Razer Hydra mods etc...

[OzOnE2k10]

Firstly, congrats to @tbowren for getting good positional tracking working this early on.
Looks like it will be must-have addition to the Rift.

I thought I'd make a new thread about general mods to the Razer Hydra.
(I didn't want to hijack @tbowren's thread. This one also keeps things separate from the "Rift mods" thread.)

The following is just me throwing some ideas around about modding the Hydra so more than the standard 2 controllers could be used.
Although, earlier @geekmaster mentioned that the Hydra API appears to support 4 controllers?

I've seen on other sites where the Hydra has been modded to fit the controller coils in a very small case, so I'll look into this later on too.

Here's a link to Jan's page, for some great background info on how the Hydra works...
http://janoc.rd-h.com/archives/212

I found some Hydra board pics as well. Where, you might ask?, well on this here forum of course...
http://www.mtbs3d.com/phpbb/viewtopic.p ... 528#p83722

OK, so the MCU in the base is just a Philips LPC thingy (basic ARM chip), and is doing the USB interface duty etc.
It also has the nice looking DSP from Analog Devices, so yep, actually more complex than I first gave it credit for. lol

You can also see some of the traces between the DSP and MCU near the middle of the board, so it likely uses SPI to transfer the data.

There are separate coil driver chips (above the yellow caps), so hopefully they are also responsible for setting the "transmit" frequency, and can be modified.
Those 100V caps suggest that it drives the coils with a fair amount of power, so it may be a tad critical about the exact resonant freq?

So far, we'll assume that only 2 controllers can be used with the stock base station, but I'm now eager to find out if it even uses all the pins on the controller connector?

I'll have to wait until my Hydra arrives to do some proper tests.
It's looking feasible to get more controllers working though, and there are a few different methods to try.

OzOnE.

[MSat]

I believe I read somewhere on Sixense's site that the Hydra only supports two controllers, it was their wireless prototype hardware that supported 4.

I had wondered if the electromagnet and the controllers could be extended myself. If there were to be any problems, I think it might stem from the digital electronics in the controllers, but you can do without them if you use VRPN instead of the Hydra SDK. Another issue is that longer wires for the controller might pick up RF signals, but could probably be mitigated with shielded cables.

The caps are probably needed to provide sufficient current for each coil's fast rise time based on Jan's scope readings, considering that it's powered via USB. Since it appears to be pulsed-DC, I also don't think they make up an LC network.

[OzOnE2k10]

The caps are probably needed to provide sufficient current for each coil's fast rise time based on Jan's scope readings

Ahh, that makes more sense. I forgot that it's USB powered.
I suppose there is sufficient back-EMF generated to warrant the 100V caps?

I was thinking more that the smaller caps / resistors around the drivers might be for setting a resonant circuit.
Like you say though, it's more likely that the DSP just drives these directly with digital pulses.

EDIT: When you say "pulse driven", do you mean just the 2ms pulses, or do you think the ~13KHz freq of each pulse is driven directly by the DSP as well?
Is it safe to assume that the controller coils are tuned to the intermediate freq, and not just responding to the raw 2ms pulses?

(It's starting to appear more likely that the DSP generates the KHz range pulses directly, in which case, the only way to change the freq is to change the DSP's clock crystal?)

So, the coils are probably just MOSFET driven?

OK, so it sounds like it does only support 2 controllers as standard.
What do you think might be a good strategy for modding the Hydra so more controllers could be used?

I might try underclocking the DSP when mine arrives to see if the pulse freq changes too.
Hopefully the serial link to the MCU is SPI (and so might still work at a lower clock, as it's synchronous).

If I get impatient for the Hydra to arrive, I'll ask Dycus to send me the super high-res photos of the board.
If we could just identify the chips, it will tell us a huge amount.

Anyone fancy unscrewing their base unit to see what the main chips are?
I'm guessing there are a few ADC chips on the left, I just can't read what the coil drivers are.

[Krenzo]

There are more images here: http://www.mtbs3d.com/phpBB/viewtopic.php?f=138&t=15431

[OzOnE2k10]

Cool, thanks.

I don't know why, but I never seem to find these threads when I search for them?

[jf031]

Cool, thanks.

