An update:
My initial indication of a problem was glitches in the display, such as "_"
characters showing up on parts of the screen where they didn't belong.
These glitches were stored in video RAM and were thus visible until the
screen image in RAM was updated.
My guess for the most likely culprit was power supply noise, and indeed
there is a LOT of noise on the supplies, but it is not line noise. In fact,
it is nearly all switching noise from the digital circuitry. If I hold the
microcontrollers in reset, the supplies get very quiet. When the
microcontrollers are running, there is significant noise on all the
supplies:
-15V rail: 1.2Vpp
+5V rail: 1.1Vpp
+12V rail: 0.5Vpp
+15V rail: 0.7Vpp
The noise on each of the supplies is all the same noise (triggering on one
channel produces a solid lock on the others). It is asynchronous to the
14.xxx MHz pixel clock of the NS405, so I believe it is all from the 8032
(12 MHz XTAL). The lowest observable frequency in the noise is around 2 MHz
(only looking in the time domain at the moment).
I've isolated this down to just the main board; the CRT, front panel, and
collector supply are out of the picture.
The main board contains the 8032, NS405, base/gate drive circuit (DAC), and
emitter test circuit (ADC).
Supplies are all linear, using 78xx and 79xx regulators.
At this high of frequency, electrolytics aren't contributing much to the
bypass, right? Each rail only has 220uF after the regulator. I've checked
half of them, and while they've aged they aren't trash yet. Anybody seen a
ceramic bypass cap go bad?
Dave Casey
On Thu, Sep 24, 2020 at 1:11 AM Dave Casey via groups.io <polara413=
[email protected]> wrote:
I'm not so sure they are. DC leakage current is well below 100uA at 20V.
I'll check them again at 40.
When I say the cans are domed, they may have been made that way as these
are sealed at the bottom and don't have the usual "fault lines" in the
domed end.
I will also check ESR. Replacements are standing by, but I'm not convinced
these were the source of the noise, especially since it's not line noise.
These are all pre-regulator caps, so I'll check the smaller ones after the
linear regulators to see if anything is suspicious there. Low capacitance
on the output and some fast switching in the video circuits could be to
blame. I should also see if the noise is in sync with the video.
Dave Casey
On Wed, Sep 23, 2020 at 9:43 PM Richard R. Pope <mechanic_2@...>
wrote:
Dave,
Those electrolytics are defective. They have been running hot. Time
to replace them. I hope that this helps!
GOD Bless and Thanks,
rich!
On 9/23/2020 9:39 PM, Dave Casey wrote:
I spoke so highly of my 571 last month. When I recently went to use
it, I
was getting some artifacts in the display so I have set out to
investigate.
Unfortunately you can tell this wasn't designed in Beaverton because
you
have to remove the CRT to get the main board out. There are some large
Roedenstein electrolytics used for supply smoothing that have a bit of
a
dome to the end of them. Not sure if these were flat when new, but they
don't appear to be leaky (neither physically nor electrically). There
is
significant noise (100-200mVpp) on the supplies, but it's not line
ripple.
It appears to be around 2 MHz and is two similarly disruptive pulses
around
200ns out of phase.
In the process of disassembly, I have pulled and saved the three EPROM
images. Two run the 8032 controller, and one runs the NS405 "terminal
on
a
chip" that is used to draw the display and printer output.
I haven't exhaustively perused the images, but the graphics ROM has
some
ASCII easter eggs I noticed. After the usual Tektronix copyright
statement
is "Written by: Folkert ten Brummelhuis (started). Sjoerd J. Punter"
and
soon after "[Makke yn Frysl?n]" which is Frisian for "Made in
Friesland",
the province in which Heerenveen is situated. Later in the ROM the
phrase
"Joke Mirjam Esther" appears twice, about (but not exactly) 64 bytes
apart.
I believe these are three given names commonly used in the region,
perhaps
the names of those endeared to one or both programmers.
Dave Casey
On Fri, Aug 21, 2020 at 1:31 PM Dave Casey via groups.io <polara413=
[email protected]> wrote:
I stand corrected; can confirm, my S/N also starts with H. Thanks!
Dave Casey
On Fri, Aug 21, 2020 at 5:12 AM zenith5106 <hahi@...> wrote:
On Fri, Aug 21, 2020 at 05:49 AM, Dave Casey wrote:
Built by Sony-Tektronix.
No,not Sony-Tek. It was built by Tek Netherlands (SN begin with H)
and
as
far as I remember it was also designed by them.
/H?kan
