FlexMod P3 Replacement?

[ARG]

Ported the schematic to eagle, made an SMD version.

Little under 15$ in parts (single quantities)

Could be made more compact, with 0402's, an OP amp in a SOT-23-8 package, and generally better layout, so I'm going to work on it some more.

 

[AnachronicMan]

Hi,

I just read over this thread and was wondering how it was going?
Has anyone built a working prototype of this driver yet?

thanks,
AM

 

[KRNAZNBOY]

If ARG or Wolfman can send me the parts and board files, i can prototype a few

I am interested

 

[Wolfman29]

I'm really busy with school lately, which is why I have been so inactive. On the other hand, ARG may be interested. Unless you want to wait till summer, then I will probably try new driver designs as well as continue promoting and selling the BlitzBuck (which is still a thing, if you guys are interested!).

 

[Wolfman29]

Good We'll leave this to you then!

 

[COMMANDR]

Very cool project you got there. I would be in for a GB of 1 or 2 if this gets off the ground.

 

[LSRFAQ]

Let me make this real simple for you guys.

Your dealing with a delusion.

LostGallifrayan/TheDoctor lifted that current source design from a Ascii schematic post in an electronics chat group. He then trimmed out the negitive supply, and a current/sink source that used a second opamp and a TL431. As published, it will NEVER work. If you build it, it will flip through two current levels, with a huge spike, and very slowly, no where near 1 MHZ as claimed. I built one just to prove its a failure.

I've challenged him on this in the "other" forum, and btw he has never given the person who posted the full design any credit.

He sent it to Sam for the FAQ and Sam never really checked it. I've brought it to Sam's attention that it will never work correctly.

To get the 317 to correctly modulate, you need to both sink and source current through the ADJ pin using a complex feedback loop. Right now this design only sinks current. To both sink and source, you need split supplies.

So without about 3 times the parts count and a negative rail, the LM317 will never correctly analog modulate anywhere near zero current.

The origional design was a thought experiment done by a bunch of engineers, it was neither built nor tested.The principle engineer in that group posted a ASCII design. It works in SPICE modeling, but not in the real world.



Sorry,

Steve

 

[hakzaw1]

+rep to Steve

First off- I owe Steve and about 49 others an 'I'm sorry' for not including them along with Sam as authorS of the 'laser bible' - I have done a few edits of this error but should have known better.

I know nothing of value to add to this discussion except... 99.9% of the time Steve is correct- he has been at all this a very long time-- he was one of the two alllowed to get and test the new Pangolin ScannerMAX 506 (if you have not heard about these amazing scanners--- you will
- I have spent some time with Steve at several SELEM events - he is for real-

Len

 

[Meatball]

I suppose the op-amp and current sources inside an LM317 were not at all meant for fast feedback changes. One could instead look at using a discreetly built LM317, but then you'd have to deal with using matched transistors for the 3 current sources within. Otherwise, the voltage reference within will be most likely useless.

I'd rather see a circuit more closely modeled after the P3. I've tried reading the various packages on it, but the SMD parts don't give away their functions very easily. Without the added safety features, its a high side feedback driven PMOS device in the least - but with very fast amplifiers made with either a large negative feedback, or just fancier parts.

Let me make this real simple for you guys.

Your dealing with a delusion.

LostGallifrayan/TheDoctor lifted that current source design from a Ascii schematic post in an electronics chat group. He then trimmed out the negitive supply, and a current/sink source that used a second opamp and a TL431. As published, it will NEVER work. If you build it, it will flip through two current levels, with a huge spike, and very slowly, no where near 1 MHZ as claimed. I built one just to prove its a failure.

I've challenged him on this in the "other" forum, and btw he has never given the person who posted the full design any credit.

He sent it to Sam for the FAQ and Sam never really checked it. I've brought it to Sam's attention that it will never work correctly.

To get the 317 to correctly modulate, you need to both sink and source current through the ADJ pin using a complex feedback loop. Right now this design only sinks current. To both sink and source, you need split supplies.

So without about 3 times the parts count and a negative rail, the LM317 will never correctly analog modulate anywhere near zero current.

The origional design was a thought experiment done by a bunch of engineers, it was neither built nor tested.The principle engineer in that group posted a ASCII design. It works in SPICE modeling, but not in the real world.



Sorry,

Steve

Whose SPICE model would fail as such? How horrible.

 

[Buffo]

This thread appears to be dead at the moment (since the last post I see was back in May) but at the risk of being accused to be "far too late to the party", I feel I should throw my $.02 into this discussion, since it's pretty clear that there is some disinformation being spread.

Crow's driver design (which is based around the commonly-available LM317 voltage regulator) was planned to be an inexpensive, no-frills driver for the intermediate laser hobbyist who was at least experienced enough to both 1) assemble the driver on their own with nothing more than the schematic, the board layout, and a parts list, and 2) understand that the driver offered no extra protection for stupid mistakes like hot-plugging the laser diode or reversing the polarity of the power supply.

It is not a be-all, end-all driver. There is no such thing. Some people need a driver with crazy protections built-in for all sorts of unforeseen problems, including stray static electricity, reversed-polarity, and other general errors that can be loosely grouped into a category I like to call "operator stupidity".

