else. I'm inlined to believe that he knows what he's talking about, which means many of us may have our lasers connected incorrectly.

How about it, Bill? Care to comment on this one?

Adam

[buffo]

Well crap. Once again, my assumptions have led me astray!

I spoke with Bill last night and I asked him specifically about the color modulation connections. And guess what? My projector is wired incorrectly! (I'm guessing the same goes for a couple other members here as well...)

jmilan, you were right. Your modulation wires for the red laser should be connected to R+ and R-. Same for green and blue. Do not connect them to pin 25 (the ground pin). As Bill posted in another thread, that ground pin should only be connected to the ground of the power supply in the projector. Nothing else.

In the case of the QM-2000, it really doesn't matter, since the negative color signals are held at ground potential in all the newer QM-2000 models anyway. But using pin 25 as the negative modulation connection is incorrect, and could lead to issues with other controllers.

Also, my assumption about the R-, G- and B- pins being used for inverted color signals was *also* incorrect. The R+ and R- were originally set up for differential signaling, but these days no one has a differential receiver for color signals in their projector... It's all single ended. Bill told me that there *was* a controller (Lasergraph, I think) that used to output +2.5 volts on the R+ and -2.5 volts on the R-, but that was an exception. Nevertheless, the proper way to connect the modulation leads is to use both the R+ and R- pins. (Ditto for green and blue)

So there you have it, folks. Once again I discover that something I *thought* I understood fairly well turns out to be completely wrong. Thanks to Bill for clearing it all up!

Now if you'll excuse me, I've got to pull the back off my projector and move a few wires around...

Adam

[jmilan]

Adam, when he mentions that PIN 25 should be connected only to the power supply of the proyector, what does it mean?

I have one power supply for the scanners, one for each laser and one for the FB3 board.

Should pin 25 be connected to each one of them? or just the scanners or FB3?

cheers,

Joel

[DZ]

Also, my assumption about the R-, G- and B- pins being used for inverted color signals was *also* incorrect. The R+ and R- were originally set up for differential signaling, but these days no one has a differential receiver for color signals in their projector...

Interesting, then it could be said that the QM2K's DB-25 output connector does not conform to the ILDA standard. Unless it has been changed. It reads, "When compared to the ground level of the differential driver, the voltage level of the normal signal line shall be +2.5V and the voltage level of the inverted signal shall be -2.5V." Use to be that the QM2K was able to supply a differential signal for color, when running single ended, you manually grounded the -RGB lines so the QM2K would output 0 to 5V on +RGB.

As Bill posted in another thread, that ground pin should only be connected to the ground of the power supply in the projector. Nothing else.

This is even more interesting. Pangolin's help file reads, and I quote, "
You do not need to run separate ground wires for each signal. It is sufficient to run one ground wire to your scanner amp or projector. Of course, if you have a number of different devices, such as a scanner amp, PCAOM driver, and shutter driver, each must have its own connection to the QM2OOO ground." So whats with the change?

Ok, I think I've found the other thread that has been refered to, http://www.photonlexicon.com/forums/...6372#post46372, quite interesting that the way I read it, it contradicts the information in the manual. Which I've always tried to follow. Granted, I've only started using differential amps for the past few weeks, and so far I notice no problems, in fact, they seem to work better than the original DT-40's as far as image quality is concerned. Maybe an artifact of Bill's modification of the DT-40pro's that I have, which included thicker mirrors. I'm not sure I understand how running the ground from the DB-25 to each amp and each laser is a problem. To prevent a ground loop problem from occuring on a wire only a few inches long? If you're running proper gauge wire and you're connections are sound, I'm not sure I see how a problem can occur. If you're running 20 feet of ground in a projector, then I could understand how there could be a potential issue....

[carmangary]

What about the shield of the cables? Does it hook to pin 25? Or the power supply ground? I assume not both? If not, then I guess it makes no difference which one it is connected to. Or is a shield even needed?

[DZ]

carmangary, to quote again from what I though was the laser bible, "Ideally, the cable going from the QM2000 to the projector will have its shield (wire mesh or foil wrap around the cable) connected to the ground line (pin 25)."

[buffo]

With regard to the ground connection to the amps: The way Bill explained it, it's not so much a ground loop in the traditional sense (in that you're picking up noise), but rather a ground "lift" that affects the input signal to the amp. Let me try to paraphrase what Bill explained to me last night:

Say you have pin 25 connected to the ground pin on the differential input of your x and y amps. Now, let's say you send a strong signal to the X amp only. (Let's assume you're scanning a horizontal line.) There should be no movement in the Y axis. But the X galvo is going to take off at high speed. This places a load on the scanner power supply, and it can cause the ground to lift (that is, rize above zero potential) for a brief moment.

When this happens, if that ground is connected to the ground terminal on the scanner amp signal input, then the amp will interpret this lifting ground as a signal. Thus, the Y scanner will move slightly. You'll see this as a wavering in your otherwise straight horizontal line. This is why it's a bad idea to connect to the ground terminal on the differential signal input. It's designed to look at the difference between X+ and X-. Nothing more.

To demonstrate the problem, Bill suggested loading up the quadrature square wave pattern and then displaying it with the Y axiz size set to zero. Ideally you'd only see a thin flat line with two bright dots at each end. But if you have the ground connected to the signal input on the amps and you're getting any ground lift from the power supply, then you'll notice some wavering in the line. Likewise, you can collapse the X axis and only scan a vertical line and then look for any horizontal wavering in the vertical line. Disconnecting the ground on the input should cure the problem in both cases.

Bottom line, he said that pin 25 should be connected to the power supply(s) ground in the projector and nothing else. As for the ILDA cable shield, I thought it was already connected to pin 25 on the QM-2000 side???

