In production !

Yes yes finally my project reached the production stage. It's been 4 weeks now that the PCBs are mounted in the electrical closet. I had a few problems in the beginning but these are solved now. Let me show a few pictures to let you see how the final production is implemented live. The pictures are sorted in time as the work progressed from the old system towards the new cortex based system:

Old system based on multiple AVRs in master slave.

Old system shutdown and empty closet waiting for the new boards.

New boards mounted.

New boards fully wired
and ready to go.

Completed.

I2C2USB board integrated
in final design.

Dimmer board in action.

So after 2 years I can say that this project is finally finished. I still have some things todo but as for now I'm running with version 2.1 and this already seems to be pretty stable. I have to work on the communication making the final GUI that connects to the I2C2USB server (refer to my dedicated blog to read more). This will be done in javaFX some new scripting language I'm busy to learn as we speak. I learned many new things doing this project and I experimented a lot with small proof of concepts. This way of working works pretty good I must say. First gather the little blocks you need, study on them and finally bring them together in complete system. Of course I couldn't have done the job on my own therefore many thanks to my best friend Ronny. He did a great job in designing the hardware for me. We always had some good conversations (between some Belgium bears ;-)  on several technical topics in a way that the software and hardware pieces melted together quickly and easily. So again I couldn't have done it without him. Great job my friend !

This blog will probably be one of my last here in this topic which was focused on the overall hardware design. Maybe I will start a new one building the GUI stuff that still has to be done. Seems to be a good idea. So keep checking my blogs once and while.

I'm a very very little little bit bit wiser now ;-) 

wonderfull results

Quit some time since I wrote here but as you could guess lots of things to do. I managed to write my software for the board presented in this blog. I have a first version running now as we speak. It was sometimes not that easy though believe me. I had many little difficulties to beat which were hard to find. I will write of these in my cortex blog when I have some more time. I wanted to finish the development phase first. Below you can see a picture of the board running an integration test in a final phase towards production.

So the step to final production isn't that far away. I think it's feasible to have this milestone for the beginning of October. I still have to integrate the communication PCB, design by a friend of mine. It's also based on the cortex core which will speed up the code writing because I can reuse most of my code written already. One of the latests cortex designs of ST is the communication line (107). These types offer the Ethernet gateway possibility. I will not spend a separate blog on it but I 'll write some major topics here along the road. Next week I post a picture of this PCB and will explain the functionalities in short. So stay tuned as we now are 80% ready and you definitely don't want to miss the other 20% ;-)

complete !

Finally all is ready !

The result is not bad at all we only have just two minor issues:

• pin 4 of the DIP switch was not connected to the board solder island (re solder)
• led order is not parallel with power outlets (can be mapped in software)

I'm now started with converting my proof of concepts I did on the olimex board to our brand new board. I test every block code as I move on. Small video below shows the testing of the timers. At the moment this not yet synchronized with the zero cross and FIR it just tests the timers and nvic module. Next step will be testing the ADC and FIR for the zero cross detection.

Also the button test board (see picture) is now connected so we have a nice test kit for 12 in and outputs. Having these helper tools is indispensable as preparation for the final production release.

Button test board:

Power test with timers:

first test succeedded

Running leds as integration test:

Next test will be with power connected.

jtag testing

At the moment we do the first testing with raisonance JTAG interface:

the core is on

As promised the pictures of the cortex core:

sequel components

This time the trafo is added and other related stuff. The trafo is not used for supply reasons but to feed a reference signal for FIR filter ADC input signal.

Capacities and diodes used to work with plus/min voltages of the trafo:

What is left to do is the cortex mcu and some related resistors, capacities and crystals. So let's say ready for 75%. Next blog will be pictures focued on the cortex. Hopefully we can do the integration tests in couple of days from now. Of course these results will be posted in this blog topic.

high voltage power tests

First initial power test is done now without having any of the low voltage components on the board for safety reasons. All 12 outlet points work as it should be without any high voltage leaks through the board raster layer. The raster layer is used to better conduct the heat dissipation on the board when having high current loads on the triacs. It also gives a special effect to the surface outer view of course. The triac trigger was manually provided by means of a 3V dc voltage - (through a current limiting resistor on the pcb) - connected to the optocoupler.