I don't know why, but I never seem to find these threads when I search for them?

Because if you use search while in this forum, it defaults to only searching this forum. The linked topic, for instance, is in the VR/AR Research & Development forum.

[OzOnE2k10]

Oh dear. Yep, that'll be it. lol

Right, I've noticed something interesting on those photos...

There is an 8-to-1 mux chip (HEF4051) and it's common input comes from U11.
The first three outputs go to the coil driver chips.

The mux allows the DSP to switch the signal from U11 to each of the coil drivers.

So, I'd love to know what U11 is. It's probably an SPI DAC though?
If it's not a DAC, then it could be a simple oscillator which would be easy to mod.

I can't read the codes on the coil drivers either (U13 & U14).
It looks something like "A1Fu052", but I can't find any datasheets, even using every combination of the code / partial code.

I think the IC package is a QFN or VQFN. No joy hunting on the Web though.

[MSat]

Hmm.. So I found a 125kHz grounded coil LC oscillator schematic @ http://discovercircuits.com/DJ-Circuits ... erosc3.htm and even though it's powered by a 5V source, the leg between the coil and the cap is 100V. Even if the Hydra doesn't use a sinusoidal oscillator circuit, I suppose it's close enough? TBH, inductors are still a bit of voodoo to me

The DSP definitely controls the coils as each one is operated in succession, so they can't be connected to a free-running oscillator circuit. I tore my Hydra open to get a better look, but didn't get much out of it. I couldn't identify U13 and U14 (ICs closest to the caps) based on their markings, but I'd guess they're some sort of OP amps/buffers/inverters. Those connect to the next IC which is an analog MUX and then to another unidentified IC which I'm guessing is a buffer or something (but possibly a DAC?). Now that I think of it, and without trying to trace it out (total PITA on this board), perhaps that last IC is actually a pulse generator, and the DSP only controls the MUX? If so, that means the coil frequency is set by that chip, and not the DSP. It would seem a bit silly, but the entire circuit is a bit strange to me, so who knows?


Without major hacking or trickery, I don't think there's any way to use multiple controllers or even two Hydras in close proximity because at the very least their coil sequencing would need to be synchronised and that seems well out of the realm of possibility.

[MSat]

Oh dear. Yep, that'll be it. lol

Right, I've noticed something interesting on those photos...

There is an 8-to-1 mux chip (HEF4051) and it's common input comes from U11.
The first three outputs go to the coil driver chips.

The mux allows the DSP to switch the signal from U11 to each of the coil drivers.

So, I'd love to know what U11 is. It's probably an SPI DAC though?
If it's not a DAC, then it could be a simple oscillator which would be easy to mod.

I can't read the codes on the coil drivers either (U13 & U14).
It looks something like "A1Fu052", but I can't find any datasheets, even using every combination of the code / partial code.

I think the IC package is a QFN or VQFN. No joy hunting on the Web though.

Haha! Looks like we're thinking along the same lines!

I'm really starting to think that U11 is an oscillator. But changing the frequency won't do you any good (probably break it) for a couple different reasons.

[Krenzo]

MSat, did you ever determine the effect of the Hydra's magnetic field on a 3 axis magnetometer? I wonder how the Rift is going to behave with the Hydra when Oculus finally activates the Rift's magnetometer.

[OzOnE2k10]

perhaps that last IC is actually a pulse generator, and the DSP only controls the MUX?

Yep, that's what I'm thinking...

U11 could be a DAC, or opamp, but that would imply that the DSP is generating the raw ~13KHz waveform?
I don't know if it's necessary for the tracking to work in the KHz range during the pulses, or only needs to know the relative / average power of each 2ms pulse?

If it only needs the relative powers, then it's still possible that U11 is a simple oscillator / pulse gen.
Changing the frequency might actually work, but it will probably need some notch filters fitted to all the controllers.

I can see that the DSP definitely switches the output of U11 between the three coil drivers.

That LC circuit can generate a fairly high voltage due to the back-EMF.
This is basically where the field in the coil collapses when the current is switched OFF (IIRC).
The voltage can be quite high (although not always the current), and this is why diodes are put in parallel with various inductive devices such as relays and motors to protect the driver circuit from the back-EMF.

