What's the hi-side current sense chip du jour?

The one I did wasn't "burp mode" because the switching speed slowed down smoothly. The frequency just fell and the on time stayed at the minimum

(Giant Magneto Resistive) The company was NVE. The price was too high and it seems they aren't in digikey

Digikey seems to have only two parts. A Maxim for

I wouldn't use a Maxim part either. I designed in one of their parts many years ago. I had to design it back out.

e able

age

r.

ent

of

tion

citor

to

ry

If you want to insure the inductor is through dumping, you can use a comparator. Assuming a simple boost circuit, the voltage at the intersection of the diode and inductor will drop beneath the input voltage. There are self-timed DC/DCs using this technique. That is, charge the inductor to a set limit using a current sense, then observe that the current is fully dumped using the comparator. It has the advantage that you have flexibility in picking the inductor, plus no stability issues, but otherwise it's not that sexy of a design.

rail is shorted to ground.

it'll be tough because of wide voltage range there.

at 0V supply ;-)

ain't enough there. Need "beyond positive rail" which

know which of those are truly "common", meaning there

ADC and then issue warnings such as "Hey, you are at

like one that I'd set at 80% for warning and the

stock is lowish and I've seen clients get burned by

how it is after allocations or uncertain leadtimes

promising:

not familiar with what kind of I-sense device has the

bypass cap to the tune of two cents :-)

What about these:

I plan to use them in one of our products. There are versions up to 30A.

rail is shorted to ground.

it'll be tough because of wide voltage range there.

at 0V supply ;-)

ain't enough there. Need "beyond positive rail" which

know which of those are truly "common", meaning there

ADC and then issue warnings such as "Hey, you are at

like one that I'd set at 80% for warning and the

stock is lowish and I've seen clients get burned by

know how it is after allocations or uncertain leadtimes

promising:

not familiar with what kind of I-sense device has the

bypass cap to the tune of two cents :-)

Yes, that is an option and the price is ok. However, Allegro seems to pretty much dominate the budget-class Hall sensing area and I had a not so pleasant experience with them regarding motor drivers a while ago. I'll try to find some that are sufficiently form-fit-function compatible and there are at least two mfgs.

I usually try to do that as well, less audio noise. People don't like it if stuff hisses ;-)

"The M" pointed out some cheap ones from Allegro, aroudn $1.50 at Digikey. But single-sourced, so ...

Probably because of the typical unobtanium effect :-)

Since more than 20 years "design-out" is all I ever did with Maxim parts. Mostly because their production planning seems skewed or hardly existing. OTOH that has brought me some design work ;-)

The inductor is a whole 'nother matter. Finding non-custom low cost two-winding inductors for SEPIC with more than 5 amps turns out to be less easy than I thought. Then again nobody ever said engineering life would be easy.

You can do SEPIC with two separate inductors. I think it takes the same amount of copper and iron (or ferrite.)

John

able

Yes (although it does take a little more), but then it causes a huge racket on the input rail. I use two inductors with small SEPICs but prefer not to when it's more than 10W. In this case a lot more.

VSl.23175$ snipped-for-privacy@nlpi069.nbdc.sbc.com...

the rail is shorted to ground.

nt)

case it'll be tough because of wide voltage range there.

working at 0V supply ;-)

so RR ain't enough there. Need "beyond positive rail" which

ke to know which of those are truly "common", meaning there

nto an ADC and then issue warnings such as "Hey, you are at

be ok, like one that I'd set at 80% for warning and the

gikey stock is lowish and I've seen clients get burned by

you know how it is after allocations or uncertain leadtimes

e promising:

I am not familiar with what kind of I-sense device has the

a wee bypass cap to the tune of two cents :-)

buffer.

Different groups in companies handle different product lines, though the generally have the same fab. My point is just because one part of a product line sucks, the entire product line may not be bad. Generally if the fab is crap, the company won't last, so it is usually the case that the designs are the culprit. So don't judge the hall effect product line by the motor driver line.

No, I don't judge them. I am just concerned about the single-sourced status of them. Hall sensors from Honeywell and others can be prohibitively expensive and once I've picked a hall I'd be stuck.

Buy a high-end 3-axis Honeywell sensor... you can get that ultimate thrill knowing that yours truly designed the control chip ;-)

...Jim Thompson

You designed the HMC1043? Then I'd need the matching set of wheels around it, Jaguar XJS or something. Leather, champagne cooler, the works.

I can't counter with anything from the automotive world but when you get a real critical coronary artery blockage rest assured that yours truly has ... well, let's not go there, I hope that won't happen and they never have to thread this thing up your arteries.

I'm actually not sure what commercial-grade part numbers it may be in. I know it's in military-grade compasses.

I already did the "thread" and stent bit, July 1998 :-(

...Jim Thompson

ip

They like a high pitch whine even worse.

Their switcher driver chips (I forget the number) also had a problem when the gate of the MOSFET yanked the gate drive pin below ground. A little capacitively coupled glitch and the chip lost its mind for a few microseconds.

Or the occasional whiff of high amperage smell ;-)

Whoops, a wimpy substrate structure?

I am currently wrestling with "unorthodoxies" in LTSpice. Gate drive remaining at 1.8V for 50nsec after FET turn-off, causing humongous virtual losses. It's a totally flat line, looks like a model fault or something and kind of thwarts any meaningful efficiency analysis.

On Thu, 28 May 2009 08:57:39 -0700, Joerg wrote: [snip]

Ever hear of "gate charge"? As in...

See page 3 first.

Anytime a simulator shows you "unorthodoxies", it is appropriate to think thru your problem carefully _before_ blaming the simulator ;-)

...Jim Thompson

Well, yeah, but: Quote "RC values based on the gate-to-source capacitance normally lead to a gate drive that is hopelessly inadequate."

I don't do that. It's one of the reasons why I am not a believer in gate resistors :-)

I've designed and built bucket loads of switchers and when driving them from fast single-digit ohms sources there ain't no 50nsec pedestal even when stepping up more than 10:1. But in Spice there was. Hmm ...

You're seriously missing the point that you must inject X-amount of gate charge before you can get off of that pedestal.

The simulator is NOT lying to you. You're just out to lunch understanding this, as usual [Duck & Run :-]

...Jim Thompson

I should have pointed out... your driver is probably current-limited.

...Jim Thompson