USB-->RS232--> 2 channel, 12bit +-10V analog ? (Laser Projector)
FlyingFishFinger
Posts: 461
Hi-
I posted about something like this before, but I was in an extreme time crunch back then ( we didn't make the deadline).
We now have alot more time to work on version 2 of our custom laser projector interface and I've got time to build something proper.
Problem is, I'm not really sure where to start. I think it'd be the easiest to use one of the available USB-> serial converters for the first step, but the voltage conversion has me baffled, as does what kind of DAC chip to use.
Anyone have experience with custom DACs?
Thanks
Rafael
PS: I forgot to mention that this thing needs to be pretty fast...about 10Khz ><
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You've got to play the game.
You can't win.
You can't break even, except on a very cold day.
It doesn't get that cold.
~Laws of Thermodynamics~
Post Edited (FlyingFishFinger) : 3/31/2009 8:33:42 PM GMT
I posted about something like this before, but I was in an extreme time crunch back then ( we didn't make the deadline).
We now have alot more time to work on version 2 of our custom laser projector interface and I've got time to build something proper.
Problem is, I'm not really sure where to start. I think it'd be the easiest to use one of the available USB-> serial converters for the first step, but the voltage conversion has me baffled, as does what kind of DAC chip to use.
Anyone have experience with custom DACs?
Thanks
Rafael
PS: I forgot to mention that this thing needs to be pretty fast...about 10Khz ><
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You've got to play the game.
You can't win.
You can't break even, except on a very cold day.
It doesn't get that cold.
~Laws of Thermodynamics~
Post Edited (FlyingFishFinger) : 3/31/2009 8:33:42 PM GMT
Comments
For instance, even knowing whether you mean usb > serial or serial > usb could make a great difference in the answer.
Cheers,
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tom Sisk
http://www.siskconsult.com
·
Don't you just love deadlines?
The last time you posted... you reminded me of a project that I started and then abandoned 20 years ago... involving a LaserMedia Stingray.
I went down to the basement of my old office and there it was still sitting on the shelf. I plugged it in, turned it on, and it promptly blew a power fuse. So I cracked it open, and inside are 5 boards; three of them apparently involve PC communications and higher level control functions. Two of the boards control 3 galvanometers. The unit has an open head galvanometer architecture. So, eacj axis has its own control lines, which are completely available... and I think I found the correct documentation.
There is a single axis board controlling the X axis (g120DT galvanometer)... and the other board controls the y axis (g120DT) and z axis, which has an unknown galvanometer... only because the label is hidden. I have no documentation on the boards themselves only about the galvanometers.
The galvanometers are from General Scanning Inc... a company that has since been absorbed but is still reachable. The G120DT galvanometers are of the closed loop variety. The control boards for the galvanometers output TTL signals and in looking at them, I have to believe that they could both be replaced by a single Propeller board, with a very simple circuit.
In reviewing the literature, including the MEMS approach to building tiny galvanometers, I am beginning to wonder if in fact the entire architecture is a mistake. The limiting features are completely governed by the laws of physics. The physical reality is that it is fairly tough to start and stop a mirror very precisely and very quickly and also know exactly where it is at all times. If you do it too fast... the mirror breaks. If you do it too slow, you end up with a reduced display capacity, and everything is affected by temperature and component wear.
On the other hand... if you accelerate mirrors in a single direction and do it carefully... you can get mirrors to rotate at incredible speeds, without breaking. The single requirement to using this fact in display applications would be in knowing exactly where the mirror is... an encoding and control problem, which has multiple solutions. Multi-sided (polygonal mirrors) already exist. So, for example a hexagonal mirror can be(and often is) used to the same effect as rotating a single mirror six times faster.
My thought was that if a person really wanted to reduce cost and control requirements, he could use multiple mirrors(and motors) for each axis... with the beam being rotated in a single axis by consecutive moving mirrors. Effectively multiplexing the problem and removing the most serious design constraints. You could pick a mirror design and just about any motor and design around it.
Just a thought... but if it is correct, your bill of materials could get much shorter and a lot cheaper... and in theory you could use servo's, steppers or ordinary dc motors to tackle the problem.
One commercial limitation is that it might be difficult to underwrite this technical approach because it is probably not patentable. But if you have an application that simply depends upon the technology and really doesn't require the technology to be otherwise profitable... then you might want to consider this approach.
Regards,
Rich
Post Edited (rjo_) : 4/1/2009 6:19:44 PM GMT
Perhaps I wasn't exactly clear in my description.
We have an off-the-shelf projector unit that takes in ILDA Standard signals (+-10V on each channel x, y).
@rjo_:
Your approach seems to be geared towards building a custom projection unit, if I understand correctly. That's not quite what we're looking for, since the unit is already geared to take in the aforementioned analog voltages.
The guy who hooked us up with the unit uses a really old laptop with a modified one of these things in a dock:
http://www.advantech.com.tw/products/30-kS-s-12-bit-16-ch-SE-Input-ISA-Multifunction-Card/mod_1-2MLI5J.aspx.
That's a full DAQ system that seems to be meant for desktops.
What we'd like is a small box that will plug into USB and give us two channels of +-10V 12bit analog output.
Our programmer works in C and Java, so whatever we use needs to be addressable or have libraries for those languages.
I was thinking that if we're building the thing, we should use a USB-->RS232 converter, since that should be easliy programmable, then use some DAC chip to actually do the conversion. I'm not sure if that's a good approach, but whatever works...
I hope that's a little clearer [noparse]:)[/noparse]
Rafael
PS: If we wanted to purchase something off the shelf, we might use something like this:
http://www.accesio.com/go.cgi?p=../usb/usb-da12-8e.html
But that's $350something,...I'm wondering if there's a cheaper, DIY solution.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You've got to play the game.
You can't win.
You can't break even, except on a very cold day.
It doesn't get that cold.
~Laws of Thermodynamics~
Post Edited (FlyingFishFinger) : 4/1/2009 6:53:05 PM GMT
Sorry for being off topic. You were actually very clear[noparse]:)[/noparse]
Thanks for the link.
It sounds like this might turn into a commercial product. I would certainly want one.
This is also OT... but if you turn a projection system like this around and change the logic a little... it can be used as part of a variety of different medical imaging systems, which is a particular interest of mine. Part of the problem is cost. By the time the systems get produced, they have a defined imaging logic and a very high cost. So, there is a huge need, but in many areas, there is simply no way to actually satisfy that need.
Regards,
Rich
I hadn't thought it's useful enough to turn into a product...all we really need is a controller for an already present projector [noparse]:)[/noparse]
Rafael
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You've got to play the game.
You can't win.
You can't break even, except on a very cold day.
It doesn't get that cold.
~Laws of Thermodynamics~
not 10 V OUT but with a power op-amp this should be easy ..
Peter ..
PS I made my galvo scanner this way , and wrote custum audio files to run my scan,...