Saturday, May 17, 2014

Posted v0.1.1 or source and v0.1.1 of Regualtor Guide

Today I have finished the bench testing and final changes I wanted to do in the source.  Follow the Link to Files tabs above for the latest copy of the firmware source.  Major changes in this include:
  • Review and scrubbing of all ASCII commands, trying to have an eye towards an external 'smart' application to make it simple to configure this regulator.
  • Improved Tach Mode support.
  • Added new Battery protection capability 'Adaptive Accept'. 
  • Added new Alternator protection capability for Stalled detection.

Tech Mode is when you have an alternator sourced tech in use (use to be common on diesel engines).  These Tachs rely on stator pulses to signal RPMs.  Well, if the battery is fully charged we turn off the alternator field, and hence stator pulses can drop to 0 as well...  So, when Tach mode is selected (via the DIP switches) a small amount of Field drive is kept - even when the battery is full.  Hopefully this will be below the threshold needed to actually start charging the battery, but above the level needed to be seen by your tachometer.  Worst case, there IS a risk of overcharging the battery..  An improvement in the source is I now look to see if the CPU is seeing Stator pulses, and use that to 'decide' at what level of PWM we need to use for a floor.  (Alternatively, you can fix the value via the $SCT: command).

The new feature Adaptive Accept will cap the amount of time we stay in Acceptance phase to 5 times the amount of time we were in Bulk.  An example of where this comes in handy: if you  start the engine with a fully charged battery  the voltage will quickly raise to the Bulk limit.  If enabled, the Adaptive Accept will then quickly transition from Accept phase into Float. (as opposed to just sitting there, boiling out your battery).    Adaptive Accept is enabled in one of two ways:
  1. User has configured it by setting the Exit Accept Amps = -1
  2. We have never been able to measure Amps..
 The 5x factor is defined in the source code via   #define ADPT_ACPT_TIME_FACTOR.  There is no way to disable the protection trigger #2 above, outside of modifying the source code.  The ability to measure Amps is a prime function of this project, and if we loose that ability it should be corrected.  At the same time, it is good to take steps to protect the battery in case of a failure.

Also added is a feature which will monitor the RPMs of the engine.  If we notice the RPMs stop, then we will assume the engine has stalled.  In this case we will limit the drive of the field to 50% of its max capability.  As RPMs increase, the field drive cap is raised.  See $SCO: command, the PFB - Pull Back Factor value for more details.

Both of these capabilities depend on the regulator being correctly installed, with all its sensors attached.  It is this rich ability to monitor and measure things which gives is the ability to better manage the system for more efficiency and better protection of both the battery and the alternator.

Also posted is a .PDF file with the latest documentation of how to install, configure, and operate the Arduino Alternator Regulator.  Most all the answers can be found here, and if not, look into the source code...  Look under the Reference tab above for this .pdf file.

Monday, May 12, 2014

1st photos of assembled PCBs from China

This morning I awoke to a cloudy (I can never think of that word without hearing the NOAA computer voice pronunciation in my head "Claudiieeeee") skies and a few high resolution Emails of a sample PCB set that has been assembled in China!

What excitement, and then what - to be honest - disappointment...  It appears these hand-soldering of small runs in China is kind of a new thing, and this one is off to a very poor start.  Here is a lower resolution copy of one of the photos:

Back in my Seko-Epson days, we would call this NG.  How many examples of problems can you find?

A quick review if it showed to me soo many problems:
  • Missing components
  • One component installed backwards
  • workman ship issue with connectors on right hand side (not assembled together before soldering)
  • (The missing Bluetooth was agreed to, there was some issue importing RN41 modules into China..)
Most troubling was to very very poor quality of through hole soldering.  Insufficient quantity, insufficient heat.  Some solder flash and potential shorts.  Total unacceptable.

I am really disappointed at this.  Have asked them to review and rework all PCBs, but this should not have been done in the 1st place.  Will see what they come back with. . . . .

Wednesday, May 7, 2014

Short run of Arduino Alternator Regulators underway

I received an Email from someone today that made me realize I have not updated the Blog with what is going on right now..  Sorry, guess I have been too busy Email back and forth with a small group of about 10 people, so for anyone else, this is what is cooking:

  1. About a month ago I posted onto several Email lists about this regulator, and asked if perhaps there was some interest in getting together a group build.
  2. There was.
  3. Right now, 20x sets of Rev  0.1.3 PCBs have been fabed and are having the SMT parts mounted.
  4. 10x of those PCBs will continue to have all the through-hole parts soldered in.
  5. All are being done in China ( so no more hand-soldering for me!

I expect to received the assembled (and partially assembled) boards in a couple of weeks.  At which time I will flash in the firmware, spray them with conformal coating, and send them off to the group of folks.  Along with USB programming adapters, mounting hardware, and a heat-sink.  People will need to drill out the heat-sink, bolt it together, source their own Amp Shunt and Temp probes and do the installs.  But hey, we got a thing going!

And if anyone is interested in one of the partly assembled PCBs, drop me a note.  The bare PCB costs around $9, plus about $7 for the small SMT parts that are getting professionally re-flow soldered on.  Finished cost (after purchasing the remaining parts at say Mouser) will add perhaps another $60 or so, but you will have to hand solder those parts..  Then around another $20 for the heat-sink, screws, USB cable, etc..

OK, there is an update of the one or two readers of this blog :-)