How much input voltage can a SSC P7 handle?
- Thread starter overdog
- Start date
Help Support Candle Power Flashlight Forum
Important to know is that LEDs are current driven, not voltage driven. The maximum rated current you can drive a SSC P7 at is 2800mA, which is -depending on the bin- reached at typically 3.6V. If you connect the LED directly to 5.5 to 6V the LED will go 
:mecry: instantly, so better get yourself a good driver or make sure you don't apply too much voltage to the LED.
:mecry: instantly, so better get yourself a good driver or make sure you don't apply too much voltage to the LED.
Last edited:
Gunner12
Flashaholic
As said by Erasmus, the Vf of the LED depends on the LED. If your input source is around 6v, then you should get a buck driver.
Thank you Erasmus and Gunner12,
this is the save way to drive a P7, but has someone ever tried to go further - I had a Mag 3D, used 3x 1,5V batteries= 4,5V and the P7 felt comfortable with it. May the P7 has also variuos current operating points like the SST-90? O.K., Luminus does not say, that you´ll receive a constant current of 9Ah by driving the Led at 5,6V- their recommended fV is also 3,6V...
...may this also works with P7 ???
this is the save way to drive a P7, but has someone ever tried to go further - I had a Mag 3D, used 3x 1,5V batteries= 4,5V and the P7 felt comfortable with it. May the P7 has also variuos current operating points like the SST-90? O.K., Luminus does not say, that you´ll receive a constant current of 9Ah by driving the Led at 5,6V- their recommended fV is also 3,6V...
...may this also works with P7 ???
Last edited:
Al Combs
Enlightened
- Joined
- Jul 2, 2007
- Messages
- 872
Your Mag3D was using alkalines, yes? Their high internal resistance caused a voltage drop under load. If you had put a meter across the LED contacts, you would have seen your 4.5 volts drop to ≈ 3.6 volts. Gunner12 is correct about the need for a buck regulator to run a P7 off a 6 volt battery. Take a look at jtr1962's epic "White lumen testing" thread. Here's the post on the C binned SSC-P7 he tested over a year ago. The chart will give you an idea why it's impossible to drive a P7 to 6 volts without going. It's important to remember he's using heatsinks in these tests far better than an ordinary flashlight. You would never be able to drive a P7 in an ordinary flashlight to 7 amps without destroying it. Plus his test intervals are fairly short. Just enough to see how bright you can make it.
Were did you hear you can drive an SST-90 @ 5.6 volts? Again jtr1962 tested an SST-90 at the bottom of post#336. He reached 3.2 amps at 3.48 volts and 9 amps at just 3.84 volts. That's a difference of only 0.36 volts to almost triple the current. Even if the current didn't rise exponentially with voltage (which it does), in a linear curve you'd be at almost 53 amps by the time you reached 5.6 volts. That's almost 300 watts of power! Sorry, I'm just trying to talk you out of a really really really bad idea. :sick2:
The LED's logarithmic current rise with a change in voltage is helpful in understanding why a constant current regulator is so important. Even if you had a good voltage regulator, it would never be good enough. When an LED get hot, its Vf actually drops. The only type of device that can anticipate this drop in Vf is a constant current regulator. They actually vary the voltage while monitoring the current used.
. It's important to remember he's using heatsinks in these tests far better than an ordinary flashlight. You would never be able to drive a P7 in an ordinary flashlight to 7 amps without destroying it. Plus his test intervals are fairly short. Just enough to see how bright you can make it.Were did you hear you can drive an SST-90 @ 5.6 volts? Again jtr1962 tested an SST-90 at the bottom of post#336. He reached 3.2 amps at 3.48 volts and 9 amps at just 3.84 volts. That's a difference of only 0.36 volts to almost triple the current. Even if the current didn't rise exponentially with voltage (which it does), in a linear curve you'd be at almost 53 amps by the time you reached 5.6 volts. That's almost 300 watts of power! Sorry, I'm just trying to talk you out of a really really really bad idea. :sick2:
The LED's logarithmic current rise with a change in voltage is helpful in understanding why a constant current regulator is so important. Even if you had a good voltage regulator, it would never be good enough. When an LED get hot, its Vf actually drops. The only type of device that can anticipate this drop in Vf is a constant current regulator. They actually vary the voltage while monitoring the current used.
