Heat Generation for Reci tubes & Peltier Performance

[medicdude]

Does anyone have any data on the heat produced by the reci tubes at full power (or near full power)? I live in New england and I'm going to build a garage laser, so I figure I'll need a chiller of some sort for hot summer (~100F worst case external temp) operation but I have no idea what thermal capacity chiller I need. I was considering using some peltier devices instead of a chiller.

[medicdude]

Going off some basic math, an 80W tube consumes 25mA (100%) x 18kV = 450 Watts of electrical power, producing 80W of laser power, leaves 370W of power lost to thermal conversion. Some of that is probably radiated from the tube, and your laser probably won't be operating continuously at 100%, so ballpark around 300W of thermal cooling required?

[Tech_Marco]

For Reci 80W (W2) tube, I suggest to use a 800W water cooler like the CW5000
When you calculate the heat dissipation, you don't want to use the min. value. The 28mA is the max current it should go and the voltage I'll take the average of triggering and the operation voltage which is about 546Watt or even number 550W. Well, that is the power to 'offset' the heat but we want to keep it below room temperature, that's why a 800W chiller is required. Better use colder cooling system than regret.


Marco

[medicdude]

I was looking at the peltier coolers, using 6 peltiers and 2 120x40mm aluminum waterblocks to transfer heat from the laser water to a secondary water system and a showerhead to transfer it to ambient, simple evaporative cooler. I'll be sensing water flows and water temps for auto-shutdown if things get out of spec. I want peltiers for cheap cost, low noise, precise control over temperature, and easy expansion by just adding additional modules in series.

[Tech_Marco]

It is not going to work well as the efficiency of a peltier is very low. I tried it many times with different design but all didn't come up good. The bigger one I built was 500W peltier with 2 radiator and two water blocks, 2x 8" fan, 2x 5" fan, and 3 water pumps. I could be able to cool a 100W tube from room temperature (37'C) to (34'C) with continuously working 30~40min and stablize at 34'C. I made up the room temperature to 37'C as warehouse by turned on my house heater and did the test in a closet. I had one extra portable heater to help to warm up the room. The system was noisey and cost too high to build it. Not worthed

See the picture that I built about a year ago. It is not the biggest one I built.


Marco

[Tech_Marco]

Found my giant peltier cooler!

[medicdude]

Thanks Marco!

Do you think the problem was with the cooling system not cold enough or with the peltiers themselves? Any idea if the newer 'high efficiency' 70-90W peltiers are any better? Maybe the problem is with the aluminum blocks themselves too thick or not enough surface area?

I was thinking of building a DIY evaporative cooler, similar to a natural draft counterflow cooling tower, a showerhead (actual or DIY) that sprays into a 5 gal bucket with air holes drilled in the side near the bottom. A few of these would easily cool your secondary system to wet-bulb temperature, making it a lot easier for the peltiers. Even a single one may provide plenty of cooling. A standard showerhead wouldn't be ideal, you want the smallest drops you can get (not a stream). A re-purposed strainer may be better, or even a piece of acrylic with a ton of laser holes drilled into it.

[medicdude]

If a custom showerplate is feasible, here's an improved design, hole in the center, outer edge of the 'showerplate' would be sealed to the 5gal bucket.
Another option blowing compressed air in the bottom, with the benefit of the drier air that usually comes out of a compressor, at the cost of noise and the compressor, but you wouldn't need an exhaust fan.

You'd want to make the showerplate out of some non-polar material, so water doesn't get stuck in really small showerholes. I don't remember if acrylic is nonpolar. Thin steel or carbon steel would probably be easiest for lasercutting.

[medicdude]

Tetech.com is an american company and they offer a nice range of peltier devices with good documentation. Some listed here: http://www.tetech.com/Peltier-Thermoele ... mance.html

Here's the specifications for the 172Wqmax model ($38): http://www.tetech.com/temodules/graphs/ ... .4-0.8.pdf
At 10°deltaC it'll remove ~50W at a COP of 3 (17W waste for 50W cooling) and ~110W at a COP of 1 (110 watts waste heat + 110W cooling). The full 172W has a COP around 0.5 (waste heat is 2x cooling provided)

[Tech_Marco]

