Tektyte:LogIT USB & PoE Kickstarter

[tekthead]

Hey guys,

I am an engineer based in Melbourne, Australia who has created a new type of voltage, current, and power logging device specifically for USB and Power over Ethernet (PoE) devices.

https://www.kickstarter.com/projects/642519995/tektyte-logit-specialised-circuit-testers

The challenge of measuring USB and Power over Ethernet voltages and currents inline is normally a time consuming and often expensive exercise that leads to a complicated test bed setup for some pretty straight-forward and desirable results. The LogIT devices make the job much simpler and in doing so also make the information these devices measure, more accessible to both the professional engineer and DIY enthusiast alike. The kickstarter video clearly shows that devices like those made by Arduino and Raspberry Pi Foundation can be monitored for voltage, current, and power effortlessly. Additional features found only in professional multimeters have been included as standard such as logging to microSD card and streaming real-time data to a PC for live monitoring of devices while they are connected and being tested. What makes these devices different from a standard multimeter is that they are easy to connect and support the reduced range of positive voltages found in the USB and PoE standards. Not having to cope with wide voltage range measurement or 5 digits of accuracy make these devices much more cost effective while also being an invaluable tool potentially used more often than a multimeter for embedded electronics developers. One task that the LogIT tools perform especially well is battery charging measurement and battery life estimation. The free software week are developing has a useful battery life calculator built in that takes real measurement data and calculates the expected life for different standard battery sizes or a user selectable custom battery capacity. This helps designers optimise the battery type for a give application.

I hope people wont be scared away by the 12 bit resolution. Internally it uses a programmable gain amplifier to achieve a much higher effective resolution.

If you would like to be involved in a discussion about how we can make this device better please post. If it is a really good idea I will send you a t-shirt as a thank you.

We are currently looking into plug based screw terminals instead of spring terminals for larger wire gauge support, USB3 data pass-through, and a higher resolution display.

If you want to support what we are doing here in Australia please join the twitter @tektyte, facebook page, or mailing list at www.tektyte.com

Again, your thoughts and feedback would be greatly appreciated.

[snoopy]

This looks like a really useful tool. You should reach your target for sure

Well done !

cheers

[codeboy2k]

Just watched the video.  Looks like a nice tool.  I have a special jig I already made for measuring USB voltage and current and it works for me.

As a standalone product, with what looks like pretty decent software , it looks pretty good ! Great work.

The only thing that would make it really worth my money is full USB isolation.  If you can do that (and still measure DUT voltage & current  on the isolated side) then it's a fantastic bargain. Standalone USB isolators are about half the cost of your device, but don't offer the measurement capability.  If you can add in isolation for less than the cost of a standalone isolation box then it's a deal

Good job, nonetheless.

[tekthead]

Just watched the video.  Looks like a nice tool.  I have a special jig I already made for measuring USB voltage and current and it works for me.

As a standalone product, with what looks like pretty decent software , it looks pretty good ! Great work.

The only thing that would make it really worth my money is full USB isolation.  If you can do that (and still measure DUT voltage & current  on the isolated side) then it's a fantastic bargain. Standalone USB isolators are about half the cost of your device, but don't offer the measurement capability.  If you can add in isolation for less than the cost of a standalone isolation box then it's a deal

Good job, nonetheless.

Thanks for the positive feedback.

Actually, this device IS fully isolated. The measurement system runs off it's own internal battery supply and there is both a 1W isolated DCDC converter (for battery charging/PC power) and a separate serial isolation IC for a total isolation between the PC measurement/arduino TTL ports and the USB DUT. The complete microcontroller system floats at the DUT ground potential. The isolation hasn't been tested fully but the PCB clearances should allow for a total breakdown due to creep-age and clearance of about 1.5kV. We are working on a barrage of test methods to be applied to the final release PCB which include breakdown testing.

The PoE version has to be isolated as the -ve voltage is actually often -48V relative to system ground and the +ev power voltage is 0V!!!

We use the same floating microcontroller configuration but a much more sophisticated two board topology with a 24bit 6 simultaneous sampling channel system to achieve the more complicated task of transformer balance correction and fault detection. PoE power measurement if done properly is very tricky.

[codeboy2k]

Oh sweet! It wasn't clear that it was fully isolated (or maybe I missed that part? )  I did get that there was an isolated serial port already. 

