Anyone have General Scanning Z1913 Galvanometer specs?

Gentlemen,

I bought a couple of GSI Lumonics LightWriters (50W YAG markers) off ebay and am now stuck, unable to revive these units, because the guy who sold them omitted the little yellow security dongles that enable the installed Windows 95 ProLase software. (And now he "can't find them". What a ripoff.)

So before I give up totally, I'm thinking of repurposing the galvo heads using OpenAPC/Beamconstruct, or since I've spent the past six years integrating various machinery using Linux CNC, I might go that route.

These machines each contain a pair of Measurement Computing (MC) PCI-CTR05 and PCIM-DDA06/16 cards that drive an incredibly ancient card-plane driver board called Scanner Interface DG300 which in turn drive the X and Y galvo mirrors which are of type "General Scanning, model Z1913."

Attachment 50039Attachment 50040Attachment 50041Attachment 50042

I've been looking unsuccessfully for specifications and pinouts for the above Z1913 galvos - anyone here familiar with these units? They have DB15 connectors. I'd like to know the specs and pinouts so I can attempt to drive them without destroying them. I'll try first to drive them through the card (using only +/-10V) but if I must, I'll drive them directly.

The only control going to these galvos is the (presumed) +/-10V signal from the MC cards to the DG300 scanner control backplane card - I imagine this card controls the closed loop, or temperature compensation, or whatever. I won't know until I can get some specs on these galvos.

Anyone?

Torin...

The drive wires on GS glavos are typically the same. They are the yellow, black, green, and red wires found in the unshielded cable. They should match other GS scanners but you can put an ohm-meter on them to be sure (there are two coils in the stator). As far as impedance, drive voltage, current and such, you'll probably have to wait for one of the other users here that may have info for that. User: "mixedgas" typically has literature on almost anything in the laser field and as long as he is not under NDA for this, he will generally share. As far as the feedback wires, I am not sure about the pinouts for those. I would think they would match other GS galvos but I couldn't be sure. If you're just wanting to drive them "open loop", the drive wires should be simple to hook up.

What I'm seeing is yet another modified G330DT Galvo that has either better temperature stability or pointing accuracy improvements, or better position sensing linearity.. So it is a semi-custom G330 with closed loop feedback, with just enough mods to rate a "Z" customization number... Plus the external "T" option for heat tapes that stabilize the position sensors..
~

The amps are oddball though. I see some analog "660" style circuitry in there, with parallel input latched DACS driving the closed loop amps. But the output stage is way different and there is what looks like bus address decoding logic at the bottom..
~
I can get you the galvo pinouts, but you'll have to reverse engineer the strange amps... I've seen some strange galvo drivers, and those are just plain weird... I can give you some schematic fragments in time, but the DAC chips and the data latches are your problem... You could remove the DAC chips and inject an analog signal after the I/V converter opamp....
~
What you have is slow but very, very accurate, and probably somewhat corrected for tangent error...

~
~
~
Here's a A660 schematic fragment..
~
http://www.laserfx.com/Backstage.Las...Pinouts26.html
~
The position sensors in the Galvo emit tiny DC currents overlaid on massive RF noise... In this case there will be an AGC signal feedback to the Galvo's sensor for added stability... Beware, what is black and red on a position sensor is not what you would think of in terms of normal EIA color code.,,,
~
!
I need to dig you up a GSI Galvo engineering support brochure and scan it, maybe by tomorrow...

Steve

Mixedgas, and Absolum7691, thank you for your time and effort. I greatly appreciate your help. Yes, thank you for the pinouts of these galvos.

The next issue is to get the lamp to fire. It's driven by an MC 150-40 HV PS which is easy enough to modulate, but I need to do more research on driving the Q-switch. It might be automatic if left alone, driven off its existing RF driver and power supply. So I think I just need to turn on the lamp and see if the system is still aligned. If the lamp doesn't fire, since these lamps require lots of current and heat, I might like to explore a second option: Whether I can modify the laser head and pump it with a diode instead to reduce the resources necessary to power/cool them.

