Originally posted by Finderskeepers
View Post
Announcement
Collapse
No announcement yet.
SD2000
Collapse
X
-
Originally posted by ZED View PostHey Detecto your version of diptrace is rather old mate,get the most recent version and it will produce gerber files.
Zed
I am using DipTrace V 2.2.0.9 i think it is latest version published by diptrace, i can produce gerber files
Originally posted by Finderskeepers View PostMy finished project.
Something like this
Did you tested your SD2000?
Cheers
Detecto
Comment
-
Controls sockets
I think it would be good idea maybe to use sockets on PCB for all of the controls pots,buttons etc instead of soldering the wires directly to PCB
It would make it easier to disassemble the box ,disconnect all the wires and remove PCB
And maybe the same to connect main PCB with the small PCB
Less of desoldering when you decide to mod it.
Comment
-
Originally posted by javahitman View PostI think it would be good idea maybe to use sockets on PCB for all of the controls pots,buttons etc instead of soldering the wires directly to PCB
It would make it easier to disassemble the box ,disconnect all the wires and remove PCB
And maybe the same to connect main PCB with the small PCB
Less of desoldering when you decide to mod it.
It is not smart to add any sockets on cct like this (think of parasitic capacitance and interference) you can use connectors for wires.
What tupe of sockets we are talking abouth?
Cheers
Detecto
Comment
-
Thats the box. When selecting the elec. capacitors, make sure that thay are no longer than 30mm. The spacing between the boards is approx. 33mm. Start to plan where you want to place the controls and socket(s) on the end panels. I also put aluminium foil at the back of both panels and had the enclosure sheilded. As for connectors, you can use pcb pins and sockets, jaycar product no. hp-1250 (pins), hp-1260(sockets). Problem i found was that they stand tall and had trouble fitting both boards in the box. As a result, i trimmed down the posts and soldered the wires to it. Yeah, it's a bit tricky and care is needed to put both in at the same time.
Comment
-
Originally posted by detecto View PostIt is not smart to add any sockets on cct like this (think of parasitic capacitance and interference) you can use connectors for wires.
What tupe of sockets we are talking abouth?
Cheers
Detecto
I thought they are called sockets too
Comment
-
Originally posted by Finderskeepers View PostThats the box. When selecting the elec. capacitors, make sure that thay are no longer than 30mm. The spacing between the boards is approx. 33mm. Start to plan where you want to place the controls and socket(s) on the end panels. I also put aluminium foil at the back of both panels and had the enclosure sheilded. As for connectors, you can use pcb pins and sockets, jaycar product no. hp-1250 (pins), hp-1260(sockets). Problem i found was that they stand tall and had trouble fitting both boards in the box. As a result, i trimmed down the posts and soldered the wires to it. Yeah, it's a bit tricky and care is needed to put both in at the same time.
I thought they have a microchip in them to only work with original minelab control boxes
Comment
-
Hi Detecto,
WOW! Now that is one very nice board layout, looks better than the original one!
With the byv28-200, when you are considering the capacitance, make sure you take into account that there will be 180v sitting on the mosfet side of it during the receive period. From memory this lowers the C to about 10pf or somewhere around there.
And the buk-455's be careful changing them for something lower capacitance/on resistance. This will upset the ground balance for some reason and with a standard logic cct like that, there is no way to adjust the timing to suit. So initially stick with the original parts otherwise for some reason it may appear not to work.
It will be interesting to see how your board performs for noise too, should be good! Although I would have done it with a micro instead of the logic cct, but I see thats your next plan after getting this one up and running! Good luck!
Finderskeepers, your setup looks pretty neat too. Good stuff! One day we should meet up for a swing and compare units just for interest sake.
Cheers Mick
Comment
-
HI Javahitman,
The coils for the ML Pi's don't have any fancy code lock thingys in them! They are just a coil, however, I would suggest it is a good investment to buy a ready made coil rather than try to make one. Coil building materials are very hard to find, lets take the coax cable for instance. It is made up of a litz core of at least 32 strands of .2mm and the outer shielding is litz again with 92 strands of slightly smaller wire. This is like this to reduce the capacitance and ability to hold eddy currents, while still being able to handle the required current. Try and find a coax available like that! Then there is the coil, also uses litz and needs to be secured properly and shielded and ensured that the windings and shielding don't move in the housing as this will cause false noise which is annoying! Then there is the coil shell!
