PCB design for another box

[FROMI]

Hello, I would like to change the pvb for an old Futaba T9CAP and
would be so please to get the KiCAD project in ZIP file as far as the
dowloaded version having problems with BZR 3256 (the sd card-
cache.lib, sdcard.sch, is not an eeschema library, sdcard.pro
linne1560=expected and similar [already found the reason, but very
time consuming to get all the file]). Please could somebody send the
frozen verzion or later? Thanks in advance.
Regards, Ferenc

[alain bregand]

I got the same trouble and Bryan sent these files
have a good day

-----Message d'origine-----
From: FROMI
Sent: Tuesday, May 15, 2012 5:25 PM
To: gruvin9x-discuss
Subject: PCB design for another box

[Ferenc Roman]

problem caused by the specific file handling of the Google project...if you check the list of files, and the link under the listed item, it will not show the way to the file but to the editor of the file. So, to download correctly, press the link. It will move you to the editor view and there you can download the file itself under View raw file. (right click, save as...). but you will need libraries etc. Thats why it will be nice to have zip files for the frozen brunches available for download.
Sincerely,
F.R.

P.S.: here is the folder for files:
http://gruvin9x.googlecode.com/svn/archive/pcb-v4.1a-frozen/

[Bryan]

Even better, install Subversion and then "check out" the entire directory. Instructions for doing that (command line version) are on the standard Google "Source" page. You'd need to change the source path is all. For example ...

svn checkout https://gruvin9x.googlecode.com/svn/archive/pcb-v4.1a-frozen g9x-pcb-v4.1

... or if for some reason you want to check out (downlaod) the entire project, including Wiki etc ...

svn checkout https://gruvin9x.googlecode.com/svn g9x

That will create a directory named g9x-pcb-v4.1, containing all the files and subdirectories.

Oh and -- those other two files I posted here are not PCB v4.1. It's only the very original, reverse engineered Schematic of the factory controller.

Bryan.

[Bryan]

I messed up sentence order. :-/ Here it is again ...

On 16 May 2012 18:10, Bryan <gru...@gmail.com> wrote:

Even better, install Subversion and then "check out" the entire directory. Instructions for doing that (command line version) are on the standard Google "Source" page. You'd need to change the source path is all. For example ...

svn checkout https://gruvin9x.googlecode.com/svn/archive/pcb-v4.1a-frozen g9x-pcb-v4.1

[Ferenc Roman]

Thanks, I will try.

Have you heard may be that anybody done an adtatiob of the gruvin board for Futaba FF8, FF9(T8CAP, 8U, T9CAP)?
Once it done I'd rather pay a beer or two ;)

[Bryan]

On 16 May 2012 19:42, Ferenc Roman <fro...@gmail.com> wrote:

Thanks, I will try.

Have you heard may be that anybody done an adtatiob of the gruvin board for Futaba FF8, FF9(T8CAP, 8U, T9CAP)?
Once it done I'd rather pay a beer or two ;)

Not heard of anything like this sorry. I have heard of people retro-wiring a Spektrum DX5i though. But no change to PCB design. They just stuck in in there -- even leaving the original PCB in place.

I haven't seen the Futaba FF8 in this part of the world, by the way. But I don't really get out much these days either.

The Futaba 18MZ would be a nice platform to work with. But I'm guessing it would cost far to much to then go pulling apart and putting our board in. :p The 8J looks more realistic in this regard. Probably still a bit to costly ... though I for one would enjoy knowing I had higher quality gimbals, etc. (Not that I have personally suffered a gimbal failure on a Chinese '9X yet. Wish me luck!)

Bryan.

[x3mfly]

Or 9CAP :-) i have one. there are a lot of  8/9 series futaba sold. and people was happy, but at this moment Gruvin has done excellent replacement for my Futaba 9CAP and only thing don't like me it is case of turnigy 9x, Futaba case looks solid and has sliders, gimballs feels better etc.

среда, 16 мая 2012 г., 12:44:38 UTC+4 пользователь Gruvin написал:

[Ferenc Roman]

Privet Maxim, have you tried to replace or even pre investigated the situation of possibilities?
FF8 has a little bit week switches, ff9 better but heavier and has sliders. Thats good. I have one 9ZAP but that but my heart hurt thinking about dissembling that. :) Really robust radio. So, it will be a an ff8 or ff9 if I'll have time. Hope, here I will have some help to do that right. :)
Anybody heard about eepe or compagnio for Android?:)

[Ferenc Roman]

I think, hundreds of FF8, FF9 users will be happy to replace their native boards to Gruvin :)
What do you think?

