cheaper large instrument displays - mast mount

Backlight already working. Just need to choose a cheaper MOSFET than the one I found in my bits box.

Not got the new op-amps yet for contrast.

Need both 5V and 3.3V power supplies but this is not difficult. Just need to get the price down.

For temperature testing I need to know the values of Vo to get the contrast set at a readable level for the hot and cold temperature limits.

Circuit design can be taken from the Disco board datasheet. It's intended by ST to be a reference design to be copied.

A remaining piece of work not mentioned is software flashing. The method used on the Disco boards (JTAG) is not suitable. However, this processor can be flashed serially using a free Windows app from ST. To do this a small amount of extra componentry is needed. There are 3 options...

1) Just a connector block on the PCB. This is the cheapest. All the extra components will be on a small flashing PCB that is plugged in to flash the software. The disadvantage of this method is that the user will not be able to update the software without obtaining a flashing device. This could be made or bought, for example...

http://uk.rs-online.com/web/p/interface-development-kits/0406568/

2) A RS232 line driver and a jumper on the PCB and a 9 pin connector. The user will be able to flash the software for any updates. This will work fine if they have a proper RS232 port on their computer. I don't know if this method works via RS232 to USB converters so it can be done from a computer without a serial port. This needs investigating. It is the cheapest option for user flashing.

3) Put a FTDI USB to UART chip and a USB socket on the display's PCB. These cost about £4.50. Anyone could then flash the software with just a USB cable.

http://uk.rs-online.com/web/p/universal-asynchronous-receivers-transmitters/0406580/

http://uk.rs-online.com/web/p/type-b-usb-connectors/5298218/
 
I've got a couple of ST JTAG adaptors - that would be my preferred approach.

Bearing in mind the eventual cost of the casing - I don't think shaving a few pence here and there is a big deal. Switching regulators is the best way to go for the 5V and 3.3V supplies. Wide input range and minimum power dissipation.

MOSFET for backlight control won't be a problem, and I would say nothing further needs to be done at this stage.

A thought for the future - all electronoc devices offered for sale in the EU must comply with the EMC (Electromagnetic Compatibility) directive. This includes ensuring immunity to interference and and prevention of unwanted RF radiation. Testing for compliance can be rather expensive.
 
Feedback wanted

I guess Rohs compliance as well.

I am not sure on the flashing method..I like the reliability of jag and 2 wire..although end user flashing would be a first for instrument displays and would would be another first as far as I know...however to get to the USB/serial (usb via ftdi would be better, serial is getting harder and harder for end users) we would either need another waterproof opening or the end user would need to take the unit to bits

On the rationale above there are pros and cons...maybe some feedback on this would be useful

End user to be able to self upgrade software using USB (may entail opening the case.....yes or no?
 
Interesting...iec945 refers to solas compatibility and emc for marine navigation devices...major manufacturers state that leisure equipment is not subject to this with a disclaimer

I have checked a couple of manufacturers manuals and they claim they meet emc but don't say against which standard..I will keep looking
 
Garmin meet the following -the only company I could find with a decleration

Application of Council Directive: 1999/5/EC, 2004/108/EC
Standard to which Conformity is Declared: EN 60945: 2002 Maritime Navigation and Radio Communications Equipment – General Requirements
EN 60950-1:2001 Safety of Information Technology Equipment
 
modular approach ...

I wonder if it is worth adopting a modular approach for the separate functions, so that you end up with 3 / 4 / 5 small boards into one base, rather than an a fully integrated single circuit layout ?

The ability to repair/ upgrade a plug-in board might be a good sales point, rather than having to junk with a whole PCB.

Do we need to carry out some serious preliminary work on the question of electro-magnetic compliance to ensure that it is built-in to the development. Looks like some project management flow / objectives might be useful to clarify who is dealing with what, and to draw up a rough critical path.


Or KISS ? :)
 
ShaunG said:
End user to be able to self upgrade software using USB (may entail opening the case.....yes or no?
Click to expand...

Open the case, move a jumper, plug in a USB cable, open flashing Windows application, load file, click download, disconnect cable, replace jumper, replace cover. I think that would be within the capability of everyone.

With user upgrade if someone reported that their particular NMEA message was not supported or wanted the text in a different language it would be easy to make the changes and send a new image to flash.
 
Last edited:
AngusMcDoon said:
Opening the case, moving a jumper, plug in a USB cable, open flashing Windows application, load file, click download, disconnect cable, replace jumper, replace cover. I think that would be within the capability of everyone.

