JavaLaurence

In the timeline, there is mention of a PTC Bricklet. Is this a typo for RTC (Real Time Clock).. ? If not, what does PTC stand for? Thx.

I'm probably not the first to remark this, but I noticed that most products are called after the physical property that they measure, e.g. current, distance, humidity, temperature. Whereas a minority is called after the measuring device itself: Barometer (measures AirPressure), LinearPoti (measures Resistance). Obviously from a product naming point of view this isn't a big deal, but from an API design point of view, it is a shame, as consistency of semantics helps readability a lot. It also makes interoperability with third-party APIs dealing with measurement units a bit awkward.

Like I explained in an email I just wrote TF, I would very much appreciate tackling the two blocking issues I have in one go. The other issue being Wifi connectivity. Plz have a read of the Wifi thread I started before proposing some final action plan. Thx.

I measured with and without the Power-Down as part of a stack. So I have measured it stand-alone, and with a Master Brick and my Wifi Extension (as the picture clearly showed). In both configurations, the measured output voltage is identical. Also note that in the stack config, Brickv shows a stack voltage and small current consumption. So clearly, as part of the stack, the Step-Down has some "effect" on the stack.. but it fails to power the whole stack when I pull the USB cable out. Also if I take my cold stack, and plug the DC supply in, then absolutely nothing happens. No LEDs, however briefly, light up.

With the latest API we can read the custom characters' bitmap. Is it technically possible to also read the normal characters' bitmaps? I know this isn't currently possible with the 2.0.1 API, but could it be supported? Such an API would help me create a high-quality emulator for the display. BTW, there's also an error in the Java API docs: public short getCustomCharacter(short index) should be public short[] getCustomCharacter(short index) HTH.

Yes, I did measure the green connector, not the black. The transformer is OK because it used to power a Wifi router until very recently. It's rated as producing 12V (see attachments). I now measured the output voltage with a stack, and I'm afraid I get the same reading (see screenshot of setup).

PS. When I plug the DC supply into the Power-Down as part of a stack, the AC-DC transformer block changes its faint chirping noise to a higher-pitched noise. Unfortunately I don't know squat about electronics .. but sounds like a clue..?

OK, made myself a crappy cable, so was able to measure the output voltage: looks like 5.6-5.7V. Definitely not a clean 5.0, though my meter is analog,, so could be off by a few tenths. Is the output level OK?

:-) I'm actually using my only (very cheapo) multimeter's cables to connect my little DC adapter (bought from yourselves) to the Step-Down PS. So to do any measuring, I'd need to make a cable first. While I'm on the subject, I really think you should add a cable to your catalog that can connect a DC adapter to the Step-Down. And yes, I only used the black connector. And I didn't swap + and - around. I'll try to make a cable, so that I can use my multimeter and measure the green connector's voltage...

If you're like me, you like reading what other people are putting together (or trying to, in my case) with TinkerForge components. Here's my story.. hope it gets your creative juices going too. At home we've got an old cellar which suffers from rising groundwater. Our cheap submerged pump has a flaky floating switch which sometimes doesn't switch on when the water level rises, and sometimes doesn't switch off when the pump has pumped every last drop out. When I discovered TinkerForge.com (thanks Serge!), I started dreaming up a "high tech" solution: a sensor to detect rising water level, and a relay to switch the pump on/off. Provided the logic to drive these two components would run 24/24, I would never have a flooded cellar again, or worry about the pump burning itself out after hours of pumping air instead of water... Looking deeper into TF's product catalog and documentation, I realised that my system would have to be connected to my PC wirelessly (USB cables don't reach from our top floor to the cellar, and my girlfriend would kill me if I tried to hook things up with wires). Secondly, no USB means an external power supply, which in turn means TF's Step-Down Power Supply. So far, my solution therefore requires: 1x Master Brick, 1x Wifi Extension, 1x Step-Down Power Supply, 1x Distance IR, 1x Dual Relay. For good measure I also purchased the LCD20x4 display, so that the system could be monitored and interacted with (in the kitchen above the cellar). Being a software guy with very little hardware experience, I set about coding before I even had my physical components. What I did was to create software equivalents of the Distance IR, Dual Relay and LCD. That way, I could start coding the core system logic and worry about the physical world later. Maybe one day I will virtualize the rest of the product catalog, and give you all my emulator platform ;-) The result so far is that I've got my virtual LCD showing time, water level and pump status, and the pump activates above a certain water level, and deactivates below another. So I've got the guts of the system done and dusted, but I've yet to begin the physical installation. So far the Wifi Extension and the Step-Down Power Supply are giving me problems. But I'll assume I can move beyond these with a bit of help from TF. Once the whole thing is physically installed and working as intended I'll update this post here. In the mean time, have fun! Laurence

Does the extension work? No it doesn't. So far I haven't tried anything at the API level, since I figured that I should first be able to check things out via brickv. So if Brickv cannot show the extension's status, I don't want to try anything else. Currently I'm completely stuck on two different fronts: power supply and wifi connectivity. Both are key to the water pump solution I'm trying to put together. I think that, in your own interests, you should include much more self-diagnosing AND logging capabilities in the general TF architecture. So that customer service is more efficient, with a lot less "Did you try X" and "Can you try Y".. and more "Easy, you made a simple mistake: do Z and all will be well, and by the way, we'll improve our documentation so that future customers won't fall into the same trap." I've now invested quite a bit of time writing software for my stack, so I'm determined to get this working 100%. But I will need your assistance. Maybe a new Step-Down will need to be mailed to me, unless you have other ideas to address that particular issue.

