programmability.
For a few cents more you can upgrade to a pin-compatible 64MHz chip!
Best regards, Spehro Pefhany
-- "it\'s the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
I thought we were discussing why the *battery chargers you bought* might refuse to charge? How many amps were those supposed to be?
Not me, I would have (and have done) exactly what you did. I have a nice
5A current limited bench supply that works great for that.But then *I'm* not going to sue TTi if my battery starts bubbling hydrogen :)
-- John Devereux
On a sunny day (Tue, 29 Sep 2009 16:09:58 -0700) it happened Joerg wrote in :
fireworks.
If you looked at the diagram, and also the picture of the test setup here:
There are 2 possibilities with this circuit, when controlled externally via RS232, you can set the reference voltage of the comparator that compares the output peak current to an internal reference, so cycle by cycle current limit. Or you can use a potentiometer, and set that reference voltage by hand, This is why I mentioned lab supply. as this provides the current limit in that application. Because of the cycle by cycle current limit things work safely.
In a true 'current mode' supply the reference voltage of the cycle by cycle current comparator is set by comparing the output voltage against a reference voltage, so you in fact set the output voltage by regulating the output current. The advantage is that you only have a single pole in your feedback loop.
The PIC example does not do that however, because it is using an on / off stabilisation loop for the voltage regulation, one that does not require a complicated feedback filter either, and is lightning fast (much faster then you can do with a PID in software). So in fact has the advantages of a true current mode, while possibly being faster.
I think, but have not tried, if you wanted, you could clamp down the reference of the current comparator with the output of the voltage comparator, making a real current mode, but I see no advantage here at all. Just a difference in wiring.
As to how much power, that is an incredible misconception. That only depends on the external circuit, not on the controller. Sure you will need other switching components, with different drivers, but for the control loop it makes no difference if you regulate 1 mW or 1 kW. If you were really fast, like Lucky Luke, who shoots faster then light (his shadow), then you could do it by hand..
fireworks.
transformer I used.
I had noticed that. However, your switcher is running at a low PWM frequency, else the long wires would have fuzzed it up and caused a kablouie. Hard to see in the photo but your CT seems to go into the PIC via 1.8K. That alone would be too slow for most of my switchers, considering the usual 5-10pF input capacitance plus internal circuit delays.
comparator
cycle current limit.
application.
Well, yeah, as I said at 100kHz or so that is easier than at a MHz.
current comparator
fact set the output
stabilisation loop
filter either,
faster.
Are you using a hysteretic scheme? That of course makes things easy but results in a lot of ripple.
of the current comparator
see no advantage here at all.
The control loop isn't what makes things expensive there. It's the size of the magnetics that does. Plus in my cases there often just ain't enough space to run things at 100kHz. When doing a one-off bench supply it's ok.
shadow),
Done it by hand a few times, meaning sans PWM chip and sans uC :-)
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
On a sunny day (Wed, 30 Sep 2009 08:45:21 -0700) it happened Joerg wrote in :
Yes hysteretic (for voltage control). How much ripple really depends on how fast it reacts. As it reacts very fast, it can change the drive PWM to any value within a PWM period. There is sort of a small analog area created. If you look at the PWM on the scope, once it is stabilised at the set point, you see all sort of values. The ripple on the output then depends on the output cap and other system parameters. The PWM looks irregular, but the output is really OK. All depends on what you want to use it for, if it is to supply an audio amp, then you may have to pay more attention to ripple then when it has to supply a bulb filament, or charge a battery. When charging a battery it will be in current limit mode anyways (low ripple cycle by cycle correction), only once the battery is is full will it go into hysteretic mode, and then there is almost no current. In industrial applications often ripple in the order of magnitude of some percent is allowed, this one does better then that.
That's a fairly easy concept, and possibly the only one you can really use on a regular uC.
period.
you see all sort of values.
parameters.
Irregular is ok, we even do that on purpose sometimes, for example to dodge an EMI bullet. But hysteretic means long time-constant filters and that can screw up the load change response. Which is often unacceptable on my designs.
then you
filament, or charge a battery.
cycle by cycle correction),
there is almost no current.
percent is allowed,
Not in mine. Usually analog signals must be carried in the same bundle and things have to be super quiet. Once I had to run a switcher locked at half the video clock rate. Now that was outright painful.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
And I thought we still were.