Now, in the case of a projector with multiple power supplies (like most of us have), if you connect all the grounds of all the power supplies to this pin, then you do technically create a small ground loop inside the projector. But that really isn't a big deal, because as David already mentioned, we're talking about a loop of only a few inches of wire. And though this could induce a signal shift in your modulation circuit due to ground lift (because the negative input on most laser modulation circuits is already common with the ground of the laser power supply; it's wired that way inside the driver), the effect should not be nearly as pronounced. (Think about it - will you notice a slight, intermittent change in brightness of only a few percent? Probably not. But you'll certainly notice a few percent inaccuracy when scaning a horizontal line that shouldn't have any vertical component.)

Bill acknowledged that with modern solid state lasers, it's nearly impossible to isolate the lasers from each other. Usually the laser head is grounded to the driver power supply ground. (Remember, this ground is also common with the modulation signal ground.) And since the head bolts to the metal optical plate, it's common with all the other laser heads. Then the optical plate is bolted to the scanner block, which is in contact with the metal case of the scanner, which is *also* conencted to the power supply ground through the scanner amp! So everything is already tied together in a big loop.

But you are far less likely to have any signal interference from a ground lift on the modulation circuits, because 1) you're not talking about large changes in current like you are with respect to what the scanner power supply has to deliver to the amps, and 2) any lift in the ground caused by the scanner power supply will at best cause a slight variation in the color intensity that is unlikely to be noticed.

I suppose the ultimate solution would be to build a single power supply that could run everything in your projector, and build it robust enough that even with everything on at full power and the scanners at max acceleration, you still have plenty of headroom such that the ground stays stable. But, again, that's not the way most of us have built our projectors. (At the very least, most of us have a second, separate PSU just for the scanners, if not multiple PSU's for the lasers as well.)

Bill did say that earlier versions of the QM-2000 *did* support the +2.5v to -2.5 volt color signaling, but since most users were wiring their projectors for single ended color signaling anyway, it made better sense to just tie the negative color signals to ground internally on the board. He said that all QM-2000's from about serial number 5000 (or thereabouts) and up have the negative color pins at ground potential.

Adam

[Pangolin]

Hi guys,

Interesting discussion here. I already made some comments in another post, which you can read here:
http://www.photonlexicon.com/forums/showthread.php?t=4152


With respect to what DZ pointed out -- the earlier recommendations made in our help file, I have only four words: "Hey man, live and learn...". OK, five words .

When I did the help file documentation years ago, it was done out of haste and logic. The "logic" part was this. Scanner amps have three input connections -- Signal +, Signal - and Ground. Therefore, you have to use all three, right?. Makes perfect sense, doesn't it?

Well, unfortunately, if you do that, you could get yourself into trouble, in a couple of ways... One way is this "ground lift" scenario that Buffo and I discussed on the phone. This is a very real phenomenon (in digital logic, it is referred to as "ground bounce" and if you look it up on the internet, I am sure you will find references). Another is ground loops from the AC. Both are a little bit different, but both might be (regrettably) called "ground loops".

In an absolutely ideal world, you would only use the Signal + and Signal - of a scanner amp (or other equipment). Unfortunately, this is not optimal in most cases BECAUSE the input system of scanner amps does not have what's called an infinitely high CMRR, or "Common Mode Rejection Ratio". Since amps don't have an infinite CMRR, this means that, the further signals get away from "ground" (or zero volts), the worst the input system is at rejecting noise. There is another factor too, which is that the input system also doesn't have an infinitely large acceptable input voltage swing. So, for example, lets say you are in Europe, and you have your computer plugged in "one way" without a ground, and your projector plugged in "another way" without a ground, and you have 220VAC between the two pieces of equipment, the scanner amp really can't handle that kind of common-mode signal. They can only handle about at +/-20V common mode signal, and even with that high of a common mode signal, the CMRR is severely limited, just due to cheap construction techniques of most amps.

Everything in the last paragraph is a bunch of technical mumbo-jumbo. The point is, scanner amps need "help" to bring the common mode signals (the Signal + and Signal -) within a range they can handle, and it can get this help from the "ground".

Now, here's the rub with respect to why Cambridge makes the recommendation they do in their manual, and why there are three connections on a scanner amp. If you had only a single scanner amp, then you might get away by connecting your "signal ground" to the scanner amp's ground. But for everyone here doing two-axis laser graphics and beam shows, you have two amps, not just one, and connecting your "signal ground" to the ground connection of both scanner amps will cause this kind of "ground loop" (I hate that term because it is way over used), and the X amp will experience "ground bounce" when the Y amp undergoes hard acceleration, and vice-versa. The result is crappy images (to some extent), the extent depends on the gauge of the wire, the degree of acceleration, and other factors. You can avoid all of this by making only one single ground connection *somewhere*, and the best place to do that is the common ground connection at the power supply itself.

After all, if I am full of bullshit, why is it that you will find discussions on the internet about "single-point grounding scheme" and "ground bounce"? These are somewhat advanced topics that most people won't consciously encounter -- and REALLY easy to overlook. Remember how many people on PL think that "ground is ground"? And remember the discussion in another topic about "Digital ground" and "Analog ground" on the QM32? If you take a meter, you will measure close to zero ohms between the two. So, what's the difference? There is a big difference!!

The best grounding approach is a single-point grounding scheme (otherwise known as “star” grounding scheme). But in the end, most of the folks here are hobbyists, and part of the fun is learning yourself.

As I wrote in a previous post, you can follow my recommendations or not. It won't hurt my feelings or my reputation one way or another. And doing the experiments is part of the fun and life experiences!

Best regards,

William Benner

[carmangary]

Nevermind. Found my own answer.