GPIO:  HD44780 chip from old Fax Machine.

Check out this tutorial from Adafruit:

http://learn.adafruit.com/drive-a-16x2-lcd-directly-with-a-raspberry-pi

Adding a LCD to any project immediately kicks it up a notch. This tutorial explains how to connect a inexpensive HDD44780 compatible LCD to the raspberry pi using 6 GPIOs. While there are other ways to connect using I2C or the UART this is the most direct method that get right down to the bare metal.

Hey Drew,

The problem I am having is that the Screen is already mounted on a PCB.  I don't know which pins to send to Raspberry Pi.  See the image above.

Also, it connected to the rest of the fax machine via a 16 pin flat ribbon cable (which the socket is acessable from the rear - not pictured). 

Can I use any of these pins to acess the display?  And is the power for the display the same as what it was connected to on the fax machine (the two pins on the top left)?

Fergus,

Multimeter tracing time!    Bung the multimeter onto the continuity trace setting (often the diode test setting, normally has a speaker or sound type icon of vertical bars increasing in length left to right).

Next find the power traces on the board, put one probe on a pin at the connector and probe a corresponding trace. The power traces are likely to be the thickest ones on the PCB and often the earth/ground one is the one going around the edge of a PCB (devices run off low dc voltages not AC boards which can have different isolated grounds), not which pin number corresponds with positive and negative - tip - find a large capacitor that electrolytic capacitor and use the - pin to see if it traces back to your connector, if it does you have probably found the ground pin, likewise for the positive pin of the capacitor. SMD caps will have a small + or a white/orange band denoting the+ side.

Once you have found the power terminals, check to see if any of the other connector terminals also return most likely to ground.

Hitachi do have a datasheet available for the HD44780 and with no voltage applied to the board you could use that to trace any continuity between the connector and inputs of the actual chip.

bodgy wrote:

 

Fergus,

 

Multimeter tracing time!    Bung the multimeter onto the continuity trace setting (often the diode test setting, normally has a speaker or sound type icon of vertical bars increasing in length left to right).

 

Next find the power traces on the board, put one probe on a pin at the connector and probe a corresponding trace. The power traces are likely to be the thickest ones on the PCB and often the earth/ground one is the one going around the edge of a PCB (devices run off low dc voltages not AC boards which can have different isolated grounds), not which pin number corresponds with positive and negative - tip - find a large capacitor that electrolytic capacitor and use the - pin to see if it traces back to your connector, if it does you have probably found the ground pin, likewise for the positive pin of the capacitor. SMD caps will have a small + or a white/orange band denoting the+ side.

 

Once you have found the power terminals, check to see if any of the other connector terminals also return most likely to ground.

 

Hitachi do have a datasheet available for the HD44780 and with no voltage applied to the board you could use that to trace any continuity between the connector and inputs of the actual chip.

I have worked out that the 16 pin connect on the back of the board is for the display interface.  The two connections on the top left is a ground connector.  All I have to is desolder the 16 pins of the connect on the motherboard and away I go.  (any advice for desoldering 16 pins?)

Fergus, be interested to hear if you get any advice for desoldering, I am completely useless at that. I did pose a question in the soldering forum and it was suggested that plenty of flux is the way to go, then use de-soldering braid. Haven't had a chance to try yet, but trying to use a solder sucker is about the only way I have managed even a half decent job. Watching youtube videos and it seems so easy !

Steve

I would advise against attempting to desolder the connector unless you can lay your hands on a desoldering gun. Manual pneumatic ones can work, and solder braid desoldering is very likely to fail and cause you frustration and possibly a damaged board. I should also point out that desoldering using a desoldering gun can also ruin boards if not carefully used.

I would really recommend the tracing route for the moment.

It ought to be possible so long as you have access to the HD44780 chip pins which you do. You have already discovered the power rails, so there are just 8 data pins, 3 status pins plus possibly the contrast control pins, these probably do come out to the connector but might not and you may possibly have the backlight control coming out to the connector.

Page 4 of the http://www.sparkfun.com/datasheets/LCD/HD44780.pdf datasheet shows that DB0:7 are on pins 39...46 of the 44780 chip and it also shows that RS, RW, E are on pins 36....38 . Once you've found and mapped these on a piece of paper that leaves just 3 pins unaccounted for on your 16 pin connector. If two of these remaining pins trace to one of the power planes the third one is likely to be Vo which is the contrast control. If they don't then they could be the backlight switch.

If RW pin doesn't seem to be represented on the connector check to see if there is a direct connection between that pin on the 44780 chip and ground.

