[contactirfu]

Hi Kreutz,

Still working! wow how minature can solderable electronic parts get!

Great goin.

RGDS
Irfan

[CountZero]

a lot smaller... take a look at this: for example http://www.panasonic.com/industrial/...AOA0000CE1.pdf

I bet hand soldering is a bit tricky on the 01005 parts, the smallest I have done is 0402 and that is doable but not my first choice

[kreutz]

a lot smaller... take a look at this: for example http://www.panasonic.com/industrial/...AOA0000CE1.pdf

I bet hand soldering is a bit tricky on the 01005 parts, the smallest I have done is 0402 and that is doable but not my first choice

0402 are also doable without much problems, better with some mechanical aid and vacuum pick, like the manual pick and place stations. Not my choice either, for tweezers and hands free placement...

I improvised an squeegee using a plastic business card, it worked. My real squeegee should arrive today, the older one is rusted and unusable..

[contactirfu]

I thought that smd's were adhesively placed in the right location and then a solder paste is applied and ket in the soldering oven. is it not so? Phill from pminmo.com had given me a few links- lost long back....

any one has few tips on the www?

[kreutz]

I thought that smd's were adhesively placed in the right location and then a solder paste is applied and ket in the soldering oven. is it not so? Phill from pminmo.com had given me a few links- lost long back....

any one has few tips on the www?

That could be done in order to reflow two sided boards with components on both sides, the adhesive is applied to the bottom side (generally, the side with less components, or less amount of heavy components).

What I do:

The solder paste will keep the components attached due to superficial tension, the components will also auto-align themselves within certain limits (if you control vibration and shocks).

So, first you apply solder paste, then the components (to the side with less components), reflow, then apply solder paste to the other side, place the components, and reflow again. The superficial tension of the solder will hold the components on the bottom side during the second reflow cycle, and there is no need for adhesive (unless there are heavy components on that side). On the other hand, you reflow twice the same board, so it takes more time.

Adhesive is applied generally to heavy components (e.g. connectors, big inductors, etc) in order to keep them in place.

[contactirfu]

That could be done in order to reflow two sided boards with components on both sides, the adhesive is applied to the bottom side (generally, the side with less components, or less amount of heavy components).

What I do:

The solder paste will keep the components attached due to superficial tension, the components will also auto-align themselves within certain limits (if you control vibration and shocks).

So, first you apply solder paste, then the components (to the side with less components), reflow, then apply solder paste to the other side, place the components, and reflow again. The superficial tension of the solder will hold the components on the bottom side during the second reflow cycle, and there is no need for adhesive (unless there are heavy components on that side). On the other hand, you reflow twice the same board, so it takes more time.

Adhesive is applied generally to heavy components (e.g. connectors, big inductors, etc) in order to keep them in place.

ThanX Kreutz, that was Informative!

[kreutz]

Reflow profile for this prototype's Reflow cycle

Preset parameters for this profile:

Soaking time 150 sec, soaking temperature 168 C, Reflow time 20 sec, Reflow temp 226 C.

At the end of the cycle, the oven will keep the temperature at 60 deg C. The power curve shows that regulation even with the oven's door opened. After a while it disconnects itself and ask for the next batch.

[contactirfu]

Reflow profile for this prototype's Reflow cycle

Preset parameters for this profile:

Soaking time 150 sec, soaking temperature 168 C, Reflow time 20 sec, Reflow temp 226 C.

At the end of the cycle, the oven will keep the temperature at 60 deg C. The power curve shows that regulation even with the oven's door opened. After a while it disconnects itself and ask for the next batch.

OH Kreutz, its overload ,but still what is soaking , and what happen's when reflow happens, don't the semiconductors get fried up or something.

will they sustain at 226C for 20 seconds, i was of an impression that most IC's barring a few power semi's to get cooked at that temperature!

I myself work for a Semiconductor Nano Fab manufacturer on the mechanical front, but still have surprises and knowledge gain at every corner when Semiconductors are concerned

Thnx for all the Knowledge base
Irfan

[kreutz]

OH Kreutz, its overload ,but still what is soaking , and what happen's when reflow happens, don't the semiconductors get fried up or something.

will they sustain at 226C for 20 seconds, i was of an impression that most IC's barring a few power semi's to get cooked at that temperature!