Could you possibly post your best guess at the part numbers on U13 / U14? I can't make them out in the photos.
More importantly, are there any readable markings on U11?

[PatimPatam]

OK, since we're at it, here's another related thread:
Razer Hydra teardown

[MSat]

MSat, did you ever determine the effect of the Hydra's magnetic field on a 3 axis magnetometer? I wonder how the Rift is going to behave with the Hydra when Oculus finally activates the Rift's magnetometer.

Just tried it using an app on my phone that displays the magnetometer's reading as a 3D vector + field strength in uT. It shows some jitter when I place the phone less than 4 inches from the coils, but even then, it's only several uT (compared to ambient field being around 60uT). So I don't think it would cause a problem even if you put the base on your head. A really sensitive magnetometer would probably be able to pick it up even at moderate distances, but it should be easy enough to filter out given the values are so low. I think it'll be fine.

[OzOnE2k10]

OK, since we're at it, here's another related thread:
Razer Hydra teardown

Maybe I should have just merged with that thread?

I've been looking over the photos again and I'm 95% sure that U11 has no direct connection to the DSP (only power rails).
In other words, it's unlikely to be a DAC, so probably just a pulse gen / opamp oscillator.

Quite easy to test, just need to put a 'scope probe on pin 3 of U12, and you should see the raw intermediate freq.

The theory is that the DSP only needs to know the relative / average power of each pulse.
It still needs the intermediate freq (IF) though, as the controller coils act as a tuned circuit.
(also, there needs to be a constantly changing signal to induce current in the receiving coils).

No idea how long until my Hydra arrives so I can test this stuff?
Oh well, I can still play with the new i5 machine which I just built today in preparation for the Rift / Hydra.

btw, pretty sure U7 is just an EEPROM (ST 25 series) which stores the Blackfin's boot code.

[MSat]

perhaps that last IC is actually a pulse generator, and the DSP only controls the MUX?

Yep, that's what I'm thinking...

U11 could be a DAC, or opamp, but that would imply that the DSP is generating the raw ~13KHz waveform?
I don't know if it's necessary for the tracking to work in the KHz range during the pulses, or only needs to know the relative / average power of each 2ms pulse?

If it only needs the relative powers, then it's still possible that U11 is a simple oscillator / pulse gen.
Changing the frequency might actually work, but it will probably need some notch filters fitted to all the controllers.

I can see that the DSP definitely switches the output of U11 between the three coil drivers.

That LC circuit can generate a fairly high voltage due to the back-EMF.
This is basically where the field in the coil collapses when the current is switched OFF (IIRC).
The voltage can be quite high (although not always the current), and this is why diodes are put in parallel with various inductive devices such as relays and motors to protect the driver circuit from the back-EMF.

Could you possibly post your best guess at the part numbers on U13 / U14? I can't make them out in the photos.
More importantly, are there any readable markings on U11?

Keep in mind that the driving coils are inducing a current into the coils in the controller. Changing the frequency of the driving coils will have an effect on the current output of the receiving coils, as once the field stabilizes it starts to collapse. Regardless, I don't see how you would be able to exploit different frequencies so that you could use multiple devices. I think the only way to make it happen is if it were possible to synchronise two bases, then you could disable the coils on one of them. I don't know if Sixense added such a provision in the firmware, but I doubt it.

Yeah, I know that high voltages can be generated in a coil when the field collapses, so it could be the reason for the high voltage caps. I just can't say with certainty if that's the reason why, as solenoid circuits don't tend to have a capacitor in series, so why would a pulsed-DC? Whatever the case may be, I don't think there's much reason to concern ourselves with it (ok ok it *is* interesting )

U13 and 14 were A1F #052. I couldn't find anything appropriate doing a web search. Those are some miserably small ICs BTW .. I already closed the base back up so I can't say much about the other IC, but my search history shows "OCTI". I believe there may have been a number as well, but apparently I didn't search for it. I don't think it would have given me any useful results.

[Krenzo]

Does anyone know what material the base station's sphere core is made of?

