7135*4 vs 7135*8 drivers

20 posts / 0 new
Last post
flydiver
Offline
Last seen: 13 hours 42 min ago
Joined: 06/19/2013 - 19:16
Posts: 694
Location: Seattle, WA
7135*4 vs 7135*8 drivers

I got in on this” Convoy S2+ buy “:http://www.aliexpress.com/store/product/Free-Shipping-Convoy-S2-Gray-Cre... purchased 2, both with XML2 T6-3B, one with 7135*4 and the other with 7135*8. This was my first purchase of anything except pretty much bargain basement LED with the exception of a couple dive lights. The lights were MUCH better made than my others. But, I really cannot visually tell much difference at all between either light.
???? – on paper/spec the *8 is pulling about 2x the amps. I’ve read that you have to significantly up the power be able to see it. I have, at least I think I have. :quest:
Folks in the buy thread were pretty excited about getting the *8 drivers, which was much of my incentive to get it, just to see what this was about. Near as I can tell what I have is a light that eats my battery faster with no other real benefit. No, I don’t have a meter. Yes, I have done all sorts of A:B visual comparisons outside, in the dark at various distances and surfaces.
I have visually verified that I do indeed have *4 and *8 drivers, at least by this picture.

Is it my eyes? Is something wrong with the light/driver? Is this pretty much ‘normal’?

Note: I tried to get an answer in the S2+ thread but the buy is over and there didn’t seem to be any feedback.

unknown00101
Offline
Last seen: 6 hours 39 min ago
Joined: 08/16/2013 - 11:53
Posts: 4248
Location: USA

You should be seeing between 200 and 400 more lumens from your 8* assuming all is correct.

You need to get yourself a decent DMM (less than $20 can get you a well-suited DMM for your flashlight addiction Smile ) to make sure everything is good with your lights, and especially to check on your Liion batteries!

WarHawk-AVG
WarHawk-AVG's picture
Offline
Last seen: 4 days 9 hours ago
Joined: 01/04/2014 - 06:47
Posts: 4954
Location: H-Town

Cree works more efficient at lower current levels

100% current flow at 1400mA will run more efficiently than 50% PWM (50% on 50% off cycle) at 2800mA

Just imagine a car, 1 with a 4 banger 1 with a V8, the four banger…stomp the gas hold it down….the V8, stomp the gas, let off the gas, stomp the gas, let off the gas (or pull the plug wires on 4 of the cylinders)

DB Custom
DB Custom's picture
Offline
Last seen: 6 hours 4 min ago
Joined: 01/13/2013 - 22:28
Posts: 15447
Location: Heart of Texas

If you have a camera, set it up on a tripod in your living room and aim it at a picture on the wall, a clock, almost anything. Turn off the lights. Now shine one light at the subject you have chosen and take a picture. Do this again with the other light. Then compare the settings the camera chose for a proper exposure. You should see the difference in shutter speed and aperture value.

Dale

scooby214
Offline
Last seen: 10 months 2 days ago
Joined: 02/25/2014 - 07:25
Posts: 136
Location: Oklahoma City

For me, with my S2+ and S3 lights, 7135*4 is the sweet spot. With the small, floody reflector, higher current output levels don’t seem to give a big boost in visible brightness (at least to my unaided eyes). The brightness boost is there, but I just don’t see it as much in actual use. A ceiling bounce test does show how the 7135*8 is brighter. My regular usage with these lights is out in the backyard, where I don’t perceive enough of a brightness boost to justify the reduced battery life and increased heat.

I do have throwier lights which seem to throw noticeably further when driven with more amps.

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 2 months ago
Joined: 07/22/2013 - 13:10
Posts: 2963
Location: Columbus, OH USA Earth

A ceiling bounce test is about the best way to compare output from two of the same lights at different power levels. Dont point the light directly at the object you want to illuminate to see if you can see the difference, shine the light straight up at the ceiling and then look around the room to see the difference in brightness.

Please help support my research by giving or sharing. GoFundMe.com/DansDrones

 

Always remember SPC Joey Riley, KIA 11/24/14.

harry218
harry218's picture
Offline
Last seen: 24 min 17 sec ago
Joined: 01/01/2014 - 13:44
Posts: 427

Cereal_killer wrote:

A ceiling bounce test is about the best way to compare output from two of the same lights at different power levels. Dont point the light directly at the object you want to illuminate to see if you can see the difference, shine the light straight up at the ceiling and then look around the room to see the difference in brightness.