Now, there's no shame in admitting that you are new to electronics and might otherwise blow up an expensive diode if you are fooling around with a driver that has no such protection. We all had to start somewhere, and to be perfectly frank, back in the early 2000's I killed quite a few red diodes (harvested from DVD burner sleds) before I learned how to work with a diode safely.

If you want a driver that has those extra features to protect your diode investment, the new driver that X-laser will be releasing in December should fit the bill, as it's based on a reference driver that Pangolin designed, and it includes LASORB protection right on the PCB. But it's also going to be much more expensive than Crow's design.

A middle ground might be the popular Flexmod driver that Andrew Kibler sells. It's got some diode protection (including protection against hot-plugging), and it even includes a rudimentary interlock feature that can be tied in with your main projector interlock for added safety. But you have to understand HOW to implement a projector interlock before this feature will do you any good. Still, if you have a need for this, the Flexmod is a driver you might want to look at. (And it's priced in the middle-ground between the X-laser driver and Crow's design.)

But for the reasonably experienced hobbyist (and especially for those with experience who are also on a tight budget), Crow's design is a solid driver that offers amazing performance even at modulation speeds that are 2 orders of magnitude faster than what is needed for a laser projector, all for a price that is very affordable. Yes, there are some current limits imposed by the LM317, but there is a high-current version of that IC available that goes up to 5 amps if you need the extra headroom. That should be enough to drive just about any visible laser diode a hobbyist would have access to.

To give you some idea as to just how solid (and affordable) the design is: back in January of this year the design was handed to a budding laser company in the northwest which was looking to start producing small, inexpensive laser projector components for sale to the hobbyist market. This driver was to be their third product, after projector cases and some custom-machined optical mounts. Sadly the company folded before this product made it to market, but there was real interest in selling it. And they would have been making a profit while selling them at $16 each!

Now, to address the claims that the driver design was "stolen", that is categorically false. First, the idea of using a voltage regulator as a constant current source is a solved problem that is taught in just about every beginner electronics class. True, Crow's design does incorporate a few interesting changes, and that's how it's able to perform at speeds up to 1Mhz without suffering from serious overshoot or ringing, but the basic concept is still well known among electrical engineers the world over. To suggest that he should "credit" someone with the idea is ridiculous. (Put another way, does anyone think it's reasonable to expect someone to credit Nikola Tesla when they use an A/C power supply for their device? I though not...)

Second, Crow has spent YEARS tweaking this driver. He has simulated it literally hundreds of times, and that was before I got involved with it in an attempt to add some extra features to it. (This ended up being a mistake, and we abandoned the project, but while we were working on it Crow continued to tinker with the design to tweak it's performance.) To insinuate that the final product was lifted from some other website is an outright lie that can be easily disproved simply by reading through the volume of e-mail Crow and I exchanged back and fourth from early 2013 through spring of 2014.

Now on to the issue of stability for this driver. The LM317, when properly biased, is a VERY STABLE regulator. That's why it's been a staple of basic electronic circuits for so long. It's a really good IC. So as long as the rest of the circuit is designed correctly, the device will perform quite well.

Several people have built copies of Crow's regulator on their own and tested them (going all the way back to 2012 at least, if not earlier), and they all agree that the performance he claims is 100% verifiable. If anyone believes otherwise, I challenge them to post oscilloscope trace pictures of the offending behavior along with a detailed explanation of their test setup. Then the exact conditions can be replicated by someone else. That's called science, folks, and there are no shortcuts.

Crow has really done his homework on this design. As I mentioned above, I actually worked with him on some modifications to the original design in an attempt to make a hybrid device that would incorporate some of the safety features that less-experienced users might want (or need). During that process, I saw lots of 'scope traces from other users, and I got very familiar with the design. I even went so far as to get the board laid-out using a different software format, in preparation for a short production run to test some new features, but we never actually ordered them.

The reason for this is that in the end we both agreed that adding those extra features really defeated the purpose of this driver. It's not going to be affordable anymore if you load the PCB down with tons of extra circuitry for TEC control, interlocks, and other bells and whistles. So it remains to this day a solid, no-frills driver that will work quite well, provided you stay within it's limits.

Try to push 3 amps through a standard LM317 and it will choke. But run it at 1 amp or less, and you can modulate it at video speeds without problems. Try that with a Flexmod! (Indeed, try it with just about any other driver... Even the X-laser driver can't reach those speeds!) But be mindful of the current limit, and for God's sake remember to heat-sink the LM317 properly! If you do, and if you take reasonable static precautions when connecting things, you should be fine.

And if you don't understand what it means to take reasonable static precautions, then this is probably not the driver for you.

Adam

PS: I should also point out that David Zurcher (DZ) and Sam Goldwasser (Yes, that Sam, of Laser FAQ fame) have both reviewed this design. They agree that there is nothing in the circuit that would cause the sorts of problems that people here seem to be predicting. I say to anyone who claims otherwise: "Show me your proof, and then let's discuss the matter." As I said above, that's how you do science. So let's do some.