Narrow-Bandwidth Television Association

Steve Anderson wrote:Is the circuit you finally used the one posted in this thread on 28th Nov. 2013? The circuit itself is dated 21st Dec. 2012...it doesn't seem possible that it's almost eight years ago! I thought maybe three or four years ago - not almost eight!

Yes, looks like it! I post my own copy of the circuit here. It's definitely analogue and as I said at the time, works very well.

As for the time passed, I agree - it's incredible. For an 'oldie' like me, the passage of time appears to be becoming logarithmic...

Steve O

OK, that looks exactly the same at first glance. The only changes required will/may be R107/R108 and a reduction of the LED/Laser supply to +12V with a soft start, quite easy using a LM317. 1A will be enough for all three channels.

Glad to see that I used a low value (1k, VR101) pot for the gain, that ensures a wide bandwidth.

Steve A.

If I remember correctly there were some changes required to the 'Black' control part. If you're certain that the circuit incorporates those changes, otherwise I would compare it with what you actually built. It's probably not a bad idea anyway, it's quite a simple circuit.

There may be a some other simple mods to the three driver circuits to set the 'threshold current'. Effectively the same as the 'black' control previously but adjusted to provide that current in each channel. It may mean slightly larger heatsinks on the BD139s, it depends on how hot they got with the LED arrangement. Do you recall? Do you remember the part No. and supplier for the heatsinks?

It may be prudent to try the drivers first with small value resistors as a load - thereby not risking the laser diodes until all is confirmed.

As it turns out in all three versions R108 is 22R, so only one to change required in the red channel. It also means allowing for their optical efficiency the light output power is very similar.

I would like to incorporate some current limiting...for obvious reasons. I'm still thinking on this...as well as the 'threshold current'...don't order, build or modify yet...as for the heatsinks, the existing ones should be adequate now I've crunched some numbers. But it would be useful to have the data for them just to check.

Steve A.

Initial go at modified driver(s). I'll have a go at breadboarding it in the next day or two...as ever I need to clear my workbench of other stuff...

Note the outer of the screened cable is connected to +12VB, make sure the screen ends cannot short to anything, including the chassis. I suggest using heat-shrink tubing...

TR105 performs the function of current limiting, a bit temperature sensitive, but over a 10-30 degree Celcius range it should be OK.

The black level/threshold current may need a mod too...watch this space...there'll be a test procedure to check all of this, all you'll need is a signal generator and a 'scope.

Steve A.

Fine business!

Watching developments eagerly.

Steve O

I forgot to mention, whatever heatsinking arrangements you have in mind, check the laser diodes cases are isolated from the internals, and have an adequate voltage rating, otherwise a change of plans may be in order. There also could be parasitic elements lurking...worth checking at this stage...if the case is plastic then that problem goes away, but somehow I doubt that's the case (pun, this time, is intended).

Steve A.

Looking at the datasheets they all have different arrangements for the case, be careful...

Yes, they seem to differ as follows:

Red: centre pin (2) cathode could be joined to case.

Green: centre pin (2) joined to case and junction of LD anode and "PD" cathode (what's that?) .

Blue: centre pin (2) joined to case, which is isolated.

Best isolate all the cases!

Steve O

For ready reference:

877-5241

161-1868

758-7810

Panrock wrote:...and "PD" cathode (what's that?).

PD=Photodiode. It's used as a feedback element into the driver circuit where absolute linearity, or optical limiting or stabilisation is required. Here they're linear enough without it, just leave it disconnected. The other two don't seem to have that 'feature' so it's not worth thinking about it here.

Panrock wrote:Best isolate all the cases!

Agreed, but without introducing too much thermal resistance between the case and heatsink. I suggest using three smaller heatinks rather than one larger one - just in case the worst happens. If you connect the heatsinks to +12VB, then if a short does happen the laser diode will simply stop working and the driver circuit won't care. But that means insulating the heatsink(s) from the chassis and ensuring shorts to other stuff can't happen.

Steve A.

Hmm, in the case of the red laser, the word 'could' worries me a bit...IF the cases are connected internally it would be better the anode rather than the cathode. The green and blue seem OK from what you've said. I'll have a review of the datasheets..

