Effienciency of LT3479

Hello,

to supply the -2.0 Volt termination voltage for 16 differential ECL channels from a 5 Volt supply, I dicided use a LT3479. The circuit is like the application sheet circuit "5V to ???5V/600mA Inverting DC/DC Converter", adapted to to -2.0Volt. However efficiency is bad, below 40 %. Simulation (scad data below) confirms this data.

What in the datasheet did I miss that this result dissapoints me? I must admit, I didn't simulate beforehand...

-- Uwe Bonnes snipped-for-privacy@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt

--------- Tel. 06151 162516 -------- Fax. 06151 164321 ---------- Version 4 SHEET 1 1308 680 WIRE -192 80 -192 64 WIRE -192 176 -192 160 WIRE -128 64 -192 64 WIRE -80 160 -80 128 WIRE -80 304 -112 304 WIRE -80 304 -80 224 WIRE -64 304 -80 304 WIRE -48 384 -48 368 WIRE -48 480 -48 464 WIRE -16 64 -48 64 WIRE -16 128 -80 128 WIRE -16 128 -16 64 WIRE -16 176 -16 128 WIRE -16 240 -16 176 WIRE 16 176 -16 176 WIRE 16 240 -16 240 WIRE 16 304 0 304 WIRE 16 368 -48 368 WIRE 64 64 -16 64 WIRE 64 112 64 64 WIRE 176 112 176 64 WIRE 176 432 176 416 WIRE 192 64 176 64 WIRE 288 64 272 64 WIRE 288 112 288 64 WIRE 352 64 288 64 WIRE 352 368 336 368 WIRE 384 176 336 176 WIRE 384 192 384 176 WIRE 432 176 384 176 WIRE 432 240 336 240 WIRE 448 368 432 368 WIRE 448 384 448 368 WIRE 448 464 448 448 WIRE 464 64 416 64 WIRE 464 144 464 128 WIRE 480 176 432 176 WIRE 528 64 464 64 WIRE 592 176 544 176 WIRE 592 304 336 304 WIRE 592 304 592 176 WIRE 656 240 432 240 WIRE 656 288 656 240 WIRE 736 64 592 64 WIRE 736 80 736 64 WIRE 736 176 592 176 WIRE 736 176 736 160 WIRE 736 192 736 176 WIRE 736 288 656 288 WIRE 736 288 736 272 WIRE 848 64 736 64 WIRE 848 96 848 64 WIRE 848 192 848 160 WIRE 928 64 848 64 WIRE 928 80 928 64 WIRE 928 192 928 160 FLAG 448 464 0 FLAG 176 432 0 FLAG -112 304 0 FLAG -48 480 0 FLAG -192 176 0 FLAG 848 192 0 FLAG 928 192 0 FLAG 928 64 OUT FLAG -192 64 IN FLAG 464 144 0 FLAG 384 192 0 SYMBOL PowerProducts\\\\LT3479 176 256 R0 SYMATTR InstName U1 SYMBOL ind 176 80 R270 WINDOW 0 32 56 VTop 0 WINDOW 3 5 56 VBottom 0 SYMATTR InstName L1 SYMATTR Value 4.7µ SYMBOL res 336 384 R270 WINDOW 0 32 56 VTop 0 WINDOW 3 0 56 VBottom 0 SYMATTR InstName R1 SYMATTR Value 10K SYMBOL cap 432 384 R0 SYMATTR InstName C1 SYMATTR Value 1n SYMBOL cap 0 288 R90 WINDOW 0 0 32 VBottom 0 WINDOW 3 32 32 VTop 0 SYMATTR InstName C2 SYMATTR Value 10n SYMBOL res -64 368 R0 SYMATTR InstName R2 SYMATTR Value 17.8K SYMBOL voltage -192 64 R0 SYMATTR InstName V1 SYMATTR Value 5 SYMBOL schottky 592 48 R90 WINDOW 0 0 32 VBottom 0 WINDOW 3 32 32 VTop 0 SYMATTR InstName D1 SYMATTR Value PMEG2020AEA SYMATTR Description Diode SYMATTR Type diode SYMBOL res 720 64 R0 SYMATTR InstName R3 SYMATTR Value 160k SYMBOL res 720 176 R0 SYMATTR InstName R4 SYMATTR Value 100k SYMBOL cap 832 96 R0 SYMATTR InstName C3 SYMATTR Value 10µ SYMBOL res 912 64 R0 SYMATTR InstName R5 SYMATTR Value 4 SYMBOL cap 416 48 R90 WINDOW 0 0 32 VBottom 0 WINDOW 3 32 32 VTop 0 SYMATTR InstName C4 SYMATTR Value 10µ SYMATTR SpiceLine Rser=0.01 SYMBOL schottky 448 64 R0 SYMATTR InstName D2 SYMATTR Value PMEG2020AEA SYMATTR Description Diode SYMATTR Type diode SYMBOL cap 416 176 R0 SYMATTR InstName C5 SYMATTR Value 100p SYMBOL schottky 480 192 R270 WINDOW 0 32 32 VTop 0 WINDOW 3 0 32 VBottom 0 SYMATTR InstName D3 SYMATTR Value BAT54 SYMBOL res -32 48 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R6 SYMATTR Value 0.1 SYMBOL cap -96 160 R0 SYMATTR InstName C6 SYMATTR Value 1m TEXT 518 346 Left 0 !.tran 10m startup TEXT 288 16 Bottom 0 ;LT3479 - 3A, 300KHz to 3.5MHz, Full Featured Step-up DC/DC Converter\\nInput: 5V Output: 12V at 0.8A TEXT -256 496 Left 0 ;Note:\\n If the simulation model is not found please update with the "Sync Release" command from the "Tools" menu.\\n It remains the customer's responsibility to verify proper and reliable operation in the actual application.\\n Component substitution and printed circuit board layout may significantly affect circuit performance or reliability\\n Contact your local sales representative for assistance. This circuit is distributed to customers only for use with LTC parts\\n Copyright © 2006 Linear Technology Inc. All rights reserved.

