My FM converter for tube radios using the dial and tube amplifier

[pip5678]

Originally Posted by murphyv310

Hi
I've done a video of the mod and results.
It's fair and Frank and shows how it sounds, tunes etc. I've also commented on audio quality.
https://youtu.be/zTBlYNAH0tU

Very useful video review, thanks for taking the time. I hope your customer will be happy.

[david winter]

FM-DW MODULE UPDATE:

I took some time yesterday and this morning to check how I could modify the calibration and tuning code so that the dial positions are considered instead of the Min/Max values of the tuning capacitor.

This change required an additional cold reset feature should the user get lost by pressing the calibration button accidentaly, by setting the dial at the wrong position, or if a mistake was done in the wiring (like leaving something connected to the tuning capacitor).

Cold reset procedure:
- Module is OFF.
- Maintain calibration button pressed.
- Turn module ON.
- Release calibration button after half a second or more. The module is now idle until powered OFF.

Calibration procedure:
- Set dial to the START of the FM band. Allow a very slight shift from the position you selected.
- Press the calibration button once. FM-DW module won't tune yet.
- Set dial to the END of the FM band. Allow a very slight shift from the position you selected.
- Press the calibraiton button once. FM-DW module should enter in tuning mode. Make sure the FM band is fully covered. Else, perform a Cold Reset and calibrate again. If calibration fails again, verify your circuit and ensure that the tuning capacitor has NOTHING else connected than the wire going to the FM-DW module.


I will update the online instructions now...

[bobhowe]

Tried hard reset tuning is still from Right to Left not really bothered as the DAC90A has not got a FM Tuning scale but I will convert where on the MW scale each radio station is .Counted 12 stations I am receiving here where I live from Radio2 which is 88.5 up to Spark Radio on 107.0 . A Bush VHF 61 only goes up to 100.0 so I would lose 5 stations where I live if I used that set so that's a result in its self from only getting 3 MW no LW stations on a DAC90A as well David very pleased

[murphyv310]

Hi.
The reset didn't work for me either with the issue the same.
Fortunately the customer can't collect today due to bad weather and is now coming on the 20th, so I've some time to be able to sort out any further info that might help.

With due respect Bob I'd expect to get better results on FM than AM especially as this radio was very good before the tuner failed and is not fixable. The background noise is not easily reproduced using a phones microphone, it's more objectional to the ear. It's not a big issue unless Radio 3 or Classic FM are your favourite stations then it becomes obtrusive on quiet passages of music.

[david winter]

How do you want that modules already shiped have their software updated to allow both dial settings ?
Modules already shipped work the Cmin / Cmax way.

Only the modules I currently have are reprogrammed to allow the new method.

[bobhowe]

Hi Trevor out of the 12 stations I am getting here possibly more it I increase the length of the wire I use for the aerial I will be only listening on a regular bases is Radio 2 and Smooth radio but a DAC90A is not exactly HI FI and my hearing is not as good as I was younger ( can not even hear 405 line whistle now ) so 3 foot away from the radio I can not hear the background noise so it does not bother me hopefully David will get round the issues soon

[bobhowe]

Originally Posted by david winter

How do you want that modules already shiped have their software updated to allow both dial settings ?
Modules already shipped work the Cmin / Cmax way.

Only the modules I currently have are reprogrammed to allow the new method.

Hi David I am not bothered about mine as my next project is a pre war set and I will convert the FM dial position on the AM scale

[david winter]

The (honestly low) background noise has never been a problem so far.
However I understand that one who wants to listen to classical music may find this annoying. This said, no one has ever reported this so far.

The only workaround is to replace the TEA5767 module by a RDA5807M one, and reprogram the module's firmware, which I can do since my software can support it by changing one line of code.

I must admit that I have tried for some time to program and test the RDA5807M, but the problem was the muting between stations, which can't really be disabled, and makes the tuning mode difficult on an old radio. Hontelé has droped the idea a long time ago for the same reason.

I have a few RDA5807M in stock, so one who REALLY wants the RDA version may order it, but I will need to charge an extra for labor.
You can even consider returning your module for updating (either using the existing TEA5767, or for using the RDA5807M), but shipping costs won't be free (5 euros registered mail to the UK).

