• Please review our updated Terms and Rules here

AT&T PC 6300/M24 monitor voltage issue?

dkedrowitsch

Experienced Member
Joined
Dec 27, 2006
Messages
102
Location
Boyertown, PA USA
So I recently purchased a pretty clean PC 6300 and matching CRT 313/H monochrome monitor. The motherboard had some slight battery damage but I've got quite a bit of experience dealing with that thanks to a number of Amiga repairs I've done. I mitigated the leak with some mild acid (Tarn-X is my preference) then replaced the charge current limiting resistor with a diode so I could safely install a coin cell holder. Besides that, the system pretty much looks and works like new--with a very quiet and healthy ST-225 being the cherry on top.

The monitor however has been a different story. At first, besides a somewhat soft focus it worked OK. But after a few minutes it developed an annoying horizontal scan jitter. Here's a short video of the issue:

https://youtu.be/JkAwy-igt6E

I went ahead and replaced every electrolytic with new 105c rated parts from Nichcon and Panasonic which restored the focus, but the horizontal jitters persist. They temporarily cleared up after the monitor was on for a few hours, but returned to stay after the next power/thermal cycle.

The cause seems pretty obvious, with my scope triggering on H-sync the 12v power source to a TDA2593 horizontal drive IC looks like an EKG readout. There's a sharp 200mv dip that coincidences with the leading edge of the drive pulse to the HOT which is then followed by a 200mv bounce as the dip recovers. These ugly transients must be the cause.

At first I assumed there might still be some decoupling issues, but then I noticed the 15 volt supply from the PC 6300 video board is actually 13.4-volts under load. 13.4 volts seems hardly enough for a 12-volt linear power supply to keep out of drop-out under heavy current spikes such as when driving a CRT yoke and flyback. And this particular CRT 313/H is a 12-volt chassis with a discrete linear power supply (zener reference, adjustment pot, bipolar output transistor, etc) that's powered by the computer PCU's 15-volt "CRT" output.

So I my questions are: Is it normal for the 15-volt "CRT" supply of an Olivetti M24/AT&T 6300 to sag down into the mid/low 13-volt range under load? If not, I may need to dig into the power supply.

Would anyone happen to have a service manual for the CRT 313/H? It's a Hantarex monitor--a sticker on the HOT heatsink says so...but unfortunately no model number can be found anywhere. It's a very different beast from the 1231 chassis most M24 CRTs seem to be based on.

Thanks!
Dieter
 
Thank you! I appreciate the links. I spent quite a bit of time comparing that 1231 manual to my 313/H and while there are some similarities, there are also some pretty significant differences.

For example the 1231 document describes an on-board SMPS for line voltage down to 12v and 50v, but my 313/H instead uses a very simple 12V linear regulator and all the deflection circuitry is of a different, lower voltage design. Also, all of the alignment pots are in different locations with different designations than the 1231. Even the neck board is different and lacking a contrast preset pot the 1231 has.

So with that being said, I guess there's no known service manual for the 313/H?

As an experiment I plan to test mine using a high-current variable lab supply to see if providing a solid 15-volts will stiffen up the 12V rail in the monitor. Hopefully that will clean up the horizontal jitters and confirm my suspicions the power supply in my 6300 is the cause. Otherwise I've got some more troubleshooting to do. <EDIT: Just as tipc recommended while I was writing this post>

But I am curious. Would anyone be willing to pop the lid on their M24/6300 and measure the voltage at the "CRT" power cable where it plugs into the display board, while powering a CRT?

Cheers!
 
Last edited:
Well, using an external power supply didn't make a difference.

I'm waiting for a replacement horizontal deflection driver IC to get here to see if that helps, but I wonder if there isn't a bad film capacitor somewhere...
 
I think this problem has been resolved!

I don't know if it was the additional decoupling I added with some 1206 SMD ceramics on the back of the PCB, removal of old varnish looking flux around the flyback area of the PCB, or just the sum of all the hours it's been powered on this week burning tin whiskers out of an old component-- but as of this afternoon the monitor has been operating flawlessly with a nice solid picture.

The focus is a bit soft compared to the amber phosphor monitors I have, but I think that might just be the green CRT "blooming" more than amber does.

Here it is still unhooded until I'm confidant it's good to go.

IMG-5875.jpg
 
Congrats! The JPEG isn't good enough to figure out if the monitor is actually blooming or not, but from a distance it seems just fine for 320x200 work. While it's still open, you should run the 6300 diagnostics to show off 640x400 test patterns and make any final adjustments.
 
Thanks! Yeah I'm actually trying to create the diagnostics floppy but the image file is a slightly different "format" than the OS images and won't write correctly. The file is 8K too short and yet the directory listing the data looks to be shifted one byte too far forward.

Is there a special tool needed to write that data to disk? I'm using Mike Brutman's trusty dskimage.com on my V20 clone PC to write the 5-1/4" disk, but turns out a corrupted mess. All the other OS disks write just fine. Disk Copy Fast is another DOS tool I use for reading/writing disk images but that won't even accept the file because it's 8K too short.

In the meantime, here's a higher res picture. Looking at this it actually looks pretty good to me. In person next to my amber monitored V20 clone, the other monitor has noticeably sharper text. I'm probably just being picky...


IMG-5874.jpg
 
Thanks! Yeah I'm actually trying to create the diagnostics floppy but the image file is a slightly different "format" than the OS images and won't write correctly. The file is 8K too short and yet the directory listing the data looks to be shifted one byte too far forward.

Did you download it via FTP in ASCII mode by mistake? Ensure Binary mode and try again. They are 100% normally-sized images.

Edit: Never mind, it seems the images on the 6300 shrine might not download correctly? I'll investigate. ftp.oldskool.org is an alternate location.

Edit2: I just tested the 6300 shrine .zip files and they're fine, so not sure how you downloaded them.
 
Back
Top