Determine Pins Outs / Power Requirment of a Head Mounted Display

[zacherynuk]

Hi folks,

I wonder if any of you would be willing to give me some guidance with this little predicament:
(You can skip to the bottom few paragraphs)

HMD Project

Basically, I have a heap of pins, which I am hoping will be 2 X RGB + 2 X v-sync & 2 X h-sync – the big missing piece of the puzzle is determining what the likely power pins are… and what the power requirement is.

Is there any way I could, at least, systematically determine the most likely power-pins?

Thanks for your time, it is very much appreciated!

Zach

 

[MrAl]

Hello there,


Offhand i can think of one way to find the power pins...You have to open up the main assembly so you can get to the circuit board though, that's the requirement.
Once you can get to the main circuit board, you can look for wires that lead to traces on the board that lead to many many components rather than just a few. These would be the power supply pins most likely because in most circuits the power and ground are common to many many connection points while things like color lines and sync lines will go to pads that only have two or three components connected to the same pad. This should get you started anyway. Sometimes the board is even marked with a plus sign or a minus sign to indicate polarity, and possibly even "GND" to indicate ground. Sometimes the ground trace runs around the whole outside of the board.
The main thing is to get to the circuit board so you can look for more clues.

To determine the power voltage requirement you can also check the board for markings, but if you cant find any and you get not get more documentation then all you can do is get a variable power supply and run the circuit board up slowly, starting from say 3v and working up until something starts to work right. 3v, to 3.5v, to 4v, to 4.5v, etc. You may have to check a lot of other pins too for each voltage step though to look for activity.


If you can find someone else who has one of these things too maybe they can be of more help too.

 

[cowana]

If you can find datasheets for any parts on the board, then that'll help you figure out what power lines are what voltage.

Andrew

 

[cr0sh]

Have you tried contacting the company? They were acquired by Rockwell Collins:

Kaiser Electro-Optics Develops Infrared Optical Elements for Transmitters

MrAl is spot-on when it comes to trying to find power and ground; if you can see any parts with polarity markings, or can find datasheets for any of the ICs on-board, those will be great clues for finding and backtracing signals. In the meantime, I would first contact Rockwell Collins (perhaps by phone as well as email), and find out if you can get a pinout and/or specification sheet (or possibly a manual). Heck, they may even have a complete interface box they might give you for a few bucks (or maybe free, just pay shipping).

Good luck - you have a nice HMD there, don't destroy it!

Forget everything I wrote - I just read your blog entry and noticed you did contact them - bummer, that! Good luck, though, still!

 

[cr0sh]

BTW - If you become frustrated with this project, and aren't able to proceed - I'm willing to purchase these HMDs from you; send me PM if things go that way. I probably won't have much more luck than you, but I am a collector of 90's vintage VR equipment. If I find something out there about hookup, though, I'll let you know!

 

[zacherynuk]

Hi, Thanks for the replies - there are two wires, which when they connect to the ribbon PCB connector each appear to connecto to miltiple PCB rails - around 5 each. These were my first guess at being the power. One is white and connected to the first pins the other is black, connected to the latter end of pins.
The PCB is multi layer, so following tracks is near impossible, but, the black wire, would apear to lead almost directly into a component labeled on the PCB as JP6 - which I imagine is a capacitor - it's tiny. Would this be indicative of +ve or -ve, do you think ? (My brain says that out of two wires, black and white, black would be +ve)

I'll see if I can get better photo's of the PCB...

Zach

 

[zacherynuk]

Hi,

I have added a couple of new close up pictures to the end of the post, if this helps anybody.

Also, in this picture, the reverse of the board, there appears to be power components at the bottom of the picture; anybody recognise anything ?

http://www.wastedspace.co.uk/cms/wp-content/uploads/2011/01/IMG00156-20101231-1730.jpg

Thanks again,

Zach

 

[MrAl]

Hello again,


The first clue i see is one of those AD8004 chips. These are amplifier chips, and have positive power at pin 4 and negative power at pin 11.
What you'll have to do is trace these pins to the connector, unless they use an on board regulator, but i would think that pin 11 leads to the negative supply lead of the connector at least. So start with that pin, pin 11, and try to trace it back to the connector.
You'll have to find some way to trace this out, perhaps with a low voltage ohm meter (1.5v or less). I just dont see any other way unless you can find markings on the board or can get better documentation.

Anybody else have any other ideas?

 

[cr0sh]

Hello again,


The first clue i see is one of those AD8004 chips.

That's what I was going to suggest; there's also those devices down at the bottom left of the pic labeled "68 16A" and "33 20V" - caps, maybe? Anyhow, it looks like at the end closest to the viewer are "+" symbols (and they are marked "C##" indicating a capacitor?), so I would be expecting that might lead to a positive lead (once again with the caveat about regulators - which is very likely in this case - but if they are caps as a part of a regulator circuit, they might be filters across the positive/negative rails); I would stick with the AD8004 chips first.

