• Please review our updated Terms and Rules here

TRS-80 Model 4P Monitor Replacement with LCD VGA Monitor

ebp

Member
Joined
Jun 1, 2020
Messages
10
Location
Sparks, Nevada, USA
Last year (2020), I bought a TRS-80 Model 4P with a black screen off eBay. The LED on the disk drive would light momentarily when powered and reset, but there was no image. After fruitlessly adjusting the brightness and contrast knobs in both directions, I decided to open it up and have a look. I spent a few days trying to diagnose the monitor. I checked power supply voltages. I verified the video connector had a 60 Hz pulse on *VSYNC, a 15.86 kHz pulse on HSYNC, and 0-5V swings on VOUT. I visually inspected the video board; no bad caps, no broken resistors, nothing unusual. I tried replacing the high voltage drive transistor on the video board, but that did nothing. There was a healthy "snap" when I discharged the tube, so high voltage seemed to be fine. I don't remember a glow inside the rear of the tube. I concluded the tube was probably bad, but I needed to verify that the motherboard was working or further effort on the tube, or another option, would be in vain. I decided to see if I could get the video signal on a monitor. That's when I learned that a VGA monitor won't do that.

Since the video signals seemed to be equivalent to an IBM CGA display card (with only one color and no intensity), I figured I could hook them up to a LCD VGA monitor. Of course, this usually doesn't work because VGA is analog RGB (0-0.7V) and CGA is digital RGB (0 and 5V TTL). But more importantly, most VGA monitors won't sync at 15.7 kHz, the horizontal scan rate of CGA and the Model 4 monitor. VGA monitors usually only go down to about 31.5 kHz. There are older VGA monitors that will sync down to 15.7 kHz, but a modern LCD VGA monitor? Not likely. Since I didn't have a VGA monitor that could sync at 15.7 kHz, there seemed to be only two options. Either use NTSC (composite video signal), which is also a 15.7 kHz horizontal scan rate, or get an up-scaler to double the horizontal scan frequency to 31.7 kHz. My LCD VGA monitor has a composite input so I decided to pursue that first. I also ordered a Gonbes 8200 up-scaler. The Gonbes is inexpensive and composite video quality is poor, so I figured if the motherboard was working, the up-scaler would be a better long term repair option.

That was all in the summer of 2020. At that time, I failed to craft a circuit that would mix the model 4P's video signals, HSYNC, VSYNC*, and VOUT, to produce a composite signal that would display on my LCD monitor. I also couldn't get the Gonbes working (it just said, "no signal"). So I stuck the whole mess in the corner and forgot about it until a few weeks ago when I decided to try again. In the interim, I had acquired a CRT monitor with composite input. CRTs are a bit more forgiving in that they don't blank the screen (well, mine doesn't) like an LCD when the video signals are non-compliant. A CRT will just roll vertically and horizontally when things aren't right. This is helpful because everything doesn't have to be perfect and you can see progress. Eventually, I found a circuit that would mix the signals and display on the CRT. (I XOR'd the syncs and mixed the result with VOUT using resistors.)

20210301_215054-sm.jpg20210301_215059-sm.jpg

SUCCESS!! With the motherboard appearing to work, I downloaded the TRS-80 Model 4 Montezuma CP/M image from the Internet and wrote it to a 5 1/4" SSDD floppy disk to see if the board would boot into CP/M and switch to 80-column mode. That worked too! The CRT showed significant tearing on the top line, but it worked!

The next step was to get the composite signal to display on an LCD VGA monitor (DELL 2007FP) using its composite video input. The circuit that had worked on the CRT however did not work on the composite input of my LCD VGA monitor. After a lot of Internet searching and experimentation, I found a blog on "Cloning a Tandy TRS-80 Model 1" which provided schematics of a composite video signal circuit. Those schematics, and the original Tandy schematics, lead me to inverting HSYNC and AND'ing it with VSYNC*. Mixing this CSYNC* signal with the VOUT using a couple of resistors finally produced a composite video signal that the LCD VGA monitor would display.

Now that I had a solution for CSYNC* (CSYNC* = HSYNC* AND VSYNC* = HSYNC NOR VSYNC), which I knew worked for composite video, I tried it with the Gonbes and it worked too! The schematic below is what finally worked with the Gonbes and my LCD VGA monitor.

Mod4-Gonbes-8200.png20210327_112436-sm.jpg20210327_112428-sm.jpg

I had purchased a GGLABS CGA2RGBv2 adapter because it has the circuitry to combine HSYNC and VSYNC and it also buffers the video signals to drive them into a 75 ohm load. I thought that the board might implement some additional secret sauce that I didn't understand, but it could not combine HSYNC and VSYNC* into a usable signal either. VSYNC must be inverted so that the adapter will produce a CSYNC* signal that is acceptable to the Gonbes.

Mod4-GGLABS-Gonbes-8200.png20210327_121909-sm.jpg20210327_121902-sm.jpg

I don't perceive a difference between my circuit and the GGLABS results so I think will prototype a board with my circuit and save the GGLABS adapter for some future CGA or EGA color project. The next step is to mount the 8" LCD VGA monitor (EYOYO 4:3 1024x768) I have into the cabinet. I'll post a follow-up when I get a bracket made.
 
Update

Update

The promised bracket pics:
20210328_213147.jpg20210328_214024.jpg20210328_214319.jpg

Mock-up of monitor in cabinet:
20210328_214245.jpg20210328_214431.jpg

Still waiting on right-angle VGA connectors to connect the monitor to the up-scaler.
 
Back
Top