[x3mfly]

No, not investigated, i think only if someone do this work :-) i think not me. my knowledge of kicad so fast from good...

but i will be happy if anyone do this :-)

среда, 16 мая 2012 г., 14:30:07 UTC+4 пользователь FROMI написал:

Privet Maxim, have you tried to replace or even pre investigated the situation of possibilities?
FF8 has a little bit week switches, ff9 better but heavier and has sliders. Thats good. I have one 9ZAP but that but my heart hurt thinking about dissembling that. :) Really robust radio. So, it will be a an ff8 or ff9 if I'll have time. Hope, here I will have some help to do that right. :)
Anybody heard about eepe or compagnio for Android?:)

[Ferenc Roman]

In 1-2 weeks I will have ff8 and 9. I will make photoes, measure. Who will join? :)

[Bryan]

On 16 May 2012 20:52, x3mfly <maxim....@gmail.com> wrote:

Or 9CAP :-) i have one. there are a lot of  8/9 series futaba sold. and people was happy, but at this moment Gruvin has done excellent replacement for my Futaba 9CAP and only thing don't like me it is case of turnigy 9x, Futaba case looks solid and has sliders, gimballs feels better etc.

I think we could all agree on that score. Futaba gimbals etc are definitely nicer.

[FROMI]

Hello, got my board, thanks but have one qwe: what mCU need to be
used?

ATMEGA2560-16AU TQFP 100A 100 16 2.7-5.5 Industrial (A) (-40°C
to 85°C)
ATMEGA2560-16AUR TQFP 100A 100 16 2.7-5.5 Industrial (A) (-40°C
to 85°C)
ATMEGA2560V-8AU TQFP 100A 100 8 1.8-5.5 Industrial (A) (-40°C
to 85°C)
ATMEGA2560V-8AUR TQFP 100A 100 8 1.8-5.5 Industrial (A) (-40°C to
85°C)

[Cam]

Either ATMEGA2560-16AUR or ATMEGA2560-16AU. The difference is only the
way they are packed from the factory, one is tape and reel, the other is
tray.

Cam.

[Ferenc Roman]

ok, thanks.

[Bryan]

I personally use the ATMEGA2560-16AU. No idea what the AUR version has that the AU does not -- could be the delivery packaging?

Bryan.

[Romolo Manfredini]

R means Rail... for Pic & Place...

[Ferenc Roman]

Hi, thanks for advices but one more is needed: I have got a soldering station and need advice what temperature is preffered for 0603 smds and MCU. Don't want to fry them. I thing approx 170 and do it fast or what is the best practice (max could be set 450Cdeg :)?

[Bryan]

@Romolo: R = rail. Packaging then. I guessed right, sort of. Thanks. :-D

I operate the soldering iron tip at 340ºC, with leaded, rossin core solder. I need about 390ºC for unleaded, along with plentiful additional flux to get good results. (Unleaded is painful. Avoid it unless you're really serious about saving the planet. I use it only for reflow work on factory made boards, since the two types do not mix well.)

I run the hot air for re-flow (isolated, initial construction) work at 360ºC. (The actual surface temperature of solder paste doesn't exceed about 250ºC, due to "chill factor" of the air flow, I guess.)

For chip removal (using air) I have found 340ºC to be safest. It takes much longer to get all pins on an MCU or LCD socket to melt -- but it's worth the wait to prevent damaging the board, since you have to keep the heat on for a LONG time in this case -- though pre-heating from beneath at 150ºC makes everything go a lot smoother and quicker. This is made worse with hand soldered chips, since the hand solder wire has a higher melting temperature than most solder paste. But it can still be done, with patience. Sustained hot air at 360ºC or above will burn the PCB. But 340ºC and below seems pretty safe for very long time periods .... especially if you keep the air moving about a bit, so as to avoid focussed hot spots, as you always should.

I can also report that I have cooked several ATmega2560 and 2561 chips at 340ºC for several minutes during removal and re-soldering to new boards and found no trouble with them afterwards. This was done for cost savings during migration from older prototype boards to newer ones. So I wouldn't worry about heat damage too much, as in reality these chips seem pretty darn hardy. That said ... an extra 20 or 30 degrees can make a big difference. So you are wise to use a temperature controlled station -- for sure. Besides, solder doesn't flow well and flux burns off far too rapidly if things are too hot.

