Maker Pro
Login Join Maker Pro
Or sign in with

Menu

Connect With Us

Network

Maker Pro

micropower boost circuit

M

Michael Robinson

I need a DC to DC boost circuit with an output defined not with respect to
ground, but with respect to the input, which is variable. Output about 15
volts above the input, over a wide range -- from less than a volt in, up to
15 or 16 volts in. But can be inefficient and only needs to put out a few
milliamps. I pasted an LTspice netlist of a possible solution.
R15 simulates a grain-of-wheat incandescent. For Q6, if the Zetex is hard
to find I might use a DSS416 or some such.
What's the consensus about the vulnerability of Q6 when Vin = 16 volts?

Version 4
SHEET 1 1388 680
WIRE 496 -160 -80 -160
WIRE 384 -144 352 -144
WIRE -80 -128 -80 -160
WIRE 352 -112 352 -144
WIRE 512 -112 352 -112
WIRE 560 -112 512 -112
WIRE 352 -80 352 -112
WIRE 448 -80 448 -144
WIRE 448 -80 416 -80
WIRE 512 -64 512 -112
WIRE -528 -16 -528 -48
WIRE -528 -16 -560 -16
WIRE -464 -16 -528 -16
WIRE -368 -16 -464 -16
WIRE -352 -16 -368 -16
WIRE -304 -16 -352 -16
WIRE -288 -16 -304 -16
WIRE -192 -16 -288 -16
WIRE -160 -16 -192 -16
WIRE 208 -16 -160 -16
WIRE 272 -16 208 -16
WIRE 288 -16 288 -112
WIRE 288 -16 272 -16
WIRE 304 -16 288 -16
WIRE 448 -16 448 -80
WIRE 448 -16 384 -16
WIRE -464 0 -464 -16
WIRE -192 0 -192 -16
WIRE 272 32 272 -16
WIRE 384 32 272 32
WIRE -368 48 -400 48
WIRE -256 48 -288 48
WIRE 272 48 272 32
WIRE 496 80 496 -160
WIRE 496 80 448 80
WIRE -464 96 -592 96
WIRE -448 96 -464 96
WIRE -384 96 -288 48
WIRE -352 96 -352 -16
WIRE -272 96 -368 48
WIRE -192 96 -208 96
WIRE -128 96 -192 96
WIRE 144 96 -128 96
WIRE 208 96 208 -16
WIRE -464 112 -464 96
WIRE -192 112 -192 96
WIRE -528 128 -528 -16
WIRE -304 128 -304 -16
WIRE -80 144 -80 -48
WIRE 16 144 -16 144
WIRE 272 144 272 128
WIRE 272 144 96 144
WIRE 352 144 352 -80
WIRE 352 144 336 144
WIRE 272 160 272 144
WIRE -464 208 -464 192
WIRE -416 208 -464 208
WIRE -352 208 -352 176
WIRE -192 208 -192 192
WIRE -192 208 -240 208
WIRE -160 208 -160 64
WIRE -160 208 -192 208
WIRE 208 208 208 176
WIRE 208 208 112 208
WIRE 208 224 208 208
WIRE -560 240 -560 64
WIRE -464 240 -464 208
WIRE -464 240 -560 240
WIRE -352 256 -304 208
WIRE -352 256 -400 256
WIRE -304 256 -352 208
WIRE -256 256 -304 256
WIRE -80 256 -80 144
WIRE -16 256 -80 256
WIRE 48 256 -16 256
WIRE 144 272 144 176
WIRE -592 320 -592 176
WIRE -528 320 -528 208
WIRE -528 320 -592 320
WIRE -464 320 -464 304
WIRE -464 320 -528 320
WIRE -384 320 -384 176
WIRE -384 320 -464 320
WIRE -352 320 -384 320
WIRE -304 320 -352 320
WIRE -272 320 -272 176
WIRE -272 320 -304 320
WIRE -192 320 -192 304
WIRE -192 320 -272 320
WIRE -128 320 -128 176
WIRE -128 320 -192 320
WIRE -80 320 -128 320
WIRE -16 320 -80 320
WIRE 112 320 112 304
WIRE 112 320 -16 320
WIRE 208 320 112 320
WIRE 272 320 272 256
WIRE 272 320 208 320
FLAG 560 -32 0
FLAG -528 -48 Vin
FLAG 512 -64 Vout
FLAG -80 320 0
SYMBOL npn -256 208 R0
SYMATTR InstName Q1
SYMATTR Value 2N2222
SYMBOL npn -400 208 M0
SYMATTR InstName Q2
SYMATTR Value 2N2222
SYMBOL pnp -400 96 R180
WINDOW 0 51 37 Left 2
WINDOW 3 31 60 Left 2
SYMATTR InstName Q3
SYMATTR Value 2N2907
SYMBOL pnp -256 96 M180
WINDOW 0 36 42 Left 2
WINDOW 3 -10 95 Left 2
SYMATTR InstName Q4
SYMATTR Value 2N2907
SYMBOL cap -240 192 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C1
SYMATTR Value 1n
SYMBOL cap -352 192 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C2
SYMATTR Value 1n
