Well I’d hoped to have batch 3 of FlexVolt production PCBs by today, but, as usual, the production house is taking twice as long as expected to assemble and ship the PCBs…  I have a tracking number, and I have seen pictures of the assembled PCBs, so at least I know I should be getting them soon!  And then testing, assembling, programming, and shipping the rest of the pre-order FlexVolts!

In other news, I have finally started to get feedback on software usage, and I have already started building v3 of the FlexVolt app for desktop computers to address reported bugs.  There is also now a FlexVolt mobile app under development!  There is the possibility of developing a sleek mobile app, then porting it to desktop apps, which might offer a nicer form factor than the current Processing-based app.   I also have a test forum up and running – it will go live once I am convinced it is reliable.

Questions?  Comments?  Bug reports?  info@flexvoltbiosensor.com

This week, after several weeks of extensive testing, FlexVolt Bluetooth models are finally rolling out!  Hooray!

This is very exciting, and has been months in the making.  The USB and Hacker (Arduino Shield) models have been shipping for a while now, and are currently in stock – yes you read correctly, Hackers and USB models now actually ship when you order them!

The Bluetooth Model took considerably longer for a number of reasons, some foreseen, some unexpected.

  • The Bluetooth FlexVolts are more complicated – more components, a switch, power management, battery charging, actually connecting to your computer, and on, and on.  This means more time for development, assembly, and testing.  Expected.
  • One of the first Lithium-ion batteries tested in a bulk order was defective!  During routine testing this battery entered a defective state commonly thought of as “thermal runaway” and overheated.  Unexpected.  If you want the details, check out the previous post.

Now it is time to build the forum, expand the software, and start posting educational materials!

setbacks

If you’ve ever tried to do anything, whether it’s packing and moving, machining a new part, or building a shed, you’ve learned that everything takes longer than you think it will.  A good rule of thumb is to estimate your time requirements, then multiply by 2.

Why?  Because of unexpected setbacks.  You find some items that are heavy/hard to move/hard to pack.  Or you make a mistake machining and have to start over or create a workaround.  Or you realize it takes a lot longer to level the ground and place shed footings, and there’s a rock where you wanted to place a footing.  A car-sized rock.

Successfully shipped – USB, Hacker

The FlexVolt Hacker Arduino Shields started shipping over a month ago and the FlexVolt USB models started shipping two weeks ago!  It took much longer than expected, but it’s still very exciting.  By now many FlexVolt customers have received USB models or Arduino shield models and are using them to measure there muscle signals!

Delayed shipping – Bluetooth

The FlexVolt Bluetooth models were destined to ship last week, but then a faulty battery stopped everything.  Lithium-Ion batteries are great – high energy density, good cycle life, minimal memory-effect – but they have been known to fail through overheating.  So when one of the assembled units got warm and the battery failed, I stopped all Bluetooth shipping to learned why.  This failure resulted in me being extremely concerned about the integrity of the rest of the batteries and of the charging circuit.  I could not ship defective products that might overheat!  So I stopped preparing for shipping, unpacked devices that were already packed and ready to ship, and extensively (even more so) tested each battery and charging system.  Thermocouples were placed inside cases to monitor temperatures during charging.  Batteries were stressed and monitored, but no problems were seen.

Eventually, it became clear that the problem was inside the failed battery, not with the design of the battery charge controller circuit.  I learned that one type of manufacturer’s defect in a Lithium-Ion battery is an internal partial short-circuit.  This short can lead to internal energy loss through heating, which can damage the battery and lead to further heating, which can damage the device!  I am currently developing and performing quality control tests for the batteries, the charging circuits + batteries, and the assembled devices, to make sure devices are not shipped with a defective battery that could overheat.  Shipping of FlexVolt Bluetooth models will resume when I am convinced the problem is solved.

FlexVolt Hardware:

Another milestone has been reached in FlexVolt development.  The USB and Bluetooth models now have fancy machined enclosures!  It took a lot of planning and geometry to fit so much into such a small package, and 2 rounds of machining and test-fitting to get everything lined up right – larger components like LEDs and jacks actually have a bit of wiggle room and they can end up with quite high offsets from the expected position after assembly!