Hello Al Combs,
thank you for very good advice, jtr1962's testings are really helpful and very interesting to me, I am sure what you say is for 100% correct- I already use converters to ensure, that no emitter gets damaged and is driven at optimal condition- in case of the SST-90 some members used 4x AA 1,2V NiMN battery to direct drive the emitter to achieve a current of about 7-10 Ah, which was impossible at a voltage of 3,7V without converter- this is about 5,6V fully loaded and nothing got damaged- how is this to explain? Does some voltage get lost on its way to the emitter? Has someone tried to direct drive a P7 with same battery configuration- 4x AA NiMN? I am interested in the results-
just if it works or not: or :thumbsup:....
thank you for very good advice, jtr1962's testings are really helpful and very interesting to me, I am sure what you say is for 100% correct- I already use converters to ensure, that no emitter gets damaged and is driven at optimal condition- in case of the SST-90 some members used 4x AA 1,2V NiMN battery to direct drive the emitter to achieve a current of about 7-10 Ah, which was impossible at a voltage of 3,7V without converter- this is about 5,6V fully loaded and nothing got damaged- how is this to explain? Does some voltage get lost on its way to the emitter? Has someone tried to direct drive a P7 with same battery configuration- 4x AA NiMN? I am interested in the results-
just if it works or not:
or :thumbsup:....
Gunner12
Flashaholic
Voltage from the battery is not constant. When a battery is pushed hard enough, the voltage will drop along with some capacity (for that recharge).
Look at the battery testing threads and you'll see what I mean.
Look at the battery testing threads and you'll see what I mean.
mdocod
Flashaholic
Hello Al Combs,
thank you for very good advice, jtr1962's testings are really helpful and very interesting to me, I am sure what you say is for 100% correct- I already use converters to ensure, that no emitter gets damaged and is driven at optimal condition- in case of the SST-90 some members used 4x AA 1,2V NiMN battery to direct drive the emitter to achieve a current of about 7-10 Ah, which was impossible at a voltage of 3,7V without converter- this is about 5,6V fully loaded and nothing got damaged- how is this to explain? Does some voltage get lost on its way to the emitter? Has someone tried to direct drive a P7 with same battery configuration- 4x AA NiMN? I am interested in the results-
just if it works or not:
AA size NIMH cells can only deliver so much current while maintaining 1.2V. WHen you start to go above ~3-4 amps, the voltage will start to drop pretty dramatically on some cells. Some cell types hold up better than others, so a direct drive configuration like this is dependent on using cells in conjunction with natural host resistance that pair up appropriately.
The LED is NOT getting 5-6V when driven by 4 NIMH AA cells, the supply voltage sags under the intense load, and some effective voltage is lost in the resistance of all of the contacts between the cells, in conjunction with the switch, and any other components that the current must pass through. In the case of such a high current low voltage build, small amounts of resistance have a larger impact on voltage.
Just because cells say 1.2V or 1.5V on the side, does not mean that they are going to deliver a perfect 1.2V or 1.5V into any load like say, a regulated bench power supply might (up to it's rated maximum). The 7-10 amps they have been able to achieve is the result of the voltage sag and current reaching an equilibrium in the circuit. The actual voltage at the LED is ~3.7V (+/- 0.3V I'd guess).