+1
flydiver
Offline
Last seen: 13 hours 42 min ago
Joined: 06/19/2013 - 19:16
Posts: 694
Location: Seattle, WA

unknown00101 wrote:
You should be seeing between 200 and 400 more lumens from your 8* assuming all is correct.

You need to get yourself a decent DMM (less than $20 can get you a well-suited DMM for your flashlight addiction Smile ) to make sure everything is good with your lights, and especially to check on your Liion batteries!

I do have a DMM. I also have a wattmeter and a Accucel-6. Batteries were fine and visual observations were the same regardless of battery. I did not do a amp test and will have to wait until the *4 comes around again. It was given as a gift.

Quote:
Cree works more efficient at lower current levels. 100% current flow at 1400mA will run more efficiently than 50% PWM (50% on 50% off cycle) at 2800mA

That’s an interesting bit of info. RC plane ESC have a bit of the same kind if issue.

Quote:
With the small, floody reflector, higher current output levels don’t seem to give a big boost in visible brightness (at least to my unaided eyes). The brightness boost is there, but I just don’t see it as much in actual use.

I kind of suspected the small floody reflector might be spreading the brightness around so it was much harder to quantify by eye.

Quote:
Do you have a camera?

Ah, a digital ‘lightmeter’. I do. Again, #2 light needs to return to compare.

Quote:
A ceiling bounce test is

Another clever trick to try.

Thank you all. I learned a few more things and have new tricks and tools to try.

locusto03
Offline
Last seen: 6 months 1 week ago
Joined: 03/03/2014 - 01:35
Posts: 294

I've read somewhere (on BLF) that differences in lumen output are less apparent the higher you go. The difference between 50 and 250 lumens is much more obvious than the difference between 800 and 1000 lumens. Along with the other advice already given, this may be factoring into what you're [not] seeing.

 

On some 3 amp drivers that I programmed, 100% brightness looks just a bit brighter than about 50% to me.

flydiver
Offline
Last seen: 13 hours 42 min ago
Joined: 06/19/2013 - 19:16
Posts: 694
Location: Seattle, WA
locusto03 wrote:

I’ve read somewhere (on BLF) that differences in lumen output are less apparent the higher you go. The difference between 50 and 250 lumens is much more obvious than the difference between 800 and 1000 lumens. Along with the other advice already given, this may be factoring into what you’re [not] seeing.

 

On some 3 amp drivers that I programmed, 100% brightness looks just a bit brighter than about 50% to me.

That seems to be my experience with this particular light.
Your example is a jump of 200 lumens in each case but in the 1st it is a 5x increase and the 2nd is only a 20% jump. Have I read here somewhere a 2x increase is needed to notice it visually reliably?

Wattmeter test: Starting voltage on a new Samsung = 3.97v
Bright – 2.76A / 3.64v
Med – 1.1A / 3.84v
Low – 0.14A /3.95v
The driver seems to be doing it’s job. I certainly can tell the difference between Lo/Med/Hi
I did a full discharge test and the driver does drop to low mode and blinks. Cell was 3.2v (unloaded) immediately after.

I should have the *4 back Monday. I’ll get results then for it.

DB Custom
DB Custom's picture
Offline
Last seen: 6 hours 4 min ago
Joined: 01/13/2013 - 22:28
Posts: 15447
Location: Heart of Texas

You do realize that 3.97V is already seriously depleted?

This from Cottonpickers chargers, he puts it on the side of his chargers and in his manuals…

4.2V 100%
4.1V 87%
4.0V 75%
3.9V 55%
3.8V 30%
3.5V Empty

Dale

RaceR86
RaceR86's picture
Offline
Last seen: 1 year 7 months ago
Joined: 02/22/2012 - 08:32
Posts: 3777
Location: Norway

DBCstm wrote:
You do realize that 3.97V is already seriously depleted? This from Cottonpickers chargers, he puts it on the side of his chargers and in his manuals... 4.2V 100% 4.1V 87% 4.0V 75% 3.9V 55% 3.8V 30% 3.5V Empty

 

3,97v is not what I would consider seriously depleted on any of my cells. 2,97v, yes.. 

