August 08, 2008

Memjet printers officially delayed until 2009

Memjet technology, radically faster 360 page per minute printers, have been delayed until 2009.

Silverbrook's technology (which will be commercialized under the business name Memjet) was supposed to be released in early 2008, according to what company executives told me then. Now, a company spokeswoman says that the "A4/letter printhead and related components" will be shipped to OEMs by the end of this year, with products slated for sometime in 2009. This is consistent with "early timetables," according to the spokeswoman.

Memjet isn't going to manufacture the printers themselves. Instead, they're going to sell the components to OEMs, who can put their own stamp on the technology.

Delays associated with new technology are nothing new. Still, in 2007, Memjet officials promised: a photo printer, which the company hoped to sell for less than $150 by the end of the year or early 2008; the 8.5-inch x 11-inch (A4) color inkjet, due to arrive at the end of 2008 for under $200; a label printer; and a large-format photo printer, expected to cost about $5,000, and capable of printing poster-sized prints at rapid speed.

The Memjet technology uses a series of individual MEMS-based inkjet nozzles, fabricated using conventional semiconductor manufacturing techniques. Each chip measures 20 millimeters across and contains 6,400 nozzles, with five color channels, the company said. A separate driver chip calculates 900 million picoliter-sized drops per second. For a standard A4 letter printer, the result is a total of 70,400 nozzles.

This site covered the memjet printer last year on several occasions.

Memjet printers were one of the developments to watch for 2008.


GoatGuy said...

Thanks for the write up!

I would like to point out that at least from a graphics-community interest point of view, all those wonderful numbers need to be converted into units that 'span the pond':

6400 nozzles, 5 colors. That would be 1280 "pixels". In 20 mm then becomes 64 pixels per millimeter, or 640 dots-per-centimeter. In inches it would be about 1,600 dots-per-inch of full color resolution.

One presumes that as many 'dots' could be generated horizontally as vertically (at least for high quality, high saturation output!)

So let's see. 640 dots/cm squared ... times 22.5 cm width times 27.5 cm length (A4 paper) equals 261,000,000 dots per full page.

Now if 360 pages per minute is the realizable goal, then that's 6 pages per second. Let's give a 10% 'inter-page gap'.

260,000,000 * 1.1 * 6 (* 5 nozzles) = 8.6 gigadrops per sec.

The little driver chip is good for 0.9 gigadrops? OK, then 9 of them or so would be required to drive the 'video' for all those drops to be deposited at that rate.

Each individual piezo nozzle would have to generate 6 (pg/sec) * 27.5 (cm) * 1.1 (interpage) * 640 (px/cm) or 116,000 drops per second.

I wonder if that is realistic. Just sounds kind of high to me. Maybe the deposition needs a 'full drop' relaxation period between drops. I certainly recall most inkjet printers have more horizontal resolution than vertical though: which either implies that the printheads are being slewed more slowly than the droprate to deposit more picoliters per square (whatever)... or something like that.

On the other hand, my Ep**n C90 printer manages to slew its print-head back and forth at well over the critical (6 x 27.5) = 175 cm/sec. It deposits a good amount of ink too ... supposedly at over 1,000 drops-per-cm.

THEREFORE, I conclude that the claims for performance for this technology are supported by 'stretching present performance devices' with high confidence. 6 pages per second, 1600 dots per inch resolution, 5 color printing, 8.6 gigadrop video rates, 9 or 10 driver chips per A4 width ... all checks out.

I'd sure like to see one of these in action!

The huge market of course isn't really the mid-market, but would be as an enabling technology for the burgeoning "digital printhouse" world. Being able to print Elephant stock or A0 (85 x 120 cm) at over 2 sheets per second, in full CMYKk resolution without plates, laser xerography, or other semi-parallel approaches would be quite a boon. Especially appealing for printing low-run custom softcover books with all sheets interleaved on an individual page basis. Maybe some of our printing professionals can enlighten what opportunities would arise in being able to print an unlimited stream of individually different sheets at full rate and resolution viz a vis book and other bound material pritning?