When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Don't buy netgear unless you'd like to be fucked by the company. I had one of their routers. Multiple emails were sent to me (I had registered the product with my email address) first advising me, then strongly recommending, and finally urgently demanding I upgrade the firmware.
I DL the software from their own website, using the link they sent. So far so good. Then I followed their instructions to upgrade. Halfway through, the thing just stopped. I tried again, but at this point the computer would no longer communicate with the router.
I used my neighbor's wifi to contact the company for resolution to the problem. They said it was "a known issue". They also said there was nothing they could or would do. They would not even provide a discount coupon for me to buy another router.
Basically, they destroyed my router, and expected me to just go buy another one.
I will never, ever buy one of their products. I suggest you do not, either.
Oh, it was well out of warranty. Several years. I took it all the way to the corporate offices, to no avail. All it served to do was make me madder each step of the way.
A 555 monostable is definitely the way to go. Very predictable and stable behavior. Use its output to drive an NPN, and sink the buzzer through the transistor.
Basically, build the circuit as described in the first link gooflophaze posted, and add the components on the right:
Edit: give me the specs on the buzzer (voltage and current) and I'll help you select an appropriate transistor.
Thanks Joe.
All the buzzers I've found are very harsh hi frequency buzzers. Is it possible to soften the square wave?
The small active piezos I've found have these specs:
10PCS 5V Electromagnetic Piezo Buzzer Continous Sound Beep Continuously Ala H3J9
Rated voltage: 5V DC
Operating voltage: 3.5V - 5.5V
Operating current: less than equal 25 mA
Resonant frequency: 2300 +/- 500 Hz
Operating temperature: -20 Celsius ~ + 70 Celsius
The buzzer is rated at 5V DC, it is widely used in computers, printers, alarm system, automatic electronic equipment and other electronic applications used as voice.
The active beep measures:
Length: approx. 9 mm
Case diameter: 11 mm
storage temperature: -30 Celsius ~ + 105 Celsius
Case diameter: approx. 11 mm
Total Length: approx. 9 mm
Lead Space: approx. 8 mm
Main color: Black
Material: silicone, copper
package content: 10 x buzzer
Note:Light shooting and different displays may cause the color of the item in the picture a little different from the real thing.The measurement allowed error is +/-1-3 cm.
Ive had terrible luck with netgear. Currently using a linksys, but those ASUSs kick ***.
Data point- I was being sold on the idea of an AC1300 or 1900 router. So just last week i tested the capacity of my $40 bottom of the line 8 years ago linksys(150mb throughput capacity, dual band). Had 2 phones playing 1080 youtube vids, firestick running HD video to TV, PC playing a 1080 youtube vid while playing PUBG and downloading another game at 1.2MBPS with a dozen plus add intensive pages open in chrome. Nothing was throttled.
At these voltage / current levels, the drive transistor is ridiculously easy. Damn near any small NPN will work. I'd probably use a PN2222 just out of laziness / habit.
If it were my design, I'd just go ahead and throw an arduino in it. Your basic functionality for driving the buzzer can be done in like, 12 lines. Toss a speaker directly on it, and you can do tone manipulation. For extra fanciness, you could make it play fur elise with a note every time the hall sensor activates...
If it were my design, I'd just go ahead and throw an arduino in it.
I'm just gonna say this, and I don't want you to feel specifically insulted, but one of my most common gripes at work is about how a lot of companies want to throw a whole operating system at problems which are easily solved with a tiny bit of hardware.
None taken. That said, it did take me a little more than a dozen lines. But I'm pretty sure this'll work - and software is so much quicker to manipulate than hardware.
And I still run WRT54GL's with dd-wrt. If I need faster than G, I'll crimp a cable. And netgear has always been fairly awful - when I was doing repair and wifi was nascient, you'd have netgear B ap's that wouldn't work well with non-netgear cards.
I never really developed any serious programming skills beyond assembly / ML. For me, working directly with registers and ALUs is intuitive, while using high-level languages and IDEs is confuddling.
I recognize that many people feel the same way about working directly with transistors, passives, and ICs. I find it easy to put myself into the mindset of discrete components, sense the way that the base of a transistor feels under certain stimuli, etc. But I'm not a Luddite, and I recognize that circuit science is secondary to coding here in the 21st century.
To be honest, one of my great fears is that there's going to come a time when we (humanity) suddenly realize that we put so much focus into software that we forgot to educate the next generation of sand-benders.
Cisco Linksys WRT120N is what I have. It is connected to the cable company's Arris DG1670, which has its own wifi symbols on it but I've never tried connecting to it because the signal seems weak.
Eh, man with a hammer and all that. More tools in the box, more options - and of course, more options for scope creep. Sparkies are a different breed - don't really think you have to worry much about them. I'm content knowing enough about hardware and software to be dangerous in both, but I don't have the desire to develop software 40 hours a week or thevinize a circuit for ***** and giggles.