I think it's a case (pun again) of wait and see, build up the lasers, add the 1N914, and the screened cable then carefully try and determine which, if any are connected to the case. I wouldn't rely on anodising, if any, on the heatsink to provide insulation. It may be that one is totally isolated, the next connected to the anode and the last connected to the cathode - sod's law.

The simplest approach maybe using three smaller heatsinks insulated from each other and anything else. If the case (therefore the heatsink) is connected to anything or nothing - it doesn't matter. The 'slope' of the I/V curve is quite flat in the operating region so any additional stray capacitance this my introduce is unlikely to have any real effect.

I must get back to the driver circuit i.e. the 'threshold current'/black control...

The package is TO-56 for the red and green laser diodes, and TO-38 for the blue. I understand TO-56 is 5.6mm across and TO-38 is 3.8mm. Rather inconveniently, heatsinks for these sizes don't seem to come up on the RS site. (Though it could be the site... or me).

Another concern: I'm hoping that proprietary collimating/line generating lenses will fit over the top of these things. The lens has clearance for a 9mm OD object. Assuming this arrangement works optically (by no means certain), some sort of adapter ring will have to be made. But surely, there won't be the room for heatsinks? Oh dear...

Ready-made red, green and blue line laser modules with electrically accessible diodes don't seem to be available. We're not spoilt for choice as with LEDs.

Hopefully, your circuit will work with a wide range of laser diodes in this power class, just in case...

Steve O

Panrock wrote:The package is TO-56 for the red and green laser diodes, and TO-38 for the blue. I understand TO-56 is 5.6mm across and TO-38 is 3.8mm. Rather inconveniently, heatsinks for these sizes don't seem to come up on the RS site. (Though it could be the site... or me).

I didn't somehow expect them to. I was thinking of getting undrilled heatsinks and appropriate drill sizes and doing it myself/yourself. Then there's the optical arrangement to consider.

Panrock wrote:Hopefully, your circuit will work with a wide range of laser diodes in this power class, just in case...

It should do, up to the dissipation limit of the BD139s which does depend on their heatsink parameters. They're good up to 60V and 1.5A, but not at the same time! Depending on whose datasheet you read, it's either 8W or 12.5W, and that's probably on an oversized heatsink.

I plan to run all three lasers at a maximum current of 150mA. The green and blue lasers run at around 6V, =0.9W, the red is somewhat less at 2.8V, =0.42W.

Steve A.

As far as I remember rightly from my collegues at the CD/DVD lab, a laserdiode at low curents behaves just like an LED. Then when you pass the threshold current, it starts lasering. As an LED the colour may differ from the monochromatic colour when lasering. And the light origines from a different point in the device.

That is the reason that I would choose for a PWM driving rather than an analogue driving. However, it is worthfull to experiment with it. May be it is good enough. And there is no reason to make it better than good enough.

Yes, the picture will originate from light scattered by a ground glass screen, so it is unlikely we will be plumbing the depths of an accurately tracked black level.

Personally, given all the optical coupling unknowns here, I'll be relieved just to get a sharply focused line.

Interestingly (as Karen Orton has pointed out), in this application only the central axis of the mirror screw is used. So a much slimmer, lighter weight column of short slats would have been sufficient. However it would be fiendishly hard to align the slats on such a device.

Steve O

Panrock wrote:...However it would be fiendishly hard to align the slats on such a device. Steve O

Is it possible to CNC a screw out of a single lump of aluminium? I'm not sure if it would be possible to get the mirror finish required from these machines, and polishing it manually would be seemingly impossible. It would also be expensive. One little error in calculations or programming and you end up with an expensive, useless lump of metal...but is a thought...

Steve A.

Here's a couple of items regarding laser diodes, they're a bit 'doom and gloom' in places, but I wouldn't get too paranoid about it..

https://www.laserdiodecontrol.com/laser ... ndamentals

Steve A.

To quote from that paper: "It’s been said that there are two types of researchers - those who have blown a laser diode and those who will."

I've blown several, so I must be an example of the third type!

In a way their sheer expense is an advantage - it encourages one to take extra care. Eye care is another issue, though I'm hopeful their native diverging beams - and their living behind a line lens - will provide some protection.

When I was experimenting with high intensity infra-red LEDs for the optical broadcasting project, I went paranoid, invested in special goggles, and had a 'workshop safety protocol' posted on the wall. I've since learned that a wavelength of 2 microns would have been the best combination for eye safety with reasonable mist penetration.

Steve O