M
MassiveProng
- Jan 1, 1970
- 0
She's an idiot!Don't. You'd be lunch.
She's an idiot!Don't. You'd be lunch.
perhaps you should have read the two links.
Once again, the mighty ass has opened his mouth to roar, but merely
drooled down his chest, instead.
Ok - how do we send 100,000 volts to Homer's PC?
But my logical skills are infinitely superior to yours.
You may be a pussy who will let two dicks like Jim and John insult you. I am
not.
Nope. You haven't been paying attention so who is retarded? I actually have
two separate trade qualifications, some advanced trade qualifications, a
degree and considerable experience in a number of strange technical or
related fields.
Jim said:Jim Thompson wrote:
[snip]
So ask a technically useful question already ;-)
...Jim Thompson
OK,
Im playing with an el-cheapo smps controller. Its peak-current-mode
control, and uses the BE junction of an NPN to detect peak current, and
turn the switch off. Of course Vbe varies strongly with T.
Normally I use a TL431 etc for voltage feedback; it has a nice stable
reference. I just choose my current sense resistor for min Vbe (max T),
and the magic of -ve feedback makes sure all is well. yay.
the circuit is pretty simple, the 431 pulls down on a pnp (crappy
mirror) which feeds current into an R, also connected to the base (there
is an R from B to Rs) so the base voltage is the sum of IRs and the
error amp output voltage.
This time, however, I want to use it as a current regulator. There is
still a setpoint, but no error amp (other than the NPN), so alas all my
-ve feedback "magic" stops working, so I have to:
a) temp-comp the setpoint (or NPN, but I want to keep I*Rs low), or
b) add some other form of current sensor, a T-stable reference and an
error amp, which I dont want to do.
so far the best I have come up with is placing a Vbe multiplier (using
the other transistor in the NPN's sot-323 package) in series with my
setpoint voltage. If I choose the Vbe multiplier scale factor to be the
reciprocal of the resistive divider formed from Vsetpoint to Rs, then
the divided-down scaled-up Vbe temperature change exactly (ish) cancels
out the dVbe of the NPN.
kinda:
Vsetpoint----[Vbe multiplier]---[R1]----+-----B of CE NPN
| \
0V--[Rs]--+---------------------[R2]----+ 0V
|
to switch Source (or E)
R1,R2 >> Rs
so I get:
Vbe1 - dVbe1 = (I*Rs)*R1/(R1+R2) + (Vs - K*(Vbe2 - dVbe2))*R2/(R1+R2)
and if the 2 BJTs are at the same temperature, then dVbe1 = dVbe2
so all I have to do is choose K = (R1+R2)/R1.
but it feels kinda icky.
Cheers
Terry
Can you post a schematic? I don't understand words very well ;-)
...Jim Thompson
Then why do you never make technical contributions to the group?
What are these trade qualifications and degree?
Jim said:Jim Thompson wrote:
[snip]
So ask a technically useful question already ;-)
...Jim Thompson
OK,
Im playing with an el-cheapo smps controller. Its peak-current-mode
control, and uses the BE junction of an NPN to detect peak current, and
turn the switch off. Of course Vbe varies strongly with T.
Normally I use a TL431 etc for voltage feedback; it has a nice stable
reference. I just choose my current sense resistor for min Vbe (max T),
and the magic of -ve feedback makes sure all is well. yay.
the circuit is pretty simple, the 431 pulls down on a pnp (crappy
mirror) which feeds current into an R, also connected to the base (there
is an R from B to Rs) so the base voltage is the sum of IRs and the
error amp output voltage.
This time, however, I want to use it as a current regulator. There is
still a setpoint, but no error amp (other than the NPN), so alas all my
-ve feedback "magic" stops working, so I have to:
a) temp-comp the setpoint (or NPN, but I want to keep I*Rs low), or
b) add some other form of current sensor, a T-stable reference and an
error amp, which I dont want to do.
so far the best I have come up with is placing a Vbe multiplier (using
the other transistor in the NPN's sot-323 package) in series with my
setpoint voltage. If I choose the Vbe multiplier scale factor to be the
reciprocal of the resistive divider formed from Vsetpoint to Rs, then
the divided-down scaled-up Vbe temperature change exactly (ish) cancels
out the dVbe of the NPN.
kinda:
Vsetpoint----[Vbe multiplier]---[R1]----+-----B of CE NPN
| \
0V--[Rs]--+---------------------[R2]----+ 0V
|
to switch Source (or E)
R1,R2 >> Rs
so I get:
Vbe1 - dVbe1 = (I*Rs)*R1/(R1+R2) + (Vs - K*(Vbe2 - dVbe2))*R2/(R1+R2)
and if the 2 BJTs are at the same temperature, then dVbe1 = dVbe2
so all I have to do is choose K = (R1+R2)/R1.
but it feels kinda icky.
Cheers
Terry
Can you post a schematic? I don't understand words very well ;-)
...Jim Thompson
Not really, I dont have access to ABSE. But I did send you an email.
Thanks,
Terry
MassiveProng said:At least it hits his chest. Yours runs down your leg.
Stop talking out of your ass, boy.
Bwuahahahahaha!
Homer said:Says who? You're the one who likes to start pissing contests.
Like I'd tell you.
the ESTx were done solely so I am legally able to mess with electrickery,
instead of having to pay for a sparky.
And I learned a good trick, which is to make the earth wire in a plug
1-2cm longer than all the rest, so its the last one to come out when
yanking on the cord.
I did, dumbfuck. You didn't.
BOTH have "to make certain" in then.
One, as the first given definition, and the other, a bit farther down.
Try again, gloss over boy.
Homer said:I'm a sparky in two countries and plan to add a third.
If you've anchored the cord right it won't come out.
Homer said:I'm a sparky in two countries and plan to add a third.
If you've anchored the cord right it won't come out.
The best trick is to remember our motto: "One flash and you're ash!"
Absolutely not.
Yes there damn well is !
Yes he did !
There's a reason that insurance policies aren't called ensurance policies. One
is simply a financial matter, the other would need the company to be a God.
Oh right. So that makes them identical in meaning does it ?
So how much do you pay for your car ensurance ?
I don't suppose you understand the difference between alternate and alternative
either ?
Graham
MassiveProng said:The ALTERNATIVE to you understanding this issue is for me to ignore
you for your utter stupidity.
One does not get ensured. Yeah, dumbfuck, I know that.
MassiveProng said:The ALTERNATIVE to you understanding this issue is for me to ignore
you for your utter stupidity.
One does not get ensured. Yeah, dumbfuck, I know that.
BUT, one can say "Ensure that you handle this problem."
Correct.
AS WELL AS "Insure that you handle this problem."
As well as: "Assure that you handle this problem."
In THAT context, either is correct, as well as the third.
Ever ?