[geekmaster]

This is the older wireless prototype, before Razer converted it to the wired Hydra design:

The Analog Devices Sharc gets a workout inside Sixense Entertainment’s 3D sensor system, which will be at the heart of a new class of game controllers. Nintendo’s Wii popularized 3D wireless controllers, but its accelerometers only provide relative motion information. Sixense’s patented TrueMotion game controller delivers absolute 3D position and orientation information (Fig. 1).



The importance of the ability to determine position and orientation cannot be stressed enough. This technology supports multiple sensors that will allow a game to determine which two players were next to each other and reaching for the same object. Alternatively, a player could use two sensors. For example, a golfing program could track the club as well as the player’s waist and hip position relative to the club. This would be impossible with an acceleration-based system.

Magnetics per form the magic. A small basestation has three small orthogonal coils that generate a weak magnetic field. Three smaller coils in each sensor detect this field. A Sharc DSP in each sensor analyzes input data from Analog Devices analog-to-digital converters attached to the coils. The initial gaming system reference platform handles up to four sensors, which is enough for two sensors for each of a pair of players.

The reference design sensors communicate using a 2.4-GHz wireless link. Each sensor delivers 60 readings/s, and they can operate for 20 hours with two AA batteries. A future move to Blackfin-based platforms will double the battery life. The sensor range is 6 to 8 feet from the USB-based base unit. Accuracy falls off outside this range, but this limitation, as well as the number of sensors and frequency, is due to cost constraints for the target market—gaming.

Higher resolution, longer range, and faster sensing are possible. Part costs for the gaming platform are on the order or $25, while a system with a 35-ft radius might be around $100. The system can handle any number of sensors because each performs its own analysis. The limiting factor is how the sensors are used. In terms of gaming, the radio connection is the limit, and different technologies can provide higher performance.

Magnetic sensing systems have a number of advantages over other sensing systems, including optical systems, because they don’t require lineof- sight operation with respect to the base unit. This enables a control unit to be moved around a player’s body without disrupting the sensor’s ability to report its position and orientation.

Magnetic materials such as iron and nickel can disrupt the sensors, but small amounts, such as a belt buckle, will not significantly affect accuracy. Likewise, even large objects that are at least a foot away will have no effect. The system requires no user calibration if these restrictions are met, though it can still operate in closer proximity if the system is calibrated. This would be significant if the technology is used in an environment such as an assembly line with large metal structures.

The magnetic field strength is 1/50th of the earth’s magnetic field. It is modulated so it is possible to use multiple base units to cover a larger area, though the initial gaming reference platform is designed for a single base unit. Sixense’s initial audience includes gamers and gaming companies, yet the technology has much more potential in applications ranging from robotics to home care. No other technology provides 3D positioning information at such a low cost.

BILL WONG

SIXENSE ENTERTAINMENT
http://www.sixense.com

But a lot of the reference design probably remains. Notice how it is powered by the "The Analog Devices Sharc" processor inside. Notice how it says it can be calibrated to work near large ferrous metal (iron or nickel) objects. I wonder if the Razer Hydra supports such calibration...

[geekmaster]

Does anyone know what material the base station's sphere core is made of?

Copper wire, enamel insulation, air core, plastic bobbin, acrylic outer shell, fancy green eye candy light guides, pixie dust and magic smoke? The most important ingredient is the magic smoke. If you let that out, it will cease to function...

[Krenzo]

Air core? No ferromagnetic material?

[MSat]

OK, since we're at it, here's another related thread:
Razer Hydra teardown

Maybe I should have just merged with that thread?

I've been looking over the photos again and I'm 95% sure that U11 has no direct connection to the DSP (only power rails).
In other words, it's unlikely to be a DAC, so probably just a pulse gen / opamp oscillator.

Quite easy to test, just need to put a 'scope probe on pin 3 of U12, and you should see the raw intermediate freq.

The theory is that the DSP only needs to know the relative / average power of each pulse.
It still needs the intermediate freq (IF) though, as the controller coils act as a tuned circuit.
(also, there needs to be a constantly changing signal to induce current in the receiving coils).

No idea how long until my Hydra arrives so I can test this stuff?
Oh well, I can still play with the new i5 machine which I just built today in preparation for the Rift / Hydra.

btw, pretty sure U7 is just an EEPROM (ST 25 series) which stores the Blackfin's boot code.

