Testing MC6821 PIAs with a potential hack using the XGecu model T48

RiddledTV

Well-known member
Joined
Aug 11, 2005
Messages
1,719
Reaction score
713
Location
Kansas
Background, I've been working on a Williams System 11 F-14 Tomcat board, which has been had quiet a few bad components. Several of the MC6821 PIAs have been bad, several more that I haven't yet desoldered are probably also bad.

I was looking into testers on the market for these chips and have only found the neoloch tester and a few other pricier options. Today, just by luck I discovered that the more expensive XGecu T56 programmer can test these chips. Since it uses the same software as my cheaper T-48, I was curious why I couldn't test on mine. It turns out that the T-48 lacks the hardware to provide Vcc to pin 20 of the programmer. So I started playing around with my programmer and applying external power to pin 20. I only have a very small sample size of 3 PIAs that I can test, but it identifies my two known bad ones as defective. It identifies the one PIA that I *think is good as normal. I have no idea if I am doing something stupid that could damage the programmer or the PIAs, but it seems like it is working.

If anyone else is feeling particularly brave and has one of these cheap programmers, plus a spare MC6821. Apply 5V external power to pin 20, ground to pin 1, and then run a logic test. I suppose it might be wise to bend pin 20 upward out of the ZIF socket so that you're not sending external power directly into the programmer (I honestly didn't think of doing that until right now).

1735850228337.png
 
my Xeltek SuperproZ programmer can test 6821s. I thought it was capable but I don't know if I ever did. it can do some TTL also. only problem is it's old school with a parallel cable lol
 
The NeoLoch tester is well worth it if you work on a lot of boards and need to test PIA and RIOT chips. Does the 6520, 6820, 6821, and 6532 chips. The Back Bit tester also does a good job at testing those chips. The Back Bit also does a lot of Atari customs, some sound and misc chips so a good tester to have. Just can't trust it for the 680x processors as those test appear to be flawed.
 
Been here, done that. The Neoloch Inquisitor rules!!

 
What's the verdict on the safety of testing 6821's with a T48 programmer supplied with external 5v as Riddled TV describes? And should the 5v pin be lifted from the programmer?
I've got a non working Qix board with all of the six 6821s missing that I want to try to get working. Those PIAs are quite expensive now, so I took a punt on some cheap ones from China (I know...). It would be great to be able to test them with my T48, before putting them on the board, but I don't want to cause more expense by frying my programmer!
Has anyone else tried this or know if this is a good idea.
Many thanks for the post Riddled TV.
 
I don't know what i'm talking about.


With that said, applying 5v to a logic line is likely not going to damage anything, it's at worst going to stick the logic line at high. That's kind of the main risk here, if you accidentally supply the voltage to a logic line, but it's only 5v which is what the logic line is looking for anyways.

I don't know what i'm talking about.
 
I suspect you do know what you're talking about, Lyons Arcade...Thanks for your reply. Definitely worth a try when I get the ICs. I'll post back when I test them.
Many thanks
 
Last edited:
I tried this hack and it works great. The Chinese ICs were all good and it identified my known faulty 6821.
Brilliant.
 
Back
Top Bottom