SYMBOL cap -384 80 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C3
SYMATTR Value 1n
SYMBOL cap -208 80 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 8 14 VTop 2
SYMATTR InstName C4
SYMATTR Value 1n
SYMBOL res -320 112 R0
WINDOW 0 1 -2 Left 2
WINDOW 3 -13 62 Left 2
SYMATTR InstName R1
SYMATTR Value 47k
SYMBOL res -400 80 R0
WINDOW 0 0 40 Left 2
WINDOW 3 -5 70 Left 2
SYMATTR InstName R3
SYMATTR Value 47k
SYMBOL res -288 80 R0
WINDOW 0 1 42 Left 2
WINDOW 3 -2 72 Left 2
SYMATTR InstName R4
SYMATTR Value 4k7
SYMBOL voltage -528 112 R0
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
WINDOW 3 -42 -248 Left 2
WINDOW 0 -13 58 Left 2
SYMATTR Value PULSE(0 1 0 100n)
SYMATTR InstName V1
SYMBOL res -480 96 R0
WINDOW 0 3 55 Left 2
WINDOW 3 13 80 Left 2
SYMATTR InstName R5
SYMATTR Value 10
SYMBOL res -208 96 R0
WINDOW 0 -16 36 Left 2
WINDOW 3 -2 70 Left 2
SYMATTR InstName R6
SYMATTR Value 10
SYMBOL diode -288 320 R180
WINDOW 0 3 -13 Left 2
WINDOW 3 -84 -14 Left 2
SYMATTR InstName D1
SYMATTR Value 1N4148
SYMBOL diode -336 320 R180
WINDOW 0 88 -13 Left 2
WINDOW 3 -1 -12 Left 2
SYMATTR InstName D4
SYMATTR Value 1N4148
SYMBOL diode -272 48 R180
WINDOW 0 8 108 Left 2
WINDOW 3 -29 82 Left 2
SYMATTR InstName D2
SYMATTR Value 1N4148
SYMBOL diode -352 48 R180
WINDOW 0 31 45 Left 2
WINDOW 3 9 94 Left 2
SYMATTR InstName D3
SYMATTR Value 1N4148
SYMBOL res -368 80 R0
WINDOW 0 6 32 Left 2
WINDOW 3 9 71 Left 2
SYMATTR InstName R2
SYMATTR Value 4k7
SYMBOL res -608 80 R0
WINDOW 0 17 38 Left 2
WINDOW 3 2 88 Left 2
SYMATTR InstName R7
SYMATTR Value 4k7
SYMBOL res -576 -32 R0
WINDOW 0 -3 53 Left 2
WINDOW 3 -12 89 Left 2
SYMATTR InstName R8
SYMATTR Value 4k7
SYMBOL res -144 80 R0
WINDOW 0 15 38 Left 2
WINDOW 3 7 73 Left 2
SYMATTR InstName R9
SYMATTR Value 4k7
SYMBOL res -176 -32 R0
SYMATTR InstName R10
SYMATTR Value 4k7
SYMBOL npn 144 224 R0
SYMATTR InstName Q5
SYMATTR Value 2N2222
SYMBOL npn 208 160 R0
WINDOW 0 56 45 Left 2
SYMATTR InstName Q6
SYMATTR Value FZT849
SYMBOL ind 256 32 R0
WINDOW 3 64 53 Left 2
SYMATTR Value 100µ
SYMATTR InstName L1
SYMATTR SpiceLine Rser=.5
SYMBOL schottky 272 160 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D5
SYMATTR Value 1N5819
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL pnp 384 80 M180
SYMATTR InstName Q8
SYMATTR Value 2N2907
SYMBOL zener 416 -96 R90
WINDOW 0 -4 32 VBottom 2
WINDOW 3 37 -14 VTop 2
SYMATTR InstName D6
SYMATTR Value BZX84C15L
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL cap 352 -128 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C6
SYMATTR Value 1µ
SYMBOL res 544 -128 R0
SYMATTR InstName R14
SYMATTR Value 4k7
SYMBOL res 192 80 R0
WINDOW 3 -254 -27 Left 2
WINDOW 0 -5 2 Left 2
SYMATTR Value R=100+sqrt(54000*V(Vin))
SYMATTR InstName R15
SYMBOL res 128 80 R0
WINDOW 0 -2 40 Left 2
WINDOW 3 -2 83 Left 2
SYMATTR InstName R16
SYMATTR Value 1k
SYMBOL res 400 -32 R90
WINDOW 0 9 81 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R11
SYMATTR Value 1k
SYMBOL npn 48 208 R0
SYMATTR InstName Q7
SYMATTR Value 2N2222
SYMBOL res -96 -144 R0
SYMATTR InstName R12
SYMATTR Value 1k
SYMBOL res 112 128 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R13
SYMATTR Value 10k
SYMBOL cap -16 128 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C7
SYMATTR Value 100p
SYMBOL diode 0 320 R180
WINDOW 0 24 64 Left 2
WINDOW 3 -17 25 Left 2
SYMATTR InstName D7
SYMATTR Value 1N4148
SYMBOL cap 448 -160 R90
WINDOW 0 27 65 VBottom 2
WINDOW 3 5 -6 VTop 2
SYMATTR InstName C5
SYMATTR Value 1n
TEXT -320 -136 Left 2 !.tran 0 5m 0
 
