Input current limits for Badboy and SOB boards

I see that the input current limit is stated as "about" 1.5A, based on some sort of "internal switch current limit". I assume that this internal switch is the IC that does the constant current regulation?

Post #10 in this link says that if you load a Badboy such that the board needs more than Iin,max (i.e., the input current limit) to deliver the power being demanded by the load, the board will deliver only as much power as it can up to its input current limit. It is also stated that you might get lucky and the board's Iin,max could be a little higher than 1.5A.

In the example given in Post #10 in the link above, is there any reason to spec out a Badboy 400 instead of a BB 750 (assuming that both boards have a max input current of 1.5A)? Is there any risk of overtaxing the BB 750 if you present it with a load that requires >1.5A input current? If not, then it might be better to go with the BB 750 since you might get lucky and get an IC that can handle slightly higher than 1.5A on the input.

How about the SOB? Same performance characteristics in terms of Iin,max behavior? Any risk of overtaxing that board?

Thank you.

Several of us are running SOB 1500's without issue. An SOB is a buck driver and input is not appreciably above output.

The BadBoy is a boost driver, and not having ever used them I can't answer what the max input would be. I use the Blue Shark and a heatsink. It is rated for higher than 1.5 amp input. Edit: The Sandwich Shoppe claims here that the Badboy boost converter has a 1.5 amp max input.

I am running a MC-E in 2S2P using a BB1000 no problems so far.
2 primary CR123 for power in a old PM-6 host with the LED module.
The BB is potted in AA epoxy. I'll write it up when I get the reflector issue fixed. The McR-27 won't go deep enough to focus as I used a Thin Cree PCB. It had better wiring holes than the 650MCPCB.

Greg G said:

Several of us are running SOB 1500's without issue. An SOB is a buck driver and input is not appreciably above output.

The BadBoy is a boost driver, and not having ever used them I can't answer what the max input would be. I use the Blue Shark and a heatsink. It is rated for higher than 1.5 amp input. Edit: The Sandwich Shoppe claims here that the Badboy boost converter has a 1.5 amp max input.

Click to expand...

Yes, I've searched on CPF as well and the 1.5A limit comes up often. But it is less clear what happens if LED power demands result in an Iin greater than Iin,max. The specs for the regulator IC IIRC actually say that 1.5A is the lower limit, 2A is the typical limit, and 2.5A is the upper limit for Iin,max. I've also seen stated that the board will simply take what the batteries can give it, up whatever the Iin,max is for the IC. My concern is if I would be overtaxing the IC in some way if I overspec the board's Iout, or if this apparent self-limiting behavior is perfectly fine (assuming also that the LED can take the current being sent to it).

Good point about buck vs boost. However, if the Vbatt:Vdrop ratio isn't sufficiently large for the SOB board, you can end up demanding a fair amount of current. When the ratio is 2:1, for example, then the input current gets nice and low.

For an MC-E in 2S2P and using 2xIMR16340, for example, Vbatt ~ 7.6V nominal, while Vdrop ~ 6.5V nominal. If I used an SOB 1500, I think I might have Iin > 1.5A. Let's see:

Power from the batts = power to the LED
board efficiency * Vbatt * Iin = Vdrop * Iboard
0.8*7.6*Iin = 6.5*1.5 = 9.75W (assuming 80% efficiency)
Thus, Iin = 1.6A

If board efficiency is higher, then Iin decreases. Same with Vbatt. If Vf is lower (lower Vdrop), then Iin also decreases.

What batt configuration are you using with your SOB 1500? As per nein166's post above, my application of interest is also to hook up an MC-E in 2S2P. I'd probably use an SOB to run 2xLi-ion (e.g., IMR16340 or 2x17670) or 4x123A.

I am also looking at 2xNiMH (C-cell sized) to drive a SSC P4 U2XW0H-bin, and thus my interest in the BB as well (say a BB 1000). Do you think that the BB Nexgen is a better choice?

