Topics

DC503 Not Working


Stephen
 

I always discover new stuff...
It seems the pin, I think, 12? of U354 Is connected to nothing.... But I pretty sure it should be...

See pict: https://groups.io/g/TekScopes/photo/250733/0?p=Created,,,20,2,0,0

There is continuity between these to points, but nothing is connected...
I’m not sure if this is related to my display issue, but I’m sure it’s not right.
Does any have one of these and tell me what should be there, if anything?

Thanks.


Roger Evans
 

My DC503 also has nothing connected at the lower of the two points in your picture. On my board there are four of these holes in a row with no wires or components attached, partky hidden by the shaft of the function switch. They are 'vias' used to transfer a PCB trace to the opposite side of the board for routing purposes. There is another row of 7 similar vias towards the connector end of the same board.

Roger


Stephen
 

On Sat, Aug 1, 2020 at 03:34 AM, Roger Evans wrote:


My DC503 also has nothing connected at the lower of the two points in your
picture. On my board there are four of these holes in a row with no wires or
components attached, partky hidden by the shaft of the function switch. They
are 'vias' used to transfer a PCB trace to the opposite side of the board for
routing purposes. There is another row of 7 similar vias towards the
connector end of the same board.

Roger

Interesting. The schematic shows pin 12 connected to a switch....


Stephen
 

I got to working on bit on this unit today again. I rechecked all the voltages, and they’re all good.
For good measure, I replaced all the tantalums on the lower board. 2 with electrolytics, the 15uF/50V and the 6.8uF, which I couldn’t find. And of course nothing has changed, I yet have to find time to modify the JAMMA extender. After all that has been changed, there is close to no improvement...


Harvey White
 

Once you get the extender, check the strobes on each of the digits.  Then check the source of those strobes.  Then check the chip that's driving those (likely) transistors.  A typical way is to have a counter driving a decoder which then drives the digits individually.  You may find that this counter does all of that in one chip.  If it's a complicated chip, then look for a reason (if no strobes) that the chip is being inhibited.

Anything where digits freeze or are irregularly displayed indicates the scanning electronics.  Anything where there's a missing segment on ALL displays indicates the common 7447 decoder or the like.  Anything where you have one segment on one display missing tends to indicate a problem on that display.

Even if the display is wonky, you might see (and perhaps should see) digits indicating counting.  The counter is very sensitive to the input settings even on a good day.

You're going to need the JAMMA extender.  My research suggests that if you have all the lower pins connected on both sides, equivalent to 1 to 13 when you turn the extender upside down, all you need to do is mark both connectors promently, and reverse the connectors.  On the JAMMA connectors I saw, pin 7 was missing on both sides.  No other pins were missing, which suggests the reversal.

I wouldn't do much more until you can get that counter out on an extender.  Make the extension wires long enough to get the plugin at least one plugin length away from the front of the frame.

Harvey

On 8/4/2020 5:59 PM, Stephen wrote:
I got to working on bit on this unit today again. I rechecked all the voltages, and they’re all good.
For good measure, I replaced all the tantalums on the lower board. 2 with electrolytics, the 15uF/50V and the 6.8uF, which I couldn’t find. And of course nothing has changed, I yet have to find time to modify the JAMMA extender. After all that has been changed, there is close to no improvement...



Stephen
 

On Tue, Aug 4, 2020 at 11:17 AM, Harvey White wrote:


Once you get the extender, check the strobes on each of the digits. 
Strobes???


Harvey White
 

Vocabulary thingie......

A pulse, likely sequential, is (at least at on time) called a strobe.

Take a display with a common anode for all the segments.  It needs to be connected to VCC (power) to light.  However, there's a limit where the display continuous power per segment is exceeded. This is not a good design.  However, if you pulse the LED (segment) with more current than the continuous rating, but for less time, then the display survives.  Due to a quirk with human vision, it appears brighter.  So what you do is connect the source of each digit to a signal (strobe) which in sequence selects each digit.

The segments drivers themselves are selected so that the current through each segment (assuming common anode with digit strobe to +vcc, and segment driver (through resistor) to ground) is the nominal peak current that they want.

