Very high gain transformerless boost converter - is it possible

On Sep 4, 2018, Tim Williams wrote (in article ):

The Honda CR-V has a CVT transmission. I have a CR-V, and it works just fine. I guess I don?t get the point of the WPI effort.

Joe Gwinn

Those are standard in power distribution as well -- tap changers. :)

Tim

I'll assume a (good) infinitely variable gearbox

Yeah I have a Nissan Rogue with a cvt. Funny but they use software to imitate gear changes. I wouldn't be uncomfortable with the constant change of ratio but their marketing folks disagree I guess.

Basically everything wrong with mass-market / consumer products is because people are... dumb. Facebook extracts ad revenue by being just annoying enough. Microsoft has annoying features because of many reasons, among them backwards compatibility, point-and-click users, and adding proprietary features to an otherwise standard environment. And cars have sounds and shifting because people want their cars to be bad. :-/

Tim

onsdag den 5. september 2018 kl. 22.25.54 UTC+2 skrev Tim Williams:

a cvt shifting in steps could possible be a matter of efficiency and wear

Clockworks?

even a toy one is going to take a couple of watts to run. Output what, a microamp? Maybe we'll see how wrong I am.

NT

gray_wolf wrote in news:QsViD.13417$ snipped-for-privacy@fx28.iad:

What follows is an LT Spice representation of it. The drive sources are out of polarity of the same square wave. My circuit values were to get about 9 or 10 amps at 36 volts. The caps are like 10,000 uF CGR or CGS quality, lo w ESR and the diodes are Shottkeys. The transistors I was thinking of 2N588

If 150 watts is desired, this circuit should supply about 200, after more c onversions that should be enough for half decent regulation without much ad o.

Version 4 SHEET 1 880 1208 WIRE 48 -288 -720 -288 WIRE 480 -96 -416 -96 WIRE -720 -64 -720 -208 WIRE -64 64 -224 64 WIRE 48 64 48 -288 WIRE 48 64 -64 64 WIRE 160 64 48 64 WIRE 336 64 160 64 WIRE -64 112 -64 64 WIRE 160 112 160 64 WIRE -544 224 -784 224 WIRE -288 224 -288 112 WIRE -288 224 -544 224 WIRE -224 224 -224 160 WIRE -64 224 -64 176 WIRE -64 224 -160 224 WIRE 160 224 160 176 WIRE 272 224 160 224 WIRE 336 224 336 160 WIRE 400 224 400 112 WIRE 480 224 480 -96 WIRE 480 224 400 224 WIRE 336 256 336 224 WIRE -224 272 -224 224 WIRE 400 304 400 224 WIRE -288 320 -288 224 WIRE -784 400 -784 304 WIRE -544 480 -544 224 WIRE 480 480 -544 480 WIRE 48 608 48 64 WIRE 48 608 -224 608 WIRE 336 608 48 608 WIRE -224 624 -224 608 WIRE -64 704 -64 224 WIRE 160 704 160 224 WIRE 160 768 160 752 WIRE -416 816 -416 -96 WIRE -416 816 -784 816 WIRE -288 816 -288 672 WIRE -288 816 -416 816 WIRE -224 816 -224 720 WIRE -64 816 -64 768 WIRE -64 816 -160 816 WIRE 160 816 160 768 WIRE 272 816 160 816 WIRE 336 816 336 704 WIRE 400 816 400 656 WIRE 480 816 480 480 WIRE 480 816 400 816 WIRE -64 880 -64 816 WIRE 160 880 160 816 WIRE -224 896 -224 816 WIRE 336 896 336 816 WIRE -288 944 -288 816 WIRE 400 944 400 816 WIRE -784 976 -784 896 WIRE -64 1200 -64 944 WIRE 160 1200 160 944 WIRE 160 1200 -64 1200 WIRE 864 1200 160 1200 FLAG -224 384 0 FLAG 336 368 0 FLAG -224 1008 0 FLAG 336 1008 0 FLAG -720 -48 0 FLAG -784 400 0 FLAG -784 976 0 SYMBOL npn 400 64 M0 SYMATTR InstName Q2 SYMBOL pnp 400 352 R180 SYMATTR InstName Q3 SYMBOL pnp -288 368 M180 SYMATTR InstName Q1 SYMBOL npn -288 64 R0 SYMATTR InstName Q4 SYMBOL npn -288 624 R0 SYMATTR InstName Q5 SYMBOL npn 400 608 M0 SYMATTR InstName Q6 SYMBOL pnp 400 992 R180 SYMATTR InstName Q7 SYMBOL pnp -288 992 M180 SYMATTR InstName Q8 SYMBOL cap -160 208 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C1 SYMBOL cap 336 208 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C2 SYMBOL cap -160 800 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C3 SYMBOL cap 336 800 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C4 SYMBOL diode -80 112 R0 SYMATTR InstName D1 SYMBOL diode 144 112 R0 SYMATTR InstName D2 SYMBOL diode 144 704 R0 SYMATTR InstName D3 SYMBOL diode -80 704 R0 SYMATTR InstName D4 SYMBOL diode 144 880 R0 SYMATTR InstName D5 SYMBOL diode -80 880 R0 SYMATTR InstName D6 SYMBOL voltage -720 -304 R0 SYMATTR InstName V1 SYMATTR Value V +12 SYMBOL voltage -784 208 R0 SYMATTR InstName V2 SYMATTR Value 0-12 SQUARE SYMBOL voltage -784 800 R0 SYMATTR InstName V3 SYMATTR Value 12-0 SQUARE TEXT 664 1168 Left 2 ;36 V OUTPUT

