Maker Pro
Maker Pro

Hybrid linear/switching supply

G

Gibbo

Jan 1, 1970
0
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input voltage.
5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive ISO
7637-2. That dictates upto -600V spikes with rise times of 1uS or so and
repetive +200V spikes with similar rise times.

So I thought about switching but due to a few problems this isn't really
workable. Mainly due to ISO 7637-2 and the high +ve voltages causing a
problem with the size of the inductor.

So I thought maybe linear at input voltages upto the first regulated
output voltage (10V), then switching regulation upto the max input V of
34V. Input voltages above that then revert to a linear reg feeding the
switcher (because voltages higher than this are only spikes so
dissipation isn't such a problem). Sounds ridiculously complicated. But
I got this....

Version 4
SHEET 1 1176 776
WIRE 48 208 -512 208
WIRE 208 208 112 208
WIRE 384 208 208 208
WIRE 512 208 384 208
WIRE 208 224 208 208
WIRE 384 288 384 208
WIRE -512 336 -512 208
WIRE 112 336 -32 336
WIRE 208 336 208 304
WIRE 208 336 112 336
WIRE 320 336 256 336
WIRE 112 352 112 336
WIRE 512 384 512 208
WIRE 944 400 640 400
WIRE 1024 400 944 400
WIRE 1120 400 1024 400
WIRE 256 416 256 336
WIRE 256 416 208 416
WIRE 384 432 384 384
WIRE 448 432 384 432
WIRE 112 448 112 432
WIRE 272 448 112 448
WIRE 944 448 944 400
WIRE 1120 448 1120 400
WIRE 528 480 512 480
WIRE 640 480 640 400
WIRE 640 480 608 480
WIRE 640 496 640 480
WIRE 656 496 640 496
WIRE 768 496 736 496
WIRE 848 496 768 496
WIRE 880 496 848 496
WIRE -32 512 -32 336
WIRE -80 528 -336 528
WIRE 272 528 272 448
WIRE 640 528 640 496
WIRE -336 544 -336 528
WIRE 112 544 112 448
WIRE 512 544 512 480
WIRE 768 544 768 496
WIRE 848 544 848 496
WIRE 944 560 944 544
WIRE -512 576 -512 416
WIRE -80 576 -144 576
WIRE -32 656 -32 592
WIRE 112 656 112 608
WIRE 272 656 272 592
WIRE 512 656 512 608
WIRE 640 656 640 608
WIRE 768 656 768 608
WIRE 848 656 848 608
WIRE 944 656 944 640
WIRE 1120 656 1120 512
WIRE -336 672 -336 624
WIRE -144 704 -144 576
WIRE 1024 704 1024 400
WIRE 1024 704 -144 704
FLAG -512 576 0
FLAG 848 656 0
FLAG 944 656 0
FLAG 112 656 0
FLAG 768 656 0
FLAG 640 656 0
FLAG -32 656 0
FLAG -336 672 0
FLAG 512 656 0
FLAG 1120 656 0
FLAG 272 656 0
SYMBOL voltage -512 320 R0
WINDOW 123 0 0 Left 0
WINDOW 39 24 132 Left 0
SYMATTR SpiceLine Rser=1m
SYMATTR InstName V1
SYMATTR Value 35
SYMBOL diode 48 224 R270
WINDOW 0 32 32 VTop 0
WINDOW 3 0 32 VBottom 0
SYMATTR InstName D2
SYMATTR Value BYG21M
SYMBOL npn 320 288 R0
SYMATTR InstName Q1
SYMATTR Value MMBTA42
SYMBOL npn 448 384 R0
SYMATTR InstName Q2
SYMATTR Value FZT658
SYMBOL npn 880 448 R0
SYMATTR InstName Q3
SYMATTR Value 2N2222
SYMBOL res 752 480 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R1
SYMATTR Value 5k
SYMBOL zener 864 608 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D3
SYMATTR Value BZX84C6V2L
SYMBOL res 928 544 R0
SYMATTR InstName R2
SYMATTR Value 360
SYMBOL cap 752 544 R0
SYMATTR InstName C1
SYMATTR Value 10n
SYMBOL res 624 512 R0
SYMATTR InstName R3
SYMATTR Value 720
SYMBOL sw -32 496 R0
SYMATTR InstName S1
SYMATTR Value cmp
SYMBOL voltage -336 528 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V2
SYMATTR Value 10.85
SYMBOL ind 512 496 R270
WINDOW 0 32 56 VTop 0
WINDOW 3 5 56 VBottom 0
SYMATTR InstName L1
SYMATTR Value 0.01
SYMATTR SpiceLine Rser=10m
SYMBOL schottky 528 608 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D6
SYMATTR Value PMEG3005AEA
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL cap 1104 448 R0
SYMATTR InstName C2
SYMATTR Value 1µ
SYMBOL res 192 208 R0
SYMATTR InstName R4
SYMATTR Value 10k
SYMBOL res 192 320 R0
SYMATTR InstName R6
SYMATTR Value 2k
SYMBOL res 96 336 R0
SYMATTR InstName R5
SYMATTR Value 10k
SYMBOL cap 256 528 R0
SYMATTR InstName C3
SYMATTR Value 1µ
SYMBOL zener 128 608 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 -22 Left 0
SYMATTR InstName D1
SYMATTR Value BZX84C39L
TEXT -664 248 Left 0 !.tran 0 100m 0
TEXT -728 712 Left 0 !.model cmp SW(RON=50 ROFF=1E6 VT=1m VH=5m)