I don't think it's like an RF circuit at all. Lets say U11 is a free-running pulse generator @ 12kHz. The DSP will feed that signal directly to a particular coil for X amount of time by way of the mux. While a given coil is being pulsed, the receiving coils will be sampled - which will reveal the induced field strength in each coil, thus the distance from the base as well as the orientation relative to the pulsed base coil can be calculated. It then repeats this step, but pulses another coil instead. Once it has orientation and distance data relative to each of the 3 base coils, it can then calculate the absolute position and orientation of each controller. I wouldn't look at it as a tuned circuit in the RF sense.

I didn't look up U7 as I also assumed it was just an eeprom.

[geekmaster]

Air core? No ferromagnetic material?

That would concentrate the elecromagnetic field too close to the base unit. The point is for it to radiate out freely to the hand held controllers, not to be tightly coupled to a nearby device like a transformer coil...

A magnetic core is a piece of magnetic material with a high permeability used to confine and guide magnetic fields in electrical, electromechanical and magnetic devices such as electromagnets, transformers, electric motors, inductors and magnetic assemblies. It is made of ferromagnetic metal such as iron, or ferrimagnetic compounds such as ferrites. The high permeability, relative to the surrounding air, causes the magnetic field lines to be concentrated in the core material.

Air core... Absolutely!

And besides, I can SEE into the ball on my Hydra base, and I can see THROUGH it too. No opaque core material...

[OzOnE2k10]

Thanks for the info, geekmaster. That definitely hints that the modulation freq can be changed so more units can be used together.

I had a search for chips matching U11, but no joy. I think it's an MSOP package?
I searched for pulse gens and opamps etc. It could even be a 555 timer, but the pinout doesn't seem to match?

solenoid circuits don't tend to have a capacitor in series, so why would a pulsed-DC?

Many circuits that drive an inductive device at relatively high currents will often have some form of protection diode.
With smaller currents it may not be necessary, but for things like solenoids the driver IC / transistor usually has an internal protection diode.

Not sure about the caps in the Hydra yet, but your earlier theory sounds correct.

I'm working on the assumption that the coils in the controllers form a tuned circuit, so they will also need to be modified if the mod freq in the base unit is changed.
They may also need a bit of extra filtering to prevent cross-talk issues. This is the theory of course, I'm probably completely wrong. lol

I don't think it will be possible / easy to synchronize the bases either.

The "ghetto" method would be to share one crystal osc between both Blackfin's, then tie the /Reset signals together.

I've messed with SHARC DSP's before for audio, but it's been a long while.
(I still have two SHARC evaluation kits, and an old TI kit here. I don't even want to think about writing a tracking algo for one. hehe).

Does anyone know what material the base station's sphere core is made of?

The same stuff as this, I think?...
[youtube]http://www.youtube.com/watch?v=DgwprntqqmE[/youtube]

[Krenzo]

That would concentrate the elecromagnetic field too close to the base unit.

Ok, that makes sense.

[OzOnE2k10]

I don't think it's like an RF circuit at all. Lets say U11 is a free-running pulse generator @ 12kHz. The DSP will feed that signal directly to a particular coil for X amount of time by way of the mux. While a given coil is being pulsed, the receiving coils will be sampled - which will reveal the induced field strength in each coil, thus the distance from the base as well as the orientation relative to the pulsed base coil can be calculated. It then repeats this step, but pulses another coil instead. Once it has orientation and distance data relative to each of the 3 base coils, it can then calculate the absolute position and orientation of each controller. I wouldn't look at it as a tuned circuit in the RF sense.

I didn't look up U7 as I also assumed it was just an eeprom.

Agreed, I think this is exactly how it works too.
Maybe I shouldn't have called it an "IF" in RF terms as such, but it does share many RF principles.

I was wondering why the outputs from the controller opamps didn't appear to be filtered much though?
You'd think that if the DSP is only sampling the average power during each pulse, the signal could have been low-pass filtered a tad?
I guess this would slow down the response times though?