Speaking of scope creep - one of the things I'm working on right now (and why I refamiliarized myself with the 555 recently) was creating a dumb alternator controller for the LFX. Needs a 128khz PWM signal, otherwise it defaults to 13.8v output on the alternator. GM's got a neat controller with desulfication, temperature compensation, cruise mode for MPG, bumps up the volts when you're running lights, etc etc. 13.8v works pretty well in the summer, but last winter we would occasionally need a jump box - looks like the ECM has a low voltage cutoff for the injectors (50-ish volts boosted) , so even though the starter can crank the engine perfectly, no fuel - no run.13.8v charges a battery pretty well in the Camaro since it's only got like 2 feet of 00awg battery cable - but our 12-ish foot 2awg run to the trunk drops that a little.
So - initial thought - make a smart controller, add a temp sensor, closed loop control by monitoring the voltage at the battery.. or, you know, a 556 for a fixed frequency and use the 2nd output on a trimpot for set it and forget it.
So - initial thought - make a smart controller, add a temp sensor, closed loop control by monitoring the voltage at the battery.. or, you know, a 556 for a fixed frequency and use the 2nd output on a trimpot for set it and forget it.
I don't get to do a lot of actual engineering these days. My job is half management / project planning, half dealing with vendors & suppliers, and half systems-level design. (Yes, that's three halves. I work a lot.)
But lately, I have been doing some cool stuff with the Pi. By fortunate coincidence, my VP is a hardcore coder, and tinkers with Pis quite a lot. So he writes the code, I build the hardware to interface them with the outside world, and as a result, we've now got some really awesome remote-telemetry monitoring going on, all viewable from anywhere on a web browser.
The main UPS:
(Just noticed that the calibration seems to have drifted a tad.)
Using Raspberry Pis for production equipment? The sketchiness of micro-USB power supplies and SD card storage isn’t a concern? Or are these not mission-critical systems?
There two schools of thought in prod - low MTBF vs failover. You pay out the *** for low MTBF quality and get dinged 4 years down the line when it's obsolete / unsupported. Failover on cheap devices are more economical - you need more engineering resources on the software side, but it's incredibly effective. And in Joe's case, it looks like they're largely just remote reporting nodes - I imagine the application is actually hosted in the DC. If it ***** the bed - toss in a new one for $30. And of course - being able to roll your own solution vs. a vendors almost-fills-my-need-but-not-quite is a boon.
Not arguing about the principle of the thing, just that Pis have two very specific failure points. I mean, you could do exactly what they’re doing but just virtualize it rather than running it on something that relies on a smartphone charger and an SD card. Even if you wanted a hardware solution, you can pick up old thin clients for $30 on eBay and do the same thing with them.
But sure, if it’s just a non-mission critical app and it doesn’t matter if it’s down for a day or two before you can replace the hardware, and you already had a Pi kicking around, that makes enough sense.
These aren't mission-critical systems at all, just status & telemetry monitoring. And as gooflophaze suspects, we don't store much data locally. The SD card just hosts the OS and application, and the devices all push their data out to a central server every few minutes. We treat them as though they're disposable, but in the whole time I've been here we've only had one power supply failure (I bought a replacement at the Walgreens a few blocks from Sears Tower) and zero SD / Pi failures.
And yeah, this is one of those situations where there was literally no good off-the-shelf solution, so Dave & I kinda said "Well, **** it, lets just build one." And it's been working wonderfully.
We also use a lot of Pis as RTMP video receivers for the remote cameras we have scattered throughout the city. (Weather & Traffic shots, mostly.) They output fullscreen HDMI at 1080i, which we convert to SDI with BlackMagic Micro Converters that cost $39. It's kind of hilarious how cheap "broadcast grade" equipment is getting.
As an aside, SD cards really are amazing. We used to roll HDCAM-SR tapes for all of our ISO records. The tapes themsalves cost about $200 each, the recorders are upwards of $70k, and the maintenance on them is a ******* nightmare. About a year ago, I installed the first bank of ten SD card recorders ($695 each) and bought a bunch of 256GB SD cards from Amazon. Not only is this method fantastically less expensive, it's also hugely more convenient for the editors, since they can just pop the SD cards right into their Premiere / Final Cut computers and read the ProRes files directly, rather than having to import the video in from tape in real-time. I'm presently on a quest to replace every single tape machine in the facility with an SD card recorder.
Still struggling with my nb2 artech 2560. Trying to get it thru emissisons for over a year now. There are three readiness monitors not set. Egr, o2 monitor and cat monitor. It wont pass unless all the readiness monitors show "set."
I read of a similar issue on miata.gag. His fix was replacing both o2 sensors. I did that and have driven many cycles. I have read and followed the tricks for reaching the conditions needed to set, and still no luck.
are you running the factory ECU to control the fuel and satisfy the O2 sensors?
tricking the system is incredibly hard. The ECU wants to see a very specific lean-rich pattern for each O2 sensor (they are different) to show that both sensors and the catalyst are working properly. If either is out of balance, it will either never become ready or throw a code.
06-22-2018, 10:18 AM
0
