Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Voltage mode LED Driver IC's are often best?

Status
Not open for further replies.
G

grizedale

Advanced Member level 3
Joined
Jun 13, 2011
Messages
838
Helped
17
Reputation
34
Reaction score
17
Trophy points
1,298
Activity points
8,804
Hello,

Do you agree with the following?... and if so, please send a copy to IC companys.

Please can you send a copy of this to the semiconductor companys that make PWM controllers for driving LEDs…..maybe then they will get the message and see the GREAT ERROR in their ways.

I am speaking of driving long series LED strings from low input voltages, where high duty cycle operation is required.
(eg vin = 10V, Vout = ~48V , Iout = 350mA)
-so topology is boost, buckboost or sepic.



Why on this earth do all semiconductor companys provide “CURRENT MODE” solutions for the above??

It is BLATANTLY obvious that VOLTAGE MODE is what is required.

At high duty cycles , current mode controllers need lots of slope compensation, which means the FET source current sense resistor gets downsized, (to make way for the added slope compaensatory ramp)……………..

……….-this in turn leads to overcurrent problems at Start-up, Power ON/OFF/ON and durint LED flashing ON/OFF sequences.


Why do no semiconductor companies do Voltage mode IC’s for this?

There is no requirement for fast transient response when driving LEDs…….since even if leds are flashed on and off at 50% duty and 4Hz, a 20ms start-up time for the LED driver isn’t going to be visibly noticeable.

A voltage mode converter with a well damped feedback loop can do the job best.

Do you agree?
 

No that wont work.. LED's are matched voltage within a batch but have wide variance between batches. So your 48V string will not regulate to 350mA.
There is a variance on Vthreshold as well as ESR. The brightness is not dependant on voltage variation but rather current control.

Thats why.
 
grizedale said:
At high duty cycles , current mode controllers need lots of slope compensation, which means the FET source current sense resistor gets downsized, (to make way for the added slope compaensatory ramp)……………..

……….-this in turn leads to overcurrent problems at Start-up, Power ON/OFF/ON and durint LED flashing ON/OFF sequences.
Click to expand...
Could you give some details here? How does slope compensation cause problems at startup? If anything, current mode control should be much more immune to input surge current than voltage mode control.
 

Sunnyskyguy:
Sorry i dont understand you.
-There is no reason a voltage mode converter cant regulate a LED current accurately to any choosen value.

mtwieg:
I think the first step is to appreciate that a current mode converter with much slope compensation has a degraded (smaller ohmic value) fet source current sense resistor.
-once you realise that then you realise that the FET source current limit is not as low as it could be, and overcurrents can occur.
With voltage mode, you can make your fet source current limit much lower, because you dont have the slope ramp to also think about