J

Joerg

Michael said:
I need a DC to DC boost circuit with an output defined not with respect to
ground, but with respect to the input, which is variable. Output about 15
volts above the input, over a wide range -- from less than a volt in, up to
15 or 16 volts in. But can be inefficient and only needs to put out a few
milliamps. I pasted an LTspice netlist of a possible solution.
Click to expand...


Wow, did you knead it afterwards or compress it with vise grips? :)

R15 simulates a grain-of-wheat incandescent. For Q6, if the Zetex is hard
to find I might use a DSS416 or some such.
What's the consensus about the vulnerability of Q6 when Vin = 16 volts?
Click to expand...

A 30V transistor is the wrong choice here, it'll like go *PHUT* and I'd
suggest looking at some with at least 50V rating. At 16V input it sees
around 32V and you need some margin. Also, mind ESD. There should be a
capacitor on the VIN side.

[...]
 
M

Michael Robinson

Tim Wescott said:
Snarky comments about the layout of the schematic aside, why don't you
want to use a chip for the switcher, or a chip + op-amp? You're going to
consume a lot of board space with what you have here.
Click to expand...

Can you suggest a chip which will work in a range from less than a volt to
fifteen volts.
 
M

Michael Robinson

Joerg said:
Michael said:
I need a DC to DC boost circuit with an output defined not with respect
to
ground, but with respect to the input, which is variable. Output about
15
volts above the input, over a wide range -- from less than a volt in, up
to
15 or 16 volts in. But can be inefficient and only needs to put out a
few
milliamps. I pasted an LTspice netlist of a possible solution.
Click to expand...


Wow, did you knead it afterwards or compress it with vise grips? :)

R15 simulates a grain-of-wheat incandescent. For Q6, if the Zetex is
hard
to find I might use a DSS416 or some such.
What's the consensus about the vulnerability of Q6 when Vin = 16 volts?
Click to expand...

A 30V transistor is the wrong choice here, it'll like go *PHUT* and I'd
suggest looking at some with at least 50V rating. At 16V input it sees
around 32V and you need some margin. Also, mind ESD. There should be a
capacitor on the VIN side.

[...]
Click to expand...
DSS416 has a 60 volt rating. And thanks for reminding me, I forgot the
bypass.
 
J

Joerg

Tim said:
The LT8410 does what you want, if by "a few mA" you mean 5 or so, and if
you can stand the difference between Vin and Vboost to be close to, but
not exactly, your set value:
Click to expand...

I don't think they work at less than 2.2V and abs max is 16V. Close call
for the abs max because Michael needs to use it up to there.