For the BB 1000 driving an SSC P4, I calculate:

PLED ~ 3.5V*1A ~ 3.5W

Assuming 80% BB efficiency,

Iin ~ 1.8A

Since Iin,max seems to vary (1.5A appears to be a conservative limit, with 2A typical), let's assume that the BB 1000 will work (assuming proper heat management).

Thus, if we assume Iin,max ~ 2A, then

Vin ~ 2.19V

So, it looks like 2xNiMH would even run in regulation with a BB 1000 (if the BB can take 1.8A on the input side).

For nein166's MC-E in 2S2P,

PMC-E ~ 6.5W

If we assume 80% BB efficiency as before, and Vin ~ 5V (for 2x123A), then

Iin ~ 1.6A

So, there is at least one data point saying you can exceed 1.5A input. nein166, is your BB 1000 potted or heat otherwise managed somehow?

What tailcap current draw are you guys actually seeing?

For my SOB 1000 driving a SSC P4, I've measured 0.5A draw, suggesting a real world board efficiency higher than 80%.

My thinking was to pot the driver cavity of my AW tower module where the SOB or BB would be installed, filling the cavity with thermal compound, also covering the board with compound, and then squishing the board into the cavity. Thus, the board and its components would have a thermal path through the compound to the tower module/heat sink (and to the metal flashlight bezel since the tower will be in tight contact with the bezel's reflector).

Does that sound like it might work in terms of heat management?

You didn't mention wanting to run an MC-E off an SOB in your first post. I wouldn't even try it. You need a Shark Buck.

I've never tried to run a Seoul P4 @ 1000ma off only 2 NiMH batteries, so I don't know which board would be best. I don't like to be in a situation where I have to boost the voltage that much. I would prefer to use a FiveMega or Modamag battery holder that holds 6-8 AA's and use a good buck converter, like an SOB.

About 1.6A at the batteries, gotta take care as when the cells get low I start to hear a buzz from the circuit. The 2xCR123 were at 5.6v but still only 1.5A I put fresh cells in and the output doubled.

Greg G said:

You didn't mention wanting to run an MC-E off an SOB in your first post. I wouldn't even try it. You need a Shark Buck.

I've never tried to run a Seoul P4 @ 1000ma off only 2 NiMH batteries, so I don't know which board would be best. I don't like to be in a situation where I have to boost the voltage that much. I would prefer to use a FiveMega or Modamag battery holder that holds 6-8 AA's and use a good buck converter, like an SOB.

Click to expand...

Can't fit a Shark in an AW tower. Need an 0.55" diam driver.

I wouldn't use an SOB 1500 anyway. Generating ~10W is a lot for such a small board. I'd consider an SOB 1000, which I estimate would generate about 6.6W and have to dissipate a bit over a watt as heat. That's still a lot, but probably about half the heat dissipation of a 1500.

nein's experience with his BB 1000 gives me some hope. If I cook the SOB 1000, maybe I'll try something a little lower in output.

I conservatively rate most of the converters.

In some cases you can drive them harder with lower efficiencies assuming you can keep it from going nuclear (heat wise).

The BB IC specifications is: 2A. But, the manufacturer re-did the die several years ago and the idle current specification was reduced. This resulted in the internal power transistor not being driven as hard and it would not sustain the 2A drive level. The package was also limiting the total power you can get out of this part along with the inductor ripple current reduces the total design to around 1.5A. But, the BB is very very inefficient over 1A.

The NewBB is way more efficient for a 3V to single LED driver. If at all possible use the newbb. No heat and the most efficient driver on the market.

As for the SOB. The IC is rated for 2A. The IC packaging and as mounted on the SOB combined with a possible issue with high value feedback resistor network has made the current SOB rating a safe 1.2A. Between 1.2A output and 1.5A there is a 50/50 chance it won't work due the high value resistor network interference with proper operation. I have had plans to investigate this and use lower value resistors for the network. This same IC is used in the SPY light and can easily do 1.5A with no issue. The SOB should do 1.5A, but, due to some high Z resistors may cause stability issues above 1.2A.