-Eric
Hello Eric, this sounds logical to me- why I think about such things is, I built a triple P7, which was planed to be powered by 4x14670 battery and Der Wichtel-buck converter in combination with D2Flex, but the buck converter pushed the batteries too hard (I guess), that battery protection kicked in- (with 4x18650 everything in combination worded very very good!)- So now I run the triple P7 only with D2Flex and direct drive with 3x14670 and one dummy battery in battery holder: And I am thinking about to put in the 4´th battery: may Li-Ion´s have lower internal resistance and will give too much voltage on the 3 P7´s, so that I´ll damage them- before I do this I better ask here about some experiences with this idea- could this work to direct drive 3x P7 with 4x14670 Li-Ions?
:thinking::thinking::thinking:
:thinking::thinking::thinking:
Al Combs
Enlightened
- Joined
- Jul 2, 2007
- Messages
- 872
The LED without a regulator was inadvisably just about money...:thumbsdow Going beyond 2C on a Cobalt Li-Ion is potentially deadly. Check out mdocod's sigline and BatteryUniversity.
Check out mdocod's sigline and BatteryUniversity.
mdocod
Flashaholic
Hello overdog,
The reason the 4xNIMH cell configuration works in driving a SST-90 is because of the very high discharge rate for the size of the cells... The fact that the ST-90 can handle that current is important. Consider, a normal AA NIMH cell will have around 2AH capacity, when delivering 7-10 amps, it's running in the 4-5+C discharge range, which, while hard on cells, is not unsafe per-say with NIMH cells because they are safe and tolerant of abuse. Lithium Cobalt cells should not be discharged at a rate higher than 2C (for most cells, some should not exceed 1.5C, and I personally always recommend trying to keep all LiCo cells running in the 1.5C or lower arena as 2C rates have been shown to dramatically reduce cycle life). With that in mind, a 1.1AH 14670 should not be discharged faster than around 2 amps. However, if we did take 4 unprotected 14670s, and try to drive say, 3 SST-90s, it would likely "work" in so far as the drive current would reach whatever the maximum current the cells could deliver is. Without a protection circuit, I'd estimate a similar scenario as with the 4 NIMH cells driving a single LED, the cells would deliver as much as they can produce through the effective resistance of the configuration, which would result in a discharge rate far above the safe rate, possibly upwards of the same 7-10 amp range, obviously, if the LEDs in question were P7s, they would be severely over-driven and likely fail quickly.
As you found, the protection in the cell doesn't even allow for these levels of discharge, so the above scenario is based on a hypothetical unprotected application.
Now.. Lets explore the concept of using 4 14670s to drive 3 P7s in a proper manner. This can be done if you use an appropriate driver and take into consideration the total power consumption from the cells and modify the output to be at or below the 2C limit for the cells.. Here's how this is done:
The total stored energy of a cell or battery of cells can be stated in watt-hours. A very rough way to estimate the watt-hour "rating" of a pack, is to multiply the capacity of the cell by the nominal operating voltage. For LiCo cells, 3.7V is about the average operating voltage under most common loads. De-rating capacity a bit in these estimates is always a good idea. So, a 4x14670 pack is roughly 4x3.7Vx1AH=14.8WH. Power flowing from the pack is stated in watts. A 1C discharge rate for the pack would be right about 14.8W, lets call it 15W. A 2C discharge rate would be about 30W. For a 3 cell pack, the stored energy would be ~11WH, and the maximum safe discharge rate would be around 22W.
You'll see where I'm going with this in a minute...
The power consumption of a device like an LED can be expressed in watts also. A P7 running at 2.8A with a Vf of around 3.6V would have an operating power of around 10W.
Now, if we use a true current regulator, lets just assume a 100% efficient scenario that can boost and buck without limitations (doesn't exist but for this state of the comparison lets assume this for the time being)...
3 P7 LEDs are going to draw 30W if driven at 2.8A, with the 3 cell pack, this is over the maximum safe discharge rate, however, if we bump it up to 4 cells, and have the regulator in place limiting current across the LEDs to 2.8A, the power consumption will still be 30W, but the extra cell in the pack will reduce the load on each cell just enough to be operating safely.