I myself work for a Semiconductor Nano Fab manufacturer on the mechanical front, but still have surprises and knowledge gain at every corner when Semiconductors are concerned

Thnx for all the Knowledge base
Irfan

They can take more than that for a short time during reflow. I am using a solder paste with 63/37 Sn/Pb ratio, it melts about 185 C. Recommended maximum reflow temp is 230 deg C. Newer solder pastes, with no Lead, require higher reflow temperatures (250 C)

Soaking time is the time needed for all the board and components to reach an uniform base temperature so the thermal stress on the component's encapsulation and PCB is kept to a minimum when the reflow temperature is reached afterward.

Some components require an extended temperature treatment in order to control cracking due to absorbed humidity, that is the case for many PLCC, BGA, etc. That process is called Baking period or Baking time (could last for days), and is done prior to the reflow cycle.

[kreutz]

Ready for first set of Tests.

The PCB is ready for the tests without high voltage, (Low voltage power supplies, logic and translator's tests), after finished with those tests, we attach the power chips and heat-sink in order to begin with the motor driving tests.

[kreutz]

FYI

The low voltage tests, logic tests, translators, CPLD and the protection circuits have been tested. I made a few customizations to the Mardus-Kreutz translator's firmware, so this board needs its own firmware Rev B03a, it won't work with the Mardus-Kreutz boards and vice-versa, if somebody installs the wrong firmware, the board in question will simply quit working, and the Error LED will light up, no damage to the boards or chips.

So far the RF PC input isolation works perfectly fine too.

[contactirfu]

FYI

The low voltage tests, logic tests, translators, CPLD and the protection circuits have been tested. I made a few customizations to the Mardus-Kreutz translator's firmware, so this board needs its own firmware Rev B03a, it won't work with the Mardus-Kreutz boards and vice-versa, if somebody installs the wrong firmware, the board in question will simply quit working, and the Error LED will light up, no damage to the boards or chips.

So far the RF PC input isolation works perfectly fine too.

Great going Kreutz. Looking forward to the Finale!

[kreutz]

Early Motor tests

I made a few tests tonight, nothing impressive yet. Using a Keling Technologies' motor model KL23H2100-30-4B (495 oz-in, 3Amp RMS) Tested at 24 V and 35 volts, my settings are still very conservative on the board, I need to tune up the blanking time, and need to replace the heat-sink by the proper one. The current heat-sink is not made for vertical air flow, so it heats up considerably even with only one axis on.

Early on the thread we discussed the L6203 heating behavior, it really gets hot. The other problem is motor noise, and that one is going to be addressed when tuning the current regulation loop. I want to clarify, that this tuning is neither necessary (after development) nor even an available option for the end user.

More modifications will be done to the firmware, due to the differences in timing required for switching the bipolar board in slow and fast decays modes (not available on the unipolar board).

Most of the required logic for Decay mode control is contained into the CPLD, but the translator is the source for the switching signals and micro-step reference signals and some delays added for unipolar dead-band protection will be modified also.

[kreutz]

Test setup photos

[contactirfu]

HI Kreutz,

Nice pics, howz the firmware coming up.

Did those "L's" get heated up so much u had to put 2 fans!

Great going

Regards
Irfan

[kreutz]

HI Kreutz,

Nice pics, howz the firmware coming up.

Did those "L's" get heated up so much u had to put 2 fans!

Great going

Regards
Irfan

They really get hot easily, the fins on the heat-sink run horizontally, I placed the two little fans blowing down (used to cool hard-drives) and the thermal problem is temporarily solved, not very efficiently, at least I don't worry too much now while running the tests. So far, I am not impressed at all by the L6203s.

I haven't had much time to play with the firmware modifications.

Regards,

Kreutz.

[pminmo]

I've noticed the same on the L6203's. Major heat dissipation.

[contactirfu]

I've noticed the same on the L6203's. Major heat dissipation.

Kreutz

the L6203's still perform even when hot- dont they! Well Kreutz cant we go the Mosfet way, somehing similar to a GECkO, think you mighthave already given a thought in teh earlier part of the thread. but just looking around.

Regards
Irfan

[kreutz]

Kreutz

the L6203's still perform even when hot- dont they! Well Kreutz cant we go the Mosfet way, somehing similar to a GECkO, think you mighthave already given a thought in teh earlier part of the thread. but just looking around.

Regards
Irfan

They do work well, so far. I am limiting my tests under 4 amps peak until I gather enough information to run a few destructive tests... going the discrete mosfet output way won't be an option for this 4 axes board. The beauty of the board's idea was the integrated output simplicity.

[kreutz]

I've noticed the same on the L6203's. Major heat dissipation.

Thanks, Phil, it is comforting to hear I am not the only one having this issue...