Homemade High Power Diode Laser

koliaco · Post by **koliaco** » Tue Feb 24, 2009 12:02 am

Colin Kaminski wrote:Wow, those are big caps. What is the voltage rating on them?

The biggest one is 4.7uF and 250V rating - I only chose it over smaller choice cause it has stability and high frequency feature. others are in 50v and 25v for smaller one-0.47uF

Hans · Post by **Hans** » Tue Feb 24, 2009 9:10 am

Have you tested the circuit yet? How did you determine the correct values for the caps and the resistors in the PID section of the circuit?

Here are some tips for testing:
- Use a rather large cooling fin on the other side of the Peltier.
- Use shielded coax cables for both the NTC sensor and the Peltier. (I got this tip from Colin a long time ago)
- Place the NTC sensor as close to the Peltier as you can manage. That will minimize delay between temperature changes and sensing.
- Shield all connections to the NTC/Peltier/diode from moisture. When the diode-side of the Peltier gets cool water will condense on that side.
- If you test the circuit without the housing of the diode on the temperature controlled side of the Peltier, do put some equal thermal mass in contact with it instead. Otherwise you will get wild temperature swings that might fool you into thinking that the circuit does not work properly.

koliaco · Post by **koliaco** » Tue Feb 24, 2009 11:32 am

Hans wrote:Have you tested the circuit yet? How did you determine the correct values for the caps and the resistors in the PID section of the circuit?

Here are some tips for testing:
- Use a rather large cooling fin on the other side of the Peltier.
- Use shielded coax cables for both the NTC sensor and the Peltier. (I got this tip from Colin a long time ago)
- Place the NTC sensor as close to the Peltier as you can manage. That will minimize delay between temperature changes and sensing.
- Shield all connections to the NTC/Peltier/diode from moisture. When the diode-side of the Peltier gets cool water will condense on that side.
- If you test the circuit without the housing of the diode on the temperature controlled side of the Peltier, do put some equal thermal mass in contact with it instead. Otherwise you will get wild temperature swings that might fool you into thinking that the circuit does not work properly.

No I havent yet, still waiting for the regulated 5v 5A power supply from Surplusshed to arrive. the values were taken from MAX datasheet for basic operation, as I am quite new to these things and dont yet understand how they work.

How do you determine which way the peltier should be positioned? if looking from back side of LD where you see pins (as in picture) should the negative lead be on the right hand side or left?

Oh and thanks for your suggestion about using coax cable. do you actuallly cut off the wire from peltier and solder the coax one? About the NTC I dont think I will be able to do it, its just too tiny -1.5mm in dia for the head and the leads are heavy isomid insulated Nickel lead-wires. in my design the whole things is placed together in an aluminum housing and plate underneath, so would coax cable be necessary? since I can make zero distance from peltier to driver samw for ntc.

I placed the thermistor so that its distance to LD is equal the distance to peltier, thinking that if its too close to peltier it will have lag from LD temperature change, its just my assumption, will have to somehow test it later. Am planning to use foil foam (used to cover car's window) to cover arround the housing.

Hans · Post by **Hans** » Tue Feb 24, 2009 5:48 pm

How do you determine which way the peltier should be positioned?

The easiest way is to not solder the Peltier but to use a plug. If the Peltier is connected in reverse, the temperature will shoot in one direction or the other. That way you can see it is wrong.

do you actuallly cut off the wire from peltier and solder the coax one?

Yes, that's best. The peltier is driven with high frequency pulses. So, it acts a little like a transmitter. If you don't use coax, you can get rid of most of the interference by wrapping the leads a few times around a circular ferrite core.

I placed the thermistor so that its distance to LD is equal the distance to peltier, thinking that if its too close to peltier it will have lag from LD temperature change, its just my assumption, will have to somehow test it later.

Your guess is as good as mine. I chose for short distance to the peltier for fast response time. But then, my controller is a PI and not a full PID. It also helps to insulate the diode housing from the surrounding air as much as possible. Perhaps build a air tight chamber around the temperature controlled side and fill that with really dry air to prevent condensation when the diode gets too cold.