Test labo :

As we provided 12 small lamps (40W) for our test here so we will also introduce 12 mini buttons to connect the input connectors. We need these buttons to complete our integration test. Now we can safely continue in putting the cortexes and others low voltage components on the board. When that is ready the first integration tests can be done using my code described in my previous blog.

The shining continues !

sequel adding components on pcb

Now we added the triacs making the power sections complete. Also the opto coupling is already on board as you can see:

Meanwhile I wrote some test code to perform some basic testing. So it will be easy to see if everything lines up with the initial drawings as it should be. This will be used whenever the cortexes are present on the board. I wrote a separte topic in my cortex blog to highlight these tests. So refer to following link if you want to read more :

http://guyvo-cortex.blogspot.com/

(topic integration tests )

Still to come to the board:

• cortexes
• leds
• JTAG
• ADC circuit

soldering is started

Components are all arrived now so in the next coming weeks they will be proper placed on the pcb. Of course a couple of basic board integration tests must be made in parallel with this process. Like testing the inputs and outputs end to end. The two identical pcbs will be ready till end of august maybe a bit sooner. As soon they are ready I can start to port may proof of concept code build for the olimex development kit towards these final cortexes. This will be covered in detail in my cortex blog.

A first shot of the fuses , coils and connectors:

Stay tuned to follow the progress in next blogs.

PCB ready in black and white

I 'm not going to write much here I think the pictures will speak for themselfs ;-)

Just a special thanks to the creators of this beauty:

PCB Design     : manes6969 at telenet dot be

PCB Production : http://www.eurocircuits.com/

Component view 0:

Component view 1:

Component view 2:

Component view 3:

Total size front view:

Total size back view:

size idea

PCB desgin printed out on paper to give you an idea on the size of the board. The measurement can be read on the orange meter in cm. ( ca 42cm )

( click on picture to see full size with me in the shadow ;-)

final

Minor bugs and symbol changes.

3 PCBs are ordered this weekend.

Estimated arrival time is 10 days from now.

Next post the live pictures without components  !

hex release

Alignment symbols improved
now 99,9% sure this wil be the final release.

echo release

Symbols added.

delta release

After a few check iterations.

almost ready

PCB Alfa release.

power schematics

DC supply,

ADC input 

and triac layouts.

cortex schematics

Cortex IO pin layout.

all components on the board

Next step is ready. All of the used components are now on the board. The first routing is succeeded and it's only done on 2 layer. Apart from some minor details and printing symbols which ar not yet on the board this will be pretty close to the final board layout. Hopefully ready end of this mounth.

first early drawing

The PCB will be around 45cm long and 20cm height. Measurements are taken into account to fit properly in my dedicated electrical closet. This can be changed as we go along the road of drawing this beauty. Remember that it's not me doing this it's actually my friend. I just give some useful tips ;-)

We also increased the maximum power of each dimmer to around 1000W. The maximum power is bottlenecked by the coils against disturbanty because the triac can have 3000W easily. This implies that the central power lines on the PCB will be about 1 cm width (see on the picture the big blue one) which can take 16 amps so pretty secure. We will also try to make it on a simple double layer to reduce the PCB cost. As you can see all the power (AC) connections are grouped on the right of the board. Our intention is to group the DC connections on the left. This makes the board very suitable for connecting it and moreover the sensitive signals are proper separated from the manes. I took black as the color and the printed tekst will be white. A pitty yellow is not available tough. The purpose is to mount the board with plastic or Teflon distance tubes using 4 or 5 mm thread inside and some nuts. Because of the precense of highvoltage on the board we must make a shield over it to avoid direct contact. That we 'll do using a transparant plexy glass which we are going to mount above the PCB using the same threads and distance tubes so it becomes very safe and nice to look at. I'll continue in this blog letting you see the evolution of the board with some extra guided words.

Kick off

My first initial concept drawing containing:

- 12 triac based dimmers each up to 500W
- CAB free filtered ( like the professional NIKO ones )
- User programmable (min/max , switch or button, speed)
- I2C bus connectivity

Drawing is not wysiwyg it's just a first idea of what must come where. So much can be turned around. Stay tuned to see how this beautiful project gets his final shape.