The OFFICIAL MSM5540RS/M5540 Gx700 40pin IC Thread! NUDIES

Hi all ... interesting and useful thread. After regretting giving away my Sansui T-80 when I moved, I recently purchased another with a view of repairing it. The digital display is only partially working and no quartz lock. It has an F-3000 digital board and an F-3010 pre-scaler board. My investigation has concluded I have a faulty MSM5540..... The signal level works (IC04), but only a few of the 7 segment display segments work, one bar stuck on permanently. Others will sometimes work at half level whilst tuning. If I unplug the feed from the pre-scaler board I get no change when tuning AM/FM thus it looks like the Pre Scaler board is okay. I tried changing the 6Mhz crystal just in case and re-soldered many joints that looked okay. Having looked around it's difficult/expensive to buy a replacement MSM5540 IC, cheaper just to buy another tuner but I would like to restore this one. The options are trying to build a replacement 7 segment driver with new chips but even then I don't understand how the quartz locked circuit interfaces with this and can't find a MSM5540 data sheet. Can anyone offer any assistance?
 
Last edited:
Update: As I said above LED display P1 segment "a" was permanently on which is fed from MSM5540 pin 10. The others were mostly dead/erratic. I thought I would try unsoldering it at the display as it was obviously receiving a faulty output. Hey presto the rest of the display burst into life! I soldered it back and the display started playing up again so this time I unsoldered it at the actual pin 10. I wanted to take a photo of it faulty so I temporarily reconnected it and this time the display worked correctly, but for the fact this segment was stuck on. I tried to work out if I could feed this segment from elsewhere on the chip but no combination worked 100%. I then realised for FM this segment is always required so I thought I would try feeding it from the Mhz/FM/TUNE output of MSM5540 pin 28. I put in a link and now the display works perfectly for FM and has just 1 segment missing on AM.... not bad for a faulty chip that is too expensive to replace. All I need to do now is see if I can get the quartz lock working........
 
This is how I bought her....
15626390_10154113671520205_1801898264977162881_o.jpg

Made a draft drawing of the LED display
15591182_10154113671565205_4548640548393411541_o.jpg

After disconnecting MSM5540 PIN10
15540807_10154113671575205_8658112397689109181_o.jpg

Adding link to enable disconnected segment to display when FM selected
15272277_10154113671585205_5580496199178299984_o.jpg

FM now displays correctly...
15626258_10154113671525205_8277246715461891695_o.jpg

AM now only has the one segment missing....
15585128_10154113671495205_3003329485983767486_o.jpg

I may correct this by adding a logic chip
 
Last edited:
You should consider starting your own thread, with a title more specific to the problem you are solving. This puts us a few steps closer to solving the enigma of the 5540 chip.
 
As for the non functioning Quartz lock, from what I can gather (No data sheet) MSM5540 pin 35 should output a voltage when on fm signal? I have sometimes had 2.8v when on signal and 0v for no signal, other times nothing. A decent output here should turn on TR21 allowing the 5v DC through to the emitter of TR10. To guarantee a voltage here for testing, I temporarily lifted R53 (seems to be a lot of negative voltages the other side of it). When the FM signal is centre, (via Plug 12) the base of TR10 goes from 12v (AM/FM no signal) to around 0.1v which should switch TR10 on, again allowing the 5v through to the Centre LED "0" lighting it. With R53 removed the centre "0" lit when on Centre. This should also drive TR12 base via D07 switching it on, in turn it should switch TR13 off, driving the Locked LED via connector 13. TR12 seems to need a lot more than the 0.8v I am getting on the base to switch it, thus no Locked LED.
  • I had a problem that the centre LED "0" was permanently dimly lit on FM and AM (should be off) and the only way I could correct this was remove C53. Odd.
  • I have replaced most capacitors and transistors on the F3000 board.
  • After reconnecting R53 I tried removing TR21 and supplying D32 directly with the 5v supply. Now when on centre, TR10 still turns on and this time there is enough voltage to drive the centre "0" LED and TR12/TR13 ensuring the Locked LED comes on too.
Ideally I would like a working F-3000 unit to enable me to monitor what should happen to enable more fault finding as I still don't fully understand how it should work (I am not an fm tuner engineer). So what I have is a partially working circuit now which is an improvement!
15726808_10154135444820205_8304694347827901565_n.jpg


15774646_10154135452340205_1309142170880121876_o.jpg
 
Steve alerted me to this thread. I started in earnest about 3 years ago thinking about how to replace the MS5540 chip with a small pcb like this one in the picture, which can plug into the socket with no other circuit changes needed. I have designed things like this one before for work and school, but similar to you I am not a tuner person. My specialty has always been digital. designing a small digital circuit to drive the display is definitely the easy part. its the quartz lock thats the trouble. I spent 6 months probing my working G-4700 and my broken G-4700 both with a F-3000 display board and did not really figure out what was going on with the quartz lock. Then we had a baby and I haven't touched actual electronics in almost a year. Good luck if you decide to tackle this. I will bookmark and provide input if I have anything to offer.
47526.jpg
 
You can't edit your posts because you aren't a subscriber, (can you afford £18.50 ? ;)). As for the images were they pasted into the posts? - only Admins could remove them, (or 'you' if you were a subscriber), and there would be no reason to do that.

You could re-post them? - but this time upload them to AK, (cut and paste, or drag and drop). It would be nice if you could do this, for sense, as your posts on this subject are valuable and depended on the images being there. ;)

The expression on the face of the cat in your avatar says it all. :)
 
Last edited:
Hi....you all are obviously in tune with this MSM5540 chip. Wondering if you could help me out. I have a Sansui 8700gb receiver on the bench. It uses this very chip. Everything works except the digital display does not show the correct frequency. The dial pointer does point correct. Getting FM nicely, including stereo and quartz lock. Just the digital readout is not correct. Same problem with AM. From what I can gather, looks like the chip itself needs replacement. Am I correct on this? The thought did occur to replace the crystal???
 
From what I can gather, looks like the chip itself needs replacement. Am I correct on this?
Maybe, it depends what you mean by not showing the correct frequency, are the segments all there?... in other words, just works as it should but shows a slightly incorrect frequency readout. Or do you mean that it is wildly out? or stuck at a certain frequency?
The thought did occur to replace the crystal???
That is one solution to one of the possible display problems, but I am not sure of the indications that you should do this.

Several people more expert on this than me, I am only parroting what I have read on here, never dealt with one of these in person.

BTW: Welcome to AK and the Exclusively Sansui Forum. ;)
 
Hi....you all are obviously in tune with this MSM5540 chip. Wondering if you could help me out. I have a Sansui 8700gb receiver on the bench. It uses this very chip. Everything works except the digital display does not show the correct frequency. The dial pointer does point correct. Getting FM nicely, including stereo and quartz lock. Just the digital readout is not correct. Same problem with AM. From what I can gather, looks like the chip itself needs replacement. Am I correct on this? The thought did occur to replace the crystal???
How far out is it? In most cases, the LO may just need adjustment, or the pre-scaler may be faulty, not necessary to immediately assume the 5540 is the culprit.
More detail, and we may be able to help.
 
Thank you for the prompt response. When the dial pointer is at 87.5 the digital readout shows 80. When the dial pointer is in the middle at 98, the digital readout shows 187. When the dial pointer is at 109.5, the digital readout shows 296. It counts up as I move the FM dial to the right in perfect numerical order with all segments working....simply not correct. I'm quite sharp with audio electronics, but I'm afraid I am a bit behind when it comes to this arrangement. For example, not sure what the pre-scaler does.
 
It does sound like it might be related to the pre-scaler...but who would know with these things, its pretty complex and with little helpful information in the SM it can get tricky.

Look at the flow diagram for the tuner and you will see how the pre-scaler is integrated into the digital readout.
Unfortunately there is not a lot of information about this, so you might have to reverse engineer and work backwards.
The pre-scaler is underneath the tuner board in a metal box. Pretty easy to find.
A frequency counter would be good so you can sniff about the pre-scaler output..
 
I decided to look into this prescaler circuit. It contains an AN6821, which is a frequency divider, and a 74LS90, which is a counter. Looks to me like a circuit that simply takes the FM front end oscillator frequency, and outputs a lower frequency of some proportion to the input frequency. Indeed, as I scope the prescaler output on pin 7 of the prescaler board, I see a sawtooth wave. This wave does vary as I spin the tuner dial from full left to full right. Full left I measure .943mhz, and full right, I measure 1.136mhz. This is roughly a .2mhz difference. Figuring the FM range is a difference of 20mhz, this doesn't seem right. But then, I am far from an expert on how this works. I'm including here, a scope trace of the prescaler output. Any thoughts?
 

Attachments

  • IMG_0674.JPG
    IMG_0674.JPG
    78.6 KB · Views: 9
Doing a little simple math, I see the proportion is roughly the same from one end of the dial to the other. The scaler output is generating roughly 1/94th the frequency of the dial pointer. This leads me to believe the prescaler is working correctly???? Although 1/94th seem a bit odd???
 
reading through this entire thread again, I see lots of talk about that crystal. Also, I am convinced, as mentioned above, that the prescaler divides the input frequency by 100, not my 94. My 94 is likely an inacuracy in reading the time scale division on my scope. Anyway, I am convinced the prescaler is working, now that I know how it works, and what it's supposed to do. So, back to the MSM5540. The crystal....I don't have a 6.5536 on hand, so will order one.
 
The fact that you are getting intelligible display (albeit incorrect) rather than missing or incorrectly locked-on segments leads me to think the MSM5540 itself is ok, and rather that there is an issue with the data on which it relying to generate the display output, or a timing problem (which points to the crystal as a possible cause).

The typical failure mode of the MSM5540 is to generate "garbage" on the display. The present case is a notably different symptom.
 
Sansuiman, Indeed, makes sense. Looking forward to receiving the new crystal. I'll let you all know how that goes.
 
Back
Top Bottom