Driver needed: Cree Q3-5A, 1000mA, 2NiMhs

[rayman]

I want to mod a MiniMag with a Cree Q3-5A and want to drive it with 2 NiMhs at 1000mA. I searched at the Sandwich Shoppe and found three drivers:

• Badboy 1000mA Step Up Converter Board
• BB Nexgen 1000mA
• GD 1000 Boost/Buck Conveter

So which is the best for my application?

thanks
rayman

 

[Justin Case]

This post by Wayne seems to suggest that the BB Nexgen is the best choice, at least when choosing between the BB and BB Nexgen. Not sure why you need to consider the GD if you are going to run only with 2xNiMH. The buck capability seems superfluous. If you are also considering possibly bucking down a 1xLi-ion, then the GD makes more sense. But why would you want to use only a single AA Li-ion (plus presumably a AA spacer) instead of 2xNiMH in your Mini Mag host?

 

[rayman]

Yes, you are right that would just shorten the runtime. Thanks for your link.

rayman

 

[metlarules]

Crees data sheet says max current for neutral white is 700mah. I don't know that you have enough heat sinking ability in a mini mag to run it at 1000mah and have it live long. Click the link below and look in the characteristics section.
http://www.cree.com/products/pdf/XLamp7090XR-E.pdf

 

[mudman cj]

I believe the issue with driving neutral white emitters over 700mA is that higher currents lead to degradation of the phosphors. But who knows at what rate that will occur at 1000mA? :thinking: :shrug:

 

[yellow]

inside a MM one cant even get enough heat away for 500 mA,
(but do we care if the led gets to 50 % after some 200-500 hours?)


When I used those drivers, MMs and BBs worked good every time, while three Nexgens gave bad results (I wasnt even able to understand why). 2 died for no reason I could cee then, one simply did not give the power.
Before and after, normal BBs still worked great with any project, I feel them being more "rugged".

Try one of both.

 

[Justin Case]

There is this homemade AA Mini Mag heat sink, which the OP appears to have already purchased. The link shows the heat sink equipped with a Luxeon K2 and an SOB 1500.

 

[TexLite]

I want to mod a MiniMag with a Cree Q3-5A and want to drive it with 2 NiMhs at 1000mA. I searched at the Sandwich Shoppe and found three drivers:

• Badboy 1000mA Step Up Converter Board
• BB Nexgen 1000mA
• GD 1000 Boost/Buck Conveter

So which is the best for my application?

thanks
rayman

The GD will be the best choice. According to Wayne, while the BB/NG may work with fresh cells, once the light has been cycled and the battery voltage drops the BB/NG won't start again in regulation.

Also, with the above configuration(4.35W), the input current at 2.4V will be 1.8A, above the 1.5A max recommended input current for the BB/NG series.

-Michael

 

[Justin Case]

The switch current spec for the TI TPS61030 regulator IC used in the BBNG is 4000ma (typical) and 4500ma (max). Not sure what the effective input current limit is when the IC is implemented on a driver board.

You are correct that the BB input current limit is more like 1.5A, plus Wayne states that the BB gets inefficient above 1A.

Regardless of the switch current though, I'm not sure why the GD is the preferred board. How does the GD overcome the issue of cell voltage sag and not starting up in regulation for non-fresh cells? Also, isn't the GD also limited to the same ~1.5A input current limit?

I'm using a BBNG 1000 driving a Seoul P4 U2SW0H-bin LED and it works pretty well. However, I measure 1.2A at the tailcap when using one SF123A cell. If the BBNG is 80% efficient and assuming 2.5V for the 123A under load, that translates to about 700ma drive current and 3.4V Vf. For 90% efficiency, it could be more like 800ma drive current and 3.45V Vf. So it looks like I'm not quite sending the full 1000ma to the LED.

 

[TexLite]

The BB/NG were originally rated for 2A input, see here, but the IC was changed by the manufacturer which resulted in the current capacity being reduced to 1.5A.

Unless I'm mistaken, the GD dosen't have the same 1.5A input cap that the BB/NG do. It has been used in similar setups with success.

There were a few threads a year or two ago, along with many individual member posts when many of us were looking for a 1000mA driver with the same setup. That was the outcome IIRC, with some of the information coming from Wayne and some from other members. The search function won't allow a two or three letter search, and Google's shotgun approach isn't much help, but I'll see what I can come up with later when I have more time.

In your setup, I'd go with a GD1000, they have great regulation and will completely drain a cell.

The VIP1000 was another great two cell boost driver,but its no longer in production.

-Michael

 

[Justin Case]

It may be that the BBNG is not running in full regulation, hence my getting less than 1000ma output for a Seoul P4/BBNG1000/1xSF123A setup (which, to stay on-topic, should be close to the voltage under load for 2xNiMH):

http://www.candlepowerforums.com/vb/showthread.php?p=763314

Interestingly, Wayne states that safe mode input current is 1A while I measured 1.2A for my setup.

Perhaps using an RCR123A instead of the SF123A primary would kick the board into full regulation. Similarly, maybe you can use 3xNiMH instead of 2xNiMH to get full regulation.

Anyone know if the BBNG can take an AW16340 driving a Seoul P4 U2SW0H-bin LED? The Shoppe product description says that the BBNG can take 0.7V of input overvoltage above Vf, so it seems possible (though close to a fully charged Li-ion's 4.2V voltage). But I hate to fry a functioning driver board/LED setup. I suppose one could use a less-than-fully-charged Li-ion -- say 4.0V resting voltage.

Yes, I have also seen comments by Wayne and others that the BB used to have the 2A input current limit, and now it is ~1.5A. But it is unclear to me that this limit applies to the BBNG. The link you provided says that 1.5A is the safe mode current limit for the BBNG. That is different from the max limit for the input current. The TI regulator IC in the BBNG implements a soft start to protect the IC from shorts. The boost switch current is set to 40% of the nominal value to avoid high peak currents at the battery during startup.

40% of 4000ma turns out to be 1.6A, which is probably where the more conservative safe mode 1.5A figure comes from.

The first link listed below states that the IC in the GD board is spec'ed at 1.2A input current.

Another issue is heat generation and dissipation. Even if one of these 14mm boards can take >1.5A input current, the extra heat generated at these higher input currents has to go somewhere. So now thermal management becomes the limiting factor.

Here are some links:

http://www.candlepowerforums.com/vb/showthread.php?t=221244

http://www.candlepowerforums.com/vb/showpost.php?p=2851722&postcount=8

http://www.candlepowerforums.com/vb/showthread.php?t=215680

http://www.candlepowerforums.com/vb/showthread.php?p=2171782

http://www.candlepowerforums.com/vb/showthread.php?p=2551390

http://www.candlepowerforums.com/vb/showthread.php?t=159052

http://www.candlepowerforums.com/vb/showthread.php?p=2414048