Thankfully the electronics side of things is fairly generic!
Cheers Mick
- Likes 1
Comment
-
Originally posted by javahitman View PostSorry if I didn't phrase it correctly , I was thinking of using connectors for wires
I thought they are called sockets too
I am week with English
I didn't practice writing and speeching English from high scool and it is normal proces to forgot things.
Cheers
DetectoLast edited by Carl-NC; 07-16-2012, 02:56 PM.
Comment
-
Originally posted by Mechanic View PostHi Detecto,
WOW! Now that is one very nice board layout, looks better than the original one!
With the byv28-200, when you are considering the capacitance, make sure you take into account that there will be 180v sitting on the mosfet side of it during the receive period. From memory this lowers the C to about 10pf or somewhere around there.
And the buk-455's be careful changing them for something lower capacitance/on resistance. This will upset the ground balance for some reason and with a standard logic cct like that, there is no way to adjust the timing to suit. So initially stick with the original parts otherwise for some reason it may appear not to work.
It will be interesting to see how your board performs for noise too, should be good! Although I would have done it with a micro instead of the logic cct, but I see thats your next plan after getting this one up and running! Good luck!
Finderskeepers, your setup looks pretty neat too. Good stuff! One day we should meet up for a swing and compare units just for interest sake.
Cheers Mick
Original SD2000 don't have ground plane, and i have a little filing that engineers from Minelab almost randomlu placed components and afther that used some autoruter,
ofcourse they crucial parts ruted manualy
I am pretty shure that my PCB will have less noise from original!
About BYV28-200 that part left a lot of place for experimenting, and abuot FET-s its hooded job.
Regards
Detecto
Comment
-
Hi Detecto,
The original one does have a ground plane, although it is not very plainit has lots of slots in it where they routed traces across the board, yours looks much more plain!
I don't know that the engineer's that originally worked on the 2000 just randomly placed components, there would have been a lot of thought gone into it. At the end of the day though they did not need to make the machine real low noise as it was a giant leap in front of what had come before it. Heck they could have probably run it from a bread board and still had excellent results in the field!!!!The amount of gold they found with the 2000 was amazing!
As for the fets being a hooded jobI will let you in on my observations. Ages ago I first tried irf740 cause thats what everybody was using for their home brew detectors at the time. These fets worked, however their on resistance was much higher than the buk-455's, so me being me, I decided it would be better to push more current into the coil, so then opted for some other ones that were used in the 2200(I don't remember the part number anymore) they were lower resistance. I then found that the I ran out of ground balance in a lot of locations and could not use the detector properly. So the stock ones went back in. My thoughts are that the lower R fets allowed more current into the coil, causing the flyback to last longer which in turn made the first sample closer to the flyback with a higher amplitude ground signal. Probably if I could have had an extra 1us delay before the first sample everything would have been fine!
Make sure you use good low esr caps for the tx tank and 5V power supplies.
Cheers Mick
Comment
-
Originally posted by detecto View PostThanks satdaveuk
You can find DipTrace here===== http://www.diptrace.com/download.php
LoL i should be pay from DipTrace for marketing :P
Generaly i don't want to share project right now because concept may have error.
Building SD2000 is expencive project like Finderskeepers say he spent abouth 470 US dollars, and i don't know what kind of component he used... with my calculations abouth 450~600 USD will be enough for building, I would not like to someone start building SD2000 and on the end if something is not right what? throw money!!
To prevent that i will build one working device and when i assure that everything is OK i will share than.
Schematic is fully correct without discussion!
Here is what you can import and export from and into DipTrace
Cheers
Detecto
Shall download and see how I get on, like the fact it does a few more than just its own format unlike others that only do there own.
Looks more user friendly by the what I see with the menus there.
With these programs I do everything *** backwards, tend to jump in head first get to know it with trial and error then read the instructions.
Alot will laugh but it works for me
Thanks again
Regards
Comment
-
HI Javahitman,
Do you have a spare computer?
There is a free full 30 day trial.
System restore is your BEST friend....
However if you were to just download a large layout(larger than your pin limit), you can re-arrange components and traces, just not add any more components.
I think there would be a couple more than 300 pins on the main board of the 2000
Cheers Mick
Comment
Comment