Thrunite TC20 - XHP70 - 3800lm

BTW, on this particular light I was comparing Newlumen and Robo819’s results. I wish they could test on the same battery.

On High:
Robo819 1710
Newlumen 1550
Difference of 160. That’s 9.4% lower than 1710 and 10.3% higher than 1550.

I’m not sure what turbo readings to compare. Maybe:

Robo819 3560
Newlumen 3090
Difference of 470. That’s 13% lower than 3560 and 15% higher than 3090.

So maybe Robo819’s sphere is only 10% to 15% higher than the TA Tube, not 20% to 30% like I said earlier.

Are there any diffusers in the tube or is it a straight shot from end to end?

No diffuser in it. Dale would probably know more than me about how the calibration was all done though , because he is friends with the one who made the tubes.
Don’t know if you been here long enough to remember “manxbuggy1” or not but he made the tubes.

Edit: one thing I didn’t point out is that these tubes are made with a 1 piece P-trap and not made with several 90’s pieced together also. I don’t know if that plays a factor or not but thought I would mention it.

I think i might have did something wrong in my previous test… i am not sure… i will make sure the light is align with the disc… i will report back this afternoon…

I welcome this discussion here as it pertains to the TC20.

I’m not doubting the validity/accuracy of DB, Tom E and Robo’s tubes or anyone else’s, I’d just like to know why my TC20 doesn’t appear to have near the rated output.

I remember reading years ago about manxbuggy1 making those tubes and I’ve always trusted the readings posted by DB, Tom E and Robo.

Since I’m getting what appears to be proper figures for all the other lights I’ve tested, I’m thinking my TC20 may have something wrong with it. As soon as I get some time I’m going to test it further with a Lii-50A and also test the lower modes.

.

I am curious if Dale or Robo got similar figures to mine for these two lights? (If they have them):

(My readings):

BLF Q8 (stock)…………4865lm
Sofirn C8F (stock)……2338lm

(Both tested @ 30 seconds with fully charged 30Q cells

This is what I remember, the foam cutouts and the light sits on a piece of glass? (I vaguely remember reading it may have been a special type of glass?)

I like this setup, it seems it would be a bit easier to use and every light tested sets at the same place.

.
With the TA tube we have to hold the lights in through a open hole and move it up and down to get the “best” reading (no glass to set the light on)

There’s no foam so some light escapes around the opening if the head of the light isn’t exactly the same diameter as the hole.

I’m not complaining, it’s not really a big deal, but the glass/foam version would seem easier to use, especially for testing longer runtimes, 5-10 + minutes, or even 30-60 minutes, can’t really hold the light for that long.

I did the turbo mode test last night and got 3192 lumen…

I tested it again this morning with fully charge battery and make sure the light is align with the disc… I got 3220 lumen…

So my last two turbo mode test are 3192 and 3220 lumen.

I think it’s pretty clear that the TC20 specs are overrated. The driver is just not producing the power to get 3800 lumen at 30 secs.

The NW emitter might be a bit lower in output with this type driver. I think Robo819’s tc20 was CW. Does Thrunite mention what emitters they use? They might be using a NW that is 1 or 2 bins lower in output over the CW. Who knows?

The factory protected battery might also be playing a role.

2960 lumen is pretty darn close to 3087. It’s only 4% less. TA’s tubes are only rated at plus or minus 5% so if Newlumen’s is reading 2% high and yours is reading 2% low, then that could also explain it.

I would not worry too much. We will never get exact numbers.

Do see if a better battery helps.

Geez, now you mess up my numbers. Lol.

Okay, so now Beamos light is measuring about 8% lower than Newlumen.

What is Crees 70.2 deviation per bin, plus or minus 7%. We are still in that neighborhood.

A good battery might get Beamos numbers higher. Even a good old 30Q will probably get higher output than the protected 26650. At least a lower amp draw, that is for sure.

Yes mine has the CW emitter in it………and if you have been arguing that my test are wrong at 30 seconds then that is correct……

I got 3791 AT START and 3560 at the 30 second mark , as noted in my charts I put in this thread and in my review.

Yes we have the foam that covers the opening where the light sits and various size holes in those pieces so that the head of the light sits down on the glass and the foam puts each and every light to the same spot in the center and I don’t touch anything when taking the reading , other than pushing the switch to get the next mode.

I don’t have the Q8 , but my 30 second reading on the Sofirn C8F was 2551 with the Sofirn cell.
I actually got higher readings with the Sofirn cell than the Efest purple high drain 3000mah and slightly lower than the LG HG2.

I was saying the Thrunite specs of 3800 are just plain overated/exagerated. I’m sure you agree.