Does this help?

Wayne

Do you think a BB Nexgen 750, 850 (if there is a resistor combo that gives this output), or 1000 would run in full regulation using 2xNiMH to drive a Seoul P4 U2SW0H-bin? I would also assume that the BB NG 1000 might require too high of an input current. What happens then? Does the 1000 simply take what the batteries can deliver? Any downsides to that, such that I should spec out an 850 or a 750 instead? Thanks.

the NGBB has a startup current limiting feature for shorted outputs.

Since the IC was designed to regulate output voltage they added a protection feature to keep the chip from killing itself when it starts up.

This startup kicks in when power is first applied and output regulation is not met.

The output is reduces or restricted. This is not really a bad thing.

For a 2AA battery configuration on fresh cells and with todays lower Vf LEDs you probably can get 1A regulation when you power first power up more than probably 40% of battery life.

If you power it up and it comes up in full regulation it will maintain full regulation till there is no more juice. Under this constant on situation you would get full regulation till the batteries give out or the IC hits ~1.8V.

The rule of thumb on the Luxeon3 series or era was 750mA tops if you turn the light on/off all the time and needed full regulation. Nowdays this number could be higher and it now depends mainly on the Vf of the LED at the desired current.

Even under limited mode, you will still be driving the LED at ??500mA or so which is still bright.

Wayne

Would it matter if the NiMH cells were C-cells, instead of AA, in terms of starting in full regulation? When the voltage falls below ~1.8V and presumably out of regulation, does the light still work (at dimmer level), or does it cut out completely?

Justin Case said:

Would it matter if the NiMH cells were C-cells, instead of AA, in terms of starting in full regulation? When the voltage falls below ~1.8V and presumably out of regulation, does the light still work (at dimmer level), or does it cut out completely?

Click to expand...

The determining factor on restart for any battery is the re-coup level after you turn it off. How much it recovers varies depending on if it is Alkaline, NiMH or other chemistries and how hard it was being used (temperature of cell).

I believe the newbb drops into a moon mode effect below 1.8V. This is the same principle used with the 5-7 ohm MCE2S switch.

If I recall somewhere someone noted that the internal resistance of AA cells were actually better than C cells and could handle higher discharge rates. I could be wrong on this. Don't quote me...

The most dominate factors in starting in full power is the ability of the battery to not sag for the first fraction of a second, the actual battery voltage under this load and the Vf of the LED. (applicable to the newbb only).

Wayne

dat2zip said:

The NewBB is way more efficient for a 3V to single LED driver. If at all possible use the newbb. No heat and the most efficient driver on the market.

As for the SOB. The IC is rated for 2A. The IC packaging and as mounted on the SOB combined with a possible issue with high value feedback resistor network has made the current SOB rating a safe 1.2A. Between 1.2A output and 1.5A there is a 50/50 chance it won't work due the high value resistor network interference with proper operation. I have had plans to investigate this and use lower value resistors for the network. This same IC is used in the SPY light and can easily do 1.5A with no issue. The SOB should do 1.5A, but, due to some high Z resistors may cause stability issues above 1.2A.

Does this help?

Wayne

Click to expand...

Very helpful.

I went with a BB NG 1000 to drive a Seoul P4 U2SW0H bin and the combination works great. Virtually zero heat-up of the tower I assembled for a SureFire TurboHead (I potted the driver using Fujik thermal compound from Deal Extreme). Tailcap draw using a single, fresh SF123A cell was 1.2A. Assuming a cell voltage of 2.5V under load, that translates to about 3W of power. If the BB NG is 90% efficient under these conditions, that means about 2.7W of power to the LED.

For the SSC P4 I vs Vf curve, the 2.7W point on the curve looks to be at about 800ma drive current and Vf~3.45V. If the BB NG is 80% efficient, then the LED is being fed about 2.4W, corresponding to about 700ma drive current and Vf~3.4V.

Either way, not bad. The light is plenty bright.