Bryan.

[Ferenc Roman]

man, i have a degree in environmental protection so - 100% serious. ;)
I've just got new pins for my old stattion:
http://www.howardelectronics.com/goot/images/RX802AS&Standweb.jpg
I will do my best with this as far as I was working with 5 mm wide tiped hand tool for years even for SMD. :)

[Bryan]

On 2 June 2012 00:23, Ferenc Roman <fro...@gmail.com> wrote:

man, i have a degree in environmental protection so - 100% serious. ;)

I certainly care about the environment. But for development boards that will not leave my ownership, I can control the disposal side of things properly and maybe save an endangered species, after a unleaded solder dry joint cause a plane crash and fire. (Perhaps only 98% serious here. :p) Unleaded solder is very difficult to work worth by and and get quality results, compared to leaded. Unleaded is fine for mass production / robotic / re-flow methods.
 

I've just got new pins for my old stattion:
http://www.howardelectronics.com/goot/images/RX802AS&Standweb.jpg
I will do my best with this as far as I was working with 5 mm wide tiped hand tool for years even for SMD. :)

Ah yes. That would present the odd challenge! :-D

Bryan.
 

[Bryan]

I hate auto-correct. " ... very difficult to work with by hand ... "

[Ferenc Roman]

Hello, working on the pcb received. R31, 34, 79 are marked 10k2, do you mean 10 200? could it be replaced witth just 10 000? Thanks.

[Bryan]

The short answer is, "Yes, you can change those values."

Actually, unless you intend using the 3 spared ADC ports, you can leave them off entirely, without any harm. You should install the 3 x 5K1 resistors (R17, 33, 77) though, to protect the ADC inputs.

Bryan.

On 11 June 2012 05:30, Ferenc Roman <fro...@gmail.com> wrote:

Hello, working on the pcb received. R31, 34, 79 are marked 10k2, do you mean 10 200? could it be replaced witth just 10 000? Thanks.

--

Bryan.

[FROMI]

hi, seems like parts together.
Please help with J1 and J4 what are they for?

Thanks in advance.
Regards, Ferenc

[Bryan]

Ferenc

On 7 July 2012 09:55, FROMI <fro...@gmail.com> wrote:

hi, seems like parts together.
Please help with J1 and J4 what are they for?

They are just for selecting optional wiring for the vibrator (haptic) and LED backlight control ports.

The short version, is that we simply tie pads 1&2 together (solder-bridge or zero-ohm resistor) on both J1 and J4 (independently, of course) -- left two and upper two pads, respectively (I think). That will set things up in the way expected for current firmware versions.

Bryan.

 

[Ferenc Roman]

Ok, thanks, done.
My last Qwe is what are the test steps need to be done? I hope you
have a list of test points newbies :)

[Bryan]

On 8 July 2012 06:00, Ferenc Roman <fro...@gmail.com> wrote:

Ok, thanks, done.
My last Qwe is what are the test steps need to be done? I hope you
have a list of test points newbies :)

Actually, we're all newbies to this project. Personally, this thing is the first SMD board I ever made. Just believing that I might be able to do so was a major leap for me, after many years believing it was impossible without robotic assembly tools. (Truly. Seems dumb now. But I genuinely gave up on my long loved electronics hobby for about ten years, due to this errant belief! So much time lost over silly, personal ignorance. Ah well. Things like this happen when one lives in a bubble. :P)

To answer your question (and hope) -- no, we don't, sorry -- at least, not as far as I can recall. I guess the assumption was that someone capable of reading the schematic diagram and assembling the board should also be capable of devising their own tests. In retrospect, whilst that is likely true in your case, a pre-scribed set of procedures would have been convenient.

Not really sure what to suggest. In truth, the way I test these boards, is to simply connect everything up and see if it comes up with the right stuff on the LCD display. If it doesn't then I start looking for possible causes. If it does, then I proceed directly to stick calibration and then to the DIAG screen, to check all switches and rotary encoders. Oh and I also set the date and time (RTC/SDCARD module installed) and power cycle to see if it sticks -- then use the model backup/restore feature of open9x to ensure the SD card is working. Some time during all that, I naturally fire up an RC receiver and servos and check that it does the right stuff. 