Finding traces is a difficult task on something like this, but something I learned a while back was that if you can at least see the traces, a good way to trace them was to lay the board flat on a scanner, and scan that sucker at a very high DPI. Then use a paint program to flood fill the traces, and/or do other manipulations to it to bring out the traces (you could even do a scan of the second side, then flip/reverse and paste it as a second layer in something like photoshop/gimp - and do an "xray" view by changing the alpha transparency of the scan - which might help in tracing vias, at least).

Here's another issue: That plug you posted may actually be carrying numerous power lines; the regulation (to a certain level, at least) might be being done in whatever control box or board the HMD was to be connected to. Just something to keep in mind. I also know that some of those high-end HMDs used a completely different scan rate and such from standard VGA, and the control box translated those signals. Finally, something to verify is that there isn't any built-in 3DOF tracker on the HMD that is sending signals back. Those were usually connected to the computer driving the HMD via serial, but the conversion to RS-232 might have been in that box, with TTL serial being used on the HMD. Everything was done to make that HMD as lightweight as possible, and that meant offloading a bunch of stuff to the control box the HMD was plugged into.

Hope this helps some! Good luck!

 

[cowana]

The groundplane is usually 0V - usually the copper around mounting holes is also at this level. Therefore, check to see if the tin pads around the three mounting holes are connected if so, there's a fair chance that they are 0v.

 

[MrAl]

Hi again,

Some good points brought up here.

I also fear that there are too many wires to figure them all out. Once power and ground is found and assuming it is the only power and ground and we power up we could look at all the signal wires with a scope and see what they are doing, but unfortunately depending on the complexity of the other circuitry external to the board it could be very very difficult if not impossible to figure out what the input lines are.

 

[zacherynuk]

Thanks all for the ideas, I am aware this if this is to ever power up and show even raster I will need hard documentation, or extreme hard core luck!

I am hoping, that, as with similar headsets of the time, that the 'control box' is simply no more than a signal distributor, allowing for external monitors, single input feeds and, in some cases, video conversion from svideo / pal ntsc etc.

Sugnal wise, I expect the headset itself to conform, to the specs in this document: https://www.mindflux.com.au/products/keo/Proview%2030.pdf

So far, I have confirmed that pin 11 of the AD8004 chip connects to ground, which is the same ground as the metal around the mounting holes and the other wires marked as ground on the picture. This does not directly correlate back to the white wire, but there is a path there of fairly high resistance.

If I can get a blank raster, or at least the condensers for the backlights to fireup, I think the other wires should fall into place quite easily, depending on any built in sanity checking, a single colour and a single sync should be enough to get some feedback

The black wire, when tested on pin1 of the AD8004 chip and +ve parts of other components seems to 'pip' that is the resistance quickly builds up from very low=> to very high - within a fraction of a second.

I'd love to get my mitts on the control unit for an afternoon - there are still some left for sale at DOTMED.COM (Search for vista medical, or stereosite) but these seem to be the final hard core holder outers, who won't part with the gubbins for less than the shirt off your back, even though it's pretty worthless to anybody but hobbiests.

I think I'm gonna start pumpin some juice in and see what happens.

 

[cowana]

'The black wire, when tested on pin1 of the AD8004 chip and +ve parts of other components seems to 'pip' that is the resistance quickly builds up from very low=> to very high - within a fraction of a second.'

That sounds like a connection to ground via a series capacitor - often a decoupling capacitor.

 

[zacherynuk]

'The black wire, when tested on pin1 of the AD8004 chip and +ve parts of other components seems to 'pip' that is the resistance quickly builds up from very low=> to very high - within a fraction of a second.'

That sounds like a connection to ground via a series capacitor - often a decoupling capacitor.

Hi, Would you expect this to be on the +ve rail ? - I imagine it's for current smoothing and interference reduction.

Would you expect the Power ground to be common with the signal ground ? - Especially weird I thought, since the Audio ground is seperate (though this is classed as passthrough only and could be for interference purposes too)

Zach

 

[cowana]

I would have said that that wire was one of the power rails, but because it's black, I'm much less certain.

It could possibly be the audio ground - as this might well be connected to ground via a capacitor to remove AC from it. Hopefully someone else has an opinion!

 

[cr0sh]

You might also try contacting the webmasters and/or others at these sites (especially the first one):

V-Rtifacts

A VR Geek's Blog
Discussions - VR Geeks | Google Groups

These two are run by the same guy (who lives in Paris); one's his blog on VR stuff, the other is a mailing list/group for similar interests.

 

[zacherynuk]

Hi all,

Just a quick 'Thank you' for all your help.

Sadly I have managed to blow one of the headsets, which is a bummer... but does leave me with an interesting new dilema. HMD Project - how to drive the LCD's directly.

 

[zacherynuk]

Thanks Cr0sh - I have already bugged Tone at Vritafacts a fair amount! - But I may try him again...

You might also try contacting the webmasters and/or others at these sites (especially the first one):

V-Rtifacts

A VR Geek's Blog
Discussions - VR Geeks | Google Groups

These two are run by the same guy (who lives in Paris); one's his blog on VR stuff, the other is a mailing list/group for similar interests.