There are some micro-power devices that will start at less than 1V but I
haven't looked if any of them would go to 18V or more, so there'd be
some margin.

[...]
 
J

Joerg

Tim said:
Tim said:
On Sun, 22 Sep 2013 12:01:32 -0400, Michael Robinson wrote:

I need a DC to DC boost circuit with an output defined not with
respect to ground, but with respect to the input, which is variable.
Output about 15 volts above the input, over a wide range -- from less
than a volt in, up to 15 or 16 volts in. But can be inefficient and
only needs to put out a few milliamps. I pasted an LTspice netlist of
a possible solution.
R15 simulates a grain-of-wheat incandescent. For Q6, if the Zetex is
hard to find I might use a DSS416 or some such.
What's the consensus about the vulnerability of Q6 when Vin = 16
volts?
The LT8410 does what you want, if by "a few mA" you mean 5 or so, and
if you can stand the difference between Vin and Vboost to be close to,
but not exactly, your set value:
Click to expand...
I don't think they work at less than 2.2V and abs max is 16V. Close call
for the abs max because Michael needs to use it up to there.

There are some micro-power devices that will start at less than 1V but I
haven't looked if any of them would go to 18V or more, so there'd be
some margin.

[...]
Click to expand...

I totally missed Michael's "less than 1V" constraint.

You could fix the 16V in problem with an LDO, since efficiency doesn't
matter -- but not with an internal-switch device.
Click to expand...

The LT1615 probably comes close. But only close. This is one of those
jobs that in all likelihood has to be done with discretes.
 
M

Michael Robinson

Tim Wescott said:
The LT8410 does what you want, if by "a few mA" you mean 5 or so, and if
you can stand the difference between Vin and Vboost to be close to, but
not exactly, your set value:

(What a waste of a Sunday -- but it was fun)

Version 4
SHEET 1 1136 680
WIRE 768 -32 304 -32
WIRE -144 -16 -192 -16
WIRE -32 -16 -144 -16
WIRE 16 -16 -32 -16
WIRE 176 -16 96 -16
WIRE 768 -16 768 -32
WIRE -144 16 -144 -16
WIRE 176 32 176 -16
WIRE 304 32 304 -32
WIRE -32 96 -32 -16
WIRE 96 96 -32 96
WIRE 816 96 384 96
WIRE 944 96 816 96
WIRE 976 96 944 96
WIRE -144 112 -144 80
WIRE 944 112 944 96
WIRE 816 128 816 96
WIRE 480 176 384 176
WIRE 672 176 480 176
WIRE 480 192 480 176
WIRE 672 240 672 176
WIRE -352 256 -432 256
WIRE -32 256 -32 96
WIRE 96 256 -32 256
WIRE 416 256 384 256
WIRE -432 272 -432 256
WIRE 416 304 416 256
WIRE 480 304 480 272
WIRE 480 304 416 304
WIRE 480 320 480 304
WIRE 560 320 480 320
WIRE 768 320 768 48
WIRE 768 320 640 320
WIRE 816 320 816 192
WIRE 816 320 768 320
WIRE 944 320 944 192
WIRE 944 320 816 320
WIRE 976 320 944 320
WIRE 480 336 480 320
WIRE -432 384 -432 352
WIRE 240 448 240 320
WIRE 480 448 480 416
WIRE 672 448 672 304
FLAG 240 448 0
FLAG 480 448 0
FLAG 672 448 0
FLAG -432 384 0
FLAG -144 112 0
FLAG -352 256 Vin
FLAG -192 -16 Vin
FLAG 976 320 Vin
FLAG 976 96 Vout
SYMBOL PowerProducts\\LT8410 240 176 R0
SYMATTR InstName U1
SYMBOL ind 0 0 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L1
SYMATTR Value 470µ
SYMBOL cap 752 -16 R0
SYMATTR InstName C1
SYMATTR Value 100n
SYMBOL cap 800 128 R0
SYMATTR InstName C2
SYMATTR Value 100n
SYMBOL cap 656 240 R0
SYMATTR InstName C3
SYMATTR Value 100n
SYMBOL res 464 176 R0
SYMATTR InstName R1
SYMATTR Value 772k
SYMBOL res 464 320 R0
SYMATTR InstName R2
SYMATTR Value 559k
SYMBOL cap -160 16 R0
SYMATTR InstName C4
SYMATTR Value 100n
SYMBOL voltage -432 256 R0
WINDOW 123 0 0 Left 2
WINDOW 39 24 124 Left 2
SYMATTR InstName V1
SYMATTR Value PWL(0 12 400u 12 500u 5 3000u 16 5000u 5)
SYMATTR SpiceLine Rser=10
SYMBOL res 928 96 R0
SYMATTR InstName R3
SYMATTR Value 3300
SYMBOL res 656 304 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R4
SYMATTR Value 10meg
TEXT -424 432 Left 2 !.tran 5m

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Click to expand...