So if all displays were driven in parallel, there'd be one driver chip per display, no pullup transistor to VCC, and  a lot of chips.  If you multiplex the displays (please look this up), then you have a sequential pullup to VCC (called a strobe), a pulldown to ground (called the segment driver, which determines which segments on which digit are on when....) and a current which is more than needed for a single digit (perhaps # digits times current per segment, because you don't know if a segment is on or off, so you do it per segment).

So:

a scanned display turns on each digit sequentially, with more current per segment than permitted (steady state but less than pulsed) and with less chips, but more complex than a few digits. However, as the digit count goes up, the circuit is simpler. Microprocessors love this kind of thing.

Harvey

On 8/4/2020 6:46 PM, Stephen wrote:
On Tue, Aug 4, 2020 at 11:17 AM, Harvey White wrote:

Once you get the extender, check the strobes on each of the digits.
Strobes???



Stephen
 

On Tue, Aug 4, 2020 at 02:03 PM, Harvey White wrote:


Vocabulary thingie......

A pulse, likely sequential, is (at least at on time) called a strobe.

Take a display with a common anode for all the segments.  It needs to be
connected to VCC (power) to light.  However, there's a limit where the
display continuous power per segment is exceeded. This is not a good
design.  However, if you pulse the LED (segment) with more current than
the continuous rating, but for less time, then the display survives. 
Due to a quirk with human vision, it appears brighter.  So what you do
is connect the source of each digit to a signal (strobe) which in
sequence selects each digit.

The segments drivers themselves are selected so that the current through
each segment (assuming common anode with digit strobe to +vcc, and
segment driver (through resistor) to ground) is the nominal peak current
that they want.

So if all displays were driven in parallel, there'd be one driver chip
per display, no pullup transistor to VCC, and  a lot of chips.  If you
multiplex the displays (please look this up), then you have a sequential
pullup to VCC (called a strobe), a pulldown to ground (called the
segment driver, which determines which segments on which digit are on
when....) and a current which is more than needed for a single digit
(perhaps # digits times current per segment, because you don't know if a
segment is on or off, so you do it per segment).

So:

a scanned display turns on each digit sequentially, with more current
per segment than permitted (steady state but less than pulsed) and with
less chips, but more complex than a few digits. However, as the digit
count goes up, the circuit is simpler. Microprocessors love this kind of
thing.

Harvey
Ok, so you suggest I sweep each digit for a signal at its input? At the ribbon cable connections then...


Harvey White
 

That's a start place.  Knowing how it works, you can start at either end and work towards the other.

Harvey

On 8/5/2020 5:07 AM, Stephen wrote:
On Tue, Aug 4, 2020 at 02:03 PM, Harvey White wrote:

Vocabulary thingie......

A pulse, likely sequential, is (at least at on time) called a strobe.

Take a display with a common anode for all the segments.  It needs to be
connected to VCC (power) to light.  However, there's a limit where the
display continuous power per segment is exceeded. This is not a good
design.  However, if you pulse the LED (segment) with more current than
the continuous rating, but for less time, then the display survives.
Due to a quirk with human vision, it appears brighter.  So what you do
is connect the source of each digit to a signal (strobe) which in
sequence selects each digit.

The segments drivers themselves are selected so that the current through
each segment (assuming common anode with digit strobe to +vcc, and
segment driver (through resistor) to ground) is the nominal peak current
that they want.

So if all displays were driven in parallel, there'd be one driver chip
per display, no pullup transistor to VCC, and  a lot of chips.  If you
multiplex the displays (please look this up), then you have a sequential
pullup to VCC (called a strobe), a pulldown to ground (called the
segment driver, which determines which segments on which digit are on
when....) and a current which is more than needed for a single digit
(perhaps # digits times current per segment, because you don't know if a
segment is on or off, so you do it per segment).

So:

a scanned display turns on each digit sequentially, with more current
per segment than permitted (steady state but less than pulsed) and with
less chips, but more complex than a few digits. However, as the digit
count goes up, the circuit is simpler. Microprocessors love this kind of
thing.

Harvey
Ok, so you suggest I sweep each digit for a signal at its input? At the ribbon cable connections then...



Stephen
 

This is a partial copy of a post a posted while trying the extender work.

Something very unexpected and quite puzzling just happened.
As many pointed out, I forgot to connect pin 6A. I just did.
Some of you may have seen my post regarding the DC503 that is not showing more than 1 digit (0) at any one time, if at all, when I power it on. Never ever.

