I was just writing up a response for 'viewing' the ink level in the 3880 refill carts when I bumped into this update by websnail (Ron)?
Yes, it appears that we will be forced (but not defeated) to return to top filling or viewing. Previously, I had suggested the old car battery fluid level viewer--a plastic rod with a beveled end mounted into the plugcap. If the fluid level fell below the beveled end light was reflected to viewer--if the fluid level was OK, no reflection. This would work well with Canon BCI-6, CLI-8 and other similar carts as well as the larger 3800/3880 carts when they go totally black. However all we would know is that the ink level is at or below the chosen level limit. Let's now take advantage of another law of physics--capillary action. Opening the top of cart in use, I insert a piece of glass tubing til it hits the bottom of the cart--now place your finger over the top of the glass tube and raise the tube out. With calibrated eyeballs measure the fluid height in the tube, now using your teflon greased 'sliderule' compute the quantity of ink required to fill the cart and sponge properly. You will, most probably, wait for the first burst of ink to be absorbed in the foam before finishing the refill.
If you have reasonable manual dexterity and sufficient toweling, just leave the glass tube in the cart after determining how empty (or full) it is. Add ink to cart, recheck the ink level in the tube and repeat until the cart is filled to your satisfaction.
Should this direct path be blocked off, as per the maze construction in certain Epson carts, then weighing is our final resort. The least expensive simple scale is just a spring from which the cart is hung and a ruler. Weigh several new full carts, average the measurements and you are good to go--as long as the cart is not totally invisible this technique is foolproof.
If springs don't move you then a cantilever beam should make your day. Mount a thin, long strip of plastic or metal to a small block of something and construct a holder for the cart on the other end. Again weigh several full carts and average the measured drops in height. Mount your empty cart, add ink until the 'drop height' is equaled and you are done once again. Please note: this technique will not work in outer space! Oops! I'm wrong. If you deflect the spring or beam and release; a measurement of it's vibration frequency will yield the weight.
With all punditry intended I think the weight of this problem has been defeated! irv weiner