For my personal edification, what is the reason why I am apparently not quite getting 1000ma drive current?

What combination of sense resistors is recommended for an SOB 1200? I can see 3 combinations:

0.05,0.27, 1185ma
0.05, 0.22, 1227ma
0.05,0.18, 1278ma

Is there any reason to favor one pair over another? How sensitive is this resistor network interference/stability issue to the 1.2A level?

Edit: This thread seems to explain why I am apparently not getting the full 1000ma drive current. Post #3 says that the safe mode drive current is around 750ma, which seems to square with my estimates above. Still quite good drive current level and lumens output is bright. Post #22 says that the safe mode input current is 1A, whereas I measured 1.2A. Not sure if this difference is meaningful. Perhaps the difference is due to IC-to-IC variation in switch current limit (the start-up, or safe mode, current limit spec appears to be 40% of the switch current limit), and the 1A figure is simply an average, rule of thumb value?

The various posts in the linked thread also suggest that the BBNG can be driven by a single Li-ion cell. I had been hesitant to try this since I estimated that my SSC P4 U2SW0H-bin LED had a Vf ~3.4V. Thus the Vin overdrive voltage of Vf+0.7V seemed uncomfortably close to the initial voltage of 4.2V for a fully-charged Li-ion. Can the BBNG safely use a freshly charged 1xLi-ion or should I partially-deplete the cell first, say down to 4.05V or 4.10V?

Justin,

I would not have a problem with the overdrive as you have outlined. It's only a bit more than the overhead which means you might be overdriving the LED to 1.5A or something more than the regulator would be doing. When the battery voltage falls to a lower level the regulator will kick in.

Key to the BBNG to start up in full regulation may be dependant on all the components in the light. The resistance of the battery spring, contact between the light parts, LE and anode connection. The small resistances can have a significant voltage reduction as seen by the board and it's the voltage present at the converter board that makes or breaks the full regulation start up.

The li-ion combination should provide full regulation most of the time vs the standard primary CR123 battery configuration.

Wayne

Thanks. The sensitivity to small changes in overall system resistance makes sense. When I assembled the tower for the SF TurboHead, I tested the partially assembled tower by using some skinny hookup wire to connect my SF123A to the tower. Output was clearly less than when I completed the assembly and inserted the tower into an actual TH and flashlight body. No doubt, the hookup wire resistance was higher than the flashlight body's.

Other than cleaning all contacts (springs, battery terminals, flashlight body, threads), any suggestions on reducing parasitic resistances? I'm using an AW aluminum tower based on the Arcmania design, AW-supplied center and outer springs for the tower (shiny, chrome-plated springs), SSC P4 U2SW0H-bin LED glued to the tower using AA epoxy, tower inserted into a SureFire KT TurboHead, and threaded onto either a SF 3P body or a FiveMega 2xC-cell body. Tailcap is an older SureFire momentary-on unit (not a clicky). Do you think SS Krytox is better than Nyogel for thread lube and lower resistance (is that a loaded question or what )?

On my SOB 1200 question, any reason to favor one of the three sense resistor combinations that I outlined (assuming that those combinations are correct)?

Thank you again.

Input current requirements are always more demanding on a boost configuration as you are stepping up from a lower voltage to "boost" to a higher output voltage. Under a boost configuration battery current is always higher than the output current.

In a buck configuration input current max is equal to the output current. The SOB1200 would draw most of the time less battery current than 1.2A.

My comments were really around low battery voltages like 3V where a 1/10 of a volt is a significant drop from the main battery voltage.

With a large battery pack and a battery voltage of say 12V a 1/10 of a volt drop or loss has no bearing on the total design and has no effect on the how the regulator works.

If your LED is 3-4V and your battery pack is 12V it will be in regulation until the battery voltage reaches ~4.5V or so. That's a long time and more than likely a multi-cell battery pack will be dead long before it reaches this voltage.

Wayne

I think I'm confused, and I also probably confused you.

