Electrical Engineering - 2018-11-16

« first day (2941 days earlier) ← previous day next day → last day (1978 days later) »

1:21 AM

Okay, so turns out the unit was faulty. Test a second device and it showed 12V across the pins.

11 hours later…

12:18 PM

@laptop2d . . Regarding isolating the liquid cooling isolation : The liquid is delivered via non-conductive hose, with a length sufficient for the resistance of the liquid. Then other end of hose is grounded.

@W5VO I have sent 1500 amps thru standard 1/2" refrigeration copper tubing (water cooled). (approximate 0.048" wall thickness).

@JonRB . . Since bussbar is back on table, consider using standard plumbing tubing and elbows.

1 hour later…

1:29 PM

Wow, that would be weird... using "regular" solder on plumbing for once. :)

@rdtsc . . not really. Silver solder (brazing) should be used for carrying current.

2 hours later…

3:18 PM

@Marla Cool! Do you use standard tap water, or something fancier like distilled or DI ?

Distilled is good. Deionized is OK. For freeze protection use water with ethelyne glycol.

Deionized is expensive because it reacts with the copper and soon becomes not deionized. Then you have to connect a deionizer device (expensive).

@W5VO . With a copper tube you can actually send very high current provided that water flow is good enough that water doesn't boil. It is not unusual to send 5000 amps through a hollow. Copper coil for induction heating.

I've seen some people use DI water in chiller loops, and a few months later they got to replace the pump from corrosion

@W5VO. Yup

The current limit is then more about how much loss you're getting than managing the loss?

Too aggressive on the replenishment of the DI

Need to set the control to a lesser setting for DI

The wall thickness of the copper, and turbulent flow of water are variables in cooling the tube. It, pretty much yes. Enough water and you can run lots of current

Wall thickness because you could melt the outer copper while inside is still cool

3:36 PM

@Marla That sounds like an awesome failure mode to watch

It is awesome. However the more common failure is when calcium builds up on inner copper wall (thermal insulator). And steam explosion. And I don't often use the word explosion. @W5VO

Oops, not steam explosion. The outer wall melts, then steam.

Regarding the pump corrosion, that usually happens with steel impeller. Bronze impellar is required

3:52 PM

I think that was a "learning mistake"

Why not use some other, inert chemical, like mineral oil?

@rdtsc . . not sure if mineral oil is flammable

It shouldn't be. I guess that's what HV and MV transformers use nowadays. Down-side is, it is rather viscous. Might not pump all that well.

And messy to clean up if leaks.

4:17 PM

Indeed... xylene might dissolve it, untested. But it's inert, clear, and non-conductive.

6 hours later…

9:55 PM

Would EE be the right place to ask a question about a specific microprocessor?

Specifically regarding feature support.

10:11 PM

@forest Is it an embedded system processor, or something more like an i7

@W5VO ARM7TDMI and ARM946E-S (so embedded)

Specifically regarding possible internal core temperature sensors.

undocumented or documented?

and those are not vendor-specific devices

Could be either. All I know is I have a device with both processors, and the device appears to be able to detect the current temperature. The behavior of the device itself is definitely undocumented, as it has absolutely no need for temperature sensors (so it's certainly not an external sensor).

As for ARM7TDMI, according to an ARM forum, it has no temperature sensor, so if anything has it, it would likely be the ARM946E-S processor.

@W5VO What do you mean by them not being vendor-specific?

"The ARM946E-S is a synthesizable processor combining an ARM9E-S™ processor
core with a configurable memory system. It is a member of the ARM9E™ family of
high-performance, 32-bit system-on-chip processor solutions."

hm

So any temperature sensor would have been added by the vendor?

10:17 PM

yeah, I'm not sure how much of a temperature sensor you can have at the HDL level

but peripherals like that are typically vendor specific anyways. And the temperature it is measuring may even be junction temperature (of the processor)

This is for an old Nintendo DS I have which is running homebrew applications (i.e. it's hacked and is running custom code). I have never, ever seen it display the temperature, but I found one program which offhandedly mentioned the current temperature. It displays the current temperature in celsius and appears not to be entirely bogus (i.e. it gets lower when it's cold, higher if it's warm or if I wrap my hands around the system). Whatever it is is undocumented on the system level.

Could it be an external sensor on the Nintendo DS? e.g. TMP122

External meaning not in the processor cores?

If it's always higher than ambient, then it could definitely be something on-chip. It could also be embedded in some other chip

Correct. Do you happen to know if there's a separate package that reads temperature?

10:22 PM

It does go up slightly (usually by only .1 or .2 Celsius) if I do something that really works the 2D rendering engine (which is contained in the ARM9), e.g. use the stylus to move the window containing the sensor readings rapidly around the screen.

