One way to make a Copper Star for a LED - with photos

32 posts / 0 new
Last post

Pages

Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA
One way to make a Copper Star for a LED - with photos

Making a Copper Star for a led - with photos EDIT: Basically this is just an FYI idea, if anyone is wanting to know a way to make a Copper star.

I have seen the XM-L leds on Copper stars out there, on the internet and I often thought about making my own. I just thought it would be too hard (due to the small size), for me to do. Well, sooner or later I had to try it for myself.

Here are four drawings of what I came up with. This is what I wanted to make and then we will see reality.

The center pad of the led is soldered directly to the star and the outside pads are soldered to isolated copper strips. Kapton tape does the isolation, and here's an explanation of the process:

 

1 

I start with a copper round. This one is 3/4" diameter and 18ga thick. I polish it with 0000 steel wool and clean it with alcohol.
(All the copper has been polished and cleaned in the same manner).

Then I take a piece of 22ga copper sheet, (cut to the size of the led center pad) and I solder it onto the center of the star. What I did was to flow a very thin layer of solder to the sheet. I did that by flowing it on and wiping it off, with a paper towel, so it just leaves a really thin layer. Then I just set the sheet on the star and heat till they flow together.

Next I take a Kapton round and stick it to the star. I use an Exacto blade to cut out around the center pad and to cut the places where the + & - strips will be.

2

 

I take some double sided adhesive Kapton tape and put pieces down, where I removed tape on the star. These spots will be where the copper strips go.

I cut two pieces of 24ga Copper and stick them to the double sided tape, making sure of my spacing to the center pad, so that when the led is soldered onto the pads, they line up correctly.

3

 

Then I add low heat solder paste to all three pad areas of the star and set the led on the pads. I heat each outer pad, by heating right next to the led on each side. I heat the center pad, by heating from the underside of the star.

I take another piece of Kapton tape (pre-cut round) and cut out a center square opening, the same size as the led.

I stick it on the star and then cut out two spots where I want to solder the wires. Flow solder on them and I'm ready to go!

4

 

I have to take a moment and explain why the 24ga and 22ga copper sheet. I use 22ga for the center, because it's 0.005" thicker than the 24ga for the outside strips. That way, the thickness of the double sided tape has been accounted for and overall, the three pads come out to the same height, (or close enough to be called the same height). If all the copper sheet was the same gauge, the outside pads would be higher than the center pad (which is the mistake I did with the very first star) and more solder would be needed to bridge the center pad.

I wanted to know how this whole thing would work. It looks fine, but does it really work and how well does it work? Is it worth the work? For me, it was worth it, just to be able to learn how to do it successfully, but that does not give you any real data does it?

Well I found a very gracious gentleman by the forum name of Match and I asked him if he would test out my very first star, to see if it worked and how well it worked. He did a full test all the way up to five amps for me. I want to sincerely thank match for his generosity and his time. He took the time to hook this star up and test it out completely and I am so appreciative of his work.

Unfortunately Match is out of pocket right now, but here's what he told me in a PM: your led performed within ~10% of mine, only dropping off at > 5 amps. I’d say it’s a winner. I’ll email you the exact details (tested every 200ma up to 6amps) when I can get to my notebook.

When he gives me the details, I will include them here.

 

Tonight, the Reality: I did a "Down and Dirty" to show you how it goes and it's not exactly like the drawings.

Here's the components:

11

12

Above, are the sheets of copper, with the Kapton tape. Then there are the star and the three pieces of copper for the pads.

 

33

Center pad soldered to the star and some excess solder sanded away.

 

44

Kapton tape round put on, center is cut away with a New, Sharp, Exacto blade. If you use a dull one, the tape will tear apart.

 

55

Double sided tape put on and outer pads stuck in place.

 

66

Top Kapton round stuck on and I simply cut around all 3 pads. I don't see the need to do any different. It's functional. As I said, this is just a down & dirty, to show how it's done. I went for quick, not pretty

I do not have an XM-L to put on right now, so I didn't put on any solder paste. I just left the pads bare, till I get some LEDs, which will be delayed due to the hurricane over HK.

