Switching vs.Linear PSU

[planters]

If planing to drive several 445 diodes for a projector with a single bench type PSU such as those from BK, are there any fundamental problems with the switch mode supplies? Understood that they have greater voltage ripple ie. 40 mv, but this would seem insignificant when driving say 3 diodes in series at aprox.13 V. Do they pose a greater risk from esd or transient voltage spikes due to their design? Any other thoughts? They are cheaper.

[-bart-]

What driver will you be using? Good drivers don't really care what supply you use.

Crappy drivers however...

[planters]

Because the power supply will be driving several diodes in series and several series in parallel I had intended to avoid the added cost and complexity of 12 individual drivers and rely on one quality current and voltage limiting PS. I do not intend to use any drivers other than the possible insertion of a transistor switch for blanking. This switch will be controlled by the 5V signal from Q. Show. The plan will be either to simply interupt the current or to shunt the current to a dummy load (if the power supply voltage does not remain stable in the short term to the rapid switching).

[DMills]

One current source for a series string is fine (as long as it has sufficient compliance range), but I would be really leery of using a bench power supply as this current source, particularly if switching the output.

Consider that most bench power supplies are fundamentally voltage regulated devices with current limiting bolted on basically as a safety measure. If you set the voltage to say 15V and the current to say 1.2A it might all be fine sitting in current limit at 13V, 1.2A until you disconnect the output whereupon the output caps will charge to 15V.... Reconnect the output and you get 3A or so down the chain until the caps discharge, not healthy.

Seriously building a 1.2A current source (or even easier sink!), with a suitable compliance range is trivial, run it from a suitable supply and it will just work, but I would never trust a generic bench supply on its own for laser diode work (Unless it had Agilent/HP/Rode & Schwarts or similar printed on the front panel, and even then I would have to think about it).

Regards, Dan.

[planters]

Thanks for the response. At the risk of holding to an idea too firmly, it seems as if the concern re. the bench power supply could be avoided if it was used only as a voltage limiting source and the inherent resistance of the diode array limited current. As long as each diode was run conservatively, the variations in current flow between each series of diodes would be acceptable. Alternatively, if a driver was constructed I assume it would be for each series in a parallel array? Would it be trivial to construct if current output needed to be modulated or conveniently adjusted? What is the advantage if any of limiting a diode's drive current as opposed to limmiting it's drive voltage and allowing it to self limit its current? It is interesting that throughout this forum and even at LPF that so many projects use small or multiple individual drivers even for larger assemblies rather than what would seem to be a simpler solution ie. a good quality single supply. It makes me think.

[DMills]

Basically consistency.

A laser diode is NOT a resistor and they can have quite small incremental resistances (which means that once biased into conduction small voltage variations can cause huge current changes).

Draw a graph of voltage vs current for a diode some time and you will see what I mean.
Further this curve depends on temperature and varies diode to diode so you really do want some current feedback to stabilise the operating point.

For the blue diodes I don't favour separate drivers per diode necessarily, but a series string of 4 surely warrants a driver from a 24V rail.

If you really must do the bench supply thing, then at least put a non inductive power resistor in series with the string (sized to drop a few volts or so), this will reduce the slope of the V/I curve somewhat.

But a real driver is not that complex and will be better behaved.

Regards, Dan.

[planters]

Thanks, I think I'll look into a string of flexmod 3s. They are rated for up to 25W continuous and if that is reasonable then running 4 diodes in a series at aprox. 1.2 A. should work as well as provide the modulation circuit I was planning anyway. Thanks again