KD (kaidomain): Deals and new products thread

… and CRI99? Wellp, CRI90+ is 0K. :-)

Hello.

I was going to submit some quick comments about the P4000 buck driver in Kaidomain's website; I used the opportunity to review the comments on the K6 flashlight host in Kaidomain's website, but found :!!!: none. Weird, because I perfectly remember having submitted a review of it…

Concerning the K6 host I can recall I gave it a 4/5. It has excellent heatsinking capacity. The tail switch is a big forward type and there's no straightforward easy way to convert it to reverse switch, although it can be done namely by DIYing a new switch centering ring out of some material. The Omten 1217 is a great choice for it, I used one.

According to what BanL told me past september very few hosts remain. I know this because I talked with BanL about getting a new reflector for the K6, and he said no way among other things. I finally realized the K6 uses the same reflector as the Convoy L2/L21A, so I got my replacement from Convoy.

About the P4000 driver, in all honestly I have not even powered it on. However, its sense resistor stack consists out of 50mΩ plus 1Ω sense resistors in parallel, which is 47.619mΩ combined. So, when 4A of current go across the sense resistor stack the voltage drop is 190.476mV and the power loss is 761.9mW. For 5A, voltage drop is 238.095mV and power loss 1190.5mW. These are huge figures. I can understand whoever designed the driver probably thought “it is good enough”, but I don't. In my opinion a switching driver needs to be efficient, minimizing undue heat dissipation, and this driver fails in this respect. This driver uses about the same sense voltage when in high/turbo as typical AC led driver. Using a 0.2+V sense resistor stage in an AC led driver outputting tens of volts is 0K, but using something so gross in a small driver meant to output not much above 3V is… inadequate. Lazy job from the driver designers, imho.

Like said before, any chance for CRI90?

BanL, could you comment on that?

I don’t usually take stuff like this to public forums, but KD has been a real pain for me lately. They don’t notify me of shipping (also don’t change the status on the website), sometimes appear to not ship stuff for ages, and don’t appear to have shipped one particular order I made a long time back. They don’t respond to my emails to their official contact address and don’t respond to messages using their contact form on the website.

Is there some secret to getting a reply from KD, just like Neal only ever responds on exactly one channel? Should I just get Paypal involved? I’m really frustrated by now. Am I the only one?

I have had similar problem with KD two months ago with not received package. I escalated Paypal dispute and there they didn’t respond at all there. They responded to my email once. Finally Paypal refund me after I waiting one month for that from them, so just open Paypal dispute and escalate it if you don’t received answer from KD. After that case now I buying from Aliexpress KD shop. It is more expensive but I have better protection compared to PP.

This is very discouraging. Ignoring all their customers puts their business in jeopardy. I stopped ordering from KD because of it and was forced to file PayPal disputes to retrieve my money. What a massive pain in the ass and a huge waste of time! Never again! Maybe its time everyone does the same (and I hope you do). Perhaps some day they will be able to reopen and win back our trust.

Good luck BanL (I can see you lurking online!!!). I truly hope you all remain safe. For now and into the foreseeable future, I WILL NOT waste another moment with your company and warn all others to purchase elsewhere. You can stop cowering in your closet now…

I’ve now opened a case with Paypal and within seconds I had my money back. It appears KD has been put on some sort of shitlist by Paypal already.

When I had a customer request a long time ago, they had excellent skype support in real time. Not sure if they still offer that, but it was a time saver.
BanL seems very active on here as well, is there an agreement that we can PM him or something?

+1

Could you please get 4500K/5000K 90 CRI and 3500K 90 CRI of either XHP35 HI or XHP35.2 HI?

My 1st order (during CNY) 11.30 am just arrived today, 7 days from KD to the UK. Stellar service as per usual.
There is an issue on the site, which may cause problems for some - allowing you to add to basket, and buy items that are not in stock/pre order, this may (just suggesting as I’ve not noticed myself sometimes) cause some issues. May be an idea BanL to make it impossible to add OOS items to the basket, to save any confusion. :+1:

indoor sorting - ready for delivery——Delivery Office
14
2021/03/04 05:43:10
inward mc - handheld acceptance——Inward MC
13
2021/03/03 23:10:06
outward rdc - handheld acceptance——Outward RDC
12
2021/03/03 05:29:21
delivery to local courier
11
2021/03/03 08:00:00
Departed from Facility
10
2021/03/03 08:00:00
Arrived at Sort Facility
9
2021/03/02 13:19:44
Custom clearance completed
8
2021/03/02 13:05:53
Arrived at AIRPORT of Destination,clearance in process
7
2021/03/01 14:17:00
Departed from AIRPORT of Origin
6
2021/02/28 17:51:00
Arrive at international airport to abroad
5
2021/02/27 20:00:47
Departed Facility In processing center
4
2021/02/27 10:57:55
Departed Facility in SHENZHEN ——SHENZHEN
3
2021/02/26 19:53:00
Arrived at Sort Facility SHENZHEN ——SHENZHEN
2
2021/02/26 14:39:36
Shipment information received
1
2021/02/25 18:09:40

Wellp, I was going to submit a P4000 review in Kaidomain's site, but it complained “Warning: Review Text must be between 25 and 1000 characters!”. It also complained about name too long (max 25 characters), LMAO. My review is well above 3K characters long LoL, so here it goes:

With a sense resistor stack consisting of 50mΩ plus 1Ω sense resistors in parallel (47.619mΩ combined) and driving currents of 4 and 5A in high and turbo modes, by Ohm's Law (V = I × R) and electric power (P = V × I = I² × R) formulas it is easy to infer the voltage and power drops across the sense resistor stack: 190.476mV / 761.9mW at 4A and 238.095mV and 1190.5mW at 5A. These are HUGE figures, causing a lot of wasted heat and a massive drop in efficiency.

