VT132 - Stand Alone troubleshooting

[Mark Cummings]

I've just finished building the VT132 Stand Alone kit. I plan on using it as a front end for my IMSAI 8080 replica initially. Maybe I am missing something, but initial tests don't look good to me.

Before I connect it up and configure it for serial terminal use I thought I'd test it with just my IBM Model M PS/2 keyboard and a Dell VGA LCD monitor, assuming it will do something even if it's only the setup menu's.

Unfortunately I am not getting any video at all, the monitor keeps switching to standby and just wondering if I should get something even if it's not connected to anything else (such as a DTE or DCE device).

To power my VT-132 Stand Alone I am using a +12V 1A regulated SMPS wall wart plugged into the DC jack (later I will replace it with a lower 7.5VDC - 9VDC unit to reduce heat in the 7805 regulator). I don't have any jumpers on the board except for the patch strip with jumpers all straight through (1-2, 3-4, 5-6, 7-8, 9-10, 11-12).

I've measured the DC voltage at the +5V regulator and getting +12.2VDC IN and +4.94V OUT. I've also measured +3.31VDC out of U2. For current consumption I've tested this with and without a keyboard on my variable bench supply at 7.5VDC, 9VDC and 12VDC. With a keyboard connected it uses about 150mA and without it uses 32mA at all voltages, which doesn't seem excessive to me.

I've also tested for shorts between pins on the ESP32 and all the terminal and modem connectors and didn't see any shorts.

I've also tested it with two different VGA LCD monitors, but results are the same.

Do I need any other jumpers in place to get the setup screen to work, or do I have to connect it to a DTE device before it will work.

Is there any other thing I should test/do before I ask Dave for advise.

thanks,

Mark.

[Mark Cummings]

Continuing with the fault investigation, I am assuming that I don't need a terminal or modem attached and just have my PS/2 keyboard and 12V 1A PSU attached. I haven't bothered with connecting a VGA monitor since there is nothing coming back to the ESP32 in that direction.

I connected up my CRO and measured the following on the VGA pins at the 6-pin header:

pin6. H-Sync = 30mVDC

pin5. Red = 90mVDC

pin4. Green = 50mVDC

pin3. Blue = 1.5VDC

pin2. V-Sync = 3.3VDC
pin1. GND

I'm not getting any video signals on any of the VGA pins, just some noise <50mVAC. At some stage during my tests I managed to break CH1 on my CRO so I retested on CH2. 

I also tested the PS/2 keyboard signals to some degree and during/after keypresses I get some data at the junction of the voltage divider resistors R1-R3, and R2-R4 as follows:

R1-R3 = p1. K_DATA = 4.4Vp-p digital signal (several active low pulses depending on key pressed)

R2-R4 = p5. K_CLK = 4.4Vp-p digital signal (11 active low pulses)

I have also observed that the 3 keyboards LEDs light up at power up, but then extinguish shortly afterwards.

At this stage I am inclined to think that the ESP32 is not running, but not knowing anything about these ICs I don't know what to try next. I can't see a clock signal or anything else I can test other than the power supply pins and they look fine.

I have reached out to Dave @ The High Nibble for advise.

any thoughts?

Mark

[Mark Cummings]

I just thought I'd provide an update on where I am up to after trying a few suggestions from Dave.

The first suggestion was to capture the boot log and this was done by connecting the VT132 and FTDI adapter (which I had just purchased for this task and possible future use) up to my PC running Ubuntu Linux and then starting a PuTTY: session. The boot loop sequence was apparently indicating that it wasn't able to boot the firmware. I've attached a screen shot below.

The second suggestion was to test pins 16-23 for shorts to each other or shorts to ground. most of these pins are apparently for the internal flash memory. I found that I didn't have any shorts to ground or each other, which was also the case when I first soldered the ESP32 to the board. I read from another builder here that suggested to check all the pins for shorts to adjacent pins and I agree that is a wise step due to the tight pin spacing. What I did find was that 3 of my pins were not as good looking as I would like, looking a bit dry, and these turned out to be the GND pins pin 1, 15 and 38 (which I didn't know until I looked up the spec for the VT132-WROVERE). I had used a fine point iron and those joints were not as good as the others because the ground plane had sucked the heat out of my tip making for a poor connection. I re-soldered those connections and re-tested for shorts again and there were none. I then reconnected the VT132 to a PuTTY: session on my PC running Ubuntu and still got the same boot loop that I had before.

The third suggestion which I just completed after a lot of headaches installing ESP-IDF software in Linux (probably not needed) on two separate computers was to re-flash the ESP32. I am new to this, so it's been a wild ride figuring this out, but I managed to get there eventually using my Raspberry Pi to do the flash. It now looks like the ESP32 cannot detect it's internal flash memory (as Dave had indicated) and that doesn't look good to me, but I'm open to further suggestions or tests. 

Mark

[Mark Cummings]

for those following along, Dave suggested a last ditched attempt at easing the flash on the ESP32, but that also ended up in a fatal error. So next step was to remove the ESP32 and even though I have a fair bit of soldering experience and equipment, this was the first time I've tried to remove a largish SMD IC. I used LMP (Low Melt Point) solder which melts at 125degC according to the label on my packet, by adding it on top of the ESP32 pins in a big blob. I used a generous amount of liquid flux and added the LMP solder with a large chisel tip iron at around 250degC. I had already removed several surrounding components to keep the area clear. I then used my rework station at a low 200C to heat up the LMP solder both sides of the ESP32 until it let go of the PCB. The masking tape is just to keep the excess heat off the surrounding components. Maybe I should have used Kapton tape or foil tape, but this worked OK at 200C.

The hardest part was then to remove the LMP solder as it doesn't suck up very well using solder wick. Instead I used the large chisel tip and dragged it along the pad rows to draw the solder to the end and wipe it off the tip. I then added a liberal amount of 60/40 solder to mix in with the remaining LMP and that was removed with solder wick more easily but at a higher 320degC. I repeated this a couple of times. This seemed to tarnish the pads (possibly the Bismuth burning in the LMP remains). Then I used an ink eraser (RS part number 216-2706) to carefully polish down the pads but not so far as to remove the tin plating. After that it was just a lot of flux to be cleaned off with IPA. The end result looks very clean and ready to accept a new ESP32. I've attached a few images of the job.

Mark