 

 

Here are the sources for the components:

Stamped Copper rounds - Etsy (search for copper rounds)

 

Kapton tape (Double Sided Polyimide Tape - 1/4in) - Amazon, but really from

http://www.polyimidetapeonline.com/

 

Pre-cut Kapton tape Rounds (Film Shape, Round, 1 In, PK 50) - Drillspot

 http://www.drillspot.com/adhesives-sealants-and-tape/tapes/film-tape/?spc=Shape%3DRound

 

Low heat (solder it) solder from Lowes in the plumbing section, but since then I have purchased kester solder paste 

 http://www.cmlsupply.com/kester-ep-256-syringe.html

 

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

Edited by: Old-Lumens on 07/25/2012 - 03:40
Major
Major's picture
Offline
Last seen: 4 years 9 months ago
Joined: 11/10/2011 - 13:59
Posts: 232
Location: Peak district, UK

Great! thanks for listing the items used, make it easier for us beginners.

Match
Match's picture
Offline
Last seen: 3 years 11 months ago
Joined: 11/14/2010 - 11:57
Posts: 1488
Location: South Carolina, USA

True, I have been out of pocket (vacation)...but the sad reality is I'm also extremely lazy and thoughtless at times for I've had the completed test data for some time now.  Justin, please accept this as my public apology for taking so long.  Without further ado, here is the graph depicting the creation above vs a standard aluminum 20mm star:

 

scaru
scaru's picture
Offline
Last seen: 3 years 3 months ago
Joined: 03/22/2012 - 13:36
Posts: 6946
Location: Virginia

Wow! So it this using your optimal heatsink by bolting it to a big chunk of aluminium/copper?

mattthemuppet
Offline
Last seen: 2 weeks 4 days ago
Joined: 01/04/2012 - 16:20
Posts: 1271
Location: WA

wow! Very neat and ingenious piece of work OL and worth a 10% increase in output, which isn’t to be sniffed at. Is that 10% difference greater than the expected margin of error (i.e., is it due to the Cu star or just LED variance)?

Thanks to Match too, for the testing data. How did the OL Cu star compare with your XM-L reflowed onto copper?

ma_sha1
ma_sha1's picture
Offline
Last seen: 3 years 10 months ago
Joined: 05/06/2012 - 10:20
Posts: 367
Location: CT

Nice!

JohnnyMac
JohnnyMac's picture
Offline
Last seen: 2 years 4 months ago
Joined: 04/12/2011 - 16:03
Posts: 8863
Location: Eastern PA

That’s pretty cool, O-L!! Nice work.

Have you or anyone else tried the copper stars from DX yet? I’m ashamed to admit it but I have them but have never tried reflowing anything yet. I’m going to have to one day very soon. Seems like it may give the same results without the added expense or time involved. It wouldn’t give the great feeling of satisfaction of making it completely for yourself but I think I could get over that. LOL

Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA

Match wrote:

True, I have been out of pocket (vacation)...but the sad reality is I'm also extremely lazy and thoughtless at times for I've had the completed test data for some time now.  Justin, please accept this as my public apology for taking so long.  Without further ado, here is the graph depicting the creation above vs a standard aluminum 20mm star:

 

Match, absolutely no need to apologize. I think you are more like over worked and vacation was probably the first break in a long time.Wink

What type of heat sink did you use? I am figuring on Aluminum. The one you had used before, for other testing.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

DarkSide
DarkSide's picture
Offline
Last seen: 2 years 6 months ago
Joined: 06/04/2012 - 01:50
Posts: 792
Location: NYC

Handsome piece of work.

The “solder it” stuff from Kesler leaves some residue, IMO. They also make one specific for
copper . If you solder onto copper it can be a little difficult if you dont have a high output iron.
The copper solder paste makes it a bit easier to work with.

 


Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA

JohnnyMac wrote:
That’s pretty cool, O-L!! Nice work.

Have you or anyone else tried the copper stars from DX yet? I’m ashamed to admit it but I have them but have never tried reflowing anything yet. I’m going to have to one day very soon. Seems like it may give the same results without the added expense or time involved. It wouldn’t give the great feeling of satisfaction of making it completely for yourself but I think I could get over that. LOL

!http://img.dxcdn.com/productimages/sku_126699_2.jpg!


I have never had one in my hands. I would be interested as to what copper alloy they are.
I haven’t heard of any testing on those yet.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

ma_sha1
ma_sha1's picture
Offline
Last seen: 3 years 10 months ago
Joined: 05/06/2012 - 10:20
Posts: 367
Location: CT

That PCB doesn’t look like direct solder to copper, you might have to cut-out the center to expose the copper to achieve O-L level

JohnnyMac
JohnnyMac's picture
Offline
Last seen: 2 years 4 months ago
Joined: 04/12/2011 - 16:03
Posts: 8863
Location: Eastern PA

ma_sha1 wrote:
That PCB doesn’t look like direct solder to copper, you might have to cut-out the center to expose the copper to achieve O-L level

A great idea and shouldn’t be hard at all. Wink
Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA

DarkSide wrote:
Handsome piece of work.