Reply to
Uwe Bonnes
Loading thread data ...

Maybe you could tell what format your file is in ..?

Btw.. where did you get the LTC chips ..? Seems hard to get in 20..200 quantities.

Reply to
pbdelete

It is the .asc format of LT's switchercad. Thanks to LT nad Mike Engelhard for providing these tools for free.

You can credit card order them directly from LT. Click on the "Buy" link. Expect about 18 $ packaging fee and be prepared to pay local VAT/duties.

Bye

--
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply to
Uwe Bonnes

Are you some sort of thick old cunt who needs someone else to chew his custard for him but knows how to vote in Eurovision?

DNA

Reply to
Genome

Hello Uwe,

There is a mistake in the drawing of the datasheet for the negative voltage converter. They forgot the second inductor. Just add an inductor with 4.7uH between D1 and D2.

Best regards, Helmut

"Uwe Bonnes" schrieb im Newsbeitrag news:e6sm6d$7vh$ snipped-for-privacy@lnx107.hrz.tu-darmstadt.de...

Reply to
Helmut Sennewald

Hello Uwe,

40% seems a bit low but remember that you are losing a lot in the two diodes you need for an inverting DC/DC scheme. Especially at low voltages like 2V. The only way to get around that is to rig a synchronous scheme. Don't know any chips for that off the top of my head though since I do most of my converter designs with half-bridges or H-brigdes, or transistors.
--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

At 40%, you may as well just get it from -5 with a linear reg, or just a biggish resistor.

Something like 400 mA is needed here, depending on the trace impedances.

John

Reply to
John Larkin

Hello John,

Yes, or do some kind of PWM thing. But only if it has a provision that it retreats into the "good corner" when the clock fails.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

Thanks,

the additional inductor did the trick.

--
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply to
Uwe Bonnes

Thanks,

the additional inductor did the trick. I am not at 60 %.

--
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply to
Uwe Bonnes

Hello Uwe,

I wonder about your result. I have got 68.5%.

LTspice has a built-in efficiency calculation. It requires one source V1 and one load I1 in your circuit. Therefore I replaced the 4 Ohm load with a 0.5A current source. Additionally the option "steady" is required in the .TRAN command.

Best regards, Helmut

--- Efficiency Report ---

Efficiency: 68.5%

Input: 1.44W @ 5V Output: 988mW @-1.98V

Ref. Irms Ipeak Dissipation C1 0mA 0mA 0mW C2 0mA 0mA 0mW C3 108mA 189mA 0mW C4 389mA 680mA 2mW C5 0mA 0mA 0mW C6 108mA 192mA 0mW D1 511mA 688mA 166mW D2 646mA 1144mA 180mW D3 0mA 0mA 0mW L1 301mA 470mA 0mW L2 511mA 688mA 0mW R1 0mA 0mA 0µW R2 0mA 0mA 91µW R3 0mA 0mA 24µW R4 0mA 0mA 15µW R5 0mA 0mA 1µW R6 289mA 290mA 8mW U1 487mA 1145mA 97mW

Reply to
Helmut Sennewald

You can even replace D2 with the inductor, resulting in a diode drop less losses, like in the LT1611 demo circuit in switchercad.

--
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply to
Uwe Bonnes

I think the Switchercad Model overestimated efficency. Looking at the loss data, is seems like only Switch VCESAT is taken into account, but not the base current driving the switch.

For example, the original demo circuit, which you find when you go the the LT3479 page, simulates the data sheet circuit "5V to 12V/800mA 1MHz Boost Converter" At a load of 800 mA, SCAD3 calculates an efficency of 91%, while the datasheet only gives about 85% in that situation. This is about 0.75 Watt difference, or 150mA@5V. The simulation tells about an RMS current of 1.2A, so an effective beta of 10 for the bipolar switch could account for the errors. Applying this Transistor base current theory to my -2.0 volt switcher could also explain the difference between simulation and the real circuit.

Anyways, thanks for the help!

--
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
Reply to
Uwe Bonnes

Hello Uwe,

If you take alook to the startup, you will see that the output voltage is a few hundred millivolt positive in this case for the first 300us. The diode prevents this positive output voltage at the beginning. If this isn't an issue in your application, then indeed it improves efficiency a lot.

Best regards, Helmut

Reply to
Helmut Sennewald

Ac coupling helps in situations like this.

John

Reply to
John Larkin

Replacing D2 shown with an L2 makes the circuit the original Cuk converter, low ripple and high efficiency, with Vout=-Vin*Duty or D=50% in this case. I would consider putting D2 in parallel with C3 , cathode to GND, so no steady state current, and C2 can be initialized to its steady state 5V in shutdown. Then the feedback shows Vout/5+Vref*4/5=0 or Vout=-4*Vref=-5V.

Reply to
Fred Bloggs

Hello John,

Yep, plus an angst resistor. That's how I do it on switchers.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

Joerg wrote: ...

German "Angst"?

-- Uwe Bonnes snipped-for-privacy@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt

--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Reply to
Uwe Bonnes

Hello Uwe,

Yes. It's a pretty common word in the US now along with lots of others:

formatting link

My wife and I just finished stacking the firewood for next winter. So, we are looking forward to cooling off with a nice Hefeweizen. Oh wait, that was another one of those words.

--
Regards, Joerg

http://www.analogconsultants.com
Reply to
Joerg

Good idea! I have some in the fridge. It's already 101°F outside :-(

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.