Power Supply for a 15W CO2 Laser?

phenol said:

i dont think that the rectifier potted in the output tranformer of those psu's is too much different in terms of topology from a tripler

Click to expand...

I must be misunderstanding something, because a diode is quite different from a tripler last time I checked. Also, rated voltages for diodes range by several orders of magnitude.

So you believe that they obtain the 30-40kV required to start the tube directly from the secondary followed by some sort of a rectifier without any multiplication?
either way, the tripler in argument had 10kV diodes inside and was in fact a multiplier by 5 provided the input voltage has equal positive and negative halfwaves, which in fact it did. This, of course, doesnt mean that it actually multiplied the input voltage by 5 when loaded with 8mA or so; it is rated for 1.3mA max at 27kv when used as a tripler (with positively going input pulses). anyway, it is dead and resting in peace in the trash. I use a real PSU and it seems fine apart from the fact that it is not regulated.

phenol said:

So you believe that they obtain the 30-40kV required to start the tube directly from the secondary followed by some sort of a rectifier without any multiplication?

Click to expand...

Good question. I don't know. I'm not sure a ~20mA multiplier would FIT in the flyback's case though. Try putting some DC through the secondary. If current flows, there's no multiplier.

I am the new owner of this tube and power supply.

A bit of info:

1) the P and L terminals need to be jumpered together and then collectively switched to the G (ground) terminal for On/Off operation. If the P terminal is permanently connected to G the psu will output several kV at 2.9mA regardless of if the L terminal is brought High or Low.

2) this SMPS PSU is a basic analog PWM DC-Chopper flyback topology so it will readily accept 120V mains input even though it is made for 240V. The power supply's logic is unaffected (as it all runs on regulated low voltage), and the output current and voltage are reduced by roughly half. This is perfect for a small 15W tube since it is normally a 35W psu.

It should be noted that the main capacitors on the smps board are rated for ONLY 200V. They may be in series, but I would have to completely remove the board from the housing to see the traces to determine if this is the case. Even if they are in series that is 400V maximum rating on a board which could see 361Vpk, not a lot of engineering headroom.

so, did you get any lasing out of that setup?

Yup. I'll post it for him:

Thanks, there's that buzzing again, but i guess it's the pump. What Sigurthr said about it delivering some idle voltage with the Pump terminal permanently grounded is also true as far as my 60-80W psu is concerned. Any idea why it biases the tube? So it fires faster with the capacitors in the multiplier (if any) precharged maybe?...

Thanks cyparagon!

Lol, is that the buzzing you were commenting on earlier? That IS the little 9V pump. The sube and psu are silent.

The reason there is some idle voltage is because you have the connectors wired wrong! There does not appear to be any multiplier circuit (at least not several stages anyway, it may be a single diode-cap inside though) in this PSU. Even so, you can't precharge a CW multiplier effectively as the current available at the end of the cascade is miniscule and simple leakage would deplete it. Also, it is incredibly hard to design a multiplier capable of delivering several mA without ripple.

Here is the correct wiring scheme:



Here is the cathode current with the PSU at full power on 120Vac input. Output is about 12.8mARMS

i followed the wiring instructions in the manual that came with the psu and the pump terminal is described as a safety interlock independent of the other terminals; when left floating, laser is disabled at all times, when grounded, laser is controlled by one of the two inputs. However, when permanently hooked to gnd, the LED on the psu glows dimly even when ttl inputs are disabled. the led is connected to a loop of wire weakly coupled onto the core of one of the two hv transformers.

Right, that all makes sense, what doesn't is why your manual wanted you to permanently connect the pump interlock to ground (that is what you meant, right?). The way MrDelsic had it wired up this was this case: the P (pump interlock) was hard wired in to G (ground) and only the TTL input was switched. Wired that way there was always 2.9mA @ a few kV present on the output since the interlock was always defeated. By tying the pump interlock in with the TTL low you can switch the unit entirely on or entirely off with a single SPST switch. You could also use a DPST if you wished to retain isolation between the two circuits (for example, using TTL high switched to 5V).

The reason why the manual advised that is to "test the laser" just by using active H or L ttl signals. It makes sense that the water protect input is controlled independently by some external circuitry sensing water flow or whatever. The water pump doesnt have to be pulsed in sync with the laser, so WP input will stay low as long as the cooling loop is operational...
anyway, i tested your wiring and it works as you say- the output is fully disabled.

Ok, so I asked the CO2 tube and psu maker about that and here is what they replied (copy-pasted):

"Thank you for your mail.
I asked our engineer about your question.


In fact, the "some output voltage on the HV wire" is called (by us) as "Pre-current".
For a laser machine, the main board of control system need to send signals to power supply, then the power supplu will give the laser tube voltage and current so as to make the laser tube working.
If without the pre-current, the response time of the voltage from 0 V to 20KV or 30KV is too long. If there are some pre-current, the response time of voltage will be greatly shorten.
That is designed by laser machine manufacturer. "

Interesting, they consider it a type of di/dt and dv/dt reduction. I suppose it makes for a pulsed or cnc application, but it only makes it more dangerous for us hobbyists who don't use the tubes industrially. Good to know, though!

BTW Sigurthr, I hate brushmotor driven water pumps so I went and bought a little brushless DC water pump for my laser engraver. They're super quiet and have very good pressure ratings which is good for pumping water through the relatively small tubing they usually use for CO2 lasers.

Something like this would probably be good for your tube: Mini DC 12V 3M 4L MIN Mini DC Brushless Motor Submersible Water Oil Acid Pump | eBay

here is what i have, a fish tank pump:
Bg-Pet.com - ïîìïè Blu Power 900, BluPower 1200 Ferplast

Most small fountain and pond pumps are similar in design where there is no motor shaft seal so they are a good choice as well, but they usually don't get you as much head pressure.

My laser cutter came with a fairly large mains powered aquarium pump but the little brushless one I bought pushes more water through the restrictive tubing