Epson HX

[jacojove]

I recently got this hx20 and recapped it. But, now the screen does not work properly, you can only see it from a really sharp angle and its inverted. I tried the view angle control but it does nothing. What should I do?

 

[Jeff_Birt]

You are probably missing the -5V to the LCD. Every HX-20 I have recapped has has 2-3 traces eaten away near leaking capacitors; guessing that might be the issue you are having.

 

[jacojove]

thanks I will check

 

[jacojove]

none of the capacitors were leaking and I checked the voltages and Im getting 3v to the lcd

 

[Jeff_Birt]

I have never seen an HX-20 where the caps have not leaked. Consider yourself very lucky.

The LCD runs on 7V, not 3V. It develops this from the +5V supply using IC 2C. (This LCD does not use -5V I mentioned earlier, many do). I have seen trace corrosion from caps in this area multiple times. The -8V for RS232 is also generated in same general area.

The LCD signals also run through keyboard flex cables and the carbon flex tape KB cables love to break.

 

[Martin Hepperle]

... leaking capacitors in that context does not mean that any leakage is visible from looking at the installed capacitor ... usually the soup has dried up and the white crystals can only be seen after desoldering a cap and sometimes also by the dull grey solder joints ...

 

[jacojove]

I have never seen an HX-20 where the caps have not leaked. Consider yourself very lucky.

The LCD runs on 7V, not 3V. It develops this from the +5V supply using IC 2C. (This LCD does not use -5V I mentioned earlier, many do). I have seen trace corrosion from caps in this area multiple times. The -8V for RS232 is also generated in same general area.

The LCD signals also run through keyboard flex cables and the carbon flex tape KB cables love to break.

where would IC 2C be?

 

[jacojove]

I have never seen an HX-20 where the caps have not leaked. Consider yourself very lucky.

The LCD runs on 7V, not 3V. It develops this from the +5V supply using IC 2C. (This LCD does not use -5V I mentioned earlier, many do). I have seen trace corrosion from caps in this area multiple times. The -8V for RS232 is also generated in same general area.

The LCD signals also run through keyboard flex cables and the carbon flex tape KB cables love to break.

where would IC 2C be?

 

[Jeff_Birt]

where would IC 2C be?

The HX-20 tech manual is a great resource: https://archive.org/details/epson-hx-20-tech-manual-ocr_202011

About the 10 minute mark in this video I point out where that IC is and about the 20 minute mark I show testing all voltages at the voltage test points:

 

[jacojove]

The HX-20 tech manual is a great resource: https://archive.org/details/epson-hx-20-tech-manual-ocr_202011

About the 10 minute mark in this video I point out where that IC is and about the 20 minute mark I show testing all voltages at the voltage test points:

Thanks

 

[Psyberducky]

By complete coincidence, I recapped my HX-20 this week and ended up with the same problem. I used Mr. Birt's excellent Capacitor Map, cleaned the corrosion with a fiberglass pen and IPA, added flux and solder and then removed it with my Hakko FR301, and the new caps went in fairly smoothly.

Unfortunately, while the unit was working perfectly before, I'm now only getting 3-6V to the screen. There's a lot of variation for the first 1-60 seconds (depending how recently I turned it on), and then it just drops to a wobbly 3.5-3.8V and stays there. I hooked up VLD (Voltage going to LCD) to the scope so I could keep an eye on it, and it's not smooth at all. All continuity checked out OK, but just in case there was high resistance from corrosion or some other trace issue, I temporarily jumpered each trace with alligator clips on the LCD power circuit, while keeping an eye on VLD (which didn't improve). I spent way too long studying the schematic, and checked each of the pins of IC "2C" (it's getting 5.3V from the transistor Q8, and it's outputting a pulse, although I don't think it's outputting the correct voltage because by the time it gets to C2 (not to be confused with 2C) it's already down to 0.8V. I don't have a Chip Tester to test "2C" with, or a spare TC4049 - for reference, the modern equivalent is CD4049UBE. I did try reflowing the pins on "2C". I also changed diodes D5/D6/D7 with 1N4148 which didn't make a difference, and even ended up replacing C1 and C2 a second time just in case. Interestingly, removing C1 entirely doesn't make a difference at all. The area around it was corroded and I cleaned that up, but as I said earlier, continuity is fine for everything nearby, and I never saw any evidence of a failed trace.

In the end, I just bridged + of C6 to Pin 3 (5V) of CN7 (Expansion Connector). This turns on and off with the power switch, and it forces the screen to 5V, which is at least readable, and still adjustable with the contrast knob. It's not the ideal 7 Volts, but at least nothing appears to be overheating or shorting.

 

[Jeff_Birt]

The corrosion damage can be devilishly hard to find. I did a Tandy Model 102 this weekend wish had a tiny amount of corrosion on PCB under battery box. This is pretty common. I cleaned it out, ohmed out signals and all was well. After a few hours running a burn in program the LCD started flickering. Found the problem was a corroded via. I had previously ohmed it out and it tested. Measuring the voltage when the T102 was running showed a fluctuating voltage though. Sometimes even flexing the PCB with your meter probes can cause a bad connection to test good. Very frustrating.

 

[Jeff_Birt]

Well, I actually found an HX-20 without damage from leaking capacitors! After pulling the caps 2-3 showed signs of starting to leak but no damage to PCB was done. The NiCad did some damage though due to the corrosion seeping down the leads to the PCB. Obliterated pads for battery connector, a few small traces were discolored but not eaten through and I had to pull the charging jack to soak it in citric acid because it was corroded (but it cleaned up.)