This circuit cascaded, rather than one main voltage multiplier, should be m ore solid as far as regulation. Each multiplier in the stage would use diff erent components chosen successively for more voltage and less current, wit

ginal 12 volts is tripled twice only a doubler is needed. If push pull oper ation is maintained very little filtering will be needed.

Originally for whatever reasons, I chose the transistors for their low Vces at and the diodes were 1N6095 but I am sure better are available now. The c ap values are overkill seemingly but that is done for the low ESR. Junction drops and ESR are the only things keeping the gizmo from 100 % efficiency because I chose a quite low frequency. It would run just fine at 100 Hz, at that speed you don't need high speed anything so switching time vanishes a s a concern. No inductive load so no significant dv/dt.

I believe it is the most efficient way to do it, if one wants 150 watts. St ack those things up. Only the first stage is going to need any significant heatsinking or low Vcesat transistors or even Shottkeys, that inefficiency practically vanishes in the subsequent stages.

Maybe just one multiplier and one output stage is better, if so, show me.

Note that I would not run it at 100 Hz, that is simply too low.

Also this was for an amp in cars that would have the amp under the dash, re member these were 1970s cars, a three wire cable would go out under the hoo d to this unit and carry the square wave drive to it and bring back the 35 volts. The way the transistor bases are connected, it would run underwater. Teh unit IS actually built, if I had hosting I would upload a picture. The only thing it lacks is the transistors.

The amp was something else, knowing the assholes around I made it bulletpro of. You could connect the speaker wires to the battery, or even into an AC outlet and all that would happen is the wires would burn off. If the unit w as running at the time all that had to be done was to switch is off and bac k on to reset it. The amp was BTL so it was better than 50 WPC. Total DC co upled, clip proof and a set of tone controls that would put hair on your ch est, curl it and take it off all in one fell swoop.

We are talking late 1970sd here, this was unheard of in a car back then. S ure there were PA amps (like Bogen n shit) that had an output transformer a nd would put out some power but they sounded like shit.

There was a chance of working if you connected the lower collectors.

Don't those little arrows mean common ?

Are there wires between the collectors and the arrows?

torsdag den 6. september 2018 kl. 14.41.10 UTC+2 skrev snipped-for-privacy@gmail.com:

ut of polarity of the same square wave. My circuit values were to get about 9 or 10 amps at 36 volts. The caps are like 10,000 uF CGR or CGS quality, low ESR and the diodes are Shottkeys. The transistors I was thinking of 2N5

conversions that should be enough for half decent regulation without much ado.

looks like two Dickson multipliers in parallel

Umm, actually no. If the heatsinks are grounded it is a simple matter to leave out the insulators, assuming a metal tab or as in the case of my little box the cases of TO-3 transistors.

Somebody called it "a" Dickson convertor. I went with that. That came out of my head but like all the other things I invented I was not the first. (I got all of them with my DVD rewinder though)

Well, one could use a wristwatch-type escapement and a fluid damper. Might not be super durable!

Cheers

Phil Hobbs

There are several theoretically lossless variable-ratio gadgets. One uses a belt that can be positioned between tapered shafts. Another moves a roller at adjustable radius from a rotating disk. The latter was used in analog computers for aiming guns and torpedoes.

You can see such a computer onboard the submarine Bowfin at Pearl Harbor. It's worth the trip if you are ever in Hawaii.