Many cells are only 12-18% drained when they are down to 4v. Assuming starting point at 4,2v. And to some, starting point might be 4,15v if charger stops early. And maybe more like 4,13v if the cells have rested for several days or weeks.

Many cells have around 70% capacity left when they are at 3,9v.

Some cells have around 25% capacity left after 3,5v.

On the other side, some cells are pretty much empty at 3,6v.

Basically, on some cells the cottonpickers guide work nicely, on others it can be quite misleading. On most of my cells its quite misleading.

HKJ have graphs that shows these things nicely.

BLF LED database – collaboration spreadsheet and latest news about where to buy LEDs
http://budgetlightforum.com/node/19342

flydiver
Offline
Last seen: 13 hours 42 min ago
Joined: 06/19/2013 - 19:16
Posts: 694
Location: Seattle, WA

3.97v on a decent lion is hardly depleted. A decent cell (these are good Samsung) will hold a fairly flat discharge curve until getting close to empty. If the cell drops it’s voltage badly under load that’s an indication of a cell with high resistance and low discharge support.
What I’m really comparing was 7135*8 vs 7135*4. In that case I’m using the same cell at the same level of charge.
Proof: Same light, same test equipment. This is #2-7135*4 light that I got back for testing today. See post #9 for 7135*8 discharge test.

At 4.13v charge______3.80v charge
Hi 1.40A/3.78v______1.40A/3.60v
Md .57A/3.96v_______.56A/3.70v
Lo 0.05A/4.01v______0.05A/3.76v

You can see in both cases there was a similar drop in voltage but the amp output was essentially identical.
Compared to the 7135*8 the amp draw was about 1/2.

I did the ceiling bounce test. With that I could tell a difference. It was subtle, needed A:B comparison directly and the evaluation would not have worked otherwise.

I let both lights run for 10” on high and tested the temp: 7135*4 = 95*F, 7135*8 = 137*F

I tried the camera test but came up with the same settings results. But the way I set this one up I could visually tell the light output difference. This was only because it was an A:B comparison. In regular use I could not tell.

Thoughts – for THIS light (floody Convoy S2+) and THIS application the 7135*8 is essentially useless. In fact it is significantly detrimental to the battery life. Makes me wonder if the 7135*3 would in fact be better in effective efficiency of light output vs. battery life for real world use.

unknown00101
Offline
Last seen: 6 hours 39 min ago
Joined: 08/16/2013 - 11:53
Posts: 4248
Location: USA

flydiver wrote:
3.97v on a decent lion is hardly depleted. A decent cell (these are good Samsung) will hold a fairly flat discharge curve until getting close to empty. If the cell drops it’s voltage badly under load that’s an indication of a cell with high resistance and low discharge support.
What I’m really comparing was 7135*8 vs 7135*4. In that case I’m using the same cell at the same level of charge.
Proof: Same light, same test equipment. This is #2-7135*4 light that I got back for testing today. See post #9 for 7135*8 discharge test.

At 4.13v charge______3.80v charge
Hi 1.40A/3.78v______1.40A/3.60v
Md .57A/3.96v_______.56A/3.70v
Lo 0.05A/4.01v______0.05A/3.76v

You can see in both cases there was a similar drop in voltage but the amp output was essentially identical.
Compared to the 7135*8 the amp draw was about 1/2.

I did the ceiling bounce test. With that I could tell a difference. It was subtle, needed A:B comparison directly and the evaluation would not have worked otherwise.

I let both lights run for 10” on high and tested the temp: 7135*4 = 95*F, 7135*8 = 137*F

I tried the camera test but came up with the same settings results. But the way I set this one up I could visually tell the light output difference. This was only because it was an A:B comparison. In regular use I could not tell.

Thoughts – for THIS light (floody Convoy S2+) and THIS application the 7135*8 is essentially useless. In fact it is significantly detrimental to the battery life. Makes me wonder if the 7135*3 would in fact be better in effective efficiency of light output vs. battery life for real world use.

Are you testing comparing these indoors at close range? I find outdoor tests at further distances to be much easier to see output differences. The results are sometimes surprising (this goes for beam profile as well).