I guess that's all kind of unprofessional and hobbyist-like. Then again, that would be because I am in fact only a hobbyist, with limited professional experience. Hmmm. (A genius, not an engineer! :P)

Now that I'm made to think about it ... I suppose some kind of, "Install only the following components, then test voltages from regulator. If OK, then proceed with the following components ..." construction regime might have been in order??? Frankly, now that the Chinese factory is building units for less than we can all buy just the components, it's doubtful we will bother doing such a thing.

OK ... so in practical terms, for you, right here, right now ... I guess it might make sense to install everything except the LCD screen cable, switch on and check for smoke. If no smoke (I have had smoke, twice -- but that's another story!) then check that the 5V and 3.3V regulators are working as they should. In practice, you'll be wanting to see between 4.88-to-5.12V and 3.28-to-3.32V, respectively. After that, all I can think of is the above mentioned acid test -- looking for the right tuff on the LCD display. (Obviously, you need to first program the fuses and Flash some open9x firmware. More details can be provided, if needed.)

This is not to say that all my builds have 'just worked'. Far from it. The biggest problem has always been with the soldering of the MCU and LCD sockets -- specifically in the form of solder bridges and (more commonly) invisible;y dry joints -- the type that look fine even at 400x magnification, under a USB microscope! In reality though, that was all before I "perfected" my "reflow flux and drag soldering" technique, which now appears to make really good joints every time. (There's some YouTube videos linked in the Wiki somewhere, covering that -- I think.) But it wasn't like that in the beginning, so the odd poor joint did come about now and then.

In any case, if your LCD screen doesn't produce the right (or any) display, then continuity testing with a multimeter or something (power off) is the only advice I can give. It worked for me in all cases -- eventually. The trick I finally learned was being aware not to put pressure on device pins themselves -- directly over the PCB pad -- because that can make a connection seem OK that actually isn't, without said pressure literally making the connection at the time. That one cost me several hours of frustration on more than one occasion, before I finally realised what was going on. After that, my first action after seeing no display became to re-solder the LCD socket and MCU -- another coating of flux and dragging the iron over them. This inevitably fixes the problem, without tedious testing in-between.

Having said all that ... with just a little luck, your build will Just Work™ :-D Let us know!

Bryan.

[Ferenc Roman]

All parts together, works fine, but I am worried about stability,
thats why asking. I thought you worked out some tests where test pads
are used :). From 1980s I remember for a schematic of a Russian TV and
there were points and with signals displayed. I think it will be
utopia to have that made :). I think that need deep knowledge of
analog technique and noise reduction. Ok, let us see what will happen
with this radio under full power. powered from USB its beeping
randomly so something goes wrong or suffering of low V.
For flashing I used Win Environment with AVRDUDE and AVR-O-BURN. This
is the simplest combination. Unfortunately was not able to set
correctly the USBasp on UBUNTU, but that means I was not patient
enough. May be next time, when Open9x will be installed. Now I am
going to put all together in my old 9x.

Beeping resolved by calibration, smoke collected into a bug and placed inside.
One more qwe:
"Several cases of 'ground looping' resolved. (Use single, common tie
points instead of 'loops' whenever possible.)"
Could you show those points. As for me unpredicted behavior is worse
that totally dead TX...
Thanks for all in advance.
Regards,
FErenc

[Bryan]

On 9 July 2012 06:31, Ferenc Roman <fro...@gmail.com> wrote:

All parts together, works fine, but I am worried about stability,
thats why asking. I thought you worked out some tests where test pads
are used :).

We did discuss making a "board of nails" test rig for a while. But it never eventuated. I am re-considering it again now though, since I'll have 50 boards here to test very soon. *gulp* The problem is that I believe it will take longer to make the test rig than to test even 50 boards manually. The benefit of having a test rig would be the potential of doing longer term burn in tests ... maybe. A lot of work to set all that up though.

From 1980s I remember for a schematic of a Russian TV and
there were points and with signals displayed. I think it will be
utopia to have that made :). I think that need deep knowledge of
analog technique and noise reduction.

Indeed. ;-)

 

Ok, let us see what will happen
with this radio under full power. powered from USB its beeping
randomly so something goes wrong or suffering of low  V.

That should just be the low battery alarm (in the SETUP menu, accessed using [LEFT long] button) which is slightly random sounding, at times, but otherwise pretty regular in interval and pitch. I've never heard random beeps other than this, when powered from USB.