For the SOB, the application would be to drive something like an MC-E in 2S2P. I thought that the 1.2A-1.5A stability issue was for the output current (not the input current), which I thought would be the case for an SOB 1200.

I hadn't thought about the issue that in a buck configuration, max input current equals the output current. I was overly focused on only looking at Vin > Vout (plus some headroom).

For 1.2A drive current to an MC-E in 2S2P, I'll assume a voltage drop due to the MC-E of ~6.7V (ignoring any other sources of voltage drop). If I use 2xIMR16340 and assume Vin ~7.4V, do you think this is enough voltage headroom to run the SOB in regulation?

Then, there is the driver efficiency issue. If the board is 80% efficient for these conditions, then

0.80 * 7.4V * Iin ~ 1.2A * 6.7V ~8W

Thus, Iin ~1.36A (more if the SOB efficiency is less than the assumed 80%, less if the IMR Li-ions can hold a higher voltage under load).

Will the SOB 1200 accept Iin of 1.36A (or more) in this case, or will Iin top out at Iout=1.2A?

10 watts of input power and 2 watts of power that needs to be dissipated as heat seems like a lot, so I would definitely pot the driver. The aluminum tower that I use has a lot of surface area in tight contact with the SureFire TurboHead in which the tower will be installed, so there is a fair amount of metal available for heat sinking (as long as I can get the heat away from the SOB and into the heat sink).

I'm willing to try this and hope that the SOB 1200 survives since nein166 in post #3 said he has successfully used a BB 1000 to run an MC-E in 2S2P, drawing 1.6A from 2 123A primaries. That looks roughly like 8 watts of input power. If the SOB 1200 fries, I have an SOB 1000 that I can try.

For an SOB 1000, I estimate that the Vdrop is ~6.6V. A similar calculation as above gives

80% * 7.4V * Iin ~ 6.6V * 1A ~ 6.6W

Thus, Iin ~1.15A. Again, will the SOB 1000 accept 1.15A of input current, or will it top out at Iout=1A?

If either of these configurations is problematical, I can also use 4xSF123A (let's say roughly 10V under load). Though less desirable due to larger flashlight size, I also want the flashlight to work. For an SOB 1200, I get Iin ~1A. For an SOB 1000, I get Iin ~0.83A. Both figures still assume 80% driver efficiency.

Would you recommend using either of these two configurations (SOB 1000 or SOB 1200, 4x123A primaries) over the first two (SOB 1200 and 2xIMR16340, or SOB 1000 and 2xIMR16340)?

On the resistance thing, I can see how even 0.1 ohm can matter. At 1.2A input current, that gives a voltage drop of 0.12V, which is probably significant relative to the 2.4V nominal 2xNiMH voltage. Do you think there is any benefit in using something like SS Krytox on the flashlight threads instead of Nyogel? How about cleaning the various contact surfaces with something like DeoxIT?

For the SOB. The components can handle currents of 2A.

The SOB max duty cycle is not 100%, but, IIRC 85%??? That means it will not hold regulation down to Vin = Vout. In order to remain in regulation you would need Vin to be higher by 15% or more to be fully regulated.

Therefore, if your 2S2P is 6.4V you need a minimum of 7.5V or higher.

My calculations might not be correct, but, the need to have Vin higher than Vf with some margin needs to be factored in.

The SOB will safely run 1.2A with Vin higher than Vout (+15%) up to Vin up to 16V.

As long as you meet the above requirements and can handle the heat dissipated then you are good to go. I think for a 2 cell you might have a no go due to the headroom requirements for primaries and should be fine with li-ion cells in a 2 cell configuration. For a primary 2 cell configuration it will still work, just not 100%. That may still be fine if it is providing 750mA or more initially and then slowely dimming over time.

all 4 cell configurations should be fine as long as max battery voltage stays below 16V.

Wayne

Great! I'm going to go for it and use an SOB 1200 and let you know what happens. I might compare results using 2xIMR16340, 2x17670, and 4xSF123A.