Dissipation in Q2 is about 1/3 an equivalent linear reg at max input (34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Have I missed something simple?
 
F

Fred Bartoli

Jan 1, 1970
0
Gibbo a écrit :
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input voltage.
5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive ISO
7637-2. That dictates upto -600V spikes with rise times of 1uS or so and
repetive +200V spikes with similar rise times.
Dissipation in Q2 is about 1/3 an equivalent linear reg at max input (34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Did something similar with a power mosfet, a TVS, and few other
components as a voltage limiter in front of a PFC (not the boost one
which happily withstands the main pulses, but a sepic one).
Worked like a charm.

Now for your case I'd go with a depletion mosfet, a zener and one or two
resistors:
--->|----+-------.
| |
.-. |
| | |
| | |
'-' |
| ||-+
| ||<- 700V Depletion mosfet
+----||-+ (supertex DN2470)
| |
| |
| |
z '------> limited to 50V
47V A
|
|
===
GND

You can even strengthen this further with a front R+TVS.
 
G

Gibbo

Jan 1, 1970
0
Fred said:
Gibbo a écrit :
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input voltage.
5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive
ISO 7637-2. That dictates upto -600V spikes with rise times of 1uS or
so and repetive +200V spikes with similar rise times.
Dissipation in Q2 is about 1/3 an equivalent linear reg at max input
(34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Did something similar with a power mosfet, a TVS, and few other
components as a voltage limiter in front of a PFC (not the boost one
which happily withstands the main pulses, but a sepic one).
Worked like a charm.

Now for your case I'd go with a depletion mosfet, a zener and one or two
resistors:

[snip]


You can even strengthen this further with a front R+TVS.

Yes I saw that trick on the "suriving automotive load dump" thread. Neat
idea but the exisiting circuit is already immune to load dumpt etc and
it does nothing to help dissipation in Q2 which is the main idea.

But thanks for looking.
 
J

john jardine

Jan 1, 1970
0
Gibbo said:
Fred said:
Gibbo a écrit :
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input voltage.
5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive
ISO 7637-2. That dictates upto -600V spikes with rise times of 1uS or
so and repetive +200V spikes with similar rise times.
Dissipation in Q2 is about 1/3 an equivalent linear reg at max input
(34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Did something similar with a power mosfet, a TVS, and few other
components as a voltage limiter in front of a PFC (not the boost one
which happily withstands the main pulses, but a sepic one).
Worked like a charm.

Now for your case I'd go with a depletion mosfet, a zener and one or two
resistors:

[snip]


You can even strengthen this further with a front R+TVS.

Yes I saw that trick on the "suriving automotive load dump" thread. Neat
idea but the exisiting circuit is already immune to load dumpt etc and
it does nothing to help dissipation in Q2 which is the main idea.

But thanks for looking.