In this example, we haven't taken into account the efficiency loss of the regulator. A well built and relatively expensive buck regulator would operate above 95% efficiency in most cases. So, the power consumption would actually be ~31.5W assuming 95% driver efficiency... IMO, this is pushing those cells too hard, I would suggest trying to get down around 1.5C, or ~22.5W on the 4 cell pack, or around 21W through the LEDs after efficiency loss, which translates to a drive current of about 2 amps instead of 2.8A. Which, would still be very very bright, and would actually probably be preferable since the efficiency of the LEDs would rise with the lower drive level, which would further reduce the heat that would have to be dissipated.
Hope that helps some:
-Eric
The reason the 4xNIMH cell configuration works in driving a SST-90 is because of the very high discharge rate for the size of the cells... The fact that the ST-90 can handle that current is important. Consider, a normal AA NIMH cell will have around 2AH capacity, when delivering 7-10 amps, it's running in the 4-5+C discharge range, which, while hard on cells, is not unsafe per-say with NIMH cells because they are safe and tolerant of abuse. Lithium Cobalt cells should not be discharged at a rate higher than 2C (for most cells, some should not exceed 1.5C, and I personally always recommend trying to keep all LiCo cells running in the 1.5C or lower arena as 2C rates have been shown to dramatically reduce cycle life). With that in mind, a 1.1AH 14670 should not be discharged faster than around 2 amps. However, if we did take 4 unprotected 14670s, and try to drive say, 3 SST-90s, it would likely "work" in so far as the drive current would reach whatever the maximum current the cells could deliver is. Without a protection circuit, I'd estimate a similar scenario as with the 4 NIMH cells driving a single LED, the cells would deliver as much as they can produce through the effective resistance of the configuration, which would result in a discharge rate far above the safe rate, possibly upwards of the same 7-10 amp range, obviously, if the LEDs in question were P7s, they would be severely over-driven and likely fail quickly.
As you found, the protection in the cell doesn't even allow for these levels of discharge, so the above scenario is based on a hypothetical unprotected application.
Now.. Lets explore the concept of using 4 14670s to drive 3 P7s in a proper manner. This can be done if you use an appropriate driver and take into consideration the total power consumption from the cells and modify the output to be at or below the 2C limit for the cells.. Here's how this is done:
The total stored energy of a cell or battery of cells can be stated in watt-hours. A very rough way to estimate the watt-hour "rating" of a pack, is to multiply the capacity of the cell by the nominal operating voltage. For LiCo cells, 3.7V is about the average operating voltage under most common loads. De-rating capacity a bit in these estimates is always a good idea. So, a 4x14670 pack is roughly 4x3.7Vx1AH=14.8WH. Power flowing from the pack is stated in watts. A 1C discharge rate for the pack would be right about 14.8W, lets call it 15W. A 2C discharge rate would be about 30W. For a 3 cell pack, the stored energy would be ~11WH, and the maximum safe discharge rate would be around 22W.
You'll see where I'm going with this in a minute...
The power consumption of a device like an LED can be expressed in watts also. A P7 running at 2.8A with a Vf of around 3.6V would have an operating power of around 10W.
Now, if we use a true current regulator, lets just assume a 100% efficient scenario that can boost and buck without limitations (doesn't exist but for this state of the comparison lets assume this for the time being)...
3 P7 LEDs are going to draw 30W if driven at 2.8A, with the 3 cell pack, this is over the maximum safe discharge rate, however, if we bump it up to 4 cells, and have the regulator in place limiting current across the LEDs to 2.8A, the power consumption will still be 30W, but the extra cell in the pack will reduce the load on each cell just enough to be operating safely.
In this example, we haven't taken into account the efficiency loss of the regulator. A well built and relatively expensive buck regulator would operate above 95% efficiency in most cases. So, the power consumption would actually be ~31.5W assuming 95% driver efficiency... IMO, this is pushing those cells too hard, I would suggest trying to get down around 1.5C, or ~22.5W on the 4 cell pack, or around 21W through the LEDs after efficiency loss, which translates to a drive current of about 2 amps instead of 2.8A. Which, would still be very very bright, and would actually probably be preferable since the efficiency of the LEDs would rise with the lower drive level, which would further reduce the heat that would have to be dissipated.
