Hello ZOR
There are a number of ways to do this. For starters, are you absolutely certain that the switch contacts actually "short" to ground? In fair number cases, a keyboard (particularly when there is many keys) is arranged in a matrix and doesn't actually go to ground. This is done to reduce the number of lines (pins) used on a micro. Also, if these are carbon based contacts, these usually don't actually "short", but instead offer maybe 100 Ohms or more of resistance. This may (or may not) be of importance when it comes to the actual circuit design. As a general rule, I wouldn't actually short out your switches. I would tend to suggest using a 100 Ohm resistor in the circuit. You can measure the actual resistance (often provided by a carbon pad and carbon tracks). With no power on the camera, find contact locations either side of the switch and using your DMM measure the resistance when the key is pressed. This will give you an indication of the switch "resistance".
You are on the right track, I definitely wouldn't use relays!!! In my opinion they are "too mechanical" and would tend to create more contact bounce when compared to your average key press. They are also a bulky item to use when compared to discrete or integrated components. The coil is also an inefficient use of power. Basically it's overkill for the application!!!
In my opinion you have two discrete options and one (or maybe more?) integrated options. You could (as you suggest) use regular transistors (or to reduce your component count, digital transistors). If the keys in question do actually "short" to ground, then a simple circuit could be used to interface the PICAXE with the camera. You might even be able to use the PICAXE direct (with a 100 ohm resistor), if the camera and PICAXE circuit
share the same ground. If the keys do not "short" to ground then I wouldn't use the transistor approach as it might prove difficult to correctly bias on (and maybe off) the transistors to provide correct operation.
The second discrete option is to use FETs (signal FETs would be more than sufficient). They put basically no load on the PICAXE and if you chose one with a low switch on (gate) voltage then this might be a good option to drive directly from the PICAXE. If they are used as a "switch" then I would still suggest putting a 100 Ohm in series with the Drain/Source connection. Again, problems might arise with correct FET operation if the camera switches don't operate from the switch to ground. I would have to get more information on the exact keyboard configuration and give more thought to if the FETs would switch correctly within a matrix.
The integrated option is probably the least complicated. There might be more than one way to do this. This would work particularly well if the PICAXE and camera
DON'T use the same (common) ground. If you use this idea and the PICAXE doesn't share the same ground as the camera, then the integrated component/s MUST have the same ground (and supply) as the camera to ensure correct operation. I've used a 4066 in the past to do this and it works very well. It's a quad analogue switch. It will give you isolation from the PICAXE circuit and close the switch contacts. You can also include the 100 Ohm resistor (or whatever the carbon switch reading is from your measurements) in series to ensure this doesn't change the circuit for the camera. The series resistor may (or may not) matter but it would have be factored into the original engineering.
http://www.ti.com/lit/ds/symlink/cd4066b.pdf
There are probably more ways to do this. If you need more help just ask but please do the following and post back.
1. Measure the carbon switch resistance.
2. Measure if the switches (both) actually short (allowing for carbon resistance) to ground.
3. If it isn't too complicated, try to "reverse engineer" (draw the circuit) of the keyboard section.
This additional information would be very helpful and help me/others to assist further.
Regards,
Brad