@KingDuken If there is, it's entirely undocumented. I don't see why the Nintendo DS hardware could possibly require it, as there's no way that the processors or any other circuits could ever generate enough heat to require monitoring (the battery just isn't capable of supplying enough power to overheat it).

Does it have a rechargeable battery?

850 mAh Li-ion.

@W5VO I believe the Nintendo DS does yes. @forest Anything that has perform graphical calculations like the rendering for each game you play, things could get a little hot.

It could easily be in the BMS then.

I wonder if there's homebrew software that can let you see some readings in the hardware.

10:26 PM

@KingDuken Even when stress testing the single 3D and dual 2D rendering engines, the temperature of the device in my hand does not go up perceptibly. And with a max temperature of 125 degrees C (from an ARM7TDMI datasheet at least) it's unlikely that it would ever be reached in software.

@KingDuken Not that I know of, except one outdated application from 2006 which offhandedly mentions the current temperature in a window.

@W5VO I was thinking the same but possibly some I2C stuff

But why would it even need to sense the temperature?

It doesn't act on it (you could overheat the device or freeze it and it will continue to operate until it is physically incapable of doing so, or hits the junction max).

Battery management system

A battery management system (BMS) is any electronic system that manages a rechargeable battery (cell or battery pack), such as by protecting the battery from operating outside its safe operating area, monitoring its state, calculating secondary data, reporting that data, controlling its environment, authenticating it and / or balancing it.A battery pack built together with a battery management system with an external communication data bus is a smart battery pack. A smart battery pack must be charged by a smart battery charger. == Functions == === Monitor === A BMS may monitor the state...

I thought that would be built into the battery itself?

I've seen battery management systems (BMS) specifically include temperature measurements to make sure they don't catch fire

Is the battery built in?

10:35 PM

No it's removable. I don't recall if it had anything that looked like a data link or just had a few leads that were obviously for power.

It's a very dumb system and any management on the battery's side must be done entirely within the battery (I'm not even sure if software is aware of whether or not it is in a charging or discharging state, nor can software tell if the battery is low, though a person can see it as the power light turns from green to red).

I'm willing to bet that there is something, Li batteries are pretty dangerous otherwise.

Isn't that stuff normally contained in the Li battery itself?

sometimes

It's not a smart battery, but it surely has internal protection circuitry.

11:04 PM

It's possible that I'll just have to reverse engineer the homebrew application and see exactly what it does to obtain the temperature value... I was hoping I wouldn't need to, but it seems likely that it'll be the only way to find this out.

@forest I'm afraid so - hunting for an undocumented feature is usually off-topic as a question

Fair enough.

@forest . . .mostly because an undocumented feature would be based upon opinion

Is that the real reason though? There are plenty of extremely well-known but undocumented features (like dozens for Z80 that are used all the time. In fact you can't even make a Z80 emulator that works for all but the most trivial programs without supporting a fair number of undocumented features).

Undocumented means nobody knows what is consistent, nor guaranteed

11:19 PM

Certainly not arguing with the policy itself, but I don't think that undocumented features are all that mysterious or inconsistent. It just means it wasn't published in the official documentation, not that it's not well-established. Hell, there are even some microprocessors with undocumented features that are "documented" through unofficial channels more thoroughly than an officially documented feature like the x86 A20 gate!

And thus they become documented by users. NOT guaranteed by manufacturer.

I have many times used devices outside of specifications and hope the manufacturer wouldn't change.

Looking through the search, EE has a large number of questions regarding undocumented features (ports, opcodes, etc) that appear well-received.

I won't pursue this further.

That's fine. I'm not arguing against it, just curious where the policy came from!

I guess I'll have to check on the meta.

And nice to chat with you forest. This is where we can really mix it up without boundaries. @forest

11:26 PM

Agreed. And I think I'll ask on meta since searching on it didn't help. Also... I just realized the application that checks for the temperature might not even be closed source. I might just be able to read the source to see what it does!

And opinions here are ON topic

That's what I like about chat. :P

11:46 PM

electric shock safety question -- is the "let go" current threshold for a given individual or population at all dependent on the frequency of the current (DC, 50/60Hz, 400Hz, something else?)

@Shalvenay . . Flippant answer : I don't know, but I let go of a 600v buss real quick one time. And glad to have survived

Frequency does determine how an electrical shock affects muscles.

It also determines whether or not it will bring someone's heart into asystole or fibrillation (certain frequencies are more dangerous than others).

« first day (2941 days earlier) ← previous day next day → last day (1978 days later) »

Transcript for

Nov '1816

Electrical Engineering

A place to talk with friends from the EE community about vacuu...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile