Repairing a Heathkit GC-1197 digital clock

[Fitpaul]

I have been trying to fix a Heathkit GC-1197 digital clock (for educational and sentimental reasons). The main fault I found was that diode D101 had shorted out. I replaced that but the clock still didn't work. I figured IC 101 had become damaged. I did some research and found that the IC is a FCM3817D (F3817PC). I was unable to source a replacement but noticed that an MM5316N has a very similar pinout so I bought one of these but the clock doesn't work although the display lights up and shows some numbers.

Can anyone see if the original IC can be replaced with the MM5316? I attach three files. The circuit diagram of the clock, the pinout of the original IC and the spec of the MM5316 IC.

 

[shokjok]

It should theoretically work. Switches SW1 and SW2 can be replaced with an SPDT center-off switch if the original switch(es) have failed. I doubt Heathkit can help you now, having eliminated the hobbyist market in 1992.

 

[Tony Stewart]

What is the rectifier output? Did you replace D102 as well?

It looks to be a compatible IC with 1 Hz on the digit divider. p39

The switches are momentary NO buttons. (SPST)

 

[RegularJoe]

Any luck on this? The waveform at the top looks like it is for the emitter on Q1, tied to pin 35 on the IC. This drives the timing on the circuit, so if you see digits but it does not change, I would look there. The other thing would be to confirm that you see 18V at the output of the rectifier. It might not have been D101, but rather, a problem with T1. Just my 2 cents.

 

[Fitpaul]

Thank you all for your input. I attach a circuit description from the Heathkit manual. I did not replace D102 as it checked out OK. The rectifier output is now 18v as it is supposed to be. The waveform on the emitter of Q1 is as shown on the waveform diagram with voltage of 8v. Pin 35 of the IC shows the same waveform. Like I said, random digits come on when the unit is powered up but do not change. Fast time advance switch does nothing but digits can be made to advance with the slow advance switch but only when contact is made and again when the switch is released. Holding the switch in the on position does not create slow advance like it used to. When the clock was powered up after a power cut all the digits used to flash at 1Hz rate. That doesn't happen now. Also the LDR circuit doesn't seem to work. The circuit is very simple really, what can I try to sort out the problem?

 

[augustinetez]

LM8361 is a direct sub for the 3817D, it is still available on the surplus market.

Couple of places picked at random -

in the US https://www.electronicsurplus.com/sanyo-lm8361-40-dip

in Australia https://www.wagneronline.com.au/lm/lm8361-39458/959442/pd/

There are others out there, just be careful buying from the likes of Ebay, Aliexpress etc.

I would be replacing ALL the diodes in the bridge rectifier regardless of whether they test OK. One failing will stress the others and it will only be a matter of time before another one fails.

 

[Tony Stewart]

I would verify the 5V regulator and clock signals then decide if the caps are bad or the voltage regulator or the IC or the solder joints.

 

[augustinetez]

No 5v Reg Tony - the chip runs directly from the 18v line.

 

[Tony Stewart]

Then verify the other items I listed starting clock input. The cct description page said “
The power supply is comprised of an unregulated +18-volt source and a regulated +5-volt* lamp source.
*This voltage is not filtered and may not read 5 volts on many voltmeters. This voltage also depends on the LDR resistance and will be less in dim ambient light.

 

[Fitpaul]

Thank you all for your input and suggestions. I have ordered a LM8361 as augustinetez suggested but I can't understand why the MM5316 doesn't work. It looks identical to the original F3817PC except it has an alarm incorporated. I note that pin 39 should have a 1Hz output but there is none. The timing circuit does'nt seem to work but as it is internal to the IC, I can't see anything to check. I guess that would explain why the digits don't change and the slow and fast advance switches don't work. I managed to get another MM5316 but it works the same as the one I originally got.

​

 

[Fitpaul]

Hi Guys, It's me again. I have received the LM8361 IC as suggested but the clock still doesn't work. PIN 35 has a 50Hz signal but pin 39 does not have a 1Hz output. The clock does show digits on powering up but does not advance nor does the fast or slow set work. The waveform on the emitter of Q1 is as shown in the diagram with correct voltages. The output of the bridge rectifier checks out at 18v as shown. I tried isolating pins 1 and 40 from the IC by removing R108 but still no output on pin 39. I also removed connection (AB) to base of Q209 but no help there either. Any further suggestions? It looks like a very simple circuit really.

 

[Diver300]

PIN 35 has a 50Hz signal but pin 39 does not have a 1Hz output.

What is the amplitude of the 50 Hz signal?

If you don't have a 'scope, can you measure the average voltage with a multimeter set to DC, and the AC voltage?

Measure against pin 29, and if you have any doubts as to how your multimeter behaves with a mixed DC and AC signal, put a large capacitor in series with the multimeter when measuring AC

 

[Diver300]

If you don't have a 'scope, you should check the diodes D101 - D104. It is quite possible that one of them could be open-circuit and the 18 V supply would be fine but the 50 Hz waveform would not go down to a low enough voltage each cycle to get input to turn off. It has to turn on and off 50 times a second for the clock to advance or be set.

 

[Fitpaul]

Thanks Diver300. The 50Hz signal is 12v (with a scope on pin 35). When first turned on pin 1 and 40 are 0v. When I switch the slow set switch the voltage at pin 1 and 40 goes to 18v and the digits light up. When there is a power failure, all digits should flash at 1Hz rate when power is restored. This does not happen. I will check diodes D101 - D104 again.

 

[Fitpaul]

Bingo! One of the diodes had a cold solder joint! Fixed that and now it works. Thank you so much.