Here is a voltage mode led driver schematic

voltage mode sepic.jpg

Here is a voltage mode led driver in ltspice

Code: 
Version 4
SHEET 1 3492 836
WIRE 2432 -640 2224 -640
WIRE 2432 -608 2432 -640
WIRE 2432 -496 2432 -544
WIRE 160 -432 64 -432
WIRE 720 -432 240 -432
WIRE 64 -400 64 -432
WIRE 2432 -384 2432 -432
WIRE -112 -288 -432 -288
WIRE 160 -288 -32 -288
WIRE 576 -288 240 -288
WIRE 624 -288 576 -288
WIRE 720 -288 720 -432
WIRE 720 -288 688 -288
WIRE 768 -288 720 -288
WIRE 880 -288 832 -288
WIRE 2016 -288 2016 -576
WIRE 2016 -288 880 -288
WIRE 2224 -288 2224 -640
WIRE 2224 -288 2016 -288
WIRE 2432 -272 2432 -320
WIRE 2688 -240 2608 -240
WIRE 2688 -224 2688 -240
WIRE 208 -176 112 -176
WIRE 2432 -176 2432 -208
WIRE 112 -160 112 -176
WIRE -432 -128 -432 -288
WIRE 2688 -128 2688 -160
WIRE 880 -96 880 -288
WIRE 880 -96 800 -96
WIRE -16 -80 -80 -80
WIRE 2432 -64 2432 -112
WIRE 112 -48 112 -80
WIRE -80 -32 -80 -80
WIRE 880 -32 880 -96
WIRE 2688 -16 2688 -64
WIRE 208 32 208 -176
WIRE 2096 32 2048 32
WIRE 2224 32 2160 32
WIRE 2432 32 2432 0
WIRE 576 48 576 -288
WIRE 800 48 800 -96
WIRE 1568 64 1568 0
WIRE 2688 80 2688 48
WIRE -16 96 -16 -80
WIRE 48 96 -16 96
WIRE 416 128 368 128
WIRE 528 128 496 128
WIRE 880 128 880 48
WIRE 2432 144 2432 96
WIRE -160 160 -160 96
WIRE -160 160 -272 160
WIRE 48 160 -160 160
WIRE 2048 160 2048 32
WIRE 2080 160 2048 160
WIRE 2224 160 2224 32
WIRE 2224 160 2160 160
WIRE 1568 192 1568 144
WIRE 2688 192 2688 144
WIRE 48 224 -160 224
WIRE 576 224 576 144
WIRE 672 224 672 -16
WIRE 672 224 576 224
WIRE 1072 240 1040 240
WIRE 1168 240 1136 240
WIRE 1376 240 1248 240
WIRE 2432 240 2432 208
WIRE 416 256 368 256
WIRE 48 288 -32 288
WIRE 2688 288 2688 256
WIRE 800 304 800 112
WIRE 880 304 880 192
WIRE 880 304 800 304
WIRE -160 320 -160 224
WIRE 416 336 416 256
WIRE 2144 336 2144 240
WIRE 1296 352 1296 288
WIRE 2224 352 2224 160
WIRE 2224 352 2176 352
WIRE 2272 352 2224 352
WIRE 2384 352 2352 352
WIRE 2432 352 2432 304
WIRE 2608 352 2608 -240
WIRE 2608 352 2432 352
WIRE 1376 368 1376 240
WIRE 1376 368 1328 368
WIRE 1472 368 1376 368
WIRE 1552 368 1472 368
WIRE 1712 368 1632 368
WIRE 1824 368 1712 368
WIRE 2048 368 2048 160
WIRE 2048 368 1904 368
WIRE 2112 368 2048 368
WIRE 1040 384 1040 240
WIRE 1264 384 1040 384
WIRE 2224 384 2176 384
WIRE -32 400 -32 288
WIRE 1376 400 1328 400
WIRE 1472 464 1472 368
WIRE 1488 464 1472 464
WIRE 1616 464 1568 464
WIRE 1712 464 1712 368
WIRE 1712 464 1680 464
WIRE 2224 464 2224 384
WIRE 2272 464 2224 464
WIRE 2688 464 2688 352
WIRE 2688 464 2352 464
WIRE 1376 496 1376 400
WIRE 2224 512 2224 464
WIRE 2688 528 2688 464
WIRE -432 624 -432 -48
WIRE -160 624 -160 384
WIRE -160 624 -432 624
WIRE -32 624 -32 480
WIRE -32 624 -160 624
WIRE 208 624 208 352
WIRE 208 624 -32 624
WIRE 416 624 416 416
WIRE 416 624 208 624
WIRE 576 624 576 224
WIRE 576 624 416 624
WIRE 880 624 880 304
WIRE 880 624 576 624
WIRE 1296 624 1296 416
WIRE 1296 624 880 624
WIRE 1376 624 1376 576
WIRE 1376 624 1296 624
WIRE 2144 624 2144 400
WIRE 2144 624 1376 624
WIRE 2224 624 2224 592
WIRE 2224 624 2144 624
WIRE 2384 624 2384 352
WIRE 2384 624 2224 624
WIRE 2688 624 2688 608
WIRE 2688 624 2384 624
WIRE 208 656 208 624
WIRE -272 768 -272 160
WIRE 1040 768 1040 384
WIRE 1040 768 -272 768
FLAG 112 -48 0
FLAG 208 656 0
FLAG 1568 192 0
FLAG 1568 0 5V
FLAG 2144 240 5V
FLAG 2016 -576 VOUT
FLAG 672 -16 RSEN
FLAG 64 -400 0
FLAG -80 -32 0
FLAG -160 96 COMP
FLAG 1296 288 5V
SYMBOL PowerProducts\\LT1243 208 192 R0
SYMATTR InstName U1
SYMBOL voltage -32 384 R0
WINDOW 3 -118 170 Left 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR Value PULSE(0 3 0 10n 10n 1u 7u)
SYMATTR InstName V1
SYMBOL voltage 112 -176 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -127 5 Left 2
SYMATTR SpiceLine Rser=0.001
SYMATTR InstName V2
SYMATTR Value 10
SYMBOL voltage -432 -144 R0
WINDOW 123 0 0 Left 2
WINDOW 39 24 124 Left 2
SYMATTR SpiceLine Rser=0.001
SYMATTR InstName V3
SYMATTR Value 10
SYMBOL ind2 144 -272 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L1
SYMATTR Value 33µ
SYMATTR Type ind
SYMATTR SpiceLine Rser=0.05
SYMBOL schottky 768 -272 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value MBR20100CT
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL nmos 528 48 R0
SYMATTR InstName M1
SYMATTR Value BSC057N08NS3