I think ALL the companies are exaggerated if you go by what the actual lumens are at the 30 second to 1 minute mark compared to the ANSI rated lumens they list.
Very few lights (if any) will hold the rated ANSI lumens that they list for a very long from the start.

So when you see me saying a certain light makes X amount of lumens at start on my light tube , that is just what it is that I am saying , it makes that number when I turn the light on at the very start and I always do my test with a freshly charged cell.
That is why when I test the lumens and make my charts , I always try to list the start lumens and then also what they drop to at the 30 second mark , which is a more real number to me as to what you will see from the light in the first minute to 90 seconds of use.

So yes the 3560 lumens I am showing at the 30 second mark for the TC20 , is a more realistic number than the start lumens I showed when doing my testing.

Your not telling me anything I don’t already know.

I do find it interesting that you seem to suggest that ANSI-FL1 specs are measured at turn on. They are not. The spec specifies it’s measured at 30 seconds.

Maybe you already know this, but from what your saying, I can’t tell.

I pretty much ignore turn on lumens. It seems like a useless stat. I might record it, but I rarely post it.

Ps, I do pretty much think all companies exaggerate their output numbers. Some more than others. A rare few might even be accurate. I saw a Sofirn C8A that measured right at 1800 lumen when the factory rated it at 1750 and an Astrolux C8 that measured 1380 when the factory rated it at 1300. That’s pretty impressive considering these TA Tubes seem to measure lower than existing tubes.

No I am not saying they get the ANSI readings at the start , they take X amount of lights and get an average from all those lights from what I understand , to get their numbers and that is why I show the 30 second readings also. So in reality many of the lights are at different lumens coming from the factory and there is NO absolute lumens from any certain run of lights.

Jason, you have a TA light tube? And you’re in Houston, right? I have an Emisar D1s coming that I intend to give to robo, the plan is to flash Anduril onto it so he can experience the candlelight and lightning modes from TK… I’ll get a reading on it, send it to you or robo…. maybe just send it straight to you and you can get a reading on it, then forward it to robo and he can read it too… I’ll leave the cell in it and we can all 3 get a reading on the same light, same cell, with the only variable being how hot the chargers finish it up. That would give us 3 readings, 2 boxes built by the same people… robo’s box is actually the first one manxbuggy1 and his brother built several years ago, then they built TomE’s and then mine and then Richards.

Do you know where Maukka is located? Maybe we could send him this D1s and get a 4th read on it with his sphere and see how all the numbers pan out…

Yes, I have a TA tube and am in Houston. Maukka is in Finland, though.

Why not send it to Texas_Ace? He is just north of you near Dallas and he has the lights that Maukka measured for him plus his reference tube which mine was built from. So he can do a more accurate comparison.

You might even send a cheap light to measure as long as it’s output is stable. The D1s, being a FET driver, is going to fluctuate based on voltage. It may not be the best choice for steady output.

It would be cool if you guys collaborate.

I do plan to start a thread on the different integrating spheres, just not right now. I will mention that TA’s tube differs from most in that it uses 3 diffusers to help collimate the light and give better accuracy between throwy lights and floody lights. Here is the thread on them. Texas_Ace BLF Calibrated Lumen tube / Sphere No math skills needed - Several spheres still available

In fact, I think he has a few left for sale if anyone wants to buy one.

Here is the TA Lumen Tube.

TA adds reflective tape in certain areas to get the meter both calibrated and to have the lux readout on the meter be the actual lumens! No math involved. We are still working on that last part. Right now we all have to use math to multiply the readout by .68. This is being fixed, though.

The TC20 is a boost driver, drops output quickly, steps down, has all sorts of issue maintaining output from what I can tell. Not sure how that’s any more reliable than an FET direct drive light that would drop from a full cell in a reliable straightforward manner.

When the 12” styrofoam sphere first came out here I bought one, didn’t like it, sent it to TA. He found it had limitations and that’s why he made his PVC light tubes. He made em from smaller pipe for reasons of shipping costs. He also used connectors, pieces, instead of the one-piece style P-Trap ours are made of. So yes there are some differences, but in the end the averaging of many lights to get our multiplier evens it all out. There was a great deal of thought and effort went into these, we’re not using a shoebox in the bathroom or anything like that.

I wonder what would happen if someone like Richard got in a stock of a light like the TC20, several hundred of em or whatever, and proceeded to measure a great number of the same light, what kind of differences would he see?

I’ve got it now. :wink:

I would not say the TC20 is a stable output light either, at least not on turbo. It’s lower outputs might be pretty stable. Maybe a 7135 based light would be nice and stable? IDK.

And back to the topic :slight_smile:
I decided to change XHP70.2 on MT-G2. As soon as the parcel reaches me and I swap, I will pass the report. I could not take it anymore with this tint shift.