Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
iGaging scale experimentation, et. al..
#1
[tried to post this as a follow up to the iGaging improvement discussion but wasn't able]

I've used the information here to experiment with extracting scale data in developing a custom DRO readout.  I thought to feed some info back should it be useful to others.

I've had a set of the original iGaging scales collecting dust for over a decade and figured it was time to put them to use or toss them for something better.  In the process of experimenting I found a considerable amount of last bit bobble could be reduced and nearly suppressed by assuring the 3V0 power rail at the scale's COB die was clean as possible.  At first without much thought I'd dropped 5v -> 3V with a small switching module I had handy which was a thoroughly bad idea for several reasons.  One is the inherent noise in the switcher and worse under the light load of the scale A/D (which in my measurement peaks at a whopping 16uA) the switcher is in skip-cycle mode making the rail ripple/noise far worse.

I do see a ceramic cap in on the scale PCB (silkscreen C1) on a trace meandering out from the COB blob which has about 1.3V on it and I'd hazard is a supply bypass for the A/D.  But nowhere else do I see any supply decoupling on the board.  The VDD trace runs right by C1 on its way to the COB blob and at first I thought it was a VDD bypass until I buzzed it out.  Anyway C1 connects to the large adjacent GND plane so I removed enough soldermask next to it from the GND plane and VDD trace to accept reflow of a 1206 22uF ceramic cap.  Afterwards that LSB bobble was just about gone.
In my observation the OEM iGaging LCD readout doesn't reflect any residual bobble seen in the data stream when I clock it out so either it is averaging the data and/or applying hysteresis to the value or things will improve when I finally get a linear 3V0 LDO put in use.

I haven't cut into the mini usb cabling to discover what level of engineering lies there but I don't expect much.  It appears shielded 4-wire cable with silicone jacket is now available from Aliexpress, etc.. so that's likely what I'll substitute.  You can pick up the increasingly scarce mini usb male connectors there as well.  I've in the past constructed high-flex custom cables using thin wall silicone tubing and standard silicone insulated fine strand wire.  In a pinch you could do the same here, adding a bare stranded drain wire and  length of aluminum foil spiral shield wrap scavenged from elsewhere, pulling that business through a silicone tubing jacket.   I don't see a compelling motivation to solder a cable directly to the board as you're already stuck with a larger sensor case to accommodate that getup.  And admittedly swapping a damaged cable is much less involved with the mini usb connector.

Not sure I'd invest more effort attempting further improvement however given the available real estate one could envision a small pcb with point-of-use LDO VDD regulator, clock input termination/hysteresis, and stiffer data out drive -- all of which would easily fit in a 1cm^2 board or perhaps a bit more if you'd want it to add a JST or 100mil connector to obsolete the mini usb.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)