Perhaps you could make that clearer because at your early bird pricing it's a fantastic deal

[codeboy2k]

Yes poe is usually isolated as per the standard but I am only referring to USB isolation of the PC  and the DUT.  Again it's an excellent deal with isolated USB ports

[tekthead]

Yes, yes it is. We are certainly not making these things to become rich but moreover to ensure good tools are available for more rigorous engineering and product testing. I have found many students will simply avoid thorough testing if it is "too hard". A device like this helps make qualification of USB current draw under different firmware modes of operation much simpler and it is fascinating to watch in real-time across voltage and current simultaneously.

We use a RECOM isolated SMD DC/DC in conjunction with an SiLabs capacitive barrier data transceiver for isolation, both of which have 2.5kV barrier breakdown voltages if my memory serves me correctly. This separate data and power isolation configuration actually helps us keep the cost of implementing full isolation lower than an integrated data/power isolation solution but still adds a fair bit to the total cost of these devices to make.

The display we use has an MOQ of 2240 pcs for reasonable pricing and is by far the most expensive component (check LS013B4DN04 on Mouser). The BOM has 149 components for the USB model not including housings so when you consider the number of parts on an Arduino Mega is less than half this, building these things is an interesting exercise with dubious profitability. We are most definitely subsidizing the build for the early backers but there are limits to what we can afford, hence the cap. Also, the AU dollar is in the crapper (technically speaking) which makes things even more challenging for manufacturers here.

Having said this, lean manufacturing may be easier in Australia to some extent. Our level of general knowledge often means it takes less engineers to get the same job done. For instance, I ended up building the website, building all of the KS content, designing the plastics, designing the PCB, designing the overlays, on top of designing the system architecture. Luckly, I didn't have to write all of the code for the firmware and desktop software (though I did make some groovy icons and installers). This kind of thing is what makes it happen in Australia sometimes I believe.

[tekthead]

Scratch that. I am not going to speak for every Australian engineer. Hard work and/or persistence is probably an internationally recognized key ingredient.


Thanks for the feedback. I will figure out a way to make the isolation feature clearer and post as an FAQ in the interim.

Cheers, Matt

[vargoal]

Seems like a cool project, Do you have any plans to release the software as an open source solution?

[tekthead]

Seems like a cool project, Do you have any plans to release the software as an open source solution?

Perhaps one day but there are no plans to open source the solution currently. We are still actively developing it and adding features which tie in closely with the hardware. The protocol will be released and documented for anyone to use such that the application can act as a graphing endpoint for alternate data streams. That we supply it with a share and share alike license is as far as we have currently developed the licensing model.

[EEVblog]

The display we use has an MOQ of 2240 pcs for reasonable pricing and is by far the most expensive component (check LS013B4DN04 on Mouser). The BOM has 149 components for the USB model not including housings so when you consider the number of parts on an Arduino Mega is less than half this, building these things is an interesting exercise with dubious profitability. We are most definitely subsidizing the build for the early backers but there are limits to what we can afford, hence the cap. Also, the AU dollar is in the crapper (technically speaking) which makes things even more challenging for manufacturers here.

Yep, the low Aussie battler at the moment makes for expensive parts.
I was going to say you wouldn't have much change on the early units.
Are you going to get them locally assembled?, it's not that expensive here.

[tekthead]

The display we use has an MOQ of 2240 pcs for reasonable pricing and is by far the most expensive component (check LS013B4DN04 on Mouser). The BOM has 149 components for the USB model not including housings so when you consider the number of parts on an Arduino Mega is less than half this, building these things is an interesting exercise with dubious profitability. We are most definitely subsidizing the build for the early backers but there are limits to what we can afford, hence the cap. Also, the AU dollar is in the crapper (technically speaking) which makes things even more challenging for manufacturers here.

Yep, the low Aussie battler at the moment makes for expensive parts.
I was going to say you wouldn't have much change on the early units.
Are you going to get them locally assembled?, it's not that expensive here.

It depends a bit on final numbers. If we are successful but under 150 devices we will definitely be assembling them ourselves. Over that threshold but under 500 units there are still good options in Australia. I am not really contemplating overseas assembly and testing for the final stages. PCBA is a different story. These devices have been designed with self assembly in mind so it shouldn't take very long for even the most rudimentary production line to assembly hundreds. I would prefer to do calibration in-house anyway if possible. Writing rigorous QC processes and procedures often has diminishing returns for small batches.

Besides. I love assembling kits and the rear overlay design pretty much says it all. 