As for the desoldering route, at the very least you will need some solder - apply solder to the pin and make sure it wets the surrounding pad well - to desolder anything properly solder is needed because you are relying on the flux which acts as a medium to help the solder to melt. You could use flux, I have a passion for gel flux, but this is very very expensive in hobby terms so the solder method is the better way to go.

Once you've rewetted all the connector joints, you can then either apply your desoldering gun gently to the pin and pad, make sure the solder has become liquid and gently attempt to move the pin side to side so that the melted solder makes contact with the track that will be on the other side of the board, then press the suck button, you may need to resolder and perform this more than once especially on the power plane. If none of the pins become loose the next step would be to reapply solder and then apply your soldering iron and your desodlering gun to the joint and then suck. Do this only on pins with thick traces otherwise the pad or track may well come away from the board and you're then in trouble.

Using a hand operated cold solder sucker you need to  apply your sodlering iron to the pin on the PCB (after wetting it in solder), apply the teflon tip of the sucker and release the spring catch.

Ideally for connector boards a PCB warmer gets the best results, but it it pointless getting one of those as they are not cheap and you aren't going to use it that often. Even where I work, they refuse to buy one for the workshop.

Colin

Thanks man,

i will look into what you have posted above.

Hi Colin,

I've always had problems de-soldering stuff, so to be clear (and not to hijack this thread) you use the flux from the solder core to do the job of separate flux, or is there another mechanism involved as well?

Steve

Steve,

That is correct, the job of flux in soldering is to help the solder flow (flux is based on the Latin word 'to flow'), transfer heat between the molten solder and a non melted metal object.  Solder for electronics (apart from solder paste, where it is most often added at time of use and is specific to the type of solder being used and the end purpose of the assembled product) contain flux to help aid the flow of solder and to transfer the heat from the iron tip efficiently. 

As the solder joint ages it is subjected to heat from the environment including heat that might be transferred by the component it is attached to, this causes the flux to dry out among other things. After a period of time if a soldering iron is applied straight to the joint, it will be noticed that the solder either doesn't melt around the joint or it takes a very long time or a much higher temperature is required to get to the melting point. Some of this is due to oxidisation and a film that gets layed down on the outside of the solder 'blob'. Application of flux to this joint will help burn away any oxidisation and also work as a heat transfer mechanism that enables the old solder to be able to flow again.

For electronic purposes there are flux pens, gel flux and of course the flux in solder itself.  Using solder to reinvigorate an aged solder joint often works better than just the application of flux alone on through hole components as often the solder in between the two sides of the board has become closer to concrete in texture plus there will be a metal copper sleeve inside the hole connecting the two or more layers of a PCB, this copper sleeve also has to heat to allow the solder to flow freely enough to be removed and allow the pin to become free.

So always refresh a joint with solder before attempting to desolder, it will be much easier.

A solder joint that has just been made should in most case be able to be desoldered with no additional solder naturally.  Cleaning the board thoroughly beforehand can also help, as that can remove the oxidisation layer across the board.

Colin

Hello Colin, Thank you for that, it makes perfect sense now you have explained it, but have never thought of it that way before.

I was trying to repair a Netbook power connector at the weekend and I could not get the original one off - I think another try soon using this method may prove a lot more productive.

Thanks

Steve

Don't use one of those manual desoldering pumps. When the plunger shoots back, as a reaction , the tip will move as well. You will just wipe small tracks of the pcb with it. They are good for larger solderings with wide tracks. (4 - 5 mm wide.) I use desoldering wick for the fine work. They have the disadvantage that the cupper of the desoldering also takes heat away from the solder point and the tip of your iron. So, don't use one of those small 15W irons. You will notice they have problems to melt the solder when the desoldering wick is attached to it.

Also note that some multi layer pcb's have large cupper areas in their inner layers that are used as power supply layers. If a component is attached to such a layer, you need a real powerfull soldering iron to desolder it, as the cupper layer inside the pcb takes the heat away. This often is a problem when you replace electrolitic capacitors on pc boards, as those are connected to the supply and ground layers to suppress the ripple on those. I sometimes even use a 150W plumber soldering iron for such a job. Only problem is they don't come with a fine tip to do precision work. If the heat transfer isn't good, and the solder isn't melting, don't keep trying. Stop your activity and give everything time to cool down. Add some extra fresh solder and try again afterwards.

So far my 5 cent on 27 year of desoldering experience.

Don't use one of those manual desoldering pumps. When the plunger shoots back, as a reaction , the tip will move as well. You will just wipe small tracks of the pcb with it. They are good for larger solderings with wide tracks. (4 - 5 mm wide.) I use desoldering wick for the fine work. They have the disadvantage that the cupper of the desoldering also takes heat away from the solder point and the tip of your iron. So, don't use one of those small 15W irons. You will notice they have problems to melt the solder when the desoldering wick is attached to it.