You went to a lot of trouble, Tim. The least I can do is explain why the
circuit has to run on a volt, so most likely a chip won't work.
Vcc (the armature voltage of an antique six volt generator) isn't going to
start at some higher voltage and drop to a volt, which might allow
bootstrapping. When the generator first spins up, Vcc comes up to about a
volt because of residual magnetism in the soft iron pole shoes (possibly
even a little less than a volt). Whereupon, nothing more will happen:
unless the booster produces some output, generator output will remain flat
at one volt. The booster has to take that one volt and boost it enough to
turn on a mosfet, and drive current in the field coils around the pole
shoes. Then Vcc (the generator output) will rise.
The circuit I posted only needs two thirds of a volt to run, according to
the simulator. Eight transistors and about thirty passives. It'll take up
some board space, as you pointed out, but those forty jelly beans are still
cheaper than the chip. And I think the LT8410 requires 2.5 volts.
 
M

Michael Robinson

Tim Wescott said:
BTW:

To get perfect* tracking, set

1/R4 = (1/31.85) * (1/R1 + 1/R2 + 1/R4)

To get your desired offset voltage, Vd, set

R4/R1 = Vd / Vref

Where the 31.85 and Vref come from the LT8410 data sheet.

Or:

R4 = whatever, but make it big enough for the chip (I chose 10 meg --
that's big).

Then calculate R1 from R4 and your desired Vd.

Then calculate R2 from R1, R4, and your desire for perfect tracking.

* You'll never really perfect tracking, because (a) the 31.85 comes from
the chip design, and will vary from chip to chip and probably with
temperature and aging, and (b) you'll need to set your resistor ratios
perfectly, which you can't, and (anyway) the resistor ratios will change,
and R4 needs to be big to keep the chip happy, meaning you'll have
leakage issues.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Click to expand...
Tim, my requirements are a lot more forgiving on the output side. No need
for accuracy, it can float up or down a volt without any ill effects at all.
The tough part is on the input. I can't use a chip that requires 2.5 volts
to run.
You went to a lot of trouble and the least I can do is explain why the
circuit has to run on a volt.
Vcc (the armature voltage of an antique six volt generator) isn't going to
start at some higher voltage and drop to a volt, which might allow
bootstrapping. Vcc starts at zero and comes up to about a volt because of
residual magnetism in the soft iron pole shoes (possibly even a little less
than a volt). Whereupon, nothing more will happen: unless the booster
produces some output, generator output will remain flat at one volt. The
booster has to take that one volt and boost it enough to turn on a mosfet,
and get some current going in the field coils around the pole shoes. Then
will Vcc (the generator output) rise.
The circuit I posted only needs two thirds of a volt to run, according to
the simulator.
 
J

Jasen Betts

Tim, my requirements are a lot more forgiving on the output side. No need
for accuracy, it can float up or down a volt without any ill effects at all.
The tough part is on the input. I can't use a chip that requires 2.5 volts
to run.
You went to a lot of trouble and the least I can do is explain why the
circuit has to run on a volt.
Vcc (the armature voltage of an antique six volt generator) isn't going to
start at some higher voltage and drop to a volt, which might allow
bootstrapping. Vcc starts at zero and comes up to about a volt because of
residual magnetism in the soft iron pole shoes (possibly even a little less
than a volt). Whereupon, nothing more will happen: unless the booster
produces some output, generator output will remain flat at one volt. The
booster has to take that one volt and boost it enough to turn on a mosfet,
and get some current going in the field coils around the pole shoes. Then
will Vcc (the generator output) rise.
The circuit I posted only needs two thirds of a volt to run, according to
the simulator.
Click to expand...

can't use a depletion mode mosfet or a Normally closed relay?
 
R

rickman

I don't think they work at less than 2.2V and abs max is 16V. Close call
for the abs max because Michael needs to use it up to there.

