I think it's a raster-scanned terminal, with a bitmap-like graphics board in it. Did HP ever make a vector _terminal_ (as opposed to a vector display device, they certainly made those). A useful site for older large (not handheld) HP machines is http://www.hpmuseum.net. The chap who runs it has a lot more manuals scanned that he can't host on his site, and he seems to approve of anyone who wants to preserve HP machines. He's provided me with information by private e-mail in the past. Worth asking him anyway (his contact details are on the site somewhere). -tony

HP made the (excellent) HP1350 and HP1351 vector graphics boxes, but these have no keyboard input capability. Vector quality is very very good, better than a DEC VT-11 IMO, but the VT-11 is a bit more advanced (more capable). ----- Original Message ----- From: "Tony Duell" <ard at p850ug1.demon.co.uk> To: <cctalk at classiccmp.org> Sent: Friday, August 05, 2005 7:07 PM Subject: Re: HP2648A terminal

The VT-11 has lots of line drawing modes that the 1350 does not offer. Also the 1350 (and 1351, I have both) does not do such a good job of equalizing the brightness of vectors of different lengths. I think I have to reluctantly give the nod to the DEC 1350 in terms of capability. But I think the HP's analog vector generator outperforms DEC's digital vector generator. ----- Original Message ----- From: "Tony Duell" <ard at p850ug1.demon.co.uk> To: <cctalk at classiccmp.org> Sent: Sunday, August 07, 2005 3:27 PM Subject: Re: HP2648A terminal

First off, I meant to write DEC VT-11, not 1350. As for digital versus analog vector generators, there is a vast difference. And this is not like the difference between microprogrammed and random logic designs! A digital vector generator uses counters that drive DAC's, each counter is clocked at a rate that controls the slope of the vector. Most often binary rate multipliers are used to divide a fixed clock down to the rate needed to clock the counters up or down at the rate needed to draw at the desired angle. An analog vector generator on the other hand will recharge a high quality capacitor (often compensated). Once the X and Y axis DACS have charged these caps to the vector starting point voltages, the DAC's are updated with the vector end-point values an analog switching then causes the caps to charge or discharge to the new values. In terms of screen quality, digital vector generators often have visible steps to the vectors, analog generators should never show such steps. Digital vector generators 'move' the beam around by changing the digital values to the X and Y axis DAC's over time. Analog vector generators (generally) only update the DAC's for the start and end points of the vectors. I highly recommend reading 'The secret life of vector generators", which you can find on the web, it was written by an engineer at Atari many years back. Often its possible to tell an analog vector generator from a digital one just by looking at the display quality. Some old arcade games used digital generators, others analog. IMO, digital vector generators are an exercise in sheer brute force engineering while the analog vector generators require a lot more finesse and attention to details. Clearly HP had a excellent analog design engineers, and they really worked on getting the HP 1350 and 1351 analog sections just right. DEC's focus was more on the digital side, so the VT-11 offers all sorts of graphing modes and other features that the 135x vector boxes would have to emulate in software. I think the VT-11 and HP 1350 really illustrate two very different approaches to vector graphics in many ways beyond the vector generation methods. The VT-11 takes up main memory bandwidth, while the 135x has its own display list memory. The VT-11 sits right on the main CPU bus and acts like a co-processor while the 135x is very clearly a peripheral device (HPIB interface). Sadly, the HP 135x lacks a light pen, but offers ways to logically partition the display list for easy display animation. This is harder (but not too hard) on the VT-11 (or an Imlac for that matter). Does the VT-11 have a keyboard port? ----- Original Message ----- From: "Tony Duell" <ard at p850ug1.demon.co.uk> To: <cctalk at classiccmp.org> Sent: Monday, August 08, 2005 6:05 PM Subject: Re: HP2648A terminal

Yes, thsoe are the definitions I assumed to. The HP1350 is most certainly a digital vector generator by that definition. I've just looked at the official operating/service manual. The DACs are driven by counters, clocked by rate multipliers, etc. Not really. There's a keuboard connector at the back of the monitor used with the VT11, the signals from that (2 wires IIRC) just go down the cable to one of the VT11 boards and end up on backplane pins. In the VT11 configuration (the 3 boards in a 4 slot backplane, to plug into the Unibus of a PDP11), these pins are not used. In the GT40 configuration (graphics terminal using these 3 boards, PDP11/05 processor boards, core memory, serial interdace in a 9 slot backplane, fitted into the 5.25" PDP11/05 cabinet, with a real frontpanel), the keboard uses the UART on the CPU board set, the appropriate pins are wired across the backplane. -tony

Thanks to everyone who sent comments on this terminal. I'm still digesting it all, and will get back to one or two of you for more information soon. Thanks Jim.