Noob question on scanners.

[The_Doctor]

So you never even tested a completed design? Sorry, I thought you had a proven set up. You got me all excited for a minute, LOL.

Can you point me in the direction of your post? I can run it by my friend. He has connections with a Chinese PCB maker. Maybe he can make up a prototype for testing.

I did. Just two though. I sent one to Sam Goldwasser. He checked it out on a dummy load. His report is very brief, one line added to the entry in the FAQ. The design is known good, it just hasn't been made in bulk by anyone yet so far as I know. The main thing is not to re-lay the board differently, or use other amps, and expect same performance. If you're lucky with some amp I never found, it might even get better, but most likely not. If a board with good RF performace gets laid out like a basic audio board, it will probably be limited to acting like one. But the layout in the FAQ will work ok. But it's not a Gerber, just a bitmap. Great for accurate DIY, probably not much use to a board maker firm, which is why I'm working on a revised version with Gerber files.

The original is here:
http://www.repairfaq.org/sam/laserdps.htm#dpsldd317
It's worth waiting for the revised one though, it has been adapted for a 10K input resistance because apparently that's a standard people expect in a projector system. There is also a revised layout that allows ten parallel 0.6W metal film resistors set the current. That keeps the board flat to save space, and keeps the inductance of the resistor low to prevent nasty overshoots. The design prevents those too, because the first amp stage overshoots to the zero rail, and errors are partly reversed by the second stage, almost cancelling, which is one reason it is so accurate and fast.

If you can't get the LT op-amp, use a cheaper CA3240E instead, you'll still get very good 200KHz performance, and it will be extremely cheap. Getting the board is the only tough bit right now, because so far it's all DIY.

EDIT:
You can see the built driver I have in this thread:
http://www.photonlexicon.com/forums/...-Aspheric-Lens
It's not being modulated, it's just CW in that one, driving a G71 diode. But you can see it works, doesn't kill the diode. I have no trouble casually connecting to a 12V battery or ATX PSU, or anything else convenient. Point being, it is highly resistant to power interruptions, it will not cause spikes that kill diodes.

[TechJunkie]

I'll wait for the revision. No need to use up resources till you're ready. Give me a nudge when you are ready, and I'll get the ball rolling. Even if we dont use my guy I can get you a quote so you can use it to compare pricing. Either way is fine with me.

Please note I have 0 knowledge of circuitry. I can build just about anything. I've always been a hands on man, but never took the time to learn electronics. I think I'm to old now to absorb it now.

[The_Doctor]

You'll get by if you follow that layout and text. Could be worth it as a one-off just to try. I'll stick with Buffo's help on this though, having gone as far as we have. Thanks though. If you can get a quote, plus any certainty that they accept Gerber RS174X files, that will be useful info. I already got a couple of recommendations but I can always use more.

The main sticking point is that most firms seem to offer small boards, but if I could get a larger panel, like 300x500mm, then I could get a lot more boards out of it and save space and money. Failing that I'll have to re-lay the board to optimise for some firms chosen board size. I'm trying to avoid that right now because it would conflict with the shape I'll want to fit the cases I'm buying..

[TechJunkie]

I sent out an email, so I'll wait for the reply and report back. Its not uncommon for a week to go by before he answers me back.

[mophead]

Thanks Mop, but I think 20s maybe a little to slow for what I want.

That's cool. They're not something i have or need to get rid of right now.

[JasonB]

I hope posting this does not cause problems for anyone.
Here is pricing I just received for some scansets, notice shipping is 48USD, my reseller shipped my scanner in a DHL box so I assume thats how the manufacturer shipped them to him. Also attached is a test frame from my 20KPPS (Which I paid over two hundered dollars for with very little value added other stuff, thats why I'm sharing this information).

Interestingly enough I emailed "sales@phenixtechnology.com" and received a response from eightonlight, shown below.
Hope this helps someone get good pricing on what I think is a good product for the money.

Below is the quotation for the scanner sets you requested for:
Quantity (1) 20KPPS scanner - 70USD
Quantity (1) 30KPPS scanner - 170USD
Quantity (1) 50KPPS scanner - 250USD

Shipping cost for 1 unit to USA is 48USD.

