B&F Enterprises "Digital Clock" flakes out

More
7 years 6 months ago #4856 by Dr. Optigan

wise65vw wrote: I don't mean to be condescending, but is there a chance the diode was put in backwards? Just a thought.


Best as I can recall, the new diode (the one at the bottom) is installed in the same orientation as the one it replaced. As for whether one or both of the original diodes might've been installed backwards, I cannot say for sure, but the fact that the clock is running at all (just passed the 18-hour mark), and is still in synchronization with other line-powered clocks, would seem to be evidence that something is OK.
-Adam

Please Log in or Create an account to join the conversation.

More
7 years 6 months ago #4859 by Ty_Eeberfest

Dr. Optigan wrote: Yeah, I definitely see the similarities. Any idea about the history of this design? Guessing TI had a hand in it, given the diagram in post 7, but the differences make me wonder.


Well, aside from being pretty sure the one design comes from a TI appnote, no. I'd imagine more than one person came up with roughly the same design independently since using 7490s + 7492s + some kind of reset scheme would have been an "obvious" solution with the technology available back then. Put it this way: I was active in the hobby already back then (although not much more than a kid) and if somebody said I had to design clock logic with TTL I'd have ended up doing the same thing.

The parts of the design that don't strike me as obvious - and it's interesting to note that the designs differ in these areas - are inter-stage connections and setting (related), derivation of the 1Hz master pulse, and handling of wrap-around & reset.

Probably irrelevant to your specific problem, but to clear up something that was said above about a "7490 used for both hours and ten-hours digits": 7490 and 7492 are "split" counters. ClkA and QA comprise a divide-by-2 and, independently, ClkB and QB - QD comprise a divide-by-5 (7490) or divide-by-6 (7492). So if you design your clock such that the 7492s are used as divide-by-6 (not 12) counters - as seen in the TI design - you have ClkA and QA left over; essentially a T flip-flop is left for other uses. TI design uses the leftover part of the ten-minutes 7492 to run the ten-hours digit (it's a 12 hour clock so a flip flop is all that's needed - the ten-hours digit can only ever be 0 or 1). It also uses the leftover part of the 10-seconds 7492 as the AM/PM flip flop.

By contrast, the hand drawn circuit is using the 7492s in a divide-by-12 configuration and getting away with it by ignoring the MSB (QD).

As wise65vw mentioned, the 7410 is for 13 o'clock / 24 o'clock detection and reset.

To get back to your intermittent glitch-fests, I don't have anything new to add. At the risk of stating the obvious I agree with looking closer at the 1Hz master clock generator's quality and its performance when warm or when hit with line noise or other "challenges" - because it seems like whatever is going wrong is affecting all of the counters. I also am going to stick adamantly to my original premise that everything is moot until all flux residue is gone and all solder joints and traces are painstakingly inspected and repaired wherever dubious.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
7 years 6 months ago - 7 years 6 months ago #4860 by Ty_Eeberfest
By the way - in case you didn't already discover the 3-thread conversation, the hand drawn clock does not work as drawn. If nothing else, the 13 o'clock reset scheme is incorrect and it acts as a 13-hour, not 12-hour, clock as drawn.

Reference:

www.tubeclockdb.com/forum/Builders-Forum...e-clock-problem.html

www.tubeclockdb.com/forum/Builders-Forum...from-60hz-mains.html

www.tubeclockdb.com/forum/Builders-Forum...000-clock-Help!.html

Look into it later when the dust is clearing off the crater.
Last edit: 7 years 6 months ago by Ty_Eeberfest.

Please Log in or Create an account to join the conversation.

More
7 years 6 months ago #4863 by Dr. Optigan
Ugh, another update.

The clock ran correctly at least up to the point where I went to bed. Woke up this morning, and the time reading was completely off. Came back to it later, and the display was showing nothing but zeros. Back to the drawing board, methinks... :(

At this point, I'm starting to wonder if this clock was ever reliable. The build quality, combined with the fact that all but two ICs were replaced at one (or two) points, gives me pause with regards to whether it's possible to get it working reliably at all. I'm not entirely ready to call this thing a basketcase, but I'm certainly starting to get that inclination. :pinch:

My preference would be to make a new circuit board, salvage the tubes and some of the parts from the existing one, and build a new clock within its case, but I think that would destroy its historical value. Other ideas involve a more thorough cleaning of the board, attempting to simulate possible trouble spots like line noise, and possible elimination of the IC pin holders. As usual, any ideas? I'm rather frustrated with this thing... :unsure:
-Adam

Please Log in or Create an account to join the conversation.

More
7 years 6 months ago #4864 by Ty_Eeberfest
Are all - and I mean ALL - unused reset pins on the 7490s and 7492s tied down to ground? They should be. A floating reset pin is prone to all kinds of randomness.

I will keep repeating this until you acknowledge it and hopefully do it: All this discussion is moot until you not only clean the board but inspect all solder joints under magnification and reflow as needed.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
7 years 6 months ago #4867 by Dr. Optigan

Ty_Eeberfest wrote: Are all - and I mean ALL - unused reset pins on the 7490s and 7492s tied down to ground? They should be. A floating reset pin is prone to all kinds of randomness.


I'll check that next time I have it open, and once I find the datasheets for these ICs. At the moment, its symptoms aren't quite so random: the display steadily shows all zeros, with the hours 1 flickering in and out. Pressing the hours/minutes/seconds buttons with the set/run switch held at set has no effect. Methinks something has gone seriously wrong...again... :(

Ty_Eeberfest wrote: I will keep repeating this until you acknowledge it and hopefully do it: All this discussion is moot until you not only clean the board but inspect all solder joints under magnification and reflow as needed.


Acknowledged; I'll attempt a more thorough board cleaning next time I have it open (would flux remover really do a noticeably better job than 91% rubbing alcohol?), and I'll make an attempt at looking for cracked solder joints. What I'm most worried about are the wires going from the board to the set switches and such, for they look like the solder joints for them are especially heinous, not to mention some of the trace holes looking like they've seen better days. Hoping this thing is repairable, but I definitely have my doubts... :unsure:
-Adam

Please Log in or Create an account to join the conversation.

Moderators: AccutronTy_EeberfestIan
Time to create page: 0.179 seconds

Search

Tube Suppliers

Go to top
JSN Boot template designed by JoomlaShine.com