[tekthead]

Over the longer term we will probably implement a bed of nails jig which will allow for more manageable test procedures. I have already put a large number of pogo-probe compatible test points into the design.

Btw, I really liked the way you panelised your uCurrent for integrated test wiring at the corners. Very elegant.

[EEVblog]

It depends a bit on final numbers. If we are successful but under 150 devices we will definitely be assembling them ourselves. Over that threshold but under 500 units there are still good options in Australia.

There are good options in Australia for any number unless you are maybe talking hundreds of thousands.

These devices have been designed with self assembly in mind so it shouldn't take very long for even the most rudimentary production line to assembly hundreds.

True, but is your time worth?
Assemblers here have no problem assembling small batches like 50-100, and don't cost a lot in the scheme of things.
My local assembler can even use loose tape on their new machine, doesn't need reels at all.

I would prefer to do calibration in-house anyway if possible. Writing rigorous QC processes and procedures often has diminishing returns for small batches.

True. You'd need maybe 500-1000 units to make that worthwhile I suspect.

[tekthead]

I hear what you are saying. I have a few of scenarios costed but my preference is to assembly here too so I am doing what I can to make this happen.

[tekthead]

For anyone who supports our project here in Melbourne by getting a T-shirt or LogIT device we would like to extend an invitation to an exclusive celebratory gathering in late March which includes free drinks, a chance to see were these devices are designed, wrap your hands around the prototypes, and talk with the engineers involved.

It would be great to open up our workshop and talk about our experiences with product development and Kickstarting a widget.

The funding is at about 50% after less than 5 days so it is possible we will actually make it!!

[snoopy]

The display we use has an MOQ of 2240 pcs for reasonable pricing and is by far the most expensive component (check LS013B4DN04 on Mouser). The BOM has 149 components for the USB model not including housings so when you consider the number of parts on an Arduino Mega is less than half this, building these things is an interesting exercise with dubious profitability. We are most definitely subsidizing the build for the early backers but there are limits to what we can afford, hence the cap. Also, the AU dollar is in the crapper (technically speaking) which makes things even more challenging for manufacturers here.

Yep, the low Aussie battler at the moment makes for expensive parts.
I was going to say you wouldn't have much change on the early units.
Are you going to get them locally assembled?, it's not that expensive here.

They should sell it in USD then the exchange rate variations won't kill it !!

cheers

[tekthead]

They should sell it in USD then the exchange rate variations won't kill it !!

cheers

I agree Snoopy but the Australian Kickstarter site uses AUD. US exchange rates have been factored in to an extent but it has still dropped further since costings were done which is unfortunate. Either way, the early bird tier is almost at cost without even factoring in NRE expenses.

Sometimes the costs are outweighed by the desire to make something of quality so this is all a bit of a moot point. This is more of an exercise in making a good tool without going backwards.

[jeremy]

Backed. If nothing else, I want the isolation, and I've seen isolators for at least this much. But it looks like a genuinely useful product. PoE would be cool, but I don't really have any use for it so I got the USB one

A few questions:
- Can you do min/max only on the real-time display?
- Linux? or alternatively, will you publish any file formats so that we can write our own tools?

Dave, could you list some manufacturers that you consider reasonable in Australia? I'm interested.

PS that is some expensive equipment you have in the background there. I don't suppose you filmed the video at a university or something ?

[tekthead]

Backed. If nothing else, I want the isolation, and I've seen isolators for at least this much. But it looks like a genuinely useful product. PoE would be cool, but I don't really have any use for it so I got the USB one

A few questions:
- Can you do min/max only on the real-time display?
- Linux? or alternatively, will you publish any file formats so that we can write our own tools?

Dave, could you list some manufacturers that you consider reasonable in Australia? I'm interested.

PS that is some expensive equipment you have in the background there. I don't suppose you filmed the video at a university or something ?

Hey Jeremy,

Thanks for backing the project. 

1. The PC software has real-time tracking of min-max, average, and standard deviation alongside a histogram view which pretty much does the same thing. This would be a useful feature to have on the device itself too. I will add it in.
2. Linux is doable. I will look into recompiling for Linux as well. It shouldn't be too hard as the software is written in Java. You might have to give me a bit of time to test properly though.

I am looking into the possibility of USB 3.0 on the device but the connectors are much more expensive and the Type B connector doesn't appear to come as an SMD option which is a problem because we have the screen tightly mounted on the other side of the board. I will model up the alternatives and see whether we can shoe-horn it in. Personally, I think upgrading the screen is better bang for buck as 99.9% of devices on the market are USB 2.0 anyway.

