Synchronisation of signals

[Drew Smith]

Colleagues,

We use a synchronisation pulse from one of our motion capture cameras to synchronise data collection from our force platforms and EMG system. When the number of devices we were trying to synchronise with our mocap data was small (n=3), it was fairly simple to just use BNC t-connectors to share the camera pulse. As our lab grew, we now have 8 force platforms and our EMG system, and the simple splitting of the synch pulse by means of BNC t-connectors has gotten quite unwieldy.

Can anyone point us to an off-the-shelf solution for distributing the sync pulse to multiple devices, please? I have seen some possible candidates on electronic parts websites but if someone has a working solution, it would probably save me a lot of time.

Thanks in advance.

[Edmund Cramp]

Re: Synchronisation of signals

A quick and cheap test is to connect a small loudspeaker from an old radio etc, to one of the EMG inputs, place the loudspeaker on the force-plate and drop a marker on the plate. This will generate a data file that demonstrates 3D (marker) synchronization with the force plate (marker strike) and the EMG system (the loudspeaker signal when the plate is struck is a few mV, similar to EMG surface levels). If you have a retro-reflective 3D system then a golf ball covered in reflective tape makes a good marker.

If you would like a more "professional" solution, search for a "piezoelectric hammer" - the smaller ones are used for testing hard drives, you don't want to use a larger one on the force plates to avoid damaging the plate.

[Drew Smith]

Re: Synchronisation of signals

A quick and cheap test is to connect a small loudspeaker from an old radio etc, to one of the EMG inputs, place the loudspeaker on the force-plate and drop a marker on the plate. This will generate a data file that demonstrates 3D (marker) synchronization with the force plate (marker strike) and the EMG system (the loudspeaker signal when the plate is struck is a few mV, similar to EMG surface levels). If you have a retro-reflective 3D system then a golf ball covered in reflective tape makes a good marker.

If you would like a more "professional" solution, search for a "piezoelectric hammer" - the smaller ones are used for testing hard drives, you don't want to use a larger one on the force plates to avoid damaging the plate.

Thanks for your reply, Edmund, I apologise if my posting was unclear in any way, but what I am probably looking for is some kind of bnc terminal or breakout box, where I can have the input from our camera with multiple outputs of the identical signal.

[Manvendra Singh]

Re: Synchronisation of signals

You can use a Data Acquisition System from NI and others, with Multiple Digital I/Os. Intercepting synch pulse from Camera in one of the digital input of DAQ System and then distributing(generating TTL Pulses at each output corresponding to time when pulse is being received at input from camera) from digital outputs of DA system to multiple systems with digital inputs will work.
I am editing the post to clarify that passive splitting/redistributing of synch signal has limits and is unreliable, but cheaper.

[Josh Baxter]

Re: Synchronisation of signals

This problem isn't uncommon, and I haven't found any great off-the-shelf solutions. My system is small enough to make due with a 1->4 BNC splitter <https://www.bhphotovideo.com/c/product/875355-REG/NVT_NV_EC4BNC_1_4_Coax_Splitter.html>. Or you could daisy chain all your devices together with a splitter like this <https://www.bhphotovideo.com/c/product/1020439-REG/kramer_ad_bm_2bfy_bnc_m_to_2.html>. Either of these would be pretty quick and cheap. Another option would be to build your own BNC splitter box that satisfies your criteria. BNC thru connectors are dirt cheap on most any electronics website and you could get a decent looking project box for next to nothing. Then it'd just be a matter of linking all of the connectors inside the box.

hope this helps - josh

[Edmund Cramp]

Re: Synchronisation of signals

... Then it'd just be a matter of linking all of the connectors inside the box.

This should work if you have a low impedance signal source driving a number of high impedance inputs, but you might have problems if any of your inputs are terminated or the cables are very long.

[Edmund Cramp]

Re: Synchronisation of signals

My fault for not reading the post fully I guess - it seems obvious now I know what the question is - LOL. There are a lot of CCTV signal duplicators around and, provided that the internal electronics are simply a buffer, they should work fine - in this case, the cheaper the unit, the more likely it is to have minimal signal processing and would therefore work with any kind of sync signal.