If it's not possible to add notch filters to the controllers in an RF type way, then the only way is to do TDM (as others have said).
Oh well, it was fun while it lasted. lol

If it really is just sampling the average pulse power, I don't think it would be impossible to build one from scratch.
I've read some of the links and articles on mag trackers, and although the maths is over my head, it could be a good experiment to at least build a similar circuit.

[MSat]

Sadly, I don't even think sharing a clock and resetting the DSPs at the same time would work because it seems that PC driver needs to send a command to enable the device. I realized this because when I was looking for magnetometer interference I first tried it with it plugged in a Linux box without any driver and it didn't do anything. However, when I switched to windows which does have a driver I starting getting the interference. So that's definitely not an option.

Lets say you could tune and filter all the coils (honestly, I don't think it's possible but I could be wrong - my neckbeard isn't very long yet ) you would probably also need some compensation circuitry to make the DSP happy. That's WAY too much hackery. Like you said, it probably would be easier to build it from scratch. I don't know if you're aware of it, but we already have a discussion on an open source magnetic tracker here: http://www.mtbs3d.com/phpBB/viewtopic.php?f=138&t=15431 Sadly, the math is over my head too as magnets (how do they work? heh) are all too magical.

[OzOnE2k10]

The DSP uses a PLL for it's clock too, so they'd never be completely synchronized (unless using a clockout pin).
Not looking good for a multiple Hydra mod then. I'll still might have a mess around, but probably best not to break a brand new gadget I suppose.

Yep, that was the thread I was reading earlier (and the links etc.)
The maths is all in radians and other scary trig stuff that hurts my brain.
I just about got my head around FFT / iFFT and convolution while doing audio DSP stuff a few years ago.

If you really want to screw up your brain, just go on this... lol
http://www.overunity.com/7679/selfrunni ... kapanadze/
http://www.overunity.com/10156/selfrunn ... V9vTqIph8E
http://www.youtube.com/watch?v=-MYVmtF4JYg

I got stuck into the TPU junk for years. Had enough of all that rubbish now.
(I even bought a petrol genset and HHO cell to play with, but it only ran for a few seconds due to lack of HHO production).

Anywho, I digress (as usual) hopefully the open source mag tracker will prove more fruitful.
Actually, if the 50% discount on the Hydra was made permanent, I'm sure many thousands more would be sold.
A good but cheap open-source mag tracker would fantastic though.

[WiredEarp]

@ OzOnE2k10

Maybe I should have just merged with that thread?

Probably best to keep it separate IMHO. Often if topics get merged too much, there are so many posts that its hard to find what people are specifically looking for. I think its sometimes better to have specific titles, and if necessary aggregate them into a sticky post or similar. You could maybe edit the title to make it more specific, depending on whether you are wanting general mods (like people sticking it to gloves etc) or whether you are specifically looking at multiple controller hacks etc?

Re multi sensors, I had the same idea as MSat and yourself about disabling one of the bases transmitters and getting the other sensors to pick up the first bases transmitter - I think that someone with more electronics skill/knowledge than myself could probably disable a transmitter, and synchronise that base to the other base physically, so you could use 4 receivers on the same base station.
At least until drivers were written specifically for it though, you might need to connect the 2nd base to a secondary PC and access those two sensors via VRPN, as I doubt the standard drivers or the VRPN drivers will allow you to run 4 without hacking/rewriting (since the signal would be coming from 2 Hydra bases). That said, I don't know how much work in depth work you'd want to do on the Hydra currently, as I really hope we might be due for a new version this year...

[MSat]

I do hope Sixense/Razer release a device more suitable for VR use. Or at the very least offer some alternative controllers for the Hydra.

[WiredEarp]

It would be awesome if they bought out wireless tracking 'modules' that could be snapped into different devices. Then they could sell guns, wands, etc, and people could just snap in the sensor, or other people could use them in their own solutions.

That said, I think we are more likely to see a wireless Hydra similar to the current one, possibly with more than 2 sensors, and hopefully including vibration in the controller.

[yomer]

I'm over the fence about getting the Razer Hydra at a discount. Seeing what can be done in VR settings like The Gallery and the Tuscany Demo, I do think it's a good investment. BUT, the thought of a possibility in which Razer were to release a near zero latency wireless controller is holding me back. Any thoughts?