 

For flashing I used Win Environment with AVRDUDE and AVR-O-BURN. This
is the simplest combination. Unfortunately was not able to set
correctly the USBasp on UBUNTU, but that means I was not patient
enough. May be next time, when Open9x will be installed. Now I am
going to put all together in my old 9x.

Be aware that most USBasp programmers have a bug that prevents them working correctly with the ATmega256x MCUs, because they do not reset the extended address register properly during Flash programing. At least one person (Romolo, I think) has patched a 3rd-party USBasp firmware to fix this and of course the G9X gPROG USBasp programmer also has the fix.

 

Beeping resolved by calibration, smoke collected into a bug and placed inside.

Good idea! :-D

 

One more qwe:
"Several cases of 'ground looping' resolved. (Use single, common tie
points instead of 'loops' whenever possible.)"

I believe that comment turned out to be erroneous, on my part. When the flood fill was applied to the board, all such loops vanished, since they became filled in ground planes. The corrections I thought I made were done with the flood fill view switched off and me not thinking right, no doubt late one night. I am certainly not aware of any ground loops int he real v4.1 or v4.2 designs now.

There is however a loop to be avoided external to the board. This is the one created by connecting the ground wire for Fr-Sky telemetry, since it already has another connection to ground at the power supply input to the TX module. In practice, this ground loop only caused problems when I had the RS-232 voltage level shifter circuitry wired off-board -- probably because the higher impedance transistor base inputs were much closer to the TX module. This has not been a problem since custom PCBs with ground plane and level shifter circuit included on-board. In any case and to avoid potential problems, I connect only Rx and Tx telemetry wires (marked R and T on the v4.x board) leaving ground (G) disconnected. I don't use shielded cable anywhere and have never yet experienced any instability. I have also since tested with all three wires (including ground) and not seen any problems. But I like play it safe.

Bryan.

[Ferenc Roman]

Well, thanks for comments/advices. Smoke is still inside:)

Few points to consider may be for next version:
1. USB - i think, pads need to be changed for micro USB connector.
2. j1, j4 - may be it will be better if jumpers like on HDD could be applied.
3. Jack 3.5 is an idea for vario solution that will disconnect the
main speaker if earphone is used.

For now i am up to neck with electronics. Lets build solar-powered plane.
Next time I will bother ErSky team but that will be a totally self
designed MB for one of my old-school futabas may be with color display
and touch. Plan exists the qwe is now about free time:)

Thanks for all and good luck!
;)

[Bryan]

On 9 July 2012 17:46, Ferenc Roman <fro...@gmail.com> wrote:

Well, thanks for comments/advices. Smoke is still inside:)

Few points to consider may be for next version:
1. USB - i think, pads need to be changed for micro USB connector.

Hmmm. Not feasible to have socket on the bottom of the unit, for me. (It rests on my belly! :P) That and I personally loath and hate micro-USB connectors. Mini would be OK.
 

2.  j1, j4  - may be it will be better if jumpers like on HDD could be applied.

These should have been PCB trace bridged for the v4.2 version, but forgot. It's unlikely many (or any) people will change it, me thinks.
 

3. Jack 3.5 is an idea for vario solution that will disconnect the
main speaker if earphone is used.

Not sure what you mean exactly -- in terms of the PCB itself? This can easily be done off-board, of course. Just insert the jack switch in the speaker line. I must be missing something.
 

For now i am up to neck with electronics. Lets build solar-powered plane.
Next time I will bother ErSky team but that will be a totally self
designed MB for one of my old-school futabas may be with color display
and touch. Plan exists the qwe is now about free time:)

Sounds like fun :-D
 

Thanks for all and good luck!
;)

I take it everything is working then? Cool. :-D

Bryan.

[Ferenc Roman]

3. Jack 3.5 is an idea for vario solution that will disconnect the
main speaker if earphone is used.

Not sure what you mean exactly -- in terms of the PCB itself? This can
easily be done off-board, of course. Just insert the jack switch in
the speaker line. I must be missing something.

I mean that it will be a good idea of two serialwired connectors: One
for speaker, another for earphone that has disconneting contacts for
the speaker if jack is plugged in, the same way as its done for cd
players.

BTW, what otehr firmwares are compatible with the board?

[Bryan]

On 9 July 2012 21:15, Ferenc Roman <fro...@gmail.com> wrote:

...