Also note that some multi layer pcb's have large cupper areas in their inner layers that are used as power supply layers. If a component is attached to such a layer, you need a real powerfull soldering iron to desolder it, as the cupper layer inside the pcb takes the heat away. This often is a problem when you replace electrolitic capacitors on pc boards, as those are connected to the supply and ground layers to suppress the ripple on those. I sometimes even use a 150W plumber soldering iron for such a job. Only problem is they don't come with a fine tip to do precision work. If the heat transfer isn't good, and the solder isn't melting, don't keep trying. Stop your activity and give everything time to cool down. Add some extra fresh solder and try again afterwards.

So far my 5 cent on 27 year of desoldering experience.

Could I make my own interface off the HD44800U Chip?  Say: Desolder the certain pins of the chip that I know and run a wire straight into the chip? from the Raspberry Pi?

Is that possible?  I have scene it all over the internet but is there a special techque required?

This is what I know:

  Diagram is upside-down.

I will be using http://learn.adafruit.com/drive-a-16x2-lcd-directly-with-a-raspberry-pi/wiring as the instuctions to set-up the screen.  See if a direct wire will work with the set-up in the link.

Thanks in advanced.

This is not a route I would take, but is possible, bear in mind not only is there a likelihood of breaking the pins, overheating the chip, plus trying to solder wires to the pins won't be easy if you are not experienced wit soldering nor have a fine tip and the correct size solder 0.6mm or less would be ideal.

Did you not get anywhere in tracing the pins to your connector?

The easiest thing by far would be to purchase a surplus 2 * 16 display if you cannot afford a new one.  Rockby or Oately Electronics  in Victoria and NSW respectively woul be a place to look for surplus.

If you live in SE Qld, I might be able to personally help you with your problem.  If you are at school or university, why not see if someone there could help you trace the pins?

Colin

I tried tracing but the pins on the connector don't seem to be in the correct order.  I don't know If you will be able to view the images but here a link:

https://www.dropbox.com/sh/6hj5g7gove4gj1v/KNWwcrS9Ns?m

i have uploaded my tracing progress here ^.

The pins on the connector may well not be in the same order as the pins on either the 44870 chip orthe LCD itself.

It is possible that the connector doesn't have direct connections with the LCD chips, is there another chip on the board that looks like it might be a microcontroller?

Have you searched te internet to see if there are any schematics for the fax machine posted anywhere?

As an example, I've attached a schematic of a Xerox printer graphic display board that I traced out, you'll see that on the connector there are more than just the connections to the graphics chips (not shown on this schematic).

Attachments:

There is a chip that one of the connectors goes to that is called: HC138.  Otherwise, the pins seem to go into the HD44780.

I will look for a schematic later. (or if you can find one - Panasonic Panafax UF-S2)

bodgy wrote:

 

This is not a route I would take, but is possible, bear in mind not only is there a likelihood of breaking the pins, overheating the chip, plus trying to solder wires to the pins won't be easy if you are not experienced wit soldering nor have a fine tip and the correct size solder 0.6mm or less would be ideal.

 

Did you not get anywhere in tracing the pins to your connector?

 

The easiest thing by far would be to purchase a surplus 2 * 16 display if you cannot afford a new one.  Rockby or Oately Electronics  in Victoria and NSW respectively woul be a place to look for surplus.

 

If you live in SE Qld, I might be able to personally help you with your problem.  If you are at school or university, why not see if someone there could help you trace the pins?

 

 

Colin

I thought of another idea, what if I desolder the ribbon connector off the board and replace with headers.  That way, I am not limited by the ribbon connector and can then plug-in the certain pins that need to be traced.

Also, on the personal help request: I live in NSW so that might be a bit of a strech.  That's for the offer anyway.

Here is the tracing so far:

Attachments:

That is certainly an option.

If the HC138 has 16 pins it is a 3 -> 8 line demultiplexor. In other words it has three binary data lines as input and eight data lines out.  These are used when the designer has perhaps run out of IOs from a specific chip or to save space and components.

As the 3 input pins are brought high (to the +v rail) in sequence they switch on an output. You can see that a high on an input actually takes an output low.

input       output

CBA   76543210

000 = 11111110

001 = 11111101

010 = 11111011

011 = 11110111

100 = 11101111

101 = 11011111

111 = 01111111

Assuming the 3 enable pins G1 /G2A /G2B =  100 (/x = pin works in an inverted fashion, so these two pins are taken low for an internal high or high for an internal low)

What does this mean for me and the conector?