FlexVolt_Bluetooth_Enclosure_1-1FlexVolt_Bluetooth_Enclsoure_3-1

 

 

 

 

Firmware:

What now?  Why aren’t these muscle sensors already shipping?  One milestone remains – the Firmware.  The software was released weeks ago after extensive efforts to make a good app that would work.  However, if folks find bugs in the software, or request improvements, I can modify the software and release the improved app as FlexVoltViewer version 1.2.  The Firmware, once flashed onto the 2 micro-controllers in each FlexVolt (think brains) cannot easily be changed by customers without use of a fancy programming device and accompanying software.  So this Firmware is being extensively tested, AND a lot of consideration is being put into what features should be included, such as error handling, user settings, device names, and using memory to store settings, even if some of these features are not yet available in the software.  Here’s what the programming and testing of Firmware looks like:

FlexVolt_PICprogramming_USB_3-1

The red device in the back is a programmer.  The mess of jumper wires, chips, resistors, capacitors, and LEDs on this solderless breadboard is the test setup for the FlexVolt micro-controllers (brains).  This test setup is getting a lot of attention right now, so that when FlexVolt USB/Bluetooth models ship, they are running very reliable firmware!  The current goal for finalizing FlexVolt USB/Bluetooth Firmware and shipping orders is June 20th.  This goal allows for time to distribute FlexVolt test units to local beta-testers to ensure no problems down the road!

Software:

The Software has been available for download since early June.  Check it out here.  So far, several folks have received FlexVolt Hackers (the Arduino Shield) and have, presumably, accessed the software.  I have as of yet heard no news, so I’m assuming things are running smoothly until I hear otherwise!  Send comments, questions, or bug reports to info@flexvoltbiosensor.com

500LeadCablesLeads Arrived, Hackers Started Shipping!

The custom lead cables finally arrived last week.  All 500!  These will each be tested before shipping.  A tedious effort, but worth it to ensure all shipped products function properly! HackersAndLeadCables

FlexVolt Hackers (Arduino Shields) were the first models ordered from production, to test the production house and the design basics, because the Arduino Shields are simpler electronically (fewer components, no brain).  Shown above are the assembled and tested Hackers, along with a bundle of tested lead cable sets.  Several Hackers have already shipped for pre-order customers who did not also order USB or Bluetooth models.  The Arduino Codes and Computer Apps will go up on the software page any day now.

 

PCB_Shipment2_USB_Bluetooth_HackerAnd Finally!

PCB_Shipment2_USB_BT_Hacker_BoxPCB Production Run 2, including USB, Bluetooth, and Hacker boards, arrived!  For the USB and Bluetooth models, these boards are the first PCB versions of the designs.  Everything up to this point has been breadboard and protoboard.  It’s very exciting and rewarding to see the design take form as an electronic board!

What next?  These steps remain before shipment of USB models begins:

  1. Inspect and Test – verify components and solder junctions
  2. Flash firmware and test PCB function in USB mode
  3. Machine enclosures to accommodate channel plugs, USB cable, switches, and LEDs

Along with those steps, 1 additional hitch exists for the Bluetooth models:  Bluetooth antennae have been delayed in shipping.  I am not sure if they just lost the order, or if they had none in stock when I placed the order, but they are already 2 weeks overdue.  A second batch was ordered 2 weeks ago from a backup vendor, with the hope that at least one of the shipments will arrive in the next 1-2 weeks.

Despite finalizing all designs by the end of March and setting up production orders with the PCB manufacturer and assembler, it is now looking like the first production batch will not be ready to ship to customers until the 2nd week of May.  The delay is due mainly to complications in dealing with the manufacturer (many errors and back and forth’s on details and pricing) and longer-than-expected lead times on some components (1 week backordered!).  On top of that, the cables have not been produced yet either.  I ordered 500 lead cable sets 2 weeks ago, but the manufacturing company ran out of the wire used and had to order more, which added a 2 week delay.  They tell me the cables will be finished and shipped by the 28th of April.

This delay and the frustrations of dealing with a large PCB printing/assembly company have me considering doing all assembling in-house.  This would require solder stencils and tools, solder paste (lead-free of  course), a reflow soldering oven, and pick-and-place tools to place all those tiny 0603 surface mount resistors and capacitors.  The order has already been placed for this production round – 10 units each of USB, Bluetooth, and Arduino Shield FlexVolts – but next time I will likely just be ordering blank PCBs and assembling them in my workshop.  I’ll need a large magnifying glass and steady hands, but if I assemble myself I can cut the lead time from 3-5 weeks down to 2!