Without looking too deep, seems a faster startup and less dissipation if
you lose R5, D1, C3 and (oddly) D6.
john
 
G

Gibbo

Jan 1, 1970
0
john said:
Gibbo said:
Fred said:
Gibbo a écrit :
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input voltage.
5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive
ISO 7637-2. That dictates upto -600V spikes with rise times of 1uS or
so and repetive +200V spikes with similar rise times.

<...>
Dissipation in Q2 is about 1/3 an equivalent linear reg at max input
(34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Did something similar with a power mosfet, a TVS, and few other
components as a voltage limiter in front of a PFC (not the boost one
which happily withstands the main pulses, but a sepic one).
Worked like a charm.

Now for your case I'd go with a depletion mosfet, a zener and one or two
resistors:
[snip]

You can even strengthen this further with a front R+TVS.
Yes I saw that trick on the "suriving automotive load dump" thread. Neat
idea but the exisiting circuit is already immune to load dumpt etc and
it does nothing to help dissipation in Q2 which is the main idea.

But thanks for looking.

Without looking too deep, seems a faster startup and less dissipation if
you lose R5, D1, C3 and (oddly) D6.
john

But then the inductor has to deal directly with the 200 volt spikes. The
zener reg prevents this allowing Q2 to run as a swicher at low voltages
and as a linear reg (but still with switching voltage regulation) for
the high volt spikes.

It's an odd circuit granted.
 
G

Gibbo

Jan 1, 1970
0
john said:
Gibbo said:
Fred said:
Gibbo a écrit :
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input voltage.
5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive
ISO 7637-2. That dictates upto -600V spikes with rise times of 1uS or
so and repetive +200V spikes with similar rise times.

<...>
Dissipation in Q2 is about 1/3 an equivalent linear reg at max input
(34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Did something similar with a power mosfet, a TVS, and few other
components as a voltage limiter in front of a PFC (not the boost one
which happily withstands the main pulses, but a sepic one).
Worked like a charm.

Now for your case I'd go with a depletion mosfet, a zener and one or two
resistors:
[snip]

You can even strengthen this further with a front R+TVS.
Yes I saw that trick on the "suriving automotive load dump" thread. Neat
idea but the exisiting circuit is already immune to load dumpt etc and
it does nothing to help dissipation in Q2 which is the main idea.

But thanks for looking.

Without looking too deep, seems a faster startup and less dissipation if
you lose R5, D1, C3 and (oddly) D6.
john

But I've just looked at it again and I think you might be right. Remove
the zener and associated crap and just limit the drive current to Q1 so
the inductor can't go ballistic. I think there's enough of a gap (with
hfe spread) between required minimum drive and maximum allowed to limit
the current in the inductor. I think. I'll have a look tomorrow when the
gin has worn off.
 
F

Fred Bartoli

Jan 1, 1970
0
Gibbo a écrit :
Fred said:
Gibbo a écrit :
R2 and R3 are the loads. R2 load always between 5mA and 15mA. R3 load
not always on.

Acceptable for the +10V rail to drop down to +7V on low input
voltage. 5V rail must always be at 5V

Needs to operate from 9V to 34V input. But it also needs to survive
ISO 7637-2. That dictates upto -600V spikes with rise times of 1uS or
so and repetive +200V spikes with similar rise times.
Dissipation in Q2 is about 1/3 an equivalent linear reg at max input
(34V)

Nothing like optimised, just testing the idea. Anyone done anything
similar? Any better ideas?

Did something similar with a power mosfet, a TVS, and few other
components as a voltage limiter in front of a PFC (not the boost one
which happily withstands the main pulses, but a sepic one).
Worked like a charm.

Now for your case I'd go with a depletion mosfet, a zener and one or
two resistors:

[snip]


You can even strengthen this further with a front R+TVS.

Yes I saw that trick on the "suriving automotive load dump" thread. Neat
idea but the exisiting circuit is already immune to load dumpt etc and
it does nothing to help dissipation in Q2 which is the main idea.

But thanks for looking.

Unless I didn't understand what you want, just use it as a front to a
normal "low voltage" buck. Using a zener voltage *above* your max normal
working voltage ensures the lowest dissipation.

The 42R max RDSon will give you worst case (low voltage) conduction
losses identical to your circuit.
Much lower with a rising voltage.

I just noticed your +200V (understood 600V) so RDSon can even be lowered.
 
Top