With user upgrade if someone reported that their particular NMEA message was not supported or wanted the text in a different language it would be easy to make the changes and send a new image to flash.
Click to expand...

Even I could do that !!!
 
sarabande said:
Or KISS ? :)
Click to expand...

I think we are getting carried away with ideas here.

In my opinion version 1 should be based on the Discovery board (£9.84 inc VAT) which plugs into a through-hole PCB with the few extra components required (NMEA/Seatalk interface, backlight/contrast control, power supply), use a near off-the-shelf box as possible, flog a few dozen at cost price as a demonstration kit (claimed not for actual use, no RoHS or EMC verification required) and see what the feedback is.

The kit could be a box of bits to solder up, or already soldered up and just needs the box lid screwing on - just something to keep it a 'demonstration kit' to avoid all the certification guff. The first dozen prototypes could be Veroboard based. Made properly, this manufacturing method can still make products that outlive the owner.

I've seen too many development projects fail or get bogged down because they try to do too much. This way something might actually get done.
 
Last edited:
Yes, you bypass EMC by supplying kit, rather than complete product.

ROHS is not a problem as all parts/modules that you can buy will be compliant anyway (just make sure you use lead-free solder).

You might want to consider designing the add-on PCB to use entirely through-hole parts, making it suitable for self-assembly. Perhaps that was the idea anyway?

If you're basing the "production" unit on the Discovery PCB, then it makes sense to use the usb (or whatever) gubbins this comes with it for programming and update. I use the JTAG interface because it gives a full de-bugging interface with access to all internal registers etc with minimal extra hardware (just the connector and a couple of resistors) on the board and doesn't tie up any port pins when finished with. But then, I designed the whole PCB so don't have to rely on whatever the development board gives me.

I'm guessing the Discovery board already has 3.3V regulation and only needs 5V supply?
 
mikefleetwood said:
I suspect the backlight LEDs will provide enough heat for most situations - especially in an enclosed box.
Click to expand...

Isn't that what Raymarine do in any case?

I'm sure I've read in the manual that if condensation is present on the inside of the screen just whack up the lighting to max and that clears it.
 
mikefleetwood said:
Yes, you bypass EMC by supplying kit, rather than complete product.

ROHS is not a problem as all parts/modules that you can buy will be compliant anyway (just make sure you use lead-free solder).

You might want to consider designing the add-on PCB to use entirely through-hole parts, making it suitable for self-assembly. Perhaps that was the idea anyway?

If you're basing the "production" unit on the Discovery PCB, then it makes sense to use the usb (or whatever) gubbins this comes with it for programming and update. I use the JTAG interface because it gives a full de-bugging interface with access to all internal registers etc with minimal extra hardware (just the connector and a couple of resistors) on the board and doesn't tie up any port pins when finished with. But then, I designed the whole PCB so don't have to rely on whatever the development board gives me.

I'm guessing the Discovery board already has 3.3V regulation and only needs 5V supply?
Click to expand...

Everything extra to the Discovery board can be done through-hole; that is my suggestion too.

For flashing if based on the Discovery board, then use the on-board JTAG. That is what I have done so far. My earlier FTDI suggestion was for a later custom board.

The Discovery board takes power from the USB at 5V and regulates it to 3.3V. If a Discovery board was used in version 1 then we would need to regulate 12V to 5V only. The board has an external 5V input pin.
 
AngusMcDoon said:
I think we are getting carried away with ideas here.

In my opinion version 1 should be based on the Discovery board (£9.84 inc VAT) which plugs into a through-hole PCB with the few extra components required (NMEA/Seatalk interface, backlight/contrast control, power supply), use a near off-the-shelf box as possible, flog a few dozen at cost price as a demonstration kit (claimed not for actual use, no RoHS or EMC verification required) and see what the feedback is.

The kit could be a box of bits to solder up, or already soldered up and just needs the box lid screwing on - just something to keep it a 'demonstration kit' to avoid all the certification guff. The first dozen prototypes could be Veroboard based. Made properly, this manufacturing method can still make products that outlive the owner.

I've seen too many development projects fail or get bogged down because they try to do too much. This way something might actually get done.
Click to expand...


All sounds very sensible, achievable and down to earth :)

I have a LCD on ice, I will post the results shortly
 
I've got my TL061 op-amps, but still no luck. They don't drown the negative voltage output from the display like the TL081 did, but I can't even get them to work in voltage follower mode! I can do limited contrast control by connecting the top of my pot to the DAC output and not using a op-amp at all, but I don't think it is enough range. Someone else may have to look at this. Analog is not my thing.
 
You must log in or register to reply here.

Share this page

Top