@ThePhoton if you happen to use a D21xnnB, by the way, I can share a conf_clocks.h with you that should get the USB running if your design has a watch crystal on PA1,PA2

In the near future I'm going to try getting it working with the USB SOF as well

hm, a scope with a webinterface thats almost as good as the scope controls, thats quite a neat feature...

@PlasmaHH R&S seems to still know quite well what they're doing on many of their devices

@Asmyldof I think they are also doing it at their own pace, getting stuff out when they think it is ready, instead of other companies where the products feel somewhat rushed to compete

Seems like a decent assesment

Also afaik they have good ways to connect their different devices to talk to each other, just like keysight, so there is not so much need to compete, companies will likely buy the manufacturer they are used to and can interconect instead of mixing stuff that will not be able to talk to each other

its 10 bit btw.

and the psu is available at digikey ^^

@PlasmaHH I have all kinds of brands, but then, I don't shy away from developing some code myself to run all the testing in my lab while I'm not there

@Asmyldof many companies will not do that, mostly probably due to lack of knowledge/skill/balls

Can't really disagree :-). And I now also see some of the reasons many companies don't want to have the skill or balls to do so. Unless you have a truly good developer the things become so complex and nontransparent very quickly that it becomes more and more of a hassle.

Then again, many brands support popular modelling and lab tools, such as for example Matlab and LabView, so...

Yeah, imagine if that guy is run over by a bus, hard to find a replacement that can do the same

Buss ? bus ? nice pun @PlasmaHH

@PlasmaHH The main problem I see is that even somewhat technically minded management seems incapable of discerning between a C++ code monkey and a C++ developer

@Asmyldof I have yet to se any that can do that distinction

Where the prior might start a project by writing everything in header files, because of very strict inclusion orders, for example, just of the top of my head, and the latter would set up a project with proper classes and inheritance and no magic numbers in code anywhere, preferably all numbers from external conf.

talking with management is fun... me "We haven't been able to evolve our protocol in the past decade because it is so broken that upgrading it gradually is impossible" ... mgmt: "so we went along with not upgrading the protocol in the last decade, so there is no reason to make easy upgradeability a requirement for a new protocol"

@Asmyldof Our test system uses HP-VEE (nowadays Keysight VEE-Test) to read Excel files that store all the tests required, limits, etc. Designed by a guy who retired 2 years ago, whose name is cursed for eternity.

Sorry, "Keysight VEE Pro"

How does one get an encoder signal (quadrature or even just a basic incremental) into a PC running windows 7, which only has PCI-express slots available (no full size PCI)?

I have an application written in C++ that will need to read the encoder value to track the movement of an object on a conveyor, and the PC needs to get the data somehow. Is there a kind of "go to" DAQ or encoder input module that people use?

Pros would probably use an SA for that

are there little single or dual input DIN mounted modules that act as counters with encoder inputs, and can be talked to over RS485?

surely if you make them ;)

haha

i could, i just want to buy one though :(

want want want, thats all the kids these days can do ;)

@KyranF Googling "USB motion encoder" gives this:

rls.si/e201-usb-encoder-interface-123

and a bunch of other links, but that's the best one on the first page.

Also this:

usdigital.com/products/interfaces/pc/usb/QSB

I need a few weekends off. Else I might mention having a prototype laying around that goes up to ridiculous speeds

I'm not saying this is a good idea but has anyone ever used an analog switch in place of a FET in a switcher supply?

we could ask your utility company to cut your powre for a weekend

@EwokNightmares whatever you mean exactly by "analog swtich"

@PlasmaHH a SPST cmos switch

not sure exactly the best name for them

fast switch time, low charge injection, low leakage

mosfet?

just a random question in my head

yes they're made out of mosfets but they have better characteristics for use as a reed switch substitute

so magnetically actuated?

nope, they just mimic the properties of one more closely than a power fet

they can't conduct current like a power fet

@EwokNightmares Switching time?

@ThePhoton ~30 ns

I just looked one up. I'm not seriously considering this lol

vishay.com/docs/72740/dg2715.pdf

I guess that's pretty bad

@EwokNightmares Seems kind-of long compared to an individual FET with comparable current-handlilng capability. Which would be bad for the efficiency of the switcher.

Depending, I guess, how much of the switching time is propagation delay through the buffer, and how much is the switch actually being in an intermediate state betwween on an off.

Hm, a cracked pcb in a new device, thats a new one for me... someone must have had a bad day at assembling

@ThePhoton Yes, I guess they'd be used already if they didn't have switching losses

@PlasmaHH If you're having a good day, we can always increase the quota until you have a bad one.

@ThePhoton with my current job every day is bad, so the quota is full

@ThePhoton interesting, thanks for that. I didn't know that googling "USB motion encoder" would be so useful! Doesn't seem like the usual key words for this task. I suppose USB is nice, and would work for my purposes.

@KyranF Just seemed like the interface people would be likely to design for, if response time is not critical.