I mean that it will be a good idea of two serialwired connectors: One
for speaker, another for earphone that has disconneting contacts for
the speaker if jack is plugged in, the same way as its done for cd
players.

That would be a more tidy solution to wiring it off-board, I suppose. Thanks.
 

BTW, what [other] firmwares are compatible with the board?

 
To my knowledge, open9x is the only firmware currently supporting all the most important hardware features. (gruvin9x development has cased, well short of that initial goal.)

The best way to get the .hex file you want for flashing is via the "Download" button in the preferences of companion9x.

We did send a v4.1 board to the th9x project leader, mostly as a gesture of thanks for getting this whole thing started, with no obligations or expectations attached. In short, I'm not sure if th9x has a goal to support the board or not.

I'm pretty sure I tried to contact the RadioClone fellow a couple of times, but got no response. His software was looking quite good, but running out of room on the stock board. No idea what the status of that project is at this time however.

Bryan.

[Ferenc Roman]

So, two main streams remained:
er9x and ErSky
Gruvin and Open 9x

Ok. I would like to have a little bit more fun so whom could help me
with the code? It could be usefull for the next 6 months. I would like
to adopt the menu to my needs. The other side i found that a 1 rotary
and 2 buttons are enough for the transmitter. So generally need to
rewrite the controls. Then will replace 4 buttons with rotary.
Something like FF9 has. Compact and intuitive: scroll if you vould
like to move or increase/decrease the value, press to save, menu to
enter change mode, exit to exit without saving.

BTW: how could I activate the vibra. It was continuously running while
the jumper was not soldered 1&2. Haw many pins do we have there are
not used at the moment. Thinking about scanner based switch extender.

Thanks again.

[Bryan]

On 10 July 2012 00:52, Ferenc Roman <fro...@gmail.com> wrote:

So, two main streams remained:
er9x and ErSky
Gruvin and Open 9x

Ok. I would like to have a little bit more fun so whom could help me
with the code? It could be usefull for the next 6 months. I would like
to adopt the menu to my needs. The other side i found that a 1 rotary
and 2 buttons are enough for the transmitter. So generally need to
rewrite the controls. Then will replace 4 buttons with rotary.
Something like FF9 has. Compact and intuitive: scroll if you vould
like to move or increase/decrease the value, press to save, menu to
enter change mode, exit to exit without saving.

I'm not sure who to suggest in terms of help with the code. Everyone who has ever worked on the code to date has done so by looking at what was there already and teaching themselves. I've not witnessed a single question as to how to do this or that, ever (except ones I have asked of Bertrand recently.)

Best thing I can suggest is getting involved with the the open9x team.

I will note that what you propose is a MAJOR code re-write and will take a lot of time and effort -- especially since the existing menu and screen layouts may not lend themselves well for such an input scheme. I'm sure it will be well worth it though.

BTW: how could I activate the vibra. It was continuously running while
the jumper was not soldered 1&2. Haw many pins do we have there are
not used at the moment. Thinking about scanner based switch extender.

I honestly do not know. The gruvin9x firmware never got around to including vibrator control. I believe open9x has done so though.

Bryan.

[Bryan]

Oh .. actually, I forgot. Bertrand has already coded all the menus to work with a single rotary encoder and its single push-button action! This comes standard with open9x already.

On 10 July 2012 12:58, Bryan <gru...@gmail.com> wrote:

...

[Ferenc Roman]

Yesterday I spent an hour to get into code but missing documentation means it will be not so easy. Anyway, thanks. May be I will find time to collect functions of the TX and may be a new soft release apear if I will be able to involve 2-3 eng. students. but that will be not now. :)
 See you later, alligator ;)

[Bryan]

Good luck. :-)

Of course, it wasn't easy for any of us, given the noted lack of documentation, least of all likes of myself, with limited programming skills. But the code ultimately makes its own sense. (hopefully) It's also very well known by Bertrand nowadays, since huge chunks of it are his own work. Lots of optimisation, etc. No one ever seems to have time to document it, beyond code comments. I wouldn't even know how to do so, actually.

Bryan.

On Tuesday, 10 July 2012, Ferenc Roman wrote:

Yesterday I spent an hour to get into code but missing documentation means it will be not so easy. Anyway, thanks. May be I will find time to collect functions of the TX and may be a new soft release apear if I will be able to involve 2-3 eng. students. but that will be not now. :)
 See you later, alligator ;)

--

Bryan.