I had a good look at the Step-Down board, and with bright light shining on it, I can't see any visual problems (with naked eye). When I plug the Master into the Step-Down, I ensure that the triangular markings align. I have a multimeter, but I'm afraid its probe pins are much too big for the minute leads on your designs. Instead of measuring, I would be shorting things.. Is there any other way to diagnose a rail problem? The system that I'm trying to put together cannot be powered by USB, and also has to be wireless, as the whole stack is going to live in our cellar, with the spiders and the mice. So I need this Step-Down to work as advertised, or I cannot complete the solution I have in mind. :-S

When connected to a stack via USB, and the stack contains a Wifi extension, can the stack be accessed via both USB and Wifi at the same time? Or does one medium get priority over the other? If both communication mediums are allowed in parallel, then why can't Brickv be launched twice on one machine, to approach a single stack via the two routes? When I launch brickv, it seems to detect an already running instance, and refuses to launch a second. Is that by design?

I tried running my stack without USB for the first time. I've got the boards linked together as follows: Step-Down, Master Brick, Wifi Master Extension. I've got a 13V DC transformer plugged into the Step-Down Power Supply. When the stack is hooked up via USB, I can see that the Master Brick can measure the stack's voltage as that supplied by the Step-Down. But when I pull out the USB cable, I would expect that the whole stack carries on happily as if nothing happens. Instead, everything falls dead. Did I misinterpret the function of the Step-Down Power Supply? Shouldn't it be capable of powering the bricks? When I try to start my stack without any USB, and I plug in the DC supply, then the Master Brick lights don't light up.. :-(

Hi, I managed to get the Wifi extension talking to my wireless router. The router shows the extension's MAC address as associated. The router's DHCP assigned it an IP address which I use to connect to it via brickv. The connection is successful, so I go to the WIFI Extension info panel and press Show Status. The status window comes up, but the value fields remain empty. In the mean time, the Timeouts counter keeps counting up, incrementing at a steady rate of (roughly) 0.5Hz. I don't understand this behaviour. If the Wifi extension and the router are happy (connected), then why can't any data flow over the connection? BTW, when connecting to the stack over USB, I can see that the signal strength is -84 to -88dB. Is this very weak? It must be the same strength that my iMac is getting, sitting 30 cm further away from my router's antennas, and my Mac has decent bandwidth. Help.

On the subject of exceptions, maybe you should create a hierarchy rooted in TinkerForgeException. Having an exception hierarchy allows you (and clients) to specify much shorter throws clauses or catch blocks. The same approach is used by lots of APIs throwing IOException and SQLException. Both of these have many different sub-exceptions, but clients are not obliged to specify them everywhere in their code.

What does "the port" refer to here? The DHCP server's port? Or the port number to which remote clients can connect to talk to the stack? Not clear. So far I haven't been able to let the extension see my local wireless network. I added the extensions MAC address to my router's filter table, entered the SSID and network password, Saved the Wifi Configuration, and nothing happens.. the module keeps being in Associating mode. Any hints?

In the 2.0.1 Java bindings (and maybe before, don't know), we have two public TimeOutException classes: com.tinkerforge.TimeoutException and com.tinkerforge.IPConnection.TimeOutException If your program requires both (as is typically the case), then Java requires you to fully qualify both class names so that the compiler can differentiate the two. This is pretty awkward. Historically, Java APIs have tried to avoid such name collisions. Renaming the simple name of one, or both, would solve this.

It works. Thx for fixing this. I got all my Bricks & Bricklets flashed to 2.0, and everything apart from IR Distance seems to work OK. Will investigate the IR Distance further (calibration gone due to 1.0 -> 2.0 flashing?) BTW, does step-down bricklet need flashing too?

Yes, on the Downloads page, in the Bindings and Examples section, you could add another column "Requirements" to the Language/Current Version/Changelog table. It makes sense on the Downloads page because that's the only guaranteed place where people will pass through to access the software (due to, you know, the RTFM syndrome in our industry).

Hi guys, Current documentation remains silent on what TF software requires to operate. Especially the language bindings could do with some info like, for the Java bindings, "requires version Java 1.5 or later". (the generated binding is not compatible with 1.4 and earlier - believe it or not lots of people in the embedded world still use Java 1.1 ..) HTH Laurence

Great stuff. Thanks !

The MD5 match. But I'm afraid I'll leave it to TF to get your makefile fixed. I appreciate receiving email updates on this.. AFAIMK no point in continuing in this thread.

Can you please tell me what the file sizes and MD5 hashes are for the versions of brickd that I've got so far? The file looks hopefully broken if it can't even run its main() entry point correctly. Given that it crashes means that OS X views it as an executable, but apart from that ... ?

I tried it. Same result. Looking at the source now..