[Ton van den Bogert]

Re: Synchronisation of signals

Ed, I enjoyed seeing your answer to the question that was not asked... The marker on loudspeaker drop is a great way to synchronize EMG to force plate and motion capture. I have used marker drop for mocap-force synchronization but this adds EMG.

Old style analog EMG systems are perfectly synchronized with force plate (if they go into the same ADC board), but digital and/or wireless EMG interfaces have a time delay. While they do specify the time delay (Delsys Trigno has 48 ms delay), it is good to have a way to verify this.

Thanks for sharing your experience!

Drew, while not directly related to your question, I am curious how you are recording simultaneously from 8 force plates. Your need for synchronization suggests that they do not all go into the same ADC board. Do they each have their own analog to digital conversion?

Ton van den Bogert

[Drew Smith]

Re: Synchronisation of signals

You can use a Data Acquisition System from NI and others, with Multiple Digital I/Os. Intercepting synch pulse from Camera in one of the digital input of DAQ System and then distributing(generating TTL Pulses at each output corresponding to time when pulse is being received at input from camera) from digital outputs of DA system to multiple systems with digital inputs will work.
I am editing the post to clarify that passive splitting/redistributing of synch signal has limits and is unreliable, but cheaper.

Thanks for your reply.

Interestingly enough, I spent about 25-min on the phone with someone from NI (Sales I think) who asked me many questions about what I wanted to do (and I could hear him typing away as he recorded notes on our call) then put me on hold to speak to his colleague, and, in the end, he concluded that NI had nothing for me unless I wanted to have some connection to the computer to sample the synch pulse and then use software to distribute the signal. Seemed entirely too complicated and I told him that. But, I thanked him for his time and hung up, albeit a bit confused as I thought, as you suggested, that NI would have an obvious solution.

[Drew Smith]

Re: Synchronisation of signals

...Drew, while not directly related to your question, I am curious how you are recording simultaneously from 8 force plates. Your need for synchronization suggests that they do not all go into the same ADC board. Do they each have their own analog to digital conversion?

Ton van den Bogert

When I first started working in my current lab, we had just two force platforms and a 16-channel EMG system, all connected to an A/D board. In the interim, we have gone with USB 2.0 connections from our FP amps and EMG system. We have to accept a fixed sample rate (but it is sufficiently high) but it certainly simplifies the cabling around the computer.

BTW, we don't use all 8 FP at once. If it is a treadmill protocol, then we just need two. If it is an overground protocol, then we use the six we have mounted in our floor.

[Drew Smith]

Re: Synchronisation of signals

This problem isn't uncommon, and I haven't found any great off-the-shelf solutions. My system is small enough to make due with a 1->4 BNC splitter <https://www.bhphotovideo.com/c/product/875355-REG/NVT_NV_EC4BNC_1_4_Coax_Splitter.html>. Or you could daisy chain all your devices together with a splitter like this <https://www.bhphotovideo.com/c/product/1020439-REG/kramer_ad_bm_2bfy_bnc_m_to_2.html>. Either of these would be pretty quick and cheap. Another option would be to build your own BNC splitter box that satisfies your criteria. BNC thru connectors are dirt cheap on most any electronics website and you could get a decent looking project box for next to nothing. Then it'd just be a matter of linking all of the connectors inside the box.

hope this helps - josh

Thanks, Josh. At the moment, we have four FPs and one EMG system sharing the sync pulse using a number of BNC t-connectors. With another four FPs to be added to the mix, I think it was going to get quite messy trying add another four branches off of an already crazy looking mesh of cables.

I like the split off to four outputs, and we could go the homemade route you've suggested. I am sure about a degradation of signal quality though, taking a single signal and splitting it 9 ways.

[Manvendra Singh]

Re: Synchronisation of signals

Yes, I know how it is with NI support, specially when you have some unusual application. What he was explaining is exactly the solution "....wanted to have some connection to the computer to sample the synch pulse and then use software to distribute the signal....". Check this http://www.ni.com/datasheet/pdf/en/ds-135 for example. It has 24 bi-directional Digital I/Os and through software, you can use NI software software to distribute the signal. Software options for configuration are NI-Daqmx(comes bundled with NI Hardware) and Dot Net/C/C++ or Labview(Priced extra).
Check http://www.ni.com/product-documentation/2835/en/