 

The FlexVolt Arduino Shields arrived last week.  The boards were well assembled and arrived on time, and the assembler even included the spare parts and solder mask.  DSC_0121-1After soldering the remaining components in-house, the boards went through testing and worked as designed.  So far so good!DSC_0135

The cable supply line has not been quite as smooth.  A mix-up in the initial round of samples led to poor signal, and an additional week of troubleshooting and back-and-forth with the manufacturer and a week to produce and ship another round of improved samples has finally led to a functional lead set.  I ordered the production batch today (500 lead sets!) and should get them in within 2 weeks.  I am disappointed that this delay has set back the predicted ship date for Hacker Shields from the 1st week to the 3rd week of April, but I did not want to rush through at the sacrifice of quality.

Now back to finalizing designs for the USB and Bluetooth models for production orders, and testing and finalization of firmware.

Just got word from the assembly house that the production test batch of FlexVolt Hacker Shields is assembled and shipping from the factory soon.  This is slightly ahead of schedule, which is very exciting!  Once they arrive and pass testing, they start shipping to the early supporters.  Here are images from the factory:

Printed PCBs:FlexVolt_Hacker_v1.8_PCBAnd now assembled:FlexVolt_Hacker_v1.8_Assembled

The custom lead samples also arrived this week.  Now for final testing, and if all goes well, it’s time to place a large order (500!).  Here’s what a lead set looks like plugged into a prototype Arduino Shield:

FlexVolt_LeadSample_ShieldProto

Over the next week I’ll be finalizing the Computer-side code and the Arduino Code necessary to run the FlexVolt Hacker Shield.  Time will also go into tutorials and lesson plans, so that everything is ready by the time customers start getting the Arduino Shields!

Arduino Shield – prototypes were assembled and tested this past week, and the results were good – no major, or even minor, issues were found.  The production versions of the PCBs were finalized, and orders were placed for PCB printing, components, and assembly.  The supply chain took longer than expected to establish, but things are rolling again now!  FlexVolt Arduino Shield Units should arrive around the end of March.  Assuming no issues in quality control and testing, these units should then quickly ship to fulfill pre-orders in the first week of April.

FlexVolt USB/Bluetooth – Patient protection, sensing circuitry, shielding, and power management are now finalized.  Next is packaging – how the PCB fits in the enclosure, and custom machining of ports for USB and lead wires.  Prototypes will be ordered this week for assembly and testing.

Lead wires – custom lead wires were designed, including an injection molding tool.  The tool is made, and prototype lead wires should arrive this week for testing.

In the last week of January, 2014, after the Kickstarter project ended, FlexVolt Pre-Orders became available through the webstore, and product development continued to reach the goal of shipping consumer versions of FlexVolt by April (March for Arduino Shields and initial production and testing units).

In the first week FlexVolt reached over $3,000 in pre-orders!

The first 3 weeks in February have been very busy, but were mostly devoted to fine-tuning the electronic designs and establishing the supply chain.  A lot goes into producing a consumer product!  Here are a few of the items worked on:

  • Patient safety isolation – Complete isolation between the computer and the user prevents any abnormal power surges on the computer side from coming through the USB port.  This includes digital isolators for power, ground, and signal lines, transient voltage suppression at lead connectors, and current-limiting resistors in-line with all lead connectors.
  • Supply Chain – making a one-off device is quite different from sourcing leads, electronic components, PCB printing, PCB assembly, and enclosures for even low-volume production!
  • Arduino Shield PCB – prototype done and arriving from the PCB printing facility (OSHPark for prototype boards) any day for in-house assembly and testing!
  • Bluetooth Design – Choices made for battery type, battery charge/discharge control circuitry, and power-path management (i.e. turning bluetooth on/off, charging the battery, or running on USB for both power and data communication)
  • Custom-made leads – re-designed as plug-in sets to make leads easy to remove, add, or swap.  Shielding added to help reduce 50/60Hz noise as well as other noise found in most homes these days.
  • Programming – lined up outside help for phone app development for Android and iPhone, developed and documented communications protocols (still need testing prior to public release), began to standardize functions and roll them into object classes in preparation for building a library plugin

I’ve received many great comments, either through the website or through email, ranging from requests/suggestions for specific features to offers to help with development!  As work transitions from working on the hardware and electronics to finalizing software, expect to see more open-source style activity, including software posts and establishment of user/developer forums!

Of course, please feel free to post or email any requests/suggestions/comments you may have!

Cheers,

Brendan