SYMBOL res 512 112 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 49 67 VTop 2
SYMATTR InstName R1
SYMATTR Value 1r
SYMBOL cap -176 320 R0
SYMATTR InstName C2
SYMATTR Value 4µ7
SYMATTR SpiceLine Rser=0.003
SYMBOL cap 864 128 R0
SYMATTR InstName C3
SYMATTR Value 22µ
SYMATTR SpiceLine Rser=0.005
SYMBOL voltage 416 320 R0
WINDOW 3 -100 232 Left 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR Value PULSE(0 1 0 7u 0u 0 7u)
SYMATTR InstName V4
SYMBOL Opamps\\LT1006 2144 304 M0
SYMATTR InstName U2
SYMBOL voltage 1568 48 R0
WINDOW 123 0 0 Left 2
WINDOW 39 24 44 Left 2
SYMATTR SpiceLine Rser=0.001
SYMATTR InstName V5
SYMATTR Value 5
SYMBOL LED 2416 -272 R0
SYMATTR InstName D2
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 -176 R0
SYMATTR InstName D3
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 -64 R0
SYMATTR InstName D4
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 32 R0
SYMATTR InstName D5
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 144 R0
SYMATTR InstName D6
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 240 R0
SYMATTR InstName D7
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2672 -224 R0
SYMATTR InstName D8
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2672 -128 R0
SYMATTR InstName D9
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2672 -16 R0
SYMATTR InstName D10
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2672 80 R0
SYMATTR InstName D11
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2672 192 R0
SYMATTR InstName D12
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2672 288 R0
SYMATTR InstName D13
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL res 2672 512 R0
SYMATTR InstName R2
SYMATTR Value 0.82
SYMBOL res 2176 144 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R4
SYMATTR Value 3K9
SYMBOL res 2368 336 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R5
SYMATTR Value 1K
SYMBOL res 2368 448 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R6
SYMATTR Value 1K
SYMBOL res 2208 496 R0
SYMATTR InstName R7
SYMATTR Value 3K9
SYMBOL cap 2160 16 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C1
SYMATTR Value 1n
SYMATTR SpiceLine Rser=0.001
SYMBOL LED 2416 -384 R0
SYMATTR InstName D14
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 -496 R0
SYMATTR InstName D15
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL LED 2416 -608 R0
SYMATTR InstName D16
SYMATTR Value LXHL-BW02
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL cap 688 -304 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C6
SYMATTR Value 3µ3
SYMATTR SpiceLine Rser=0.005
SYMBOL ind2 144 -416 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L2
SYMATTR Value 33µ
SYMATTR Type ind
SYMATTR SpiceLine Rser=0.05
SYMBOL res 1648 352 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R3
SYMATTR Value 22K
SYMBOL voltage 1376 480 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -51 -6 Left 2
SYMATTR InstName V6
SYMATTR Value 1.1193
SYMATTR SpiceLine Rser=0.001
SYMBOL res 1264 224 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R8
SYMATTR Value 1k
SYMBOL cap 1136 224 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C4
SYMATTR Value 150n
SYMBOL res 1584 448 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R10
SYMATTR Value 22k
SYMBOL cap 1680 448 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C5
SYMATTR Value 10n
SYMBOL Opamps\\LT1006 1296 320 M0
SYMATTR InstName U3
SYMBOL res 864 -48 R0
SYMATTR InstName R11
SYMATTR Value 0.1
SYMBOL cap 784 48 R0
SYMATTR InstName C7
SYMATTR Value 1µ
SYMATTR SpiceLine Rser=0.005
SYMBOL res 1920 352 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R9
SYMATTR Value 50
SYMBOL res -16 -304 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R12
SYMATTR Value 0.1
TEXT 360 672 Left 2 !.tran 0 150m 0 startup
TEXT 1816 -336 Left 2 !.ic V(VOUT) = 42
TEXT 1592 -640 Left 2 ;.save I(D16)\n.save I(L1)\n.save I(L2)\n.save I(M1)\n.save I(C6)\n.save I(C7)\n.save I(C3)\n.save V(VOUT)\n.save V(COMP)\n.save V(RSEN)
TEXT 160 -496 Left 2 !K L1 L2 1
TEXT 1248 144 Left 2 ;ERROR\nAMPLIFIER
TEXT 1960 -40 Left 2 ;CURRENT SENSE\nAMPLIFIER.
TEXT 392 240 Left 2 ;RAMP
TEXT 1768 248 Left 2 ;FEEDBACK LOOP\nTEST INJECTION\nRESISTOR.
TEXT 232 -72 Left 2 ;UCC35705 WOULD BE\nUSED IN ACTUAL\nCIRCUIT.
 