About the NTC I dont think I will be able to do it, its just too tiny -1.5mm in dia for the head and the leads are heavy isomid insulated Nickel lead-wires.

Ok, but what you can do is wrap Aluminum foil around the cable and hook that up to the same wire of the NTC that is connected to the ground. Now you have made your own coax cable.

koliaco · Post by **koliaco** » Wed Feb 25, 2009 4:56 am

Hans wrote:It also helps to insulate the diode housing from the surrounding air as much as possible.

I think I will be using the diode without any optics sometimes, and my diode has no glass window or anything to isolate the junction, just wondering if thats a good thing or bad? since no windows means no interference from reflected light, but its also very vulnerable to atmosphere temp change. The diode is SDL1239.

koliaco wrote:Ok, but what you can do is wrap Aluminum foil around the cable and hook that up to the same wire of the NTC that is connected to the ground. Now you have made your own coax cable.

Do you mean splitting the lead wire, cutting the ground lead off- wrap and isolate it - then connect it back? so the positive lead does not need any change? Would it work if I just wrap the whole lead with aluminum then use heat shrink? (currently the leads stick together and only split for a cm at ends for soldering)

koliaco · Post by **koliaco** » Wed Feb 25, 2009 4:57 am

Hans wrote:It also helps to insulate the diode housing from the surrounding air as much as possible.

I think I will be using the diode without any optics sometimes, and my diode has no glass window or anything to isolate the junction, just wondering if thats a good thing or bad? since no windows means no interference from reflected light, but its also very vulnerable to atmosphere temp change. The diode is SDL1239.

koliaco wrote:Ok, but what you can do is wrap Aluminum foil around the cable and hook that up to the same wire of the NTC that is connected to the ground. Now you have made your own coax cable.

Do you mean splitting the lead wire, cutting the ground lead off- wrap and isolate it - then connect it back? so the positive lead does not need any change? Would it work if I just wrap the whole lead with aluminum then use heat shrink? (currently the leads stick together and only split for a cm at ends for soldering)

Hans · Post by **Hans** » Wed Feb 25, 2009 8:35 pm

Do you mean splitting the lead wire, cutting the ground lead off- wrap and isolate it - then connect it back? so the positive lead does not need any change? Would it work if I just wrap the whole lead with aluminum then use heat shrink? (currently the leads stick together and only split for a cm at ends for soldering)

Sorry, I was wrong about that recommendation because the polarity of those changes when the TEC changes from cooling to heating. The aluminum foil wrap should be connected to the common ground of your circuit. You can put heat shrink around the aluminum wrap to make it more sturdy.

kir · Post by **kir** » Wed Mar 04, 2009 10:28 am

Couple words to Kolia about buying components. I live in Australia, quite close to you

. I usually use either digikey.com (they do sell any quantity - from one), or rs-components.com.au. The MAX1978 ic could be ordered directly from maxim site (maxim-ic.com/sales/). They have pretty expensive shipping, so if you decided to buy we could share delivery price.

a_k · Post by **a_k** » Thu Mar 05, 2009 9:57 pm

Do you mean Sony SLD1239-JL54? If so this LD does have a window.

koliaco · Post by **koliaco** » Thu Mar 12, 2009 10:03 pm

kir wrote:Couple words to Kolia about buying components. I live in Australia, quite close to you . I usually use either digikey.com (they do sell any quantity - from one), or rs-components.com.au. The MAX1978 ic could be ordered directly from maxim site (maxim-ic.com/sales/). They have pretty expensive shipping, so if you decided to buy we could share delivery price.

I wish I checked back this post earlier before, cause I just had to order the MAX1978EVKIT from them on the 03-03-09 ( and it costed $50 usd for shipping a tiny board

). the one I made in the picture did not work well, and the MAX kept heating up a lot, I spent whole day trying to figureout what went wrong but just gave up, the peltier did cool a bit but it did not seem to make ntc move towards set point.

or and thanks for your advise on rs-components.com.au, will give it a try next time, digikey prices are quite steep although they do come in very well sealed packages in 2 days. (free shipping to nz on orders over 125 nzd)