The “solder it” stuff from Kesler leaves some residue, IMO. They also make one specific for
copper . If you solder onto copper it can be a little difficult if you dont have a high output iron.
The copper solder paste makes it a bit easier to work with.

I have been trying out the other paste I linked to in the first post and I like it so far. Low melt and seems to be pretty clean. I have only used it a couple times. I plan on doing a little “copper star assembly line” when I get some leds in, then I will have more experience with it.

Edit: The solder it paste from Lowe’s was not the best. I ended up squeezing out some on a cardboard and mixing it up well before using it. It seemed to work better. I also tried the “copper solder it” from Lowe’s. I threw that stuff away. Don’t need more high melt stuff.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

NightCrawl
Offline
Last seen: 4 years 1 week ago
Joined: 01/22/2012 - 08:20
Posts: 3071
Location: Karlsruhe, Germany

That looks really great! You could also use clear coat to insulate the contacts.. thats what the manufacturers do.^^

I like your creativity.. if they offer thicker copper plates, you could make the whole heatsink direct to copper.

cheaplite
Offline
Last seen: 6 years 8 months ago
Joined: 01/08/2012 - 15:04
Posts: 215

I like the way u think. good ol yankee ingenuity. u’ve given me a few ideas for future projects. i’m kinda surprised that tape provides enough mechanical connection for the wiring pads. let me know how u center the led.

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 1 year 3 weeks ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

Once again, you impress. I’ve been wondering lately how to use a 1/2” copper cap to make a pill and you light the way. A bit of 1/2” inside to add mass and a bit outside 1-2mm longer than the cap for driver recess and follow the recipe above. Scratch lines in the first layer might help with keep the pad stable.

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

Bort
Bort's picture
Offline
Last seen: 1 week 1 day ago
Joined: 06/01/2012 - 17:15
Posts: 7460
Location: Holding the proverbial flashlight

Old-Lumens wrote:

Match wrote:

True, I have been out of pocket (vacation)…but the sad reality is I’m also extremely lazy and thoughtless at times for I’ve had the completed test data for some time now.  Justin, please accept this as my public apology for taking so long.  Without further ado, here is the graph depicting the creation above vs a standard aluminum 20mm star:



 


Match, absolutely no need to apologize. I think you are more like over worked and vacation was probably the first break in a long time.Wink

What type of heat sink did you use? I am figuring on Aluminum. The one you had used before, for other testing.


This suggests there is more gains to be had from reducing led temperature, as i had assumed they reached their lumen output ceiling due to saturation or some non heat related reason. I wonder if some kind of active cooling, say a peltier cooler, or fan connected to the heatsink behind the led would make a significant difference.

The Journal of Alternative Facts TM

"It is critical that there is a credible academic source for the growing and important discipline of alternative facts. This field of study will just keep winning, and we knew that all the best people would want to be on board. There is a real risk in the world today that people might be getting their information about science from actual scientists"

 

Chicago X
Chicago X's picture
Offline
Last seen: 6 years 9 months ago
Joined: 07/22/2011 - 16:13
Posts: 4013
Location: See Name

O-L, you hit another one out of the park.

It’s amazing what you can do, running on nothing but 140-proof stubbornSilly

Thanks also for the DIY.

http://wardogsmakingithome.org/index.html

War Dogs, Making it Home - Rescue Dogs for Returning Vets

Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA
Bort wrote:
Old-Lumens wrote:

Match wrote:

True, I have been out of pocket (vacation)…but the sad reality is I’m also extremely lazy and thoughtless at times for I’ve had the completed test data for some time now.  Justin, please accept this as my public apology for taking so long.  Without further ado, here is the graph depicting the creation above vs a standard aluminum 20mm star:

 

Match, absolutely no need to apologize. I think you are more like over worked and vacation was probably the first break in a long time.Wink

What type of heat sink did you use? I am figuring on Aluminum. The one you had used before, for other testing.

This suggests there is more gains to be had from reducing led temperature, as i had assumed they reached their lumen output ceiling due to saturation or some non heat related reason. I wonder if some kind of active cooling, say a peltier cooler, or fan connected to the heatsink behind the led would make a significant difference.