I have a second LightWriter head I can experiment with in an attempt to diode-pump the YAG rod which will greatly reduce the power requirements and allow me to cool the system with a smaller 900W refrigerating recirculator.

eBay has a 100W DPSS QCW laser array outputting at 808nm which is just the input frequency required to excite the YAG rod in these LightWriter heads - but getting the light into the Quantronix elliptical cavity will be an issue. I'm thinking I might reuse the quartz tube that held and cooled the lamp and between where it seals at the enclosure walls, I can frost the inside of the tube or fill it with acid-frosted quarts balls or cylinders to create a diffuser of sorts (perhaps an optical rod already exists for this purpose) then aim the DPSS laser right into the cavity through this diffuser rod to emulate the effect of a lamp. With a CW (I think - will have to check again tonight) rod and the higher conversion efficiency, I might be able to get the YAG rod lasing at power outputs comparable to the original lamp.

http://www.ebay.com/itm/100W-808nm-Q...SjT6lqh29bkcBg

The last issue is the CW rod. The DPSS even though q-switched itself, might not produce enough output to saturate the rod... but it's an interesting experiment for only $120 US. I suppose I could always replace this rod with a better doped rod, too.

If you're wondering why I would go from 808nm to 1064 instead of just using 808 output directly, it's because a) I think I might damage the mirrors at 808nm, and b) the system is already designed to collimate light from a lamp through the YAG rod and onto the target, meaning I can avoid having to invent/buy an array collimator.

Anyway, these are just rambling (but hopefully reasonable) thoughts which will become more coherent as I continue investigating my options.


Torin...

Regarding your YAG... here is an interesting thread/video of dual end-pumped DPSS laser with fiber coupled diodes:

http://www.photonlexicon.com/forums/...er-a-New-Video

Mixedgas,

I found a meager brochure on Camtech giving a few details on the D300 series (D330 included), but it's more of a sales brochure rather than a tech spec as it had been named. Yes, it appears to be slow - 130hz, but that's PLENTY fast for what I intend to do with it. In fact, I don't think I can drive it nearly that fast if I'm running through EMC2's gcode interpreter, but we'll see.

Any by the way, now that you mention it, the DG300 ("D"igital "G300") Scanner Controller completely makes sense. These guys (Lumonics) installed the card in the backplane merely for power, then fed the x/y signals in exactly how you explained. Unfortunately, I this card may be fried because when I found it, it was a nesting home for mice, so I had to clean it up. Now when I power the backplane, it drops one of the three-phase lines to 0V which means I have a short somewhere. After moving the backplane onto another of the three-phase bus lines, I caused the whole circuit to draw too much current and both shorted a relay in the main controller (fortunately, I have two of these controllers with relays to spare) and caused some of the wiring to smoke. If I haven't completely ruined the mainboard, I'll swap out the possibly defective DG300 backplane card for the spare I have and see if the mainframe still draws too much current. If not, I'll solder in a new relay and try again.

Anyway, this is turning out to be a fun project, and I mean this in the sarcastic sense. I just want to get the laser operational so I can use it to manufacture my product. I never wanted to spend countless hours re-integrating this laser, but I *really* need a laser for my manufacturing, so I'll wait to see the pinouts and full specs from your Galvo Engineering Document before I decide how much work this is. It could turn out to be much simpler than I estimate... stuff usually does.

Absolum7691 - great video, and I love the treatise on his choice of fiber-coupled vs diode bar array and temperature control for tuning the coupling wavelength which will be very important for my application. Overall, very informative. But now having seen it, I actually want the less-than-ideal collimation of the bar-diode because my intent is to couple it to the yag rod by diffusing it into the cavity. I actually found some fused silica in the Edmund catalog that is specifically hit with 120 or 220-grit sand-blasting to cause uniform diffusion, so I might find a glass rod with the 808nm in mind and bead-blast or possibly hand-sand its surface, mirror one end with an aluminum plug, and insert the plug and the rod into the lamp holder instead. Two acylindrical lenses should collimate the fast and slow axes well enough to direct it into the lamp/diffusion rod sleeve and then onto the gold mirrored cavity. This last part seems like the fun stuff, which I'm looking forward to.