[cms]

I have a couple of hydras that I've torn apart. I did a quick test and it seems to work...

I removed one base station's transmitter but had the other emitting like normal. I was able to get what looked like good readings from both sets independently. They were each running on their own computer, but I'm hoping it's not too hard to get vrpn working with two on the same machine.

The removal of the transmitter is really easy. To reattach, you'd need to do a bit of soldering though.

[Lookforyourhands]

Woah you guys r smart.

[OzOnE2k10]

@cms, does that mean it seems to give smooth / normal readings on the Hydra with the transmitter removed?

This is good. I'm not sure how the hell it's working, but it's good.

Don't know how the two Hydras are in sync with each other?
It looks like the DSP directly switches between the three transmit coils, so it's a bit strange.

Please keep us updated.

[MSat]

Yeah, please find out if the data is correct, and if it's repeatable. You should definitely be getting *some* sort of data, but it seems highly unlikely that it would be the *right* data. I would really love to be proven wrong

[WiredEarp]

@ cms: nice work! Often practical work yields results that seem unlikely/impossible for technical reasons. One of the reasons often great stuff is done by those who aren't massively knowledgeable about the potential problems that will arise.

I have a spare Hydra I can rip apart to disconnect the bulb on. Is it just a wire you desoldered?

From a technical perspective, I can't see how your mod would work without synchronization, UNLESS the actual synchronization is done by the handsets themselves. Perhaps they pick up the magnetic pulses and self calibrate to the signal?

[OzOnE2k10]

Agreed, often the best way is to try different things.
This trial-and-error approach tends to reveal a lot of hidden gems about a specific device.

@WiredEarp - I'd imagine it's a case of just desoldering the transmit coils from the PCB (making sure to label the wires first).
Or, you could just desolder and lift one leg of each of those big yellow caps to save removing the coils...

Then again, that could cause lots of unwanted noise in the coils etc.
Even worse is that leaving the circuit unbalanced like that could feasibly damage the driver IC's (some RF rules may still apply).
So yeah, simply disconnecting the coils would be the best bet.

If both of @cms's Hydra systems are working from one sphere, then the only conclusion (like you say) is that the DSP is actually auto-synched to the 2ms "gap" between the three pulses.
It's probably looking for the rising edge of the first pulse. If it doesn't detect sufficient power in all three consecutive pulses, it likely stops tracking temporarily.

Keep in mind though that there is no tracking related intelligence in the controllers as far as I can see.
(The three coils / mux / opamps are essentially working in the analog domain.)

In fact, the mux is apparently "always on" after the drivers have activated the Hydra.
The opamp is then simply passing the amplified coil signals directly to the DSP in the base.

So, maybe the exact sync of the pulses isn't as critical as we thought?
Would be nice if this proves to be working well.

[MSat]

The DNS my ISP directs me to is acting up again (and I can't override it) so I can't access jan's site. Can someone either give me the ip address of it, or post the scope pics here? Or just answer me this: is there any prolonged gap after the 3 pulse sequences? At the bare minimum, the DSP would need to know which coils are being pulsed - something that couldn't be determined if the gap between the coil pulse sequences was the same.

[eshan]

cms, can you elaborate on what's involved in disconnecting the bulb?

I've been contemplating the setup in the attached picture.

System 1 is powering Hydra 1, but none of its other peripherals matter.

System 2 is a backpack with a laptop, a Rift, and Hydra 2. Hydra 2 has its bulb disconnected, and has one wand attached to the head and the other in the hand as a gun controller.

As long as the wands can pick up the signal from the otherwise unrelated System 1, we should have a "wireless" 360 degree VR setup with full position tracking.

[OzOnE2k10]

@MSat... (I hope Jan doesn't mind me borrowing his 'scope shot?)


That 2ms pause in-between the pulses is probably sufficient for the DSP to see where the three pulses start.
It doesn't make much sense to me that there's no direct sync when we know the DSP has control of the transmit coil sequence?

If it indeed turns out to be auto-synching, it would be quite a surprise.

[MSat]

@eshan

The range of the Hydra is not spectacular; the controller needs to be less than 6 feet from the transmitting base (but probably closer to 3-4 feet)