To me, it’s just like a processor for a PC. Cooling is the key. Water (or nitrogen) cool a LED and the sky is the limit. Well at least we don’t know the limit for sure until that is done.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

Bort
Bort's picture
Offline
Last seen: 1 week 1 day ago
Joined: 06/01/2012 - 17:15
Posts: 7460
Location: Holding the proverbial flashlight
Old-Lumens wrote:
Bort wrote:
Old-Lumens wrote:

Match wrote:

True, I have been out of pocket (vacation)…but the sad reality is I’m also extremely lazy and thoughtless at times for I’ve had the completed test data for some time now.  Justin, please accept this as my public apology for taking so long.  Without further ado, here is the graph depicting the creation above vs a standard aluminum 20mm star:

 

Match, absolutely no need to apologize. I think you are more like over worked and vacation was probably the first break in a long time.Wink

What type of heat sink did you use? I am figuring on Aluminum. The one you had used before, for other testing.

This suggests there is more gains to be had from reducing led temperature, as i had assumed they reached their lumen output ceiling due to saturation or some non heat related reason. I wonder if some kind of active cooling, say a peltier cooler, or fan connected to the heatsink behind the led would make a significant difference.

To me, it’s just like a processor for a PC. Cooling is the key. Water (or nitrogen) cool a LED and the sky is the limit. Well at least we don’t know the limit for sure until that is done.


Lets hope whoever tries it has fire gear at the ready

The Journal of Alternative Facts TM

"It is critical that there is a credible academic source for the growing and important discipline of alternative facts. This field of study will just keep winning, and we knew that all the best people would want to be on board. There is a real risk in the world today that people might be getting their information about science from actual scientists"

 

Match
Match's picture
Offline
Last seen: 3 years 11 months ago
Joined: 11/14/2010 - 11:57
Posts: 1488
Location: South Carolina, USA

Couple of points:

- Test was performed exactly as all of my emitter tests - which is mounted to a 1lb piece of copper barstock imbedded in a 25lb block of aluminum.

- Heat is the key. The cooler an led can be kept, the more power it can be fed to produce more light.

- I have @ 10 150w peltier coolers at work that I could stack together to get an led close to cryogenic...but other than academic interest a test like that wouldn't have much bearing on what could be done in relation to a flashlight.

mattthemuppet wrote:
Thanks to Match too, for the testing data. How did the OL Cu star compare with your XM-L reflowed onto copper?

 The results from that test are here.  For Justin's benifit, I sent him a graph depicting all three tests. 

 

Bort
Bort's picture
Offline
Last seen: 1 week 1 day ago
Joined: 06/01/2012 - 17:15
Posts: 7460
Location: Holding the proverbial flashlight

Match wrote:

Couple of points:


- Test was performed exactly as all of my emitter tests – which is mounted to a 1lb piece of copper barstock imbedded in a 25lb block of aluminum.


- Heat is the key. The cooler an led can be kept, the more power it can be fed to produce more light.


- I have @ 10 150w peltier coolers at work that I could stack together to get an led close to cryogenic…but other than academic interest a test like that wouldn’t have much bearing on what could be done in relation to a flashlight.


mattthemuppet wrote:
Thanks to Match too, for the testing data. How did the OL Cu star compare with your XM-L reflowed onto copper?

 The results from that test are here.  For Justin’s benifit, I sent him a graph depicting all three tests. 


 


I agree, you can’t easily put a peltier cooler into a flashlight, so i guess its more of an academic interest. I assume at some point the wires and junctions in the led will fry from the voltage and current if you keep the temp low enough for that to happen.
We could always start a betting pool to see what amperage and lumens a U2 XM-L can take before non heat related failure Wink

The Journal of Alternative Facts TM

"It is critical that there is a credible academic source for the growing and important discipline of alternative facts. This field of study will just keep winning, and we knew that all the best people would want to be on board. There is a real risk in the world today that people might be getting their information about science from actual scientists"

 

scaru
scaru's picture
Offline
Last seen: 3 years 3 months ago
Joined: 03/22/2012 - 13:36
Posts: 6946
Location: Virginia

Bort wrote:
Match wrote:

Couple of points:

- Test was performed exactly as all of my emitter tests - which is mounted to a 1lb piece of copper barstock imbedded in a 25lb block of aluminum.

- Heat is the key. The cooler an led can be kept, the more power it can be fed to produce more light.

- I have @ 10 150w peltier coolers at work that I could stack together to get an led close to cryogenic...but other than academic interest a test like that wouldn't have much bearing on what could be done in relation to a flashlight.

mattthemuppet wrote:
Thanks to Match too, for the testing data. How did the OL Cu star compare with your XM-L reflowed onto copper?

