Boost Converter / LT SPICE

1.2 to 40 volts input isn't a specification, it's techie bashing. What sort of system would present such a requirement?

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

Pointy-headed manager,or a home-work problem set by a pointy-headed lecturer.

It could be one of those marketing department "specifications/wish lists" which basically recycle the stupidest and/or most malicious customer questions that any of the reps have been asked to respond to.

-- Bill Sloman, Nijmegen

Hi,

1) If I have a FET that can operate at lower voltage range than I might be able to get the LT3759 to work. My input voltage range is 1.2 Volts to 40 volts. If LT3759 can even work with 2.5 V than I will be happy. Output voltage is 5 V @ between 200 to 300mA.

I found the following FET from digikey with Vgs voltage is 2.5 volts minimum to turn it ON ( Comment needed here).

FET Data sheet link:

Any advice. I am looking for a FET that can be turned ON completely between the Vgs values of 2 to 3 volts. Vds = 60 Volts.

2) I do not have ANY experience with measuring the ripple at the output of the DC to DC converter and LD0. I am using LT3970. Input voltage is 9 Volts , output voltage 5 Volts, switching frequency is 600KHz , output current = 300mA. with output capacitor of 22uF. This output goes to LM1117IMP-3.3. I used a LM1117IMP-3.3 LDO to generate 3.3 V output with output capacitor value of 20uF. My load is drawing 60mA. Data Sheet linnk for LT3970 and LM1117IMP are as follows

What I did so far is that I connect the oscilloscope probe at the output of the DC to DC converter, and the second probe to output of the LM1117IMP with 200mV / DIV and 5usec / DIV settings for both channels . I can see that both outputs look like same actually they are same with some noise and ripples but I am unable to correlate with the data sheet. For example LT3970 data sheet says at page 1 that the output ripple will be 5mV peak to peak and page 12 says that with the capacitor value of 20uF the ripple will be 2mV peak to peak.

LM1117 's data sheet shows at page 8 the graph between ripple rejection vs current that at current draw less than 200mA the ripple rejection is higher number @ 20Khz, but LT 3970's switching frequency s 600KHz.

My hope was to see much cleaner signal at the output of the LDO but I see same signal at the input of the LDO and at its output. Oscilloscope probes are 10Mohm inputs and 13pF capacitance. I might be doing somethong wrong in measuring the input and outputs.

The other thing is that part LT1962 has 20uV rms output ripple. How can I compare this with LM 1117IMP's output ripple voltage and how can see the 20uV rms on the scope?

Any suggestions will be appreciated.

John

make yourself a hybrid circuit. a bipolar around the fet. The bipolar can work nicely at those low voltages.

Actually, you could most likely do it all with a bipolar..

Jamie

Hi,

Any recommendation regarding the bipolar part number ! Can I skip FET altogether. Any comments on measuring noise at the outputs of LDO. JOHN

Hi,

Would you explain briefly that what exactly did you mean by hybrid?

Thanks

john

You need to put in some context of prior messages so that us old folks can remember what the thread was about !

When I said hybrid I meant using a combination of a bipolar transistor with a power mosfet..

If that circuit you are dealing with can in deed operate at a much lower level than needed drive for a mosfet, you can couple a NPN transistor around MOSFET circuit for it to perform the same work at the lower levels.

The bipolar will start conducting < 1 volts between Base and emitter.

When the operating drive point gets to ta level where the mosfet gate can turn on, then the bipolar will just sit idle around it.

You would need to study the circuit that is doing the driving to see if you can incorporate this into the circuit. I am sure you can.

Jamie

Hi,

How about just use a BJT and get rid of MOSFET alltogether. Can you give me some schematic example or some application note.

John

Is there is reason you must use a switcher IC ? I mean if memory serves, you're trying to get 5V@300ma from 1..40 volt input? I was thinking of d basic inverter circuit with feed back regulation to throttle back the input side.

Jamie

Hi,

I am ready to use any good idea. I wanted to test the Linear technology claim abt this chip LT3759. I am trying to simulate in LT spice. I not know abt d basic inverter circuit. Any help will be appreciated> what do you think abt the switcher IC. If I can get the right MOSFET, will it work?

john

"john1987" wrote in message=20 news: snipped-for-privacy@z19g2000vbe.googlegroups.com...=

I just downloaded the latest from Linear and used their 3759 jig. I = changed=20 the output resistors for 3.3 volts and I chose a MOSFET that has a Von = of 1=20 Volt (IRLML6244). I also changed the resistors for the EM/UVLO but it = was=20 probably not needed. However, I was only able to get it to work down to =

1.6=20 V. If you need less than that, you will have to try some other tricks. = It=20 certainly seems possible.

