a****y
发帖数: 255 | 1 有人用过便宜的DC/DC converter吗? 要求 5V 到 +/-100V. |
M*******c
发帖数: 4371 | |
a****y
发帖数: 255 | 3 100V时, 电流要求5mA左右. 是的, 应该isolate. |
M*******c
发帖数: 4371 | 4 才5mA。自己作一个吗。方案可以有很多种。 最简单的好像就是flyback。用个
optocoupler就可以regulate secondary side output。
如果不需要很好的regulation，用个charge pump也行的。 |
c*******l
发帖数: 4801 | 5 老将出马，一个顶仨
【在 M*******c 的大作中提到】
: 才5mA。自己作一个吗。方案可以有很多种。 最简单的好像就是flyback。用个
: optocoupler就可以regulate secondary side output。
: 如果不需要很好的regulation，用个charge pump也行的。
|
a****y
发帖数: 255 | 6 thanks a lot. I found the attached file and each converter costs about $70.
Is it the normal price for DC / DC converter?
【在 M*******c 的大作中提到】
: 才5mA。自己作一个吗。方案可以有很多种。 最简单的好像就是flyback。用个
: optocoupler就可以regulate secondary side output。
: 如果不需要很好的regulation，用个charge pump也行的。
|
M*******c
发帖数: 4371 | 7 depends upon volume and performance(footprint and efficiency)
usualy, $1.00 per W for module is still reasonable. That is why most
companies still do discrete design.
how large is your volume? if you buy only one, I am not surprised.
.
【在 a****y 的大作中提到】
: thanks a lot. I found the attached file and each converter costs about $70.
: Is it the normal price for DC / DC converter?
|
g******u
发帖数: 3060 | 8 flyback并不好做，如果要做自己的变压器。不过应该有些现成的，可以看看coilcraft
的网站。opto应该不好做，比值太大了，duty cycle会很小。
还是买个市面上的方便合算多了。不过5V到100V的我是没见过。 |
c**d
发帖数: 57 | |
f*****0
发帖数: 489 | 10 "如果要做自己的变压器。"
take a notebook or ATX power supply. |
|
|
s*****r
发帖数: 17 | 11 If isolation is not required, a simple boost should solve your problem. If
isolation is required, flyback has the lowest cost. You might be able to use
the transformers from the adapters for cell phones. |
f*****0
发帖数: 489 | 12 a boost converter will be very hard for this kind of boosting. if you are
taking 5v and outputing 100v, your boost converter is working at 95% duty
cycle. not many boost constrollers can do that.
however, since the current load is minimum, you can use a combination of
chargepump and boost controllers to get you up there. for example, use one
boost controller for 4x and another for a 5x then you have a 100x topology.
my favorite is ncp1400 series in sot23-5 package. you can form a boost
converter |
f*****0
发帖数: 489 | 13 and all of that for less than $5. |
M*******c
发帖数: 4371 | 14 别人说了，需要plus and minus 100V。就是200V together。
你要不讲演一下怎么作这个1 to 40 ratio 的boost？
use
【在 s*****r 的大作中提到】
: If isolation is not required, a simple boost should solve your problem. If
: isolation is required, flyback has the lowest cost. You might be able to use
: the transformers from the adapters for cell phones.
|
M*******c
发帖数: 4371 | 15 才5mA。应该很好作的。OPTO也很容易design，如果你不是对transient 有很高的要求
的话。
coilcraft
【在 g******u 的大作中提到】
: flyback并不好做，如果要做自己的变压器。不过应该有些现成的，可以看看coilcraft
: 的网站。opto应该不好做，比值太大了，duty cycle会很小。
: 还是买个市面上的方便合算多了。不过5V到100V的我是没见过。
|
E*****a
发帖数: 757 | 16 老大，我的经验是抄。
这类电路多的很，随便抄个就好了
【在 M*******c 的大作中提到】
: 才5mA。应该很好作的。OPTO也很容易design，如果你不是对transient 有很高的要求
: 的话。
:
: coilcraft
|
f*****0
发帖数: 489 | 17
anything more than 10x boost is very difficult to do with just one converter
.
【在 E*****a 的大作中提到】
: 老大，我的经验是抄。
: 这类电路多的很，随便抄个就好了
|
E*****a
发帖数: 757 | 18 agree
there is max on time/min off time for most ICs
converter
【在 f*****0 的大作中提到】
:
: anything more than 10x boost is very difficult to do with just one converter
: .
|
s*****r
发帖数: 17 | 19 Albany wants +-100V dual outputs, so none of us is suggesting toward his
request.Luckily, he finds his own solution.