 The results from that test are here.  For Justin's benifit, I sent him a graph depicting all three tests. 

 

I agree, you can't easily put a peltier cooler into a flashlight, so i guess its more of an academic interest. I assume at some point the wires and junctions in the led will fry from the voltage and current if you keep the temp low enough for that to happen. We could always start a betting pool to see what amperage and lumens a U2 XM-L can take before non heat related failure ;)
CheapThrills
CheapThrills's picture
Offline
Last seen: 2 hours 28 min ago
Joined: 07/02/2011 - 10:45
Posts: 3615
Location: Suomi

This is the DIY in the proper meaning, of the words!

Great!!!

Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA

Thanks everyone, for the kind words and thanks to Match once again, for doing the testing.

I plan on using copper stars like this, in my builds this winter.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA

cheaplite wrote:
I like the way u think. good ol yankee ingenuity. u’ve given me a few ideas for future projects. i’m kinda surprised that tape provides enough mechanical connection for the wiring pads. let me know how u center the led.

I centered these just by setting the led on the pads by eye. Then I hit each outer pad with a solder gun just long enough to heat the low melt solder and last is to heat under the center pad, until it’s solder melts. If the solder is done very thin, it works pretty good.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

mattthemuppet
Offline
Last seen: 2 weeks 4 days ago
Joined: 01/04/2012 - 16:20
Posts: 1271
Location: WA
Match wrote:
The results from that test are here.  For Justin’s benifit, I sent him a graph depicting all three tests.

thanks. Looks like the only main difference between OL Cu star and direct copper mount is at very high amps (>4.5A), so within the realms of sane use OL’s method should be equivalent and a smidge easier Smile

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 1 year 3 weeks ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

I was looking up kapton tape and found a document on off gassing from the adhesive here
I don’t know if it’s relavant or not. It seems to relate to condensation of the gasses on switch contacts but I thought it might also affect reflectors adversely. Do people use kapton often on stars without problems? If so then ignore this post. This is a great idea, just want to avoid later disappointment. Thanks

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

Mr_Krabs
Offline
Last seen: 5 years 10 months ago
Joined: 07/14/2012 - 17:47
Posts: 231

I’ve done something similar, I made my own copper star by sanding a penny until it was flat. Copper is surprisingly expensive material and one convenient thing about the penny is before 1982 they were pretty much solid copper. You’re actually getting more than 1 cent worth of copper for every penny so there is no cheaper way to make a copper star than by sanding a penny flat.

Instead of soldering a base to raise the center you can actually just use the solder itself to bridge the thin gap between the kapton gasket and star. Instead of using copper sheet as conductor you can use copper foil. This way there is minimum thickness between the conductors/gasket and the star allowing the solder to fill the gap in completely. That makes it only one solder joint from the led thermal pad to the star. Solder joints are a big bottleneck for heat conduction so you want to keep them to a minimum.

Old-Lumens
Old-Lumens's picture
Offline
Last seen: 3 years 3 months ago
Joined: 11/04/2011 - 11:39
Posts: 7478
Location: Tyler, TX, USA
Rufusbduck wrote:
I was looking up kapton tape and found a document on off gassing from the adhesive here I don’t know if it’s relavant or not. It seems to relate to condensation of the gasses on switch contacts but I thought it might also affect reflectors adversely. Do people use kapton often on stars without problems? If so then ignore this post. This is a great idea, just want to avoid later disappointment. Thanks

The Polyminide tape I am using is supposed to have a Teflon Silicon adhesive, not Acrylic. Probably still has off gassing as any adhesive probably would. I imagine even the Silicone domes on leds have some form of off gassing, however minor it may be.

My PayPal address: oldlumens (insert the @ sign here) gmail.com

My YouTube Flashlight Video Channel

The BLF Modding Links Thread 

http://imageshack.com/a/img922/1374/jQ2wdL.jpg

 

Rufusbduck
Rufusbduck's picture
Offline
Last seen: 1 year 3 weeks ago
Joined: 04/04/2012 - 15:34
Posts: 10389
Location: Golden state

Not trying to be a party pooper and the article did look old. I have not used kapton and don’t know what changes have been made in the adhesive. The article just made me think of the effect super glue can have on domes and lenses. As Gilda Radner said,“never mind”.

Three Tanna leaves to give him life, nine to give him movement. But what if he eats the whole bag?

Scott

Pages