But somehow I think this may be either homework or a bar bet. I'll make = it=20 work if you split your gold star or your winnings with me. I accept = PayPal!

Circuit follows.

Paul

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

Version 4 SHEET 1 1172 680 WIRE -96 -96 -176 -96 WIRE 240 -96 -96 -96 WIRE 320 -96 240 -96 WIRE 560 -96 400 -96 WIRE 576 -96 560 -96 WIRE 672 -96 640 -96 WIRE 752 -96 672 -96 WIRE 848 -96 816 -96 WIRE 912 -96 848 -96 WIRE 992 -96 912 -96 WIRE 1072 -96 992 -96 WIRE -176 -80 -176 -96 WIRE 672 -80 672 -96 WIRE 912 -80 912 -96 WIRE 992 -80 992 -96 WIRE 1072 -80 1072 -96 WIRE -96 -64 -96 -96 WIRE 416 -48 0 -48 WIRE 240 -32 240 -96 WIRE -176 16 -176 0 WIRE 672 16 672 0 WIRE 912 16 912 -16 WIRE 992 16 992 -16 WIRE 1072 16 1072 0 WIRE -96 32 -96 16 WIRE 80 32 -96 32 WIRE 416 32 416 -48 WIRE 416 32 400 32 WIRE 432 32 416 32 WIRE 512 32 496 32 WIRE -96 48 -96 32 WIRE 0 48 0 -48 WIRE 80 128 0 128 WIRE 848 128 848 -96 WIRE 848 128 400 128 WIRE -96 144 -96 128 WIRE 560 144 560 -96 WIRE 0 224 -32 224 WIRE 80 224 64 224 WIRE 512 224 400 224 WIRE 560 256 560 240 WIRE 560 256 432 256 WIRE 560 272 560 256 WIRE 848 288 848 128 WIRE -16 320 -32 320 WIRE 80 320 64 320 WIRE 432 320 432 256 WIRE 432 320 400 320 WIRE 560 368 560 352 WIRE -16 416 -32 416 WIRE 80 416 48 416 WIRE 848 416 848 368 WIRE 848 416 400 416 WIRE 848 432 848 416 WIRE 240 496 240 480 WIRE 848 528 848 512 FLAG 240 496 0 FLAG -32 416 0 FLAG -32 320 0 FLAG -32 224 0 FLAG 512 32 0 FLAG -176 16 0 FLAG -96 144 0 FLAG -176 -96 IN FLAG 560 368 0 FLAG 672 16 0 FLAG 848 528 0 FLAG 912 16 0 FLAG 992 16 0 FLAG 1072 16 0 FLAG 1072 -96 OUT SYMBOL PowerProducts\\LT3759 240 224 R0 SYMATTR InstName U1 SYMBOL cap 48 400 R90 WINDOW 39 53 32 VTop 2 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR SpiceLine Rser=3D7.5K SYMATTR InstName C1 SYMATTR Value 22n SYMBOL res 80 304 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R1 SYMATTR Value 41.2K SYMBOL cap 64 208 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C2 SYMATTR Value .05=B5 SYMBOL voltage -176 -96 R0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value 1.6 SYMBOL cap 432 48 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName C3 SYMATTR Value 4.7=B5 SYMBOL res -112 -80 R0 SYMATTR InstName R2 SYMATTR Value 20k SYMBOL res -112 32 R0 SYMATTR InstName R3 SYMATTR Value 118K SYMBOL res -16 32 R0 SYMATTR InstName R4 SYMATTR Value 100K SYMBOL ind2 416 -80 M270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L1 SYMATTR Value 15=B5 SYMATTR Type ind SYMATTR SpiceLine Rser=3D50m SYMBOL ind2 656 -96 R0 SYMATTR InstName L2 SYMATTR Value 15=B5 SYMATTR Type ind SYMATTR SpiceLine Rser=3D50m SYMBOL nmos 512 144 R0 WINDOW 0 56 5 Left 2 WINDOW 3 60 79 Left 2 SYMATTR InstName Q1 SYMATTR Value IRLML6244 SYMBOL res 544 256 R0 SYMATTR InstName R5 SYMATTR Value 5m SYMBOL cap 576 -80 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName C4 SYMATTR Value 4.7=B5 SYMBOL schottky 752 -80 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D1 SYMATTR Value MBRS360 SYMATTR Description Diode SYMATTR Type diode SYMBOL res 832 272 R0 SYMATTR InstName R6 SYMATTR Value 17.5k SYMBOL res 832 416 R0 SYMATTR InstName R7 SYMATTR Value 15.8K SYMBOL polcap 896 -80 R0 WINDOW 3 25 53 Left 2 WINDOW 123 26 78 Left 2 SYMATTR Value 47=B5 SYMATTR Value2 x4 SYMATTR InstName C5 SYMATTR Description Capacitor SYMATTR Type cap SYMATTR SpiceLine V=3D20 Irms=3D0 Rser=3D0.15 Lser=3D0 mfg=3D"KEMET"=20 pn=3D"T495X476M020AS" type=3D"Tantalum" SYMBOL cap 976 -80 R0 SYMATTR InstName C6 SYMATTR Value 10=B5 SYMATTR SpiceLine V=3D10 Irms=3D8.919 Rser=3D0.005 Lser=3D0 = mfg=3D"KEMET"=20 pn=3D"C1210C106K8PAC" type=3D"X5R" SYMBOL res 1056 -96 R0 SYMATTR InstName Rload SYMATTR Value 12 TEXT 496 -104 Bottom 2 !K1 L1 L2 1 TEXT 912 200 Left 2 !.tran 4m startup=20