If any question, please feel free to contact me.
Thanks and regards,
Ella

Lighting Space Limited
P: + 86 750 331 5551
M: + 86 134 2279 8669
Skype: ellawin927
www.eightonlight.cn
america@eightonlight.cn

Sender: sales@phenixtechnology.com
Send Time: 2013-12-07 16:09
Receiver: america
Subject: 转发：Request for pricing

[mixedgas]

[QUOTE=The_Doctor;278455]I did. Just two though. I sent one to Sam Goldwasser. He checked it out on a dummy load. His report is very brief, one line added to the entry in the FAQ. The design is known good, it just hasn't been made in bulk by anyone yet so far as I know. The main thing is not to re-lay the board differently, or use other amps, and expect same performance. If you're lucky with some amp I never found, it might even get better, but most likely not. If a board with good RF performace gets laid out like a basic audio board, it will probably be limited to acting like one. But the layout in the FAQ will work ok. But it's not a Gerber, just a bitmap. Great for accurate DIY, probably not much use to a board maker firm, which is why I'm working on a revised version with Gerber files.

I built up two of them a few months ago. As built, from the FAQ schematic it has a error. The specified LT op-amp would not pull the LM317 feedback pin down low enough to go below 80 mA, and would not drive it above 220 mA. If you go back to Winfield Hill's LM317 controlled current source thread on S.E.D, you find you need to sink current out of the LM317 to pull it all the way down to zero. Win's way of doing that requires a negative supply for the opamp.

Sorry,

Steve

[TechJunkie]

Interestingly enough I emailed "sales@phenixtechnology.com" and received a response from eightonlight, shown below.
Hope this helps someone get good pricing on what I think is a good product for the money.

Below is the quotation for the scanner sets you requested for:
Quantity (1) 20KPPS scanner - 70USD
Quantity (1) 30KPPS scanner - 170USD
Quantity (1) 50KPPS scanner - 250USD

Shipping cost for 1 unit to USA is 48USD.

If any question, please feel free to contact me.
Thanks and regards,
Ella

Lighting Space Limited
P: + 86 750 331 5551
M: + 86 134 2279 8669
Skype: ellawin927
www.eightonlight.cn
america@eightonlight.cn

Sender: sales@phenixtechnology.com
Send Time: 2013-12-07 16:09
Receiver: america
Subject: 转发：Request for pricing

I just got the same email, =0(. Looks like the eBay price was not that far off, and included that card thing. Not sure what that is used for. If Winni cant give me a better price it may delay my project a month or more.

[TechJunkie]

Does anyone know the web page where I can plug in different color, and power values to calculate a white beam?

[The_Doctor]

I built up two of them a few months ago. As built, from the FAQ schematic it has a error. The specified LT op-amp would not pull the LM317 feedback pin down low enough to go below 80 mA, and would not drive it above 220 mA. If you go back to Winfield Hill's LM317 controlled current source thread on S.E.D, you find you need to sink current out of the LM317 to pull it all the way down to zero. Win's way of doing that requires a negative supply for the opamp.

Sorry,

Steve

It works though. I think Sam would have told me if it didn't. The upper current limit is set by a fixed resistance that needs changing by R=1.25/I where I is wanted current in amps, and R is that resistance. (Just spelling that out because not everyone seeing this might know that). My current test rig is using a G71 diode fed by the driver in CW operation, with resistance set to 39 and 4.7 ohms in parallel. 1.25 over that lot gets 298mA, and a Fluke 79 meter reads 0.296V across a series 1R sense resistor in line with the LD, put there for this purpose. It's not a lot higher than your 220mA, but it's what I asked it to do and it's within the expected 1% for the metal film resistors used. I've used a modified (slower, simpler) form of this circuit for LED drive at over an amp (and a few amps in a test with an LM338), and I can do fast strobes and such. This thing works, I promise.

About the negative voltage, it still works with no negative rail because the slight relative negative with respect to the laser diode's VF is adequate to put the total current below threshold for every laser diode I ever drove with this. You're right that it won't go to zero, but in this application it doesn't have to. It's actually better that it doesn't go that far, because the diode can be driven harder and faster if it isn't forced to 'cold start' every time it's blanked. I remember being told that setting just below threshold, for minimum drive current, was correct. If you set the resistor for a smaller maximum current, you also get a smaller minimum! I discovered that by accident, but it is totally convenient behaviour for a laser diode.

(The lack of total minimum IS a bit annoying in my dimmable DIY LED lighting, as it happens, but not enough to drive me to cure it other than to pull the plug if I really want dark).

EDIT:
Some values for min current at known resistance for hardwired max current:
8R2 for 152mA gets min of 34mA.
4R195 for 296mA gets min of 60mA.
Minimum expected threshold for G71 is 80mA, so this is ok. A sensitive diode strongly overdriven might not dim significantly below threshold, but in a projector fit for professional use it should not be strongly overdriven anyway. For the typical threshold of 130mA I suspect my circuit will still be fine even if it is (though I'm not sure yet what effect deep cooling will have on threshold currrent).