Another Quad FM4 problem

[Trigon.]

From that link: This mod came about because could not get the 4 bit CPU MP4480 or the TMS1000

There seems to be some confusion about the TMS1000/MP4480 markings.

The TMS1000 is a general purpose mask programmed microprocessor - i.e. the program is 'built-in' at manufacture, and is unique to each product.

'MP4480' indicates the program inside the processor, almost certainly unique to Quad, and is the only TMS1000 that will work.

It's not a question of 'MP4480 or the TMS1000'.

Cheers

[chartz2]

Exactly!

[NikPrice]

I've removed IC1, unfortunately I've now got a blinding migraine so I'll do some tests in the morning and post the results.

[chartz2]

Night!

[NikPrice]

Rotten headache is keeping me awake. Still, at least I can see again!

Many thanks to everyone posting advice, I really am very grateful.

What I did notice before I put everything down for the evening was that when I removed the chip (NOT a 4480!) and powered up the frequency sat at about 89MHz and stayed there. As soon as I connected pin 12 to pin 26 the display raced away again to 112MHz.

I'm going to replace the TL092 and see where that takes me.

Nik.

[NikPrice]

Had a long session on the FM4 tonight. After resoldering every joint, checking through the solder connections with a magnifier, and various buggering about, I decided to have another attempt at searching for a picture of the track side of the pcb to see if any of the tracks I had to rebuild was missing a connection somewhere.

I managed to download an image and after checking the tracks were correct I noticed in the bottom right hand corner of the picture there was a 330R resistor soldered to the track side of the pcb between the ground side of C12 and the anode of D24. Hey ho I though and grabbed the resistor box.

Tuner works a treat. Replaced IC1 with the original MP4480 (after removing certain links), and lo and behold everything is working as it should. I'd already replaced the battery and it appears the memory is working fine, all buttons are working and the tuning range is correct.

D24 is only a generic 6.2V zener at the moment, but I am waiting for a new 1N823a to be delivered.

Scoping the output shows what looks like audio. I'm guessing that the receiver is going to need realignment, and I don't have the test equipment for that.

But at least it's working. Not too sure what this 330R resistor is all about. It's not mentioned in the service manual I downloaded.

If anyone has any knowlege of this I'd be interested to hear about it.

And a big thanks to those who posted help and advice. It was all really appreciated.

Now, its 4 in the morning and I'm off to bed!

Nik.

Link to picture: https://picclick.co.uk/Quad-FM4-Tune...#&gid=1&pid=10

[chartz2]

Odd. The resistor looks like a 220 Ω by the way. Wasn’t it just required with the new PIC processor board ?
Mmm.

[chartz2]

It is across C12 in effect.

[NikPrice]

Initially I thought it was a 220R too and dialled 220 into my resistor box. When I'd finished varying the resistance to get 87.5MHz on the display it was set to 330R. I loaded the image into some photo editing software and zoomed in tight. The bands are actually orange.

It's between REF L and ground. In effect pulling it down to just under 2V. It may be a quick and dirty fix someone else has applied to the same fault, but if you follow the eBay link in the photo site the history of the tuner is detailed and it appears to have been returned to Quad for repair.

[chartz2]

Yes, probably a Quad factory fix.

[chartz2]

Another one here:




My repair is different but it appears there's an issue with the floating grounding of the TMS1000 or something like that. Also, the type of capacitor used must have an influence, because my problems started after I replaced C12 too.
Perhaps that resistor solves that, an instability of the proc. Food for thought, perhaps.

EJP, are you around to clarify this please?

[NikPrice]

Definately something to commit to memory, hopefully anyone else with the same problem will find this thread.

I'm not convinced it's an instability of the processor, as I added the resistor and tested it without the processor fitted (just the manual tune enable link), and it worked fine. I only added the processor as an afterthought just to see if it was still operable, and as luck would have it it was.

It's an interesting fault, and I don't think I'd have ever solved it without stumbling on the answer by sheer luck.

On the bright side I now have a working FM4 tuner to add to my 34/306 combo, and it cost me next to nothing.

[chartz2]

Hi Nik,

Makes sense then.
Same here, I paid € 25 for mine. I already had a grey one to keep my 34/405 company. Now and as I'm a completist, I'll have to find a broken but also minty brown 34!
I'm also looking for a 306, preferably not working as always. I'm not interested in working equipment at all! ;-)

[John_BS]

I think the 330ohms across C12 might have been added because the TL094 is not really specified to sink current down at 2 V above the negative (ground) rail. There's a few mA coming down through that zener which has to be "sunk" by the opamp. A "typical" TL094 will just about do it, but not worst-case. The 330 ohms will sink almost all the current thro D21 and allow the opamp to control the voltage more reliably.

Glad you got this working!

John
PS the voltage between 1+2 and 3 should now be close to zero!

[chartz2]

Hey hey hey, my second, untouched, FM4 has it in place too! So this was a factory fit after all! It has a Quad service sticker though, and a blue plastic battery box.

[nutteronthebus]

I have been watching this post with great interest as I have 3 FM4 with the same fault . New resistors on order so watch this space

Dave

[chartz2]

That looks very promising indeed Dave!
I’ll have one more to do myself - no lights, nothing.

[NikPrice]

Good luck Dave, I’d be interested to hear if it worked for you as well.

Nik.

[NikPrice]

The frequency span was not as wide as it should have been - around 87.5 to 106 MHz. I reduced the the value of the resistor on the back side of the PCB from 330R to 220R and, with the tuning knob tuned to the lowest frequency, realigned the lowest frequency to 87.5 by turning the core of L6. The frequency range is now 87.5 to 109.5 MHz.

The memory frequency when saved seems to jump up .05 MHz from the tune frequency, and the both tune and memory frequency appear to creep up my about .05 MHz as the reciever warms up. I'm wondering if this is due to D24 being poorer spec than thought. I've ordered another to compare.

I've also noticed that the tuning indicators don't seem to move independantly, suggesting that centre tuning is not correct.

[chartz2]

Mmm, interesting.
I also have this jump when the tuner is cold. There shouldn’t be any because AFC should take care of that.
I find manual tuning is actually more stable when the tuner is cold!