WARNING: HP 54621D to 54600A MONITOR HACKAGE MEGA-POST!So this all started back here somewhere when I had the silly idea that it might be easy-ish to convince the CDM-7SF191 from my 54600A to make believe it's a CDM-7S
X191 that my 54621D can talk to...
Uggghhh...
Okies... anybody got any tech ref on the monitors used in these HP scopes? I understand the difference between the two implementations is the 54621D and family use 32-grayscale palette, while the 54600A and family use a 2-grayscale palette.
54645A/54600A use HP 2090-0316/DataRay CDM-7SF191, the 54621D uses HP 2090-0384/DataRay CDM-7SX191. My searches in the available HP SMs and on the internet turned up bupkis.
Before I go guessing based on poking around with my scope, does anybody here have actual pinouts and or schematic for these monitors?
Tomorrow I figured I'd start looking at the CLIP package from the 54645A and see what I can find there. Maybe lookit the video chip on each and see if I can find some datasheets.
mnem
I'm would be surprized if they aren't interchangeable, they look to be standard bought-in monochrome monitors to me, the different greyscaling would be something done in the video processing, on the mainboard and the output just a video signal.
Certainly I saw the video signal on the diagrams for the 54645A, that was posted a few weeks back. I think the other connections are power (+15V?), an external pot for brightness, can't remember what else there is. The chip on the Hitachi CDM-7SF191 board in my 54615B is a uPC1379C, this is a sync signal processor IC intended for small B/W & color TVs and does the vertical & horizontal stuff, that is done by more discrete parts on the earlier scope displays (the only one that a diagram seems available for).David
Yeeah, I'm not sure. IIRC, these are actually a TTL digital input monitor, so not that simple. I need to do some research.
mnem
Neck board has an added 74S03, so maybe 16 new shades of gray.
CNC industry have had small screens for quite some time.
Possibly already have an almost ready replacement LCD.
The 7403 is on the CDM-7SF191 in the 54645A/54600A/54615B, the newer CDM-7SX191 looks to use one or two transistors instead. The larger monitor boards look to use a similar signal processor IC, layout has changed a bit, need to find better pictures.
The boards in my CDM-7SF191 from 54615B, taken before the re-capicide.
Pictures of CDM-7SX191 from ePay listing.
David
That all came to a head with this weekend of tinker-dwagon-ing...
On the Bench ALL OVER the EFFING BENCH Tonight...
A few days ago I got a bug up my arse to make a debugging breakout for the 54600A/54621D J900 connector; I needed to be able to get probes on with the monitor live and I needed to be able to insert a substitute BRIGHTNESS pot for the CDM-7SF191 monitor. You know, just in case there might be a easy way to make it work.
I started with a scrap of PCB I had that was already the right size, then added a bit of socket header and pin header, then on pins 7-12 I substituted pins from Dupont splice header to be able to use common Dupont female breadboarding jumpers to break out the signals for testing.
Here's the pinout (ribbon plug flipped hole-side up) and the finished breakout in place on the back of the 54600A.
First thing I wanted to look at was the full-bright vs half-bright signals on the 54600A; as expected, they are TTL.
After poking around pins 7-12 of the 54621D, I found only one that looked anything like signal. Rigol is displaying VSync & HSync of both scopes; looks like there's additional information carried on the 54621D's VSync, but otherwise the same TTL signals. Pins 7, 8, 9, & 12 I couldn't distinguish anything valid apart from ambient noise; pin 10 is connected to GND fill on mainboard.
Pin 11 has signal; it varies from 0.6-1.2V P-P. HP is showing a zoom of a single wave captured as close as I could manage to the same time as that on LeSiglent. I want to say VGA of some sort, but the porches aren't always where I expect them to be; maybe blanking the screen or writing text, I suspect?
Unfortunately for my hopes of modding the CDM-7SF191 to work in place of the missing CDM-7SX191, it appears the output from the 54621D is a single-channel analog signal with TTL sync. My first guess would be SVGA at 800 x 600. Either that, or the other pins are dead and that points to what killed this scope in the first place.
Later, I may drop the mainboard on the 54621D just to poke around, see if those pins at J900 go somewhere.
But hey... the Rigol and LeSiglent both got used in anger today. Made it pretty clear my thoughts of selling off the 1054Zed were ill-considered.
And so now, here I am with the results of my little "poke around the 54621D mainboard"... Comparison of monitor cable pinouts of 54621D vs 54645A/54600A. This is the pinout of the J900 connector ribbon cable AT THE MONITOR.
As we can see, they did not repurpose the BRIGHTNESS pins; they just circumvented them, then shunted pin 12 to ground through a 10K resistor. The only video is the ~0.7V P-P signal at pin 11. I drew out the circuit of Q960-961, thinking there might be some mixing going on in there, but nothing I could find coming from anywhere but the MegaZoom ASIC. And I'm guessing that the "additional information" I thought I say on the VSync was just a byproduct of both signals being generated in U411, a Quad OR Gate.
This is Q960-Q961 in the video out to J900-Pin 11 on the 54621D. Q961 base goes to the Megazoom IC through a 50Ω resistor. Q960 Emitter goes to J900 pin 11.
@factory: Based on this, I'm guessing your surmise that the only real difference between the two is that 74S03 vs the transistor amplifier on the neck board was dead on... but that makes this a bit of a bastard monitor with that weird HSync frequency.
Yes, of course I tried connecting it to a VGA monitor, but all the ones I have are relatively modern, and while they do
see the signal if both HSync and VSync are present, all I get is "Mode Not Supported". Maybe if I had some prehistoric analog unit, or the scan converter Robert was talking about, yeah. Or maybe it would sync, but the video would be out to lunch.
At this point I'm inclined to try and figure out the differences between the two neck boards next. Maybe try and ghetto-fab something
just for gits & shiggles.
mnem
*toddles off to ded*