Remember, a peltier is NOT a cooler. It is just a 'heat pump'. It suck heat from one side to another using some "PN" junction effect. The side product is "HEAT" and it take more energy to get the same amount of 'cold' to be generated. Unlike a compressor type cooling system which is applying "Potential energy" law by compressing Feon R11/R12 or R22... now R134, the peltier is only good for 'point' coolingfor electronics component only. The compressor is more efficiency than the peliter. I'll have to say it is about 2~300 better. So, forget about the peltier for laser tube cooling. You may use the peltier to cool off something small or without active load like a can of beer or a can of soda. But for bigger load like a CO2 laser or machine, it is not going to work. But the way how the peliter work is pretty cool. Fun to play around with

Marco

[medicdude]

Yes, the idea is to use peltiers to pump heat between the laser waterblock to a secondary waterblock, which will be kept at wet-bulb ambient temperature with an evaporative cooler stack.

I understand for larger systems the standard compression cycle has a higher coefficient of performance (cooling power provided:electrical power consumed), but for small and medium sized systems peltiers can be comparable or even better than compressors. I think the compressor gets better and better the colder you want it.

Peltiers can operate at and above a COP of 3, which is better than the small 800-1400W compressors (I think they are around a COP of 2-3). Obviously you'd take a slight performance hit during particularly high wet-bulb temperatures, but it's well worth it if you can get the high performance during other times, as well as the noiseless peltiers. You'd just need the water pumps, and *maybe* a low-power blower for your cooler stack.

2x to 3x the 'listed' capacity of peltiers is required to get good results, more if your cooling system requires a high deltaT to exhaust all that heat.

If wet-bulb temp was at or below your target temp you could even ignore the peltiers completely, with a second heat exchanger. It would be even better to use this as a first-pass exchanger, then use peltiers as an 'aftercooler'.

[medicdude]

Apparently I'm re-designing the wheel here

Evaporative cooler has been a popular solution for PC CPU cooling, more commonly referred to as a 'bong' cooler (usually made with PVC piping). Very similar design, but using the extra length of the PVC pipe to increase air-water contact time. Some info here: http://www.overclockers.com/forums/show ... p?t=321055

Here's my cooling system design, bypass valve is for winter operation (with antifreeze in cooler water):

[waltfl]

Hi there
your idea will not work for laser machines because you would get lots of air bubbles in the cooling stream what is deadly for laser tubes. there should never be any air bubbles at all in a laser tube
greetings
waltfl

Apparently I'm re-designing the wheel here

Evaporative cooler has been a popular solution for PC CPU cooling, more commonly referred to as a 'bong' cooler (usually made with PVC piping). Very similar design, but using the extra length of the PVC pipe to increase air-water contact time. Some info here: http://www.overclockers.com/forums/show ... p?t=321055

Here's my cooling system design, bypass valve is for winter operation (with antifreeze in cooler water):

[medicdude]

Hi there
your idea will not work for laser machines because you would get lots of air bubbles in the cooling stream what is deadly for laser tubes. there should never be any air bubbles at all in a laser tube
greetings
waltfl

That's why the heat exchanger/water blocks are there, the laser water is a sealed system, fill it once and it should stay sealed, pure, and bubbleless.

[medicdude]

So I think I'm going to drop the heat exchanger and use only the peltier block.

I've done a little bit of research; peltiers operate at their max COP around 1/4-1/3 of their rated max Qc. The lower end peltiers tend to advertise their max electrical consumption in watts (about 2x their rated WQc), so LOStore's 90W peltier is probably only rated for 45WQc, and 168W peltier is only rated for 80WQc, and that's only if they can pump across a deltaC of 0, which might explain why Marco's setup failed.

The best price per WQc I've found is for for customthermoelectric's 400WQc model at $57/u ($43/u in large bulk)http://www.shop.customthermoelectric.co ... qscsfapp03. Their models with 287 semiconductors probably have a better COP, but they're a lot pricier per WQc so I won't go that route.

I'm going to grab 6 aluminum waterblocks http://dx.com/p/m-shape-aluminum-alloy- ... QgodNyUAZQ and i'll have to thermal glue two together to fit the entire width of the 62x62mm peltiers. They won't be optimal, I'd rather custom CNC some aluminum blocks but it's not worth it for this prototype. As long as they keep the peltiers within 20C it will be fine and thats why I'm designing 1600WQc of peltiers for this prototype, for the extra head room.