When a typical AC led driver which outputs many tens of volts uses a ≈0.2V sense voltage, the impact in overall efficiency is no more than very small. When a driver meant to output a little above 3V uses ≈0.2V sense voltage, the relative impact is huge with ≈6% output power loss (!) just in the sense resistor. This is gross and inadequate in a driver like this. Lazy job from the driver designers.

Still, a driver which does its job. For 2 cells in series can be adequate.

Here it ends the text I submitted in the Kaidomain's page (limited to 1000 characters). I also add the following:

Edit.- Of course, by Ohm's Law the sense voltage gets lower in lower modes (V = I × R), and so the negative impact of it is proportionally lower in these modes. But still, lazy job tu use such a coarse sense resistor stage. Good drivers employ just ≈50mV (≈0.05V) maximum (which can still be improved by using multiple sense resistors, namely as the current output raises).

I do not mean to say that the driver cannot work, but the absurdly high sense voltage (even if not much in practice) unnecessarily restricts regulation time when powered by single cell voltage, and the increased heat dissipation on top of a (undesired by power users) “smart thermal management function” means the stuff is ultra-likely to taper down the current output rather quickly in high modes.
Single cell overdischarge cut-off, at 3V, is high. Considering that in a flashlight the driver does not see actual cell voltage, but after it goes through springs and switch losses, actual cell cut-off happens above 3V at the cell terminals; also, considering that most li-ion cell manufacturers specify 2.5V as cut-off voltage in their cell datasheets (and down to 2V in a few exceptional cases), there's no good reason for a 3V cut-off. And yes, I know that what happens at 3V is the stuff starts flashing the led (low voltage warning), with actual, real cut-off happening at 2.7V. But this is senseless, stupid if you ask me. Just imagine for a moment that I am using the flashlight for a walk or jogging: do you think I'd be happy if the flashlight starts flashing the led while I am running at night? Certainly NOT. I'd very much prefer the flashlight low power blinking just a few times to signal the low voltage warning, and then dropping to the lowest available mode. In the lowest available mode, the flashlight can still serve me while it discharges the battery down to let's say 2.5V or even 2V (so, effective cut-off at 2.5V or 2V). That can be a useful lifesaver while I am walking or running at night; blinkies… not so much.
So, in my opinion the low voltage warning can be set at 2.8V or 3V, with a few blinks to signal the warning, then dropping to the lowest mode and continuing the discharge down to 2.5V or 2V. At the lowest mode (just 5mA output in this driver), the driver could still provide a good deal of light time while it hits 2.5V or 2V. In any case, a standard white led (blue pumped phosphor converted led) has a minimum Vf of 2.5V, and this means that while powering a standard white led the led poses a hard cut-off limit: 2.5V. Red or other types of leds have a lower Vf, but this is another story.


All in all, my personal evaluation for this driver is 3/5. Which isn't bad, but considering how fearful these fellows are of anything below 5/5 or even 4/5, we'll see if this (the above part I mean) gets published in Kaidomain's site.

Thu, 03/04/2021 - 14:56

All that aside and since you’ve obviously tested it, ( I am not disputing your findings btw, it’s just over my comprehension and unlikely to have much real world effect on my usage) - how accurate are the claimed amps (Moon (5mA) > Lo (200mA) > Med (1500mA) > Hi (4000mA) Double Pressed: Turbo (5000mA)? Particularly the Hi 4000mA & Turbo 5000mA (which is what it will be on more than anything else) since I’ll be using it for an Osram KW CSLPM1.TG and importantly does it work as effectively with 1 cell as it does 2 cells? if you could give some findings on those I’d appreciate it very much since I can’t measure it. Thank you Barkuti :+1:

Seems like in a Fenix bc25r fairly smooth slicing of SST-20 just adds some artifacts to the hotspot while size stays the same and reduces Lux output from 1000Lx to 600Lx@1m.

Just received after second order this time from AliExpress KD store the K1 host. It is very nice looking host with deep SMO reflector. The light is little longer. I didn’t like the o-ring under SS front bezel. It is too thick. At bottom side of bezel there is little channel which need to hold and press oring to the glass. But included oring is too thick for that.





Hi everyone, has anyone tried this so far?
BanL, may I know does it live up to 6000-7000 OTF lumens? What is the current drain at tail?
JKK90 LUMINUS SBT-90.2 LED 7000 LUMENS WHITE 6500K 2-GROUPS MODE LED FLASHLIGHT ( 2X26650 )

Thank you.

Sbt90 lights only really emit around 6000 or less. The K75 has usually been measured to be less than 6000 by most reviewers. Anything higher is not true imo.

Yes, If this JKK90 produces somewhere around 5000 to 6000 it should be considered as great enough.
But I have some doubts about it since there is no review about this light, and high current (~20A) buck driver for SBT90.2 is rare to none currently for none branded lights, or did I miss anything?

I still got one package from Kaidomain missing from last year its been almost a year since i made that order so i guess its gone. Not there fault its an issue with the shipping company here but my money is still gone as the package was sent back to China.

My last order from Kaidomain was good they have changed the shipping method to Australia and i received the order in about 3 weeks, that’s great to see something suss was going on with Aramex shipping company.

Custom kd leds
7070 hi

P35 hi