Sears Model 10656 Clock Radio - Clock board demanding high wattage resistor?

[8769F]

Hi everyone, and thanks for reading this post. I'm doing my best to learn some electronics repair, and lately I've been working on a vintage Sears Model 10656 clock radio. It's nothing special but I like the look of it and I was hoping to fix it up. Sorry for the long post, I'm trying to include as much relevant information as I can.

Although the unit worked, it had scratchy pots and poor reception in the FM band. I opened it up to start working on it, and I noticed that there were a couple of burnt resistors on the clock board. I worked through estimating the resistance of the burnt resistors, and looked and the circuit to determine that they appear to be related to the Bright/Dim option for the LED display.

The main clock chip is a TI TMS1952NL, a 40-pin DIP, and I cannot find a datasheet for it, but after probing around it seems like it might be pin-compatible with the MM5316 which I could find a datasheet for.

The main question is, it seems to me like the resistors that burnt are being made to dissipate quite a bit of power, and yet the originals where only 1/4W each. To me this suggests either something else isn't working right in the circuit, or it was a bad design from the get-go. Problem is I'm too new to know which the answer might be so I'm hoping for some guidance here.

I measured the voltage between the points just before and after where the resistors were, and I am seeing 12V DC. The burnt resistors were in series, and I believe they were in the range of 10 ohms to 100 ohms. I've since temporarily connected two, 100 ohm 1W resistors in parallel to give me a resistance roughly in the middle of the expected range, and with a bit more ability to dissipate power. This provides a decent brightness but using some power calculators to see dissipation requirements for resistances that seem to be in line with the old resistors, and it looks like they would have been required to dissipate anywhere from 2-4W, well over the 1/4W rating of the original resistors.

Any guidance on what I should check next would be appreciated. Below are some of the things I've done so far:

A few details:

1. The clock board is powered by a transformer which has two red wires and one black coming out. When checking voltage between the red wires and black, both red wires are putting out 10.5V. I confirmed no shorting between secondary and primary.

2. If the TMS1952NL is in fact pin-compatible with the MM5316 clock chip, then the VSS pin is seeing just short of 9V DC, and VDD is -16V DC. Both pins also seem to have about 4 VAC on them. I'm checking the voltage by connecting the DMM to the incoming black transformer lead, and then probing the VSS and VDD pins with the red lead. Hopefully not making a rookie mistake there.

3. There were two electrolytic capacitors on the board which I pulled, measured and replaced. They were still border-line in spec., but since they were out for testing anyway I've changed them. The larger capacitor was replaced with one rated significantly higher for voltage (from 35V to 250V) because it's all I had in the right capacity. I've looked over each of the remaining resistors, and compared their marked values to measured values, and they are all within tolerance. Measurement was done in circuit, but I figured since the values were close to their marked values this is likely OK. Maybe I'm setting myself up for failure here?

4. I've done basic diode checks for the diodes on the boards to see that they are still only conducting in one direction, and that the drop seems reasonable.

5. I've pulled the two transistors from the board, checked them, and tried to match them up with datasheets to see if their gain was in-line with expectations and it seems to be.

Thanks again for taking the time to read my blabbering....

 

[Nigel Goodwin]

The MM5316 datasheet isn't any help at all, because it's a completely different chip, designed for fluorescent displays, not LED - and the resistors are presumably to do with the display?.

Was the clock actually working? if so, what values did the two burnt resistors read? - just because they are cooked, doesn't always mean they have changed value. The usual reason for the burn-up is poor design, fitting too small a resistors during manufacture - there's also the issue that putting resistors in series (or in parallel) tends to be very unreliable anyway, and is usually done by production (not the designers) because it's cheaper and easier.

On service courses for new TV's you could easily spot what was going to fail, usually small resistors in series or parallel, and we were pretty well always correct. We told the manufacturers, but they never cured the poor design issues.

Another possible issue is fitting a 250V capacitor in place of a 35V one, the most important value of an electrolytic is usually ESR, and high voltage ones commonly have much higher ESR's than low voltage ones.

 

[8769F]

Hi Nigel, first let me say thanks for replying, I really appreciate it. Here is some more information/responses to your points.

Clock IC - Good catch that the MM5316 was for fluorescent displays, I didn't notice that. Originally I went searching for a sheet for the TMS1952NL, couldn't find it, but found a site that mentioned that the TMS1943NL was pin-compatible with the MM5387AA (meant for LED displays) which I could find a datasheet for. I started to map out the pins on the IC, and everything seemed to match up with the MM5387AA datasheet with the exception of pin 37, which was NC in the datasheet, but connected in the circuit. The pinout for the MM5316 matches the MM5387AA with the exception that pin 37 is labelled as 'Blanking In' and so I thought maybe the pinout for the MM5316 was a closer fit.