If after all of this I ultimately decide that the integration costs aren't worth the effort and my time would be better spent making/selling product and buying a complete system (which will most probably have to be a chinese system), how much do either/both of you think these Lumonics heads are worth?

XY Galvo head with G330's
YAG (1064nm) focusing optics
Servo-actuated shutter
Q-switch with RF driver and power supply
Quantronics water-cooled YAG (4mm x 50mm rod) gold-plated elliptical cavity
Cast Iron bed with partial and fully-reflective adjustable mirrors

Note: All cables, hoses, etc, intact and still terminated with DB-connectors. (i.e. properly removed, not severed)

I have two of these systems, one is actually a complete system with enclosure, mainboard, HV power supply, closed-loop cooling system, etc., and purportedly still operational at the time I bought it. The other system was for parts and I've discarded the housing, most of the closed loop except for the pump, and kept the head and vital components. Oh, and I have two pairs each of the CTR-05 and PCIM-DDA06/16 cards which drive the heads.

Useful? Junk?

I suppose guys here and on ebay are only interested in the heads though.

Quote:

---

Originally Posted by torinwalker

Useful? Junk?

I suppose guys here and on ebay are only interested in the heads though.

---

Definitely not junk. We have quite a few members on here that run multiple YAG systems and as things move forward, some of these components are getting harder to find. What I would recommend is taking pics of what you have and putting them up in our Buy/Sell thread. I am sure you will get a few bites on some of it.

Good to know, and thank you but I'm not quite ready to give up yet. I'm still interests in trying to drive these galvos so I'll wait to see if Mixedgas comes through with those engineering notes and pin outs.

I just don't want to throw good money after bad - so once I gather all the facts, then I'll decide to keep or sell these items.

Mixedgas - did you get a chance to scan those documents?


Torin...

DB15M GALVO connector



Pin Number Wire Colour Signal Description
1 Red Drive Coil (See notes 1 & 2)
2 Yellow Drive Coil (See notes 1 & 2)
3 Black Position output (-)
4 Orange Position Output common
5 Red Heater - T-Type (See note 3)
6 N/C No Connection
7 Red Heater - T-Type (See note 3)
8 White Oscillator supply voltage (+)
9 Green Drive Coil (See notes 1 & 2)
10 Black Drive Coil (See notes 1 & 2)
11 Red Position Output (+)
12 Violet AGC Output
13 Brown Oscillator supply voltage common
14 N/C No Connection
15 Blue Thermistor - T type (See note 3)

NOTES

1/ To wire drive coils in series, connect Black to Green and drive Red to Yellow. Series wiring is recommended for use in position sensing (closed loop) mode.

2/ To wire drive coils in parallel, connect Red to Green and Black to Yellow. Use parallel wiring only when driving the galvo in open-loop mode

~
~
The easiest thing in your situation is to find the analog ground on the Amplifier card, should be fairly obvious. Then trace the positive and negative rails back from the power supply pins on the op-amps. That should go to a pair of voltage regulators some place. The power transistors in the output stage should also have their own Bipolar power rails. There will be a logic supply for the latch chips as well...
A small circuit will provide a precise, isolated, low noise 12VDC for the position sensor in each galvo..
~
~
Once you have that you can insure the card is properly powered.
~~
With those big mirrors, you really need the closed loop feedback from the amplifier working, you could drive the galvos open loop, but the small amount of additional work to get the existing hardware working is worth it. The difference between open loop and closed loop is an order of magnitude increase in speed, and about two orders of magnitude increase in position accuracy...
~~
You then need to look at the data sheet for the two DAC chips and reverse engineer the DAC chip output circuitry. The output of the DAC should go to a gain/scale and DC offset potentiometer for each axis. Once you have that data, you can lift a resistor or something and inject a audio signal generator signal (preferably a quality bench signal generator, 1 Hz and then 30 hz square wave, with adjustable DC offset) into the galvo chain. There will be a 0.1 Ohm or so current sensing resistor in the galvo coil's return lead for each galvo, a 10X scope probe goes there so you can watch what is going on... There is also a position feedback tap some place that shows the galvo's true position for tuning, the current sensing resistor reflects the PID loop's actions and the amplifier Drive, not the smoothed position.
output...