There are some micro-power devices that will start at less than 1V but I
haven't looked if any of them would go to 18V or more, so there'd be
some margin.
Click to expand...

In an automotive app this may well need a lot more than 18 volts. I'm
not certain where this is connected in the wiring, but load dumps in
autos often produce voltages well over the battery voltage, by 10's of
volts.

We were designing a PSU for a military vehicle and had a transient spec
of 100 volts! I think the spec in today's cars is some 40 or 50 volts.
 
J

Joerg

rickman said:
In an automotive app this may well need a lot more than 18 volts. I'm
not certain where this is connected in the wiring, but load dumps in
autos often produce voltages well over the battery voltage, by 10's of
volts.

We were designing a PSU for a military vehicle and had a transient spec
of 100 volts! I think the spec in today's cars is some 40 or 50 volts.
Click to expand...

The OP didn't say it's for automotive. If it is, then it either needs to
be 80V tolerant or have a surge disconnect.
 
R

rickman

The OP didn't say it's for automotive. If it is, then it either needs to
be 80V tolerant or have a surge disconnect.
Click to expand...

I guess technically he didn't say "automotive". I assumed when I read
"six volt generator" it was for a car.
 
J

Joerg

rickman said:
I guess technically he didn't say "automotive". I assumed when I read
"six volt generator" it was for a car.
Click to expand...

If so, then that must be a really, really old one. Like my 16-horse
Citroen when I was young :)

Its 6V DC generator was not capable of a serious load dump. In fact, I
rode it sans battery most of the time because 6V batteries were
expensive. Yet I still used the high beams, the radio and so on. The
crank gave me a good workout, especially in winter.
 
R

rickman

If so, then that must be a really, really old one. Like my 16-horse
Citroen when I was young :)
Click to expand...

Why else would he be building his own regulator? I figured it was for
an antique auto. The OP said, "(the armature voltage of an antique six
volt generator)".

Its 6V DC generator was not capable of a serious load dump. In fact, I
rode it sans battery most of the time because 6V batteries were
expensive. Yet I still used the high beams, the radio and so on. The
crank gave me a good workout, especially in winter.
Click to expand...

I thought it was the starter motors that produced the load dump?

But then as I think about it a bit more, this is not really a regulator
circuit as much as it is a boot circuit. In a car you have a battery to
supply the field coil current on startup. So maybe this is for a six
volt battery, but not a car. I just can't think of another reason to
use "an antique six volt generator"... maybe to power an antique six
volt radio?
 
J

Joerg

rickman said:
Why else would he be building his own regulator? I figured it was for
an antique auto. The OP said, "(the armature voltage of an antique six
volt generator)".
Click to expand...

That was after my post and my crystal ball is in the shop right now :)

I thought it was the starter motors that produced the load dump?
Click to expand...

Usually it's the alternator, the kind with rectifiers. Happens if the
battery suddenly comes off or, if operating without battery, a major
load such as the rear window defroster goes offline.

But then as I think about it a bit more, this is not really a regulator
circuit as much as it is a boot circuit. In a car you have a battery to
supply the field coil current on startup. So maybe this is for a six
volt battery, but not a car. I just can't think of another reason to
use "an antique six volt generator"... maybe to power an antique six
volt radio?
Click to expand...

I don't know if there even are 6V alternators, except maybe for smaller
two-wheelers. They were usually brushed DC generators plus an external
regulator. With the one in my Citroen 2CV I was quite impressed. It had
the old style non-electronic point contact "bzzzt" regulator. Load dumps
left it fairly unfazed, I checked that with a scope to avoid wrecking my
precious old Becker car radio with shortwave range. It contained Ge
transistors that had become nearly unobtanium by then.

Best of all, the whole car had no belts. The generator sat right on the
crankshaft in a coaxial fashion.
 
M

Michael Robinson

Tim Wescott said:
If this is all so that you can turn on the high-side of an N-channel
MOSFET, why not rethink? Why not use a PNP BJT to turn it on?

You may end up with fewer parts if you use a light-duty PNP to get things
going until a chip-based solution can kick in, and then just run with
that.

Or, you may end up with the fewest parts if you just use a heavy-duty
enough PNP as your switch, and forget about using a MOSFET entirely.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Click to expand...