Now that pin 6A is finally connected, not only the DC503 power on normally with the extender, but all digits are there!!! Not when connected inside the TM506 itself!! It’s not counting anything, though.
Everything is not lost, it gives me hope that at least something is showing now...

It’s also missing some segments. If anyone knows if the individual LED’s can be replaced, that would be nice to know of a modern replacement. I found some, but the panels look too thick, I think.


Harvey White
 

I think that someone designed a readout board using more modern (hence more available) displays.  Might have been for the DC509, but the DC503 is what triggered the memory.

Harvey

On 8/10/2020 4:20 PM, Stephen wrote:
This is a partial copy of a post a posted while trying the extender work.

Something very unexpected and quite puzzling just happened.
As many pointed out, I forgot to connect pin 6A. I just did.
Some of you may have seen my post regarding the DC503 that is not showing more than 1 digit (0) at any one time, if at all, when I power it on. Never ever.

Now that pin 6A is finally connected, not only the DC503 power on normally with the extender, but all digits are there!!! Not when connected inside the TM506 itself!! It’s not counting anything, though.
Everything is not lost, it gives me hope that at least something is showing now...

It’s also missing some segments. If anyone knows if the individual LED’s can be replaced, that would be nice to know of a modern replacement. I found some, but the panels look too thick, I think.



Stephen
 

On Mon, Aug 10, 2020 at 10:04 AM, Harvey White wrote:


I think that someone designed a readout board using more modern (hence
more available) displays.  Might have been for the DC509, but the DC503
is what triggered the memory.

Harvey
Interesting. Here on this forum? I’ll try a search then. If you remember anything else...

Thanks Harvey.

PS: Forgive the numerous typos.


Glenn Little
 

If the connector in the TM5xx has loss some of its tension and the thickness of the extender paddle is slightly thicker than the DC503 edge card connector, that would explain it.
If this is the case, the connector(s) in the TM5xx need to be replaced.

Glenn

On 8/10/2020 4:20 PM, Stephen wrote:
This is a partial copy of a post a posted while trying the extender work.

Something very unexpected and quite puzzling just happened.
As many pointed out, I forgot to connect pin 6A. I just did.
Some of you may have seen my post regarding the DC503 that is not showing more than 1 digit (0) at any one time, if at all, when I power it on. Never ever.

Now that pin 6A is finally connected, not only the DC503 power on normally with the extender, but all digits are there!!! Not when connected inside the TM506 itself!! It’s not counting anything, though.
Everything is not lost, it gives me hope that at least something is showing now...

It’s also missing some segments. If anyone knows if the individual LED’s can be replaced, that would be nice to know of a modern replacement. I found some, but the panels look too thick, I think.

--
-----------------------------------------------------------------------
Glenn Little ARRL Technical Specialist QCWA LM 28417
Amateur Callsign: WB4UIV wb4uiv@... AMSAT LM 2178
QTH: Goose Creek, SC USA (EM92xx) USSVI LM NRA LM SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"


Stephen
 

On Mon, Aug 10, 2020 at 12:32 PM, Glenn Little wrote:


If the connector in the TM5xx has loss some of its tension and the
thickness of the extender paddle is slightly thicker than the DC503 edge
card connector, that would explain it.
If this is the case, the connector(s) in the TM5xx need to be replaced.

Glenn
Yes, I thought about that too. But that would mean that all of them lost their tension. Also, all the other plugins are working properly... 🤷‍♂️
Maybe this particular plugin has a very slightly thinner backplane? I could try and thinly tin the pads maybe? Hmmmm....


Chuck Harris
 

I have once, or twice, gotten a 7000 series plugin where
the middle fiberglass layer was missing, leaving a small
slot between the top surface, and the bottom surface of
the circuit board connector. I am guessing it was a defect
in the circuit board material that just happened to be
in that relatively benign location. It did make the
card edge springy, and slightly thinner than usual.

-Chuck Harris

Stephen wrote:

On Mon, Aug 10, 2020 at 12:32 PM, Glenn Little wrote:


If the connector in the TM5xx has loss some of its tension and the
thickness of the extender paddle is slightly thicker than the DC503 edge
card connector, that would explain it.
If this is the case, the connector(s) in the TM5xx need to be replaced.

