Is Maxim now AD? If so, this set of SMPS eval boards is interesting

yes, since august last year,

I know AD bought them--what I'm getting at is, "Has Maxim stopped acting like d*ckheads by continually cancelling recently-introduced products just when you're going into production?"

Cheers

Phil Hobbs

Those Maxim parts look redundant on the ex-LTC LTM modules. Analog owns both now.

I hope they did a better job on switcher noise than Linear did. LTM8078 is ghastly. If you try the eval kit, maybe you could look at that.

TPS54302 is pretty good.

Duckheads?

That would be my primary interest. We've revamped our standard photoreceiver packaging to use a machined-from-solid brass box with a flat brass lid, communicating via feedthrough caps and a U.FL F to bulkhead SMA cable. All the switchers go in another board under the brass box, with a black powder-coated steel or aluminum lid, with a black plastic shroud covering the feedthrus and supporting a bulkhead-mount power jack. The lid can be soldered down if desired, for really tough EMI situations. Once we have the new models up on the website, I'll post a link.

Turns out that with modern laser cutting and 3D printing, the whole works is cheaper than the Hammond box we were using, besides being dramatically better electrically and much much more bling with all that gold(ish) and black. (Free-cutting brass looks more like gold than yellow sheet brass does.)

Simon unofficially calls the new version of our QL01 nanowatt photoreceiver the DK02, for Donna Karan. ;)

The new boxes are good for fast APDs, SPADs, and MPPCs/SiPMs.

Check out SendCutSend for some very interesting capabilities.

I've got that in my notes from your previous recommendation, but haven't tried it yet.

Cheers

Phil Hobbs

We did a couple of custom extrusions to replace the horrible Hammond boxes. We buy them blue anodized and blast all the artwork with a Boss laser.

I'll be in truckee later this month. I could drop in on sendcutsend just for fun.

Google says they're about 45 minutes from Truckee, so that might be a good use of half a day. They keep adding capabilities--they now do laser-cut steel (12 different alloys iirc) up to 12 mm thick, for cheap. Even titanium.

It seems to be just the sort of thing American manufacturing used to be unequalled at--apply capital and ingenuity to become the low-cost, high quality producer even in a high-wage country.

They're pretty focused on laser cutting and surface coatings--anything out of the plane costs extra. Simon knows a lot more about them than I do, and is a huge fan.

It reminds me a bit of designing with MSI logic chips BITD--with some time and creativity you can make some pretty neat things while staying within the cheap zone.

Cheers

Phil Hobbs

Why? Maxim has been around for a long time. Maxim ate a lot of companies but there's always a bigger fish in the pond.

Maxim is infamous for teasing with datasheets and samples, but when it come to buying for production you are not getting any unless, I guess, you need millions so they can be bothered to actually make some

I used about 3000 of MAX9690. They discontinued it without notice, and then they started failing in the field. We had to replace all of them.

We had to build our own:

Ah, the Motorola problem. If the auto industry placed an order for millions you went to the end of the line. We're back to the mid-80's where the salesmen would estimate delivery out a year or more with a straight face.

Government agencies are now hogging FPGAs.

We make stuff for a giant semi fab company, and we can't get FPGAs. They are using all their clout to get us parts, but even they are second in line.

If we can't ship, they can't make fab lines, so people can't make chips. That logic isn't enough to get us parts.

What are they doing with them? I knew the defense industry used quite a few but not all of them. NSA building bigger and better facial recognition tools?

"4.5-V to 28-V Input, 3-A Output, EMI Friendly Synchronous Step-Down Converter"

"EMI Friendly"? .. could mean different things. EMI is usually not our friend! Barney the purple dinosaur is editing spec sheets

It does radical spread-spectrum switching. That smears the spectrum and helps it pass FCC/CE radiated tests, but it still makes spikes that can get into analog stuff. But it is pretty good about that too; Trr is about 20 ns, could be worse.

The spectrum shaping must be done right, because all that duty-cycle thrashing doesn't make noise on the DC output.

It's rated for 3 amps out but gets pretty hot up there, and you can't heat sink the SOT23 package much. I use it up to 2 amps.

Not bad for 99 cents.

It should be possible to make a plus-to-minus converter with that one. Gotta try that.

It won't (legally) do +24 to -5, but it's close.

Actually, abs max is 30, so it might.

I never had MOTA do what Maxim does. The worst thing I encountered from them was the way they derated their otherwise very nice MC35084 quad decompensated FET op amp. The GBW and slew rate specs decreased by _half_ between the preliminary and production datasheets. That was very inconvenient, because at the time (1987) there weren't any low cost FET amps that fast.

Cheers

Phil Hobbs