That would mean something is fishy in there. It is a feedback mechanism and should "ease into" transistor conduction, then regulate. Unless, of course, you'd drive the input to the hilt in an attempt to overpower it.
-- Regards, Joerg http://www.analogconsultants.com/
How about using a accurate voltage reference and an open collector comparator instead? It probably needs some compensation though.
-- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico@nctdevpuntnl (punt=.) --------------------------------------------------------------
I was referring to the c-b junction, not the e-b.
John
Then you'd have tons and tons of gain. Could get iffy, like trying to dump speed when the end of the runway is already way too close.
-- Regards, Joerg http://www.analogconsultants.com/
be
when
cap.
Oh, ok, yes that can be fixed with a diode in the collector.
-- Regards, Joerg http://www.analogconsultants.com/
at
wI'm a bit late to the thread,
Can you arrange the circuit so that everything saturates at the same point? Throw away some gain before the integrator and add in back in later.
George H.
It's not necessarily a good thing to let it saturate. Takes longer to come out of that and worst case something in the loop can go *KANGGG* if that happens.
[...]-- Regards, Joerg http://www.analogconsultants.com/
Well, it is in an _integrator_. The transistor might take a few microseconds to come out of saturation, but it would have to be a pretty fast integrator for that to matter at all.
Cheers
Phil Hobbs
Normally yes. But Tim mentioned something being "stuck" for a while when coming back out of the pegged position. Also, 100kHz loop speed isn't exactly slow.
IME there are three main things that can rain on your parade in regulator loop designs: Too much gain in the FB/limiter path, hard saturation somewhere, and capacitive loads at the plant.
-- Regards, Joerg http://www.analogconsultants.com/
The minimum 1mA anode current is easily dealt with... but...
Has anyone actually tried the TL431 in this kind of application? I did a SPICE simulation and it seems to be glacial in turn-on- like almost
200mV overshoot at 5V with a very slow 10mV/usec ramp (almost 20usec)GBW is supposed to be ~1MHz (give or take, depending on the datasheet), which is sure nothing to write home about, but this is way way worse.. here's the model (PSPICE format):-
If it's actually this bad, it wouldn't make a very good crowbar controller for any situation where the dv/dt could be at all high. By the time it gets around to turning on, the circuitry it's supposed to protect will be well fried.
.SUBCKT TL431 7 6 11
I have never seen them act this slow. Crowbar firing of the SCR happened within a microsecond, that's where I use them most of the time.
Where did this one come from?
-- Regards, Joerg http://www.analogconsultants.com/
Texas Instruments.
Best regards, Spehro Pefhany
-- "it's the network..." "The Journey is the reward" snipped-for-privacy@interlog.com Info for manufacturers:
Interesting. Although, I had a spat with them regarding a regulator chip which went inexplicably *PHUT* and no datasheet limits were exceeded. Asked them for a detailed SPICE model, they wouldn't give it to me. Asked them to run my (simple) schematic, they wouldn't do it ... which I thought was strange.
All I know is the TL431 from TI is pretty darn good, made a really snappy crowbar. Which we needed in most cases, like one where north of $30k worth of laser diodes were running off of the supply. Ok, I installed two crowbars, just in case :-)
-- Regards, Joerg http://www.analogconsultants.com/
That model lloks nasty, try one of these...
*********************************
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
adjustable
That's a lot better than the regular version. ...Jim Thompson
[On the Road, in New York]
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
Over-kill it with diodes...
Diode in series with collector kills inverse transistor action which is probably the cause of your latch-up. ...Jim Thompson
[On the Road, in New York]
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
might
adjustable
A lot? Nah. It's just that it is the scaled down low power version. Not much to write home about.
-- Regards, Joerg http://www.analogconsultants.com/
Curious: The 2nd one is much more detailed and larger. Is bigger = better?
-- Regards, Joerg http://www.analogconsultants.com/
Yeah, I know, it's something I do at work with some circuits.
Jamie
Yes. Generally the less behavioral content, the more accurate, unless, of course if it was Yours Truly who wrote the behavioral model, then it'd be quite accurate ;-) I'm a curve-fit maniac. ...Jim Thompson
[On the Road, in New York]
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.