These generators' fields don't switch on the high side.
There are not lots of solid-state regulators available on the market for
people that want to upgrade old six-volt bikes to solid-state from the
original technology, mechanical devices.
I only know of one company that makes six volt solid-state regulators for
these old bikes, which have fields that switch on the low side (negative
ground). And I think those regulators require a battery on the bike to
work.
However, some people use magnetos, and run without a battery. Nobody makes
regulators for them, because it's a pain to design. These marvelous
suggestions you've made aren't going to solve the problem, unfrotunately...
I've come up with a design that takes extremely low voltage and boosts it,
and it uses only jelly-bean parts: eight transistors and about thirty
passives. In the original post I asked for some feedback from the people on
the forum about the transistor that switches the inductor.
 
M

Michael Robinson

Joerg said:
Tim said:
Tim Wescott wrote:
On Sun, 22 Sep 2013 12:01:32 -0400, Michael Robinson wrote:

I need a DC to DC boost circuit with an output defined not with
respect to ground, but with respect to the input, which is variable.
Output about 15 volts above the input, over a wide range -- from less
than a volt in, up to 15 or 16 volts in. But can be inefficient and
only needs to put out a few milliamps. I pasted an LTspice netlist of
a possible solution.
R15 simulates a grain-of-wheat incandescent. For Q6, if the Zetex is
hard to find I might use a DSS416 or some such.
What's the consensus about the vulnerability of Q6 when Vin = 16
volts?
The LT8410 does what you want, if by "a few mA" you mean 5 or so, and
if you can stand the difference between Vin and Vboost to be close to,
but not exactly, your set value:


I don't think they work at less than 2.2V and abs max is 16V. Close call
for the abs max because Michael needs to use it up to there.

There are some micro-power devices that will start at less than 1V but I
haven't looked if any of them would go to 18V or more, so there'd be
some margin.

[...]
Click to expand...

I totally missed Michael's "less than 1V" constraint.

You could fix the 16V in problem with an LDO, since efficiency doesn't
matter -- but not with an internal-switch device.
Click to expand...

The LT1615 probably comes close. But only close. This is one of those
jobs that in all likelihood has to be done with discretes.
Click to expand...
BINGO!
Takes eight transistors.
 
J

Joerg

Michael said:
Joerg said:
Tim said:
On Sun, 22 Sep 2013 16:32:58 -0700, Joerg wrote:

Tim Wescott wrote:
On Sun, 22 Sep 2013 12:01:32 -0400, Michael Robinson wrote:

I need a DC to DC boost circuit with an output defined not with
respect to ground, but with respect to the input, which is variable.
Output about 15 volts above the input, over a wide range -- from less
than a volt in, up to 15 or 16 volts in. But can be inefficient and
only needs to put out a few milliamps. I pasted an LTspice netlist of
a possible solution.
R15 simulates a grain-of-wheat incandescent. For Q6, if the Zetex is
hard to find I might use a DSS416 or some such.
What's the consensus about the vulnerability of Q6 when Vin = 16
volts?
The LT8410 does what you want, if by "a few mA" you mean 5 or so, and
if you can stand the difference between Vin and Vboost to be close to,
but not exactly, your set value:


I don't think they work at less than 2.2V and abs max is 16V. Close call
for the abs max because Michael needs to use it up to there.

There are some micro-power devices that will start at less than 1V but I
haven't looked if any of them would go to 18V or more, so there'd be
some margin.

[...]
I totally missed Michael's "less than 1V" constraint.

You could fix the 16V in problem with an LDO, since efficiency doesn't
matter -- but not with an internal-switch device.
Click to expand...
The LT1615 probably comes close. But only close. This is one of those
jobs that in all likelihood has to be done with discretes.
Click to expand...
BINGO!
Takes eight transistors.
Click to expand...


And cheap ones to boot :)
 
You must log in or register to reply here.

Similar threads

M
Micropower charge pump for 3-6v @ 20uA
2
Replies
31
Views
2K
Nemo
N
R
Micropower amp disabling
Replies
4
Views
1K
whit3rd
W
T
RCA ICAN-6539 by S.S. Eaton, "Micropower Crystal Oscillator Design"
Replies
7
Views
1K
TonyG
T
C
Rf Detector micropower linearizing bias generator
Replies
1
Views
828
Mark
M
A
low frequency micropower vco needed
2
Replies
34
Views
2K
Jim Thompson
J
Top