Modifying Q-switch driver for external modulation

[prodjallen]

In the manual you posted it shows 2 modes - High RF mode which is pulsed and Low RF mode which is CW.
In my 'lil green uses the 1027-25 and the Low RF mode had been factory adjusted to have approx. twice the power out as the pulsed mode making High RF operating with Q-switching and Low RF mode being blocked. Going from Low RF mode to High RF mode will give you a big first pulse.

Rgds,
Robin

I had this issue when I was working on the electronics of my Q-Switch Driver (1027-25), and that only happened when the wires were backwards on the J801. Check the polarity on pins 4,7 should be hot, and pin 8 should be neg/ground. So when there is power on pins 4,7 it should enable pulse-mode, when no power, should be CW...

[uk-laser]

That's clever, because I hadn't considered turning low RF up to full RF. It means that any FPS issues aside, the low/high RF input could be used for blanking!

I believe at full RF the driver overheats after a minute or so.
Extended blanking would not be recommended.

This trace shows the first pulse after the RF low/high transition. It's bad news, but I can't see any evidence of FPS. So occasional blanking (e.g. at the start of a show) is probably OK, but I wouldn't advise using this driver for high repetition rate blanking (e.g. blanking inside frames).

Isn't it the other edge you should be inspecting - RF collapsing after being at high power.
I'm sure there is no FPS though. If the RF could be persuaded to decay more slowly then perhaps FPS could be achieved. Since there is means of adjusting RF power with the POTs it must be possible to hack the driver to decay the RF rather than abruptly stopping.
Verification of correct FPS operation may be tricky. On cutter / welder q-switched lasers a weld line is scanned which puts individual pulses side by side. Correct setting is when the pulse train spots look identical. Where the train starts with a hole in your workpiece FPS was not set correctly.

When I get a moment I'll dig out the old Baird dummy load and have a play with the driver.

[weartronics]

I believe at full RF the driver overheats after a minute or so.
Extended blanking would not be recommended.

Yeah, Steve warned me about this too, and the RF output stage got pretty hot. A bit cheeky they don't mention this anywhere in the manual though!

Isn't it the other edge you should be inspecting - RF collapsing after being at high power.

Oh yeah... well that's the last time I keep working past midnight!

I'm sure there is no FPS though. If the RF could be persuaded to decay more slowly then perhaps FPS could be achieved. Since there is means of adjusting RF power with the POTs it must be possible to hack the driver to decay the RF rather than abruptly stopping.

t's true, you could muck around with the signal generator, keeping the RF output stage the same. But if the RF output stage can't handle 25W continuous anyway, might as well find a surplus driver with better power handling and FPS built-in.

Verification of correct FPS operation may be tricky. On cutter / welder q-switched lasers a weld line is scanned which puts individual pulses side by side. Correct setting is when the pulse train spots look identical. Where the train starts with a hole in your workpiece FPS was not set correctly.

I think a photodiode on the output could measure the pulse power accurately enough. In medical lasers, there is almost certainly a photodiode with appropriate beamsplitter already installed on the optical deck.

[Laserman532]

FPS can not be measured by photodiode. Procedure is viewing the RF through the proper splitter.

[weartronics]

FPS can not be measured by photodiode. Procedure is viewing the RF through the proper splitter.

You're right, correct operation of the FPS is verified by measuring the RF (that's what I did, looking at the wrong end of the pulse). I think what Robin's talking about is choosing the appropriate FPS decay time. We have to choose the FPS decay time long enough to prevent giant pulses but short enough to allow fast modulation. The necessary FPS decay time is a function of fluorescence lifetime, resonator design, pump power etc... I think you could probably calculate it, or just measure the effect with a photodiode.

[uk-laser]

You're right, correct operation of the FPS is verified by measuring the RF .

To be pedantic measuring the RF just confirms correct operation of the RF amplifier

It is not really first pulse suppression but first pulse(s) suppression.
The trick is to match the decay time of the RF to the remaining energy in the rod.
Get it right and the first pulse, tenth pulse and thousandth are all same amplitude give or take. Set wrong one way you get a bigger first pulse. Set wrong the other way and your first pulse(s) are too small.
For most applications other than micro-machining the exact setting is not critical.
As long as it exists in some form it will save your NLO.

A good introduction is given here:
http://www.sintecoptronics.com/qswit...PSguidance.pdf

Sintec show a linear decay on the RF however I have seen other papers which show an exponential decay. I guess the ideal setting may be somewhere between the two but that is probably overcomplicating things. (and without serious test kit or a reel of ticker-tape passing through the beam at focus fine tuning would be impossible)

Rgds,
Robin

[DMills]

The manual for that driver mentions needing a fairly significant air flow across the heatsink, which may explain some of the heat issues.

I suspect (but have not drawn the diagram) that the RF stage is actually modulated from low to high by varying the power supply voltage using that big TO247 looking part in series with the collector supply to the RF transistors (This may well be where most of the heat originates).

The subtle bit is that it is actually the depth of the modulation that needs to be controlled, switching from RF low to RF high is not a problem, it is the switching from RF CW to RF Pulsed that hurts (and potentially the switch back from RF high to low if the gain is sufficiently high).

I am thinking linear amplifier with a DBM on the input (SBL1?) or pin diode, but brewing up a new design is probably easier then modifying an existing one (The bulk power setting could be done by modulating the supply rail which leaves the RF stage in class C so little heat production).

27Mhz is a nice frequency (lots of CB radio transistors!).

Regards, Dan.

[DMills]

Humm, did you know that a pair of IRF510 will get you north of 30W at 27Mhz for a watt of drive?

I am thinking that something with say three of them (single ended driver, push pull final) and do the modulation by applying negative bias to the gate of the driver transistor should get something that would work, and it would be cheap cheap cheap.

This needs thinking about.

Regards, Dan.

[Laserman532]

To be pedantic measuring the RF just confirms correct operation of the RF amplifier

It is not really first pulse suppression but first pulse(s) suppression.
The trick is to match the decay time of the RF to the remaining energy in the rod.
Get it right and the first pulse, tenth pulse and thousandth are all same amplitude give or take. Set wrong one way you get a bigger first pulse. Set wrong the other way and your first pulse(s) are too small.
For most applications other than micro-machining the exact setting is not critical.
As long as it exists in some form it will save your NLO.

A good introduction is given here:
http://www.sintecoptronics.com/qswit...PSguidance.pdf

Sintec show a linear decay on the RF however I have seen other papers which show an exponential decay. I guess the ideal setting may be somewhere between the two but that is probably overcomplicating things. (and without serious test kit or a reel of ticker-tape passing through the beam at focus fine tuning would be impossible)

Rgds,
Robin

with link provided, now everyone has the ability to keep their components safe and continue to perform illegal unsafe shows (especially with a few here in the us) they follow this thread. we were hoping for component failure. oh well. I guess with the internets nothing can stay sacred.

[DMills]

To be fair, one you understand how an AO Q switch works, and where the giant pulse(s) originate, it is obvious what must be done to suppress them.

Nothing in that PDF says anything that someone who knows enough to use the knowledge cannot trivially work out for themselves.

Regards, Dan.