Working - Yes the clock was actually working, and still is as long as I put something in where the resistors were before. Originally when measuring the burnt resistors I could have sworn they were both shorted, but measuring them just now on the bench they both read 15 ohms. In the one picture I have of them the first band looks brown, and so I suppose it's a plausible value. In series that gets me to 30 ohms, and right now I'm subbing in the two 100 ohm 1 watt in parallel to get me 50 ohms.

ESR - In terms of ESR I have a cheapo component tester, and while maybe not super accurate I think it serves a decent ballpark. The original pulled capacitor shows as 1.3 ohms ESR, and the capacitor I subbed in is a low-ESR type, and it reads .43 ohms.

So my interpretation of what you said before is that it is in fact quite possible that these two series resistors were just under sized for their application? I can put a single 3W resistor in there to keep the temperatures down and it should work, but I didn't know if there was something more I should/could do before resorting to that approach. It just seems so... dodgey.

Thanks!

 

[Nigel Goodwin]

Hi Nigel, first let me say thanks for replying, I really appreciate it. Here is some more information/responses to your points.

Clock IC - Good catch that the MM5316 was for fluorescent displays, I didn't notice that. Originally I went searching for a sheet for the TMS1952NL, couldn't find it, but found a site that mentioned that the TMS1943NL was pin-compatible with the MM5387AA (meant for LED displays) which I could find a datasheet for. I started to map out the pins on the IC, and everything seemed to match up with the MM5387AA datasheet with the exception of pin 37, which was NC in the datasheet, but connected in the circuit. The pinout for the MM5316 matches the MM5387AA with the exception that pin 37 is labelled as 'Blanking In' and so I thought maybe the pinout for the MM5316 was a closer fit.

Working - Yes the clock was actually working, and still is as long as I put something in where the resistors were before. Originally when measuring the burnt resistors I could have sworn they were both shorted, but measuring them just now on the bench they both read 15 ohms. In the one picture I have of them the first band looks brown, and so I suppose it's a plausible value. In series that gets me to 30 ohms, and right now I'm subbing in the two 100 ohm 1 watt in parallel to get me 50 ohms.

ESR - In terms of ESR I have a cheapo component tester, and while maybe not super accurate I think it serves a decent ballpark. The original pulled capacitor shows as 1.3 ohms ESR, and the capacitor I subbed in is a low-ESR type, and it reads .43 ohms.

So my interpretation of what you said before is that it is in fact quite possible that these two series resistors were just under sized for their application? I can put a single 3W resistor in there to keep the temperatures down and it should work, but I didn't know if there was something more I should/could do before resorting to that approach. It just seems so... dodgey.

Thanks!

Yes, far better to put a decent sized single resistor in - as I said, resistors in series (or parallel) tend to fail. I presume it's some kind of current limiting for the LED display?, so it's value won't be critical.

 

[8769F]

Yes Nigel I believe it's current limiting the LED display. One end of the resistor series connects to the common anode pin of the LED display. The other end goes through a switch for a Dim/Bright setting.

I'll play with the resistance to get a display brightness that seems good, and then I'll pack a single high wattage resistor in there to keep the temperature down.

Thanks for your help!

 

[qazwiz]

with a little luck i found this ;D
does that look like your device?

note the post identifies IC as National MM5397 datasheet

 

[8769F]

Qazwiz, thanks for chiming in. Yes that is the radio, but mine is not using the same IC. I'm guessing they may have had more than one clock board for some reason. From the tracing I've done, the TMS1952NL seems to have a bit of a different pinout from the MM5397 and (almost?) identical to the MM5316.

I'm continuing to try and find the right balance between display brightness (lower resistance), and higher wattage to dissipate the heat. I had a 7W 120ohm resistor on hand and I tested it and it stays around 50C, but the display is a bit dimmer than I would like.

More testing in progress....

 

[qazwiz]

it stays around 50C, but the display is a bit dimmer than I would like.

would a thermo-resistor help?

ALSO: there's 4 versions two have different bent pins not sure what difference other two have

 

[danadak]

Tricky situation, horse before or after the apple cart.

Put a R that gives you brightness you want. Then calc current and power
of it in situ. P = (Ex E)/R. I = E / R. Current should be in range of 20 - 50 mA
for that era radio. The size R by padding power + 50%. Install on board, but
in the air, not touching board (minimize heat distribution to the other parts.