Last edited:

grizedale said:
(eg vin = 10V, Vout = ~48V , Iout = 350mA)

Why on this earth do all semiconductor companys provide “CURRENT MODE” solutions for the above??

It is BLATANTLY obvious that VOLTAGE MODE is what is required.
Click to expand...
What is required in this case is that the current through the LED string should be regulated at 350mA. One string may drop 50V at that current, while another drops 47V. When that one has warmed up, it may only drop 44V. If the voltage across the LED string is set to a constant 48V, results will be quite unpredictable.
grizedale said:
Sunnyskyguy:
Sorry i dont understand you.
-There is no reason a voltage mode converter cant regulate a LED current accurately to any choosen value.
Click to expand...
Are you suggesting that the best way to supply a constant current is to use a voltage mode converter? :???:
 

grizedale,

please correct me if i am wrong. your ltspice circuit is to regulate the output current (0.82 ohms resistor is to converter output current to a voltage level). so, you are still regulating the current instead of voltage.

what is your understanding of voltage mode and current mode control?
 

hello kind friends all.

i appreciate your time, but i beg to say many who are responding here do not know what is current mode control or what is voltage mode control.

so now i tell it...

current mode is where the pwm controller (THE PWM COMPARATOR) gets its ramp from the fet current sense signal.

voltage mode still has a ramp.....but it comes from an internal ramp generator, internal to the pwm controller.


Please i beg you to understand...........a current regulated SMPS can be EITHER current-mode or voltage-mode.

Most of the world (plus semiconductor companys) thinks that if you want to control an output current (ie led current) , then you need "current mode control"...but please, this is not so.
 
Last edited:

grizedale said:
mtwieg:
I think the first step is to appreciate that a current mode converter with much slope compensation has a degraded (smaller ohmic value) fet source current sense resistor.
Click to expand...
I don't see why this would be the case, as long as you have control of the slope of the slope compensation ramp. And I realize that there are controllers with fixed internal slope compensation, but even those can have additional external slope compensation added.
-once you realise that then you realise that the FET source current limit is not as low as it could be, and overcurrents can occur.
Click to expand...
Why would this be that case? You should be able to define the peak current limit independently of the Rsense value. If not, then that's a problem with the control IC, not with current mode control.
With voltage mode, you can make your fet source current limit much lower, because you dont have the slope ramp to also think about
Click to expand...
Yet voltage mode control has no inherent inrush limiting either, so I still don't see the point.

Also if you want to post LTspice files, just zip them and attach.
 

YOu could use an efficient stable boost voltage regulator with a simple current limiter with PWM control for average current at the cost of < 2V drop with simplicity below or as shown above with Op Amp and PID loop. I started using this in the 70's.
Fig_1.gif
 

mtwieg:
Why would this be that case? You should be able to define the peak current limit independently of the Rsense value. If not, then that's a problem with the control IC, not with current mode control.
Click to expand...