Anyway the extra ~400 lumens of a *8 at least makes me feel more powerful even if the difference is hardly noticeable. Big Smile

flydiver
Offline
Last seen: 13 hours 42 min ago
Joined: 06/19/2013 - 19:16
Posts: 694
Location: Seattle, WA

I did a number of subjective outdoor tests after dark when indoor basement (dark) comparisons did not reveal any difference I could tell. Then I started this thread.
Obviously there is a difference, but it’s subtle. I guess I’ll just have to ‘feel the power’ even if I can’t see it. Silly

locusto03
Offline
Last seen: 6 months 1 week ago
Joined: 03/03/2014 - 01:35
Posts: 294

This is the main reason I've been programming Qlite drivers with a High of about 50% and a Turbo of 100% with a 60 second timeout. The subjective difference in output isn't a whole lot for my applications. A thrower would probably show the difference in output better.

DB Custom
DB Custom's picture
Offline
Last seen: 6 hours 4 min ago
Joined: 01/13/2013 - 22:28
Posts: 15447
Location: Heart of Texas

Try this with a new build…run a hot rod light with a fully charged IMR at 4.21 or 4.22V and test amperage. Then run the light for a minute, 2 minutes, and try the amperage test again.

If you’re pushing the envelope with the new build, say 5 1/2 to 6A, the 4V cell just won’t do the job. The less you’re expecting of the cell (4A, 3A or less) then the less difference there will be. If, on the other hand, you have a build where a single cell is putting out 16A to a single emitter, well, the cell at 3.97V is simply not going to get anywhere close to 16A.

Dale

WarHawk-AVG
WarHawk-AVG's picture
Offline
Last seen: 4 days 9 hours ago
Joined: 01/04/2014 - 06:47
Posts: 4954
Location: H-Town

locusto03 wrote:

This is the main reason I’ve been programming Qlite drivers with a High of about 50% and a Turbo of 100% with a 60 second timeout. The subjective difference in output isn’t a whole lot for my applications. A thrower would probably show the difference in output better.


Ding Ding Ding…I have found the Convoy S* series pills can handle about 2 minutes of “Turbo” before throttling back to approx 50% PWM, wide open with 8*7135 it hits right at 2.8A, on high (the turbo timout fallback) it runs approx 1.29A but the light doesn’t dim considerably…I have the 8*7135 to push the light out but the auto fallback is what keeps you from burning it up, also putting all my emitters on copper sinkpads and expoxying them to the pill helps too (and some artic silver thermal grease in the threads help too)

I started with a 6*7135 and wide open it ran warm…not hot, upgraded once I got STAR V1.1 compiled to how I wanted it

nquinn
Offline
Last seen: 5 hours 27 min ago
Joined: 09/11/2017 - 21:55
Posts: 19
Location: DC

WarHawk-AVG wrote:
locusto03 wrote:

This is the main reason I’ve been programming Qlite drivers with a High of about 50% and a Turbo of 100% with a 60 second timeout. The subjective difference in output isn’t a whole lot for my applications. A thrower would probably show the difference in output better.


Ding Ding Ding…I have found the Convoy S* series pills can handle about 2 minutes of “Turbo” before throttling back to approx 50% PWM, wide open with 8*7135 it hits right at 2.8A, on high (the turbo timout fallback) it runs approx 1.29A but the light doesn’t dim considerably…I have the 8*7135 to push the light out but the auto fallback is what keeps you from burning it up, also putting all my emitters on copper sinkpads and expoxying them to the pill helps too (and some artic silver thermal grease in the threads help too)

I started with a 6*7135 and wide open it ran warm…not hot, upgraded once I got STAR V1.1 compiled to how I wanted it

If you were going to buy a new convoy s2+, would you go with 4×7135 or 6×7135? 8x seems to run too hot.

Jack Kellar
Jack Kellar's picture
Offline
Last seen: 4 hours 46 min ago
Joined: 03/09/2014 - 14:21
Posts: 950
Location: Throwing bolts at anomalies

nquinn wrote:
If you were going to buy a new convoy s2+, would you go with 4×7135 or 6×7135? 8x seems to run too hot.

Now that we have the new and amazing Biscotti firmware, I see no reason to get below 6×7135. At all.