7W R seems excessive. Old technology LED displays age, eg. 1/2 brightness
spec drops in low thousands of hours, compared to todays displays, so pushing
more current thru them burdens power supply.....might be asking, for trouble here.
7W in light of prior 1/4W or 1/2W, what if they were the correct W rating.....?

You would be better off replacing the digits if you can find a mechanical and close
electrical fit to get brightness you desire.

Is there a regulator chip for the power supply, you might want to read its case
T and calculate its die temp for the 7W case. Although modern regulators, even
one back in the day, have current limiting and thermal protection I would not run
something close to spec limit. No point in burning to death at night when radio
starts a fire in vicinity.


Regards, Dana.

 

[qazwiz]

Tricky situation, horse before or after the apple cart.

wish we had the circuit design....
unfortunately all of the circuit archival sites I've got are dark. it's a popular "manufacturer" (Sears) but no bites. 1970 should have packed in a circuit design but no go (apparently)

maybe the Canada angle ??

perhaps searching some .com.ca sites will turn up ??????????

 

[8769F]

I've continued to scour the web for more information with no luck. I've tried to draw out the schematic myself in KiCad but I'm not doing a great job of it. The board itself looks fairly simple but it doesn't take long for me to get tangled up in all the connections. As I dig in though I keep coming up with more and more questions. I would welcome insight into if anyone cares to comment. My goal from the get-go was not just to make this radio work, I wanted to really understand it.

I'm thinking to step back and start with confirming that the MM5387AA is being fed the right voltage, and then from there measure it's output to the various segments to see if the voltage and current is in the safe range for the display segments. The datasheet for the IC is here. I've attached the datasheet for the TLR2087A LED display since I can no longer seem to find a working link. For the clock IC, it looks like it requires between 24V->26V to output driving a display.

Here are some measurements I took:

Black transformer wire (ground?) to the VSS pin on the IC, 9.3VDC, 4.2VAC
Black transformer wire to VDD pin on the IC, -16VDC, 4.2VAC
VSS to VDD, 26VDC, 0-32VAC

So I'm puzzled by a few things.

1. Is it OK that there is AC showing on the pins?
2. I would have expected to see 24V-26V from the black transformer secondary wire to VSS. Am I way off base in this?
3. What the heck is going on that there is -16V on VDD. Should VDD not be essentially 0V or am I just totally screwing up how to measure it properly?

I recognize that a bunch of what I'm asking may be basic knowledge, but I'm really trying to understand how all this works and appreciate any guidance anyone wants to give.

 

[8769F]

Tricky situation, horse before or after the apple cart.

Put a R that gives you brightness you want. Then calc current and power
of it in situ. P = (Ex E)/R. I = E / R. Current should be in range of 20 - 50 mA
for that era radio. The size R by padding power + 50%. Install on board, but
in the air, not touching board (minimize heat distribution to the other parts.

7W R seems excessive. Old technology LED displays age, eg. 1/2 brightness
spec drops in low thousands of hours, compared to todays displays, so pushing
more current thru them burdens power supply.....might be asking, for trouble here.
7W in light of prior 1/4W or 1/2W, what if they were the correct W rating.....?

You would be better off replacing the digits if you can find a mechanical and close
electrical fit to get brightness you desire.

Is there a regulator chip for the power supply, you might want to read its case
T and calculate its die temp for the 7W case. Although modern regulators, even
one back in the day, have current limiting and thermal protection I would not run
something close to spec limit. No point in burning to death at night when radio
starts a fire in vicinity.


Regards, Dana.

Dana, thanks for the information. There is no voltage regulator, just a transformer which has three wires on the secondary, two red outputting 10.5VAC each, and a black. I agree with you that 7W seems wild which is why I'm trying to dig in to better understand the circuit to see if something else is off the rails. I decided to step back a bit and start with better understanding the power going into the clock IC, so that I can then get a better handle on what is coming out to the segment display.

In an ideal world I would have enough knowledge that I could build a simple replacement clock/display board with more modern components that would fit in the vintage case, but sadly I'm not there yet. #Goals

 

[8769F]

wish we had the circuit design....
unfortunately all of the circuit archival sites I've got are dark. it's a popular "manufacturer" (Sears) but no bites. 1970 should have packed in a circuit design but no go (apparently)

maybe the Canada angle ??

perhaps searching some .com.ca sites will turn up ??????????

Agreed that the web seems dark for this. I've looked for quite some time without any luck. Even finding a datasheet for the TMS1952NL would be a nice win, but sadly nothing...

 

[Nigel Goodwin]

I've tried to draw out the schematic myself in KiCad but I'm not doing a great job of it.

Use a pencil, and a number of sheets of paper - making each version more and more correct, and more readable. Only once you've got it correctly drawn out should you be considering drawing it in Kicad (or similar).