The first ones were switchable 2/6 amps. The second pair was rated 3 amps. All put out zero amps, which I thought was not very helpful. Do you think maybe I'm being too demanding?
We just loaned the Echo to a friend who ripped the bottom out of her BMW on a berm full of rocks. She was skeptical about driving such a down-scale car, but now she really likes it. She may buy it for her college-age kid.
John
On a sunny day (Wed, 30 Sep 2009 10:31:23 -0700) it happened Joerg wrote in :
period.
you see all sort of values.
parameters.
Are you referring to the loop filter? There is no loop filter.
PWM period.
you see all sort of values.
parameters.
No, to the output filter. You can't possibly let the ripple just saunter out the banana jacks of a bench supply and pollute whatever it find downstream.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
On a sunny day (Wed, 30 Sep 2009 11:17:34 -0700) it happened Joerg wrote in :
All depends, 10mV is fine with me:-)
You really got to come up with some numbers, else it all makes no sense,
They do sound pretty useless. I just like to try to understand what rationale people have for these sorts of design choices. Is it really cheaper? Afraid of being sued somehow? I'm afraid I don't quite buy your selling-more-batteries conspiracy theory, but who knows?
-- John Devereux
Ok, this is the number I typically need: Not being able to show the ripple on a scope when set to 2mV/div. Good enough?
Seriously, 10mV piping out would be disastrous when doing things like ultrasound experiments.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
I'm working on a smallish board that has four switchers in one corner and an ADC that's looking for a 1 nV spectral line in the other, range of interest straddling all the switcher frequencies. I'm planning on spread-spectrum wobulating all the switchers, just in case.
People should use delta-sigma in switchers instead of PWM.
John
The 1nV over the background of how many nV/root(Hz) ?
Synchronize the switchers away from the particular frequency of interest?
So the dirt will be in band for sure? How about combining the results of two measurements with the different switcher frequencies?
No problem except the switching losses will be an order of magnitude higher. Delta sigma has the spurious products, also.
Vladimir Vassilevsky DSP and Mixed Signal Design Consultant
On a sunny day (Wed, 30 Sep 2009 13:01:31 -0700) it happened Joerg wrote in :
LOL I have done quite interesting ultrasound Doppler experimensts. If your circuit is so bad that it cannot tolerate 10 mV ripple on the supply line, then I suggest you change to gardening.
!
Spread spectrum just smears the noise so it looks lower (and is to some extent but not a lot). Like throwing excess junk under the sofa and pretending it ain't there no more.
Since you wrote spectral line, how about rotating the switchers through three different frequencies and have the ADC always look in the quiet bands?
You could completely oversize them and run in pulse-skipping mode but it wouldn't help much.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
About 1. That's our target noise floor. The existing system is ballpark 100 nv/rtHz with huge birdies all over the place.
Never let scientists design electronics.
The signals can be all over the place.
Yup. Operating range is audio to many MHz, all at once. We digitize at
64 Ms/s and FFT and see what's there.
That might be interesting. Actually, we could determine the birdie level from occasionally running with a null sample, and subtract that out if it's not huge. But the main plan is to keep things clean.
Yeah, I suppose so. Spread-spectrum will have to do. It has its own complications. If I triangle-modulate all the switcher frequencies, I don't want the triangle to appear on the power rails either. That could easily happen. Spice? Breadboard? Take a chance? Go linear with big heat sinks?
John
Besides, you can cancel the components that are correlated with the rate of the rotation of the switchers.
Vladimir Vassilevsky DSP and Mixed Signal Design Consultant
line,
So, then, how many beamformers, audio Dopplers and color flow modules have you brought from blank sheet to market? Experiments, revenue product, two different things.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
That's how we do it in ultrasound. The first time I suggested it (and it works!) I said "... and that will be just software", almost causing a coffee mug from an attending SW engineer to come flying at me.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
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.