Glenn
Yes, I thought about that too. But that would mean that all of them lost their tension. Also, all the other plugins are working properly... 🤷‍♂️
Maybe this particular plugin has a very slightly thinner backplane? I could try and thinly tin the pads maybe? Hmmmm....




Stephen
 

On Mon, Aug 10, 2020 at 04:48 PM, Chuck Harris wrote:


I have once, or twice, gotten a 7000 series plugin where
the middle fiberglass layer was missing, leaving a small
slot between the top surface, and the bottom surface of
the circuit board connector. I am guessing it was a defect
in the circuit board material that just happened to be
in that relatively benign location. It did make the
card edge springy, and slightly thinner than usual.

-Chuck Harris
Yes, I can see how that could be a problem. It’s not the case here, but I was thinking that maybe inserting the plugin repeatedly, may have, over time, worn out the copper pads to the point where some don’t make good contact.
I went ahead and lightly tinned the first 13. That didn’t help...
I must admit that I’m quite confused... 🤷‍♂️


Dave Daniel
 

I would make six measurements with an oscilloscope:

- Measure the signal of pin 6A at the backplane where the connector is soldered to the motherboard with the extender disconnected

- measure the same signal at the connector contact (front side)

- with the extender inserted, make the same two measurements

- with the extender inserted, measure the signal on the extender contact on the plug-in end

- measure the signal at the plug-in itself (“after” the extender)

DaveD

On Aug 11, 2020, at 04:06, Stephen <stephen.nabet@...> wrote:

On Mon, Aug 10, 2020 at 04:48 PM, Chuck Harris wrote:


I have once, or twice, gotten a 7000 series plugin where
the middle fiberglass layer was missing, leaving a small
slot between the top surface, and the bottom surface of
the circuit board connector. I am guessing it was a defect
in the circuit board material that just happened to be
in that relatively benign location. It did make the
card edge springy, and slightly thinner than usual.

-Chuck Harris
Yes, I can see how that could be a problem. It’s not the case here, but I was thinking that maybe inserting the plugin repeatedly, may have, over time, worn out the copper pads to the point where some don’t make good contact.
I went ahead and lightly tinned the first 13. That didn’t help...
I must admit that I’m quite confused... 🤷‍♂️



Stephen
 

On Mon, Aug 10, 2020 at 09:24 PM, Dave Daniel wrote:


I would make six measurements with an oscilloscope:

- Measure the signal of pin 6A at the backplane where the connector is
soldered to the motherboard with the extender disconnected

- measure the same signal at the connector contact (front side)

- with the extender inserted, make the same two measurements

- with the extender inserted, measure the signal on the extender contact on
the plug-in end

- measure the signal at the plug-in itself (“after” the extender)

DaveD
I will certainly do that and report back.
Thank for the suggestion, Dave.


Dave Daniel
 

You probably also need to check the return path in the same manner.

On Aug 11, 2020, at 04:29, Stephen <stephen.nabet@...> wrote:

On Mon, Aug 10, 2020 at 09:24 PM, Dave Daniel wrote:


I would make six measurements with an oscilloscope:

- Measure the signal of pin 6A at the backplane where the connector is
soldered to the motherboard with the extender disconnected

- measure the same signal at the connector contact (front side)

- with the extender inserted, make the same two measurements

- with the extender inserted, measure the signal on the extender contact on
the plug-in end

- measure the signal at the plug-in itself (“after” the extender)

DaveD
I will certainly do that and report back.
Thank for the suggestion, Dave.



Dave Daniel
 

I’ve been fooliwing this thread but not really paying close attention.

I assume that you’ve tried the DC-503 in all mainframe slots?

On Aug 11, 2020, at 04:29, Stephen <stephen.nabet@...> wrote:

On Mon, Aug 10, 2020 at 09:24 PM, Dave Daniel wrote:


I would make six measurements with an oscilloscope:

- Measure the signal of pin 6A at the backplane where the connector is
soldered to the motherboard with the extender disconnected

- measure the same signal at the connector contact (front side)

- with the extender inserted, make the same two measurements

- with the extender inserted, measure the signal on the extender contact on
the plug-in end

- measure the signal at the plug-in itself (“after” the extender)

DaveD
I will certainly do that and report back.
Thank for the suggestion, Dave.