In my (fairly extensive) experience of drawing circuits out the first partial one is REALLY, REALLY bad, and you have to start a second time placing the components you've identified in more sensible places on the paper - then carry on in the same way, with further iterations.

This might perhaps help you:

 

[8769F]

Use a pencil, and a number of sheets of paper - making each version more and more correct, and more readable. Only once you've got it correctly drawn out should you be considering drawing it in Kicad (or similar).

In my (fairly extensive) experience of drawing circuits out the first partial one is REALLY, REALLY bad, and you have to start a second time placing the components you've identified in more sensible places on the paper - then carry on in the same way, with further iterations.

This might perhaps help you:

Thanks for the suggestion of going to paper and working through the process, I'll see where it leads me. Thanks for the sharing the datasheet for the green display, the pinouts definitely help with understanding the connections.

 

[danadak]

I've continued to scour the web for more information with no luck. I've tried to draw out the schematic myself in KiCad but I'm not doing a great job of it. The board itself looks fairly simple but it doesn't take long for me to get tangled up in all the connections. As I dig in though I keep coming up with more and more questions. I would welcome insight into if anyone cares to comment. My goal from the get-go was not just to make this radio work, I wanted to really understand it.

I'm thinking to step back and start with confirming that the MM5387AA is being fed the right voltage, and then from there measure it's output to the various segments to see if the voltage and current is in the safe range for the display segments. The datasheet for the IC is here. I've attached the datasheet for the TLR2087A LED display since I can no longer seem to find a working link. For the clock IC, it looks like it requires between 24V->26V to output driving a display.

Here are some measurements I took:

Black transformer wire (ground?) to the VSS pin on the IC, 9.3VDC, 4.2VAC
Black transformer wire to VDD pin on the IC, -16VDC, 4.2VAC
VSS to VDD, 26VDC, 0-32VAC

So I'm puzzled by a few things.

1. Is it OK that there is AC showing on the pins?
2. I would have expected to see 24V-26V from the black transformer secondary wire to VSS. Am I way off base in this?
3. What the heck is going on that there is -16V on VDD. Should VDD not be essentially 0V or am I just totally screwing up how to measure it properly?

I recognize that a bunch of what I'm asking may be basic knowledge, but I'm really trying to understand how all this works and appreciate any guidance anyone wants to give.

Vdd, Vss are referenced to each other, and have to meet -

This part is a PMOS chip, so the "nomenclature" we are all used to these days is
reversed. So as long as the span between Vss and Vdd is in spec one can "move" that
"window" of voltage around with respect to ground.

So in short Vdd < Vss, for proper operation.


Regards, Dana.

 

[8769F]

Just a quick follow-up here, I'm continuing to try and map out the circuit and better understand it. It's a slow process for sure, but I'm learning with each step. Once I have a schematic I'll post it here to hopefully assist with further discussion.

 

[Lokash20]

Stumbled across this thread and can, hopefully, contribute something useful.

Regarding the TMS1952NL, I have a bought as new early 70s Hanimex red LED digital alarm clock of the kind that keeps time by counting mains cycles and it was, until recently, still working perfectly. Then, due to a faulty socket causing a couple of brown-outs, it has started misbehaving hence, laking any schematics or IC data sheets, I've been looking for details on the chip to be able to try and repair the clock.

Anyway, long story short, despite quite good customer support, Texas Instruments, the original manufacturer of the TMS1952NL, and Rochester Electronics, who apparently now have "all" of the out of production dies and designs, weren't able to help.

Well, I'm stubborn and don't give up easily and, during a Googling session, I came up with this link which is for an NTE2060 & NTE2061 pdf data sheet on the NTE website, this chip supposedly being a drop-in replacement for the TMS1952 and the pinouts and voltages, at least, do seem to match.

Regarding my Hanimex clock, I'm currently, as Nigel Goodwin correctly suggests, iterating the circuit schematic, albeit in KiCAD (having exhausted my paper-based patience) and trying to figure out how the blinkin' (literally) clock is supposed to work.

 

[shokjok]

Could you get a datasheet for the TMS1952NL from Texas Instruments?

 

[Lokash20]

Unfortunately not.

As I touched on above, TI told me that they don't retain data sheets or manufacturing data on out of production chips and directed me to Rochester Electronics, who they said take over all the manufacturing data and information on such devices.

After much searching, Rochester Electronics also couldn't find any information regarding the TMS1952NL.

It was only after much online digging that I chanced across the NTE chips which appear to fit the bill regarding the pinouts matching the PCB track functions and also with comparison to the voltages I measured on my clock PCB.