Exactly!

....such an IC , that can do what you say , SHOULD exist.......but however, it does not.

...the fact is, pick any LED driver you like off the market, and add slope compensation to it, and you end up having to downsize your current sense resistor.

This is my point, and you have got into it now.

Please now will all readers lobby the semiconductor compenys about this?


If anybody is still not convinced that adding slope compensation means having to degrade (downsize) the fet source current sense resistor, then give a schematic with a named IC .........................but dont bother , because you wont find an ic which can add slope compensation without the user having to degrade the fet source current sense resistor.
 

grizedale said:
...the fact is, pick any LED driver you like off the market, and add slope compensation to it, and you end up having to downsize your current sense resistor.
Click to expand...
Or just increase the slope compensation ramp.... ultimately all you need is a certain ratio between the current sense ramp and the slope compensation ramp. Decrease one, increase the other, it doesn't really matter.

And I believe that any problems with overcurrent on startup or brightness variation can be dealt with by other means. Lowering the loop bandwidth, having a soft start function, lowering the bandwidth of control signals... all can help avoid overcurrent conditions without the need for a hard current limit anywhere.
 

grizedale< I hope you understand my meaning of CV PSU driving in CC mode with above simple example.

Of course better design uses independent variables Rs, I(load) and PWM ratio. I can imagine that using comparator, 2.5V ref and gain R values to determine I, PWM & Rs given any 2. Whether that is integrated or not is the issue??
At least my solution would appear to be more stable.
 

Sunnyskyguy:-
Thanks, though you are ahead of me.

I think i recognise your schem above as being like a linear regulator for led current?
 

if final target is to control the output current, why give up current mode control? it is well known that voltage mode control have difficulty in controlling boost and buck-boost converter as the RHP zero.
 

Commercial solutions exist. You need to decide which specs you need before you choose a design criteria. for example;
**broken link removed** Buck
**broken link removed** Boost

PWM & CC mode control ( like my design above ) **broken link removed**

Once you decide desired output specs, then choose appropriate design that will work.
 

kelvin_mksinst:

if final target is to control the output current, why give up current mode control? it is well known that voltage mode control have difficulty in controlling boost and buck-boost converter as the RHP zero.
Click to expand...

Thanks Kelvin, but please may i tell that in fact with led drivers, the RHPZ is no problem whatsoever....because fast transient response is never needed with led drivers, so you just damp the heck out of your control loop and youre stable like that.

My initial observation here is a genuine one.....the semiconductor companys have got it wrong......The majority of DCDC LED driver IC's should be voltage mode.

But i guess there just supplying what the customers want, and we all know that the latest "fashion" is for current mode IC's....even though for LED driver application, they are often inferior to voltage mode.

Please tell the semiconductor companys that you want voltage mode.....at the end of the day, theyre only supplying what people say they want.
 

grizedale, but you still need to sense the FET current for protection purpose, which will add the cost back. For current mode control, this is an inherent feature. You can save cost and components with using current mode control in led driving. Cost is everything in this industry.
 

grizedale said:
My initial observation here is a genuine one.....the semiconductor companys have got it wrong......The majority of DCDC LED driver IC's should be voltage mode.

But i guess there just supplying what the customers want, and we all know that the latest "fashion" is for current mode IC's....even though for LED driver application, they are often inferior to voltage mode.
Click to expand...
You still haven't actually demonstrated why this is the case though. I can find many current mode LED drivers with internal switches, sense resistors, and slope compensation which are specified for duty cycles up to and over 90%. They fit your example application fine (though not for quite that much power throughput). And there are some with external FETs which would allow for the high output power, such as the LT3755. What's wrong with such a part?
 

LT3755 is $3.7 for 5000 pieces on digikey.

-thats a bit too expensive.

LM5025C is voltage mode and is cheap, but it has an opto-coupler feedback pin insted of an internal error amp, so it needs the external opamp error amp aswell....also, the fet source sense reference is 0.5V, which is a little too high.

I agree that a voltage mode IC needs to also sense the fet source current so that it can terminate the cycle in the case of overcurrent.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top