Is 20 to 40 step up ratio from a simple boost impossible to do? You won't be
able to find a boost IC because there is no market for it. For an
argument's sake, you can always use a general PWM IC to break the duty cycle
limit (i.e. TI's UC3525).
Or think this way, is 48V to 1V impossible from a simple buck converter?
Since there is market, there are tons of ICs can do this.
【在 M*******c 的大作中提到】
: 别人说了，需要plus and minus 100V。就是200V together。
: 你要不讲演一下怎么作这个1 to 40 ratio 的boost？
:
: use
|
t*****g
发帖数: 108 | 20 我也请教下，想做一个24V（电流5A），输出12V（电流至少4.5A)，电流效率希望很高
的DC-DC变换，请问有很好的效率的CHIP吗？
谢了 |
|
|
f*****0
发帖数: 489 | 21 too many to list. if you want simplicity, try the simpleswitcher serious
from National.
4.5a is a little bit too big for internal switches so you may want to try
buck controller with external switches.
higher frequency typicall leads to higher efficiency. |
M*******c
发帖数: 4371 | 22 A flyback would solve his problem.
Boost is different from Buck on power loss distribution given the same step
ratio.
The main switch of a boost has to take both the high conduction loss and the
high switching loss. While for Buck, the top FET takes the switching lost
while the bottom FET takes the conduction loss. This is not a trivial
difference.
Of course given Albany's case (only 5mA) you could say so if you overdesign
the whold converter.
be
cycle
【在 s*****r 的大作中提到】
: Albany wants +-100V dual outputs, so none of us is suggesting toward his
: request.Luckily, he finds his own solution.
: Is 20 to 40 step up ratio from a simple boost impossible to do? You won't be
: able to find a boost IC because there is no market for it. For an
: argument's sake, you can always use a general PWM IC to break the duty cycle
: limit (i.e. TI's UC3525).
: Or think this way, is 48V to 1V impossible from a simple buck converter?
: Since there is market, there are tons of ICs can do this.
|
M*******c
发帖数: 4371 | 23 不是用opto作gate driver。是直接用opto来sense output voltage。
coilcraft
【在 g******u 的大作中提到】
: flyback并不好做，如果要做自己的变压器。不过应该有些现成的，可以看看coilcraft
: 的网站。opto应该不好做，比值太大了，duty cycle会很小。
: 还是买个市面上的方便合算多了。不过5V到100V的我是没见过。
|
s*****r
发帖数: 17 | 24 The boost is just a dual configuration of the buck. The voltage and current stresses
of these two topologies are of no difference, provided the same voltages
applied in the "boost" and the "buck" sides. The difference is simply the
current direction which has no impact on the conduction losses.
Boost is more difficult to design usually for its dynamics and many cases
for the high voltage due to the applications. The switching losses might be
higher but it is also because of the high output volta |
M*******c
发帖数: 4371 | 25 I am not your advisor nor your boss. Whatever you think is None of my business.
"The boost is just a dual configuration of the buck. The voltage and current
stresses of these two topologies are of no difference, provided the same voltages
applied in the "boost" and the "buck" sides. The difference is simply the
current direction which has no impact on the conduction losses." |
s*****r
发帖数: 17 | 26 This might be the only valid statement that can come from you. I don't feel pity
for you, but I do for your adviser and your boss.
发信人: MonkInOrc (方枪枪), 信区: EE
标 题: Re: Re: DC/DC converter
发信站: BBS 未名空间站 (Sun Jul 20 01:31:09 2008)
I am not your advisor nor your boss. Whatever you think is None of my
business.
"The boost is just a dual configuration of the buck. The voltage and current
stresses of these two topologies are of no difference, provided the same
voltages
applied in the "boost" and the |
M*******c
发帖数: 4371 | 27 严重申明！！
以下是本人不认同的观点，只是看着好玩，作茶余饭后谈资。
"The boost is just a dual configuration of the buck. The voltage and current
stresses of these two topologies are of no difference, provided the same
voltages
applied in the "boost" and the "buck" sides. The difference is simply the
current direction which has no impact on the conduction losses."
business.
current
voltages
【在 M*******c 的大作中提到】
: I am not your advisor nor your boss. Whatever you think is None of my business.
: "The boost is just a dual configuration of the buck. The voltage and current
: stresses of these two topologies are of no difference, provided the same voltages
: applied in the "boost" and the "buck" sides. The difference is simply the
: current direction which has no impact on the conduction losses."
|