Hi,

Its not a home work or bar bet, I assure you. An assignment for a junior engineer. I do not know how to open your net list in LT spice. Any pointers? I need 5 volts with atleast 200mA output with less niose as possible. Can you also comment on my questions regarding how to measure ripple voltages and noise reduction techniques in my OP. Thanks

john

The net list is actually an ASCII file which is the usual default for=20 LTSpice designs. Just create a text file which you may name John3759.asc = or=20 whatever you wish, and copy/paste the text and save the file. Then use=20 LTSpice to browse to the file and open it. It is almost identical to the =

supplied 3759.asc which comes with LTSpice in the "jigs" folder, and you =

should know how to open them and run simulations. If you are new to = LTSpice=20 I suggest you take the time to try some of the many tutorials that are=20 available so you become comfortable with the basics. I am by no means an =

expert but I've used the program successfully for many years.

Once you run a simulation (in transient mode) you need to monitor the=20 currents and voltages you are interested in. Then you will see how the=20 circuit starts up and the waveforms over time. To measure ripple = voltages=20 you just need to select a portion of the plot where the output has=20 stabilized, and then use Ctrl-Click to see the average and RMS value of = that=20 portion. You can zoom in and use the cursors to read the high and low=20 excursions of the ripple and divide that by the average to get P-P = ripple as=20 a percentage. You can probably use the math functions of the plot points = to=20 do this automatically.

I am still a little confused about your needs. I see that you want the =

3759=20 to produce 5.0 VDC and then use an LT1962 to produce 3.3 volts. Do you = need=20 both voltages for this design? And I still question your need for such a =

wide range of input voltage. And I just saw that the 3759 is rated for =

1.6=20 to 42 volts input, so you will need to boost your 1.2V to at least 1.6V=20 before it has a chance of working. That can be done, and I have made = some=20 "Joule Thief" circuits using a BJT which produce enough voltage for a = white=20 LED from as low as 0.8V. If you are really serious about this you can = search=20 for those, or I can post the LTSpice ASCII text if you can learn how = that=20 works. Then the rest of the design is up to you, unless you wish to = contract=20 my services.

We are here to help, but mostly to help you learn how to approach design =

problems. Good luck.

Paul=20

Try using the FET from page 27 of the datasheeet, or maybe the FET page

1. The one page 28 will probably give a 2.4V minimum though.

?-)

The LT3759 is only rated to work down to 1.6V. The OP's problem is not = the=20 MOSFET. I provided an ASCII file with a MOSFET that proves the = simulation=20 will work down to 1.6 volts. For the 1.2 volt input required by the OP, = a=20 boost of at least 0.4 volts is needed.

The LT1308A/B will work down to 1.2 V, but the maximum input is 10V. But =

since there is no specification for efficiency, it should work by adding = an=20

8 volt LDO regulator ahead of the IC. The only problem is that it will = have=20 to dissipate at least 300mA * 32V or almost 10 watts for 40 volts input. = Or=20 you could have a voltage sensing relay switch the input from the LT1308 = to=20 the LT3759 when the voltage exceeds the 10V limit of the 1308.

But the big question is why this application needs such a wide input = range.

Paul=20