Hope that helps some:
-Eric
Hello Eric,
even a newbie as I am can understand your detailed and really very helpful answer- thank you very much for detailed reply- so I´ll keep my fingers away from using a fourth battery for this constellation with P7´s- when I planed this light, I wanted to use a Der Wichtel buck converter in his first version, with which it was possible to regulate the current output- unfortunately he does not sell them any more- so I tried the new one and made my experiences (have to say that Yitao told me, that the converter may wont work with 4x 14670 batteries, have alredy three of them and they are highest quality!)...
Maybe someone has one Wichtelbuck version 1 left, I would be interested to buy one to finish my light as I planned it...
-best regards,
Steffen.
even a newbie as I am can understand your detailed and really very helpful answer- thank you very much for detailed reply- so I´ll keep my fingers away from using a fourth battery for this constellation with P7´s- when I planed this light, I wanted to use a Der Wichtel buck converter in his first version, with which it was possible to regulate the current output- unfortunately he does not sell them any more- so I tried the new one and made my experiences (have to say that Yitao told me, that the converter may wont work with 4x 14670 batteries, have alredy three of them and they are highest quality!)...
Maybe someone has one Wichtelbuck version 1 left, I would be interested to buy one to finish my light as I planned it...
-best regards,
Steffen.
Der Wichtel
Enlightened
If 4x14650 lithiums won't work with the buck driver then it won't work in direct drive either.
With buck driver input current is around 2A. Without driver the current will be much higher.you can remove on of the three r220 resistors, then the leds will be driven with around 1.9A. Input current will be around 1.3A which should be fine for thebatteries
With buck driver input current is around 2A. Without driver the current will be much higher.you can remove on of the three r220 resistors, then the leds will be driven with around 1.9A. Input current will be around 1.3A which should be fine for thebatteries
Last edited:
mdocod
Flashaholic
I'm curious where the cutoff on the PCB is for those 14670s. Most AW protected cells have PCB cutoffs that are actually on the high side for use in incan applications that have high current start-ups.... Perhaps the 14mm cell series is not set as high as the rest. hmmm...
...now I removed first one of the three r220 resistors: works excellent with 4x18650, but with 4x14670 only short flashing- next I removed the second r220 resistor, to shorten this, the same result again...??? To complete, now I put back one resistor and tried with 5x 14670, same again...
:shakehead ...Any Idea?
:shakehead ...Any Idea?
Last edited:
Der Wichtel
Enlightened
Alright, I think I discovered your problem.
It's the input capacitor which are charged when the driver is connected. The current during charging is pretty high but only very short. It seems like the protection IC in the 14650 batteries are very sensitive.
I'm not sure but removing one of the capacitors should solve your problem. Maybe you are even able to drive the LEDs with all 3 resistors with full power.
It's the input capacitor which are charged when the driver is connected. The current during charging is pretty high but only very short. It seems like the protection IC in the 14650 batteries are very sensitive.
I'm not sure but removing one of the capacitors should solve your problem. Maybe you are even able to drive the LEDs with all 3 resistors with full power.
Thank you,sounds good to me,
if you can give me some advice which part I`ll have to remove, I´ll try-
and I think I will not remove the third resistor- feel very comfortable with that output (I remember I asked you, how to reduce current when I ordered two or three months ago...)
if you can give me some advice which part I`ll have to remove, I´ll try-
and I think I will not remove the third resistor- feel very comfortable with that output (I remember I asked you, how to reduce current when I ordered two or three months ago...)
Der Wichtel
Enlightened
There are two capacitors on C7.
Remove both and see if it works. If yes then put one back on and see if it still works.
Remove both and see if it works. If yes then put one back on and see if it still works.