Re. the gear. Actually, my workshop is a little bit better kitted out than the universities main labs. I find is so frustrating working with crappy gear. There is some very nice gear in the RF and optics research divisions at UoM but they aren't readily accessible unless I pull strings.

It was either buy a car or get a nice scope and spectrum analyzer. I chose the latter as my Corolla suits me fine.

[Kean]

For anyone who supports our project here in Melbourne by getting a T-shirt or LogIT device we would like to extend an invitation to an exclusive celebratory gathering in late March which includes free drinks, a chance to see were these devices are designed, wrap your hands around the prototypes, and talk with the engineers involved.

Cool!  I may have to start planning another Sydney-Melbourne road trip... 

[tekthead]

Hey all. I should clarify the isolation statement just in case I have misled in any way.

There are two USB data paths in the device.

1. The primary IN/OUT data path which connects between a power source and a device under test or DUT.
2. A USB serial bridge port (micro-USB) which is both power and data isolated from the primary data path. Using a serial bridge allows us to split the serial and offer alternate TTL level comm's to Arduino devices and the like.

This IS NOT a standalone USB isolator as this would prevent the implementation of the wider voltage range measurement via the terminals (or for rapid USB power charging systems which can go up to 20V). Please don't buy it and expect the primary IN/OUT connections to be fully isolated from each other. This would make for a very expensive piece of equipment indeed and certainly not as compact. It would also limit the total current measurable to that of the isolators power delivery limits.

If you want full USB isolation I believe Analog Devices and Linear Technology all have modules. (ADUM3160, LTM2884). Is fully integrated data isolation something which is desirable in a logging USB measurement device?

Sorry for any confusion.

[Jay_Diddy_B]

Hi,

Have you looked at using the LTC2946 in this application?

link: http://www.linear.com/product/LTC2946




The current sense resistor can be in the positive or the negative lead.

Regards,

Jay_Diddy_B

[tekthead]

Hi,

Have you looked at using the LTC2946 in this application?

link: http://www.linear.com/product/LTC2946




The current sense resistor can be in the positive or the negative lead.

Regards,

Jay_Diddy_B

Hi Jay,

We looked at all options including those from LT. These guys make some great ASICs but this wasn't the best of them.

The resolution of the LT part is 12 bit, same as the solution we use but our logging device currently has a PGA in the front end that leads to an effective final resolution of approximately 15-bit.

With a 12-bit measurement of voltage across a full-scale up to 100V you have very course quantiisation which leads to poor granular measurement (the coded output looks like a step function). This is why we moved to a 24 bit ADC implementation for the PoE version. The PoE version actually has 4 current sense amps and 2 voltage sense channels to allow for cable and transformer imbalance while measuring A, B, and A + B pairs simultaneously. PoE power measurement is tricky when done properly.

This solution would be sufficient for integrated bus power measurement within a product but it is a little course for a device which is designed specifically for measurement and test. Having said this, it all depends on your application.

Also, the drift and voltage offset figures for the LT IC aren't that good. Ideally, precise current sensing is done with something like a chopper stabilized instrumentation amp configuration for current to voltage conversion but there are trade-offs for bandwidth and the like.

I am looking into direct sampling at 16 bit for the USB version but in most cases the complexity and cost of the system goes up. There are other features to consider in a device like this too.

Thanks for highlighting an alternative.

Cheers,
Matt

[Jay_Diddy_B]

Hi Matt and the group,

You raise some interesting points with the POE measurement:

1) in some implementations the spare pairs are sometimes used for power transfer, this may require current monitoring of all 8 wires.

2) The higher power applications (POE+ and POE++) use all 8 wires.

3) Back to back transformers may help, so that you can measure the currents between the center taps?

You also need to make sure that the measuring circuit doesn't interfere with the classification process.

Have you considered using a PD followed by the measurement circuit followed by a PSE? This would be a special mid-span unit?

I hear what you saying about the LTC2946 resolution. 12 bits is equivalent to using a 3 1/2 digit meter at half of full scale. There is a separate ADIN pin which can be used for the lower voltage applications.

The total unadjusted error (TUE) includes, the divider resistor tempco, the reference drift, all over the full operating temperature of the part. Like you say it depends on how much accuracy you need.

I wish you well with your campaign.

Regards,

Jay_Diddy_B