~
~

I generally charge about 100$ and shipping to reverse engineer a galvo amp's pinouts... If I have to actually run and evaluate the hardware, that gets a bit more expensive..
~
When you power on the machine, do the Galvos power up and center? Where are you located at? There are always CX-660s on Ebay...

~

Steve

Man, I love context. When I first saw the pinouts I thought, "Ok, this is useful...now I just need to understand the AGC and why I'm supplying +12." But with the A660 documentation, wow... I suddenly feel empowered to make this work. Mixedgas, thank you!

The A660 documentation is incredible. It's a wonder companies still don't design products this way - maintainable, with documentation and schematics that explain the operation. Man, with equipment and documentation like this, no one would ever need to throw away electronics... just fix it. AND, 0.1" components would stay relevant forever. I miss the good old days when everything was accessible. Of course, we're in a new revolution - young people are finding it easier (thank you arduino and other modularized microcontroller systems) and so perhaps there'll be a resurgence of electronics hobbyists - just like in the HAM era.

Anyway, this is hopeful now and gives me a fascinating new hobby to work on.

I've decided to return the 30kva transformer I bought that was intended to power these monster laser systems. I'm actually going to play with diode pumping and see if I can't get these heads to operate without all the heat and power necessary.

Wish me luck. I'll certainly be in touch to report on progress, though I must admit I'm not the best blogger. I usually get lost in the integration and end up taking more pictures toward the end. But we'll see.


Torin...

Sorry, I missed the question before posting, re: do the galvos power up and center.

I had a bit of trouble with the computer in this system - either the hard-drive failed, or the power supply failed (not enough power for some reason causing the hard-drive to tick-tick-tick, and the mainboard and video to repeatedly reset), and all probably due to the failed scanner controller that I tried to clean. Perhaps it is beyond repair and now shorting the supply. Today, Saturday, I'll give it all another go and see if I can't try to get the computer running. While there, I'll see if the galvos center as you asked. I don't see why not - the gentleman who sold it to me claimed it was working before they put it into storage, and I have no reason to doubt him. The computer is chocked full of his logos and he clearly was using it while he had it. The controls all seem to work - the shutter on the head opens/closes, the RF driver for the Q-switch shows a signal (though presently at fault - I'll figure that out), the ESKI power supply turns on (though I can't power it up on my measely 4.5kva transformer unless I can keep the current low), and notwithstanding the fault when I turn on the water pump, I think, due to low water, it all seems good. I just need to work at it a bit more.

Anyway, I'll report back for the benefit of your curiosity but not because I insist on your continuing help. You've been extremely helpful already.

Now I'm all excited again.


Torin...

Bah... went into the garage today as promised and swapped a bunch of parts between the two computers to restore the one whose hard drive was ticking. It was probably the power supply. Well afterward, while booting it up and testing to see if I could get an AEGIS security dongle to work (someone mailed me one for free), the software still didn't start because the dongle was for the wrong version THEN, I heard a pop and the breaker tripped. Not sure what it was, but something in the system was drawing too much current because afterward the third phase fuse was blown and the transformer burned my finger. This is after I pulled practically everything from the bus and only had the monitor running...

So basically, that's it. The systems are too old to attempt to restore as a lamp-driven, water-cooled monster. My only recourse is to either diode-pump the head (my preference, though cost is a concern), or dump both systems and all the parts onto eBay, which I've done. If I get any takers - reasonable offers included - I will one or both of these lasers. If not, I will keep looking for good deals on CW diode lasers in the 808nm region and see if I can't fix the better of the two heads to be diode-pumped instead.

That reminds me... I need to put up the two pairs of Measurement Computing PCI-CTR05 and PCIM-DDA06/16 cards up on ebay too. Off to do that now.


Torin...