[murphyv310]

Hi David.
Many thanks for your post. I must have misread that all was necessary was to do the reset.
As you say the background noise isn't necessarily an issue, I'm a little deaf but I thought my customer would complain. The issue Is I do a good number of restorations and some have asked me in the past if it's possible to add FM to AM only sets. The board is a solution but if the customer is fussy then I have to bear the costs etc.
I do think the board is very useful especially in areas the AM is weak or has excessive noise from SMPS and It's a useful addition.
I'm sure many would pay more for the board if there was some further refinements to add to some of the high end AM sets with push pull amplifiers and decent speakers, especially good quality radio grams, these are just my thoughts and not a criticism.

[david winter]

Trevor:

I like your video, but you insist on some points which don't deserve so much focus.
The "pops" that you report are normal if you are not correctly tuned on a station. This is due to the nature of the FM chip.
You should also ensure that the tuning capacitor is clean, and that the shaft collectors have a clean contact with ground all the way through.
Hontele's initial design generated a lot more pops. I managed to remove a good amount of them by rewriting the whole code and improving it.

The background noise brought an amazing discovery: I'm testing the module at home using a modified transistor radio. Guess what ? I have NO background noise.
Amazingly, only one station gave it, but this didn't stay as I tuned here and there, without that noise.
I thought the problem was due to some approximations in the PLL values given to the TEA chip, but after doing the calculations manually and comparing with those calculated by the module, the error is absolutley negligible.
For example, 107.1MHz means a PLL value of 13101.69 . The TEA uses non-decimal value, so the PLL must be 13102. Perfect reception there.

Using your radio, can you power the module using a 9V cell instead of the AC ? Any difference ?


Cheers

[murphyv310]

Hi David.
Initial tests were done with an alkaline PP3 9v battery with the same results. The radio has never shown the noise prior to installing the board.
Is there a possible chance the board is faulty.
I will feed it with a frequency in the FM band from my signal generator, then have a look at the output on the scope and see if this noise is apparent and at what level.
Hope that helps.

[david winter]

The module is not faulty as the background noise on tube radios has always been there.
What amazes me is that I can't hear it with my portable transistor radio.

I will try to check what can generate this noise but that's not trivial.

What will be interesting is that you scope and measure the background noise frequencies using your FM generator sending a "deaf" signal, and see if the background noise stays constant on the entire FM channel band.

I will keep checking the code, but I'm sure it's not due to the firmware.

[murphyv310]

Interesting David.
I wonder if the module works better into a low impedance load? Valve input of course is high impedance.
I'll check this out tomorrow afternoon.
I am trying to help not criticise.

[bobhowe]

Hi Trevor is it with trying the board on a transistor amplifier to see if the background noise is still there and if it goes could the Board be picking something up though the valve stage just a thought

[murphyv310]

The module shouldn't be picking up anything from the valve stages as the front end and IFs are disabled with no HT. The board is directly mounted to the steel chassis which improves screening as well. It could be impedance matching Bob, an easy thing to sort out.

[bobhowe]

Hi Trevor if its an impedance mismatch pardon of me not been as clever as you how can it be resolved

[david winter]

The problem could be more complex than thought.
I made some further tests using my transistor radio (battery operated).
I tuned on France Musique (91.70 MHz) and an operatic song was being broadcasted. I felt like the background noise was there at some moments, but not always. I could hear it only with the presence of a female voice, or with some strings in the orchestra. Kinda strange.
I tuned to another station around 88-89 MHz where there was only some talking (male). No background noise. Tuned to another station with pop music: no background noise.

I'll try to reproduce this tomorrow. Feeling tired now...

[murphyv310]

Take it easy David and don't stress yourself, feet up with a glass of wine.

Bob. The simple way is a resistor across the output to chassis say 47k. There could be a small drop in audio level but the output from the board has a good level and will drive a valve radio fine even one with just an output pentode EG EL84.

[david winter]

There is a 0.1µF / 50V capacitor before the audio output pad.
If you add that 47K resistor accross the output pad and chassis, I'm not sure if this could be dangerous on a live chassis (AC/DC sets) without changing that capacitor to a 400V type (which as you know, would take a lot more space. In such cases I always add a 0.01µF/630V in series, after what I can put that 47K resistor.... Will try this on monday, but I don't have any modified AC/DC set handy right now !

I also did some minor code cleaning and realized I could save some memory space by using a trick. Enough for today, and I was happy to find that new dial feature before all.
It would be nice that you edit your video and remove the section where you mention the dial issue, since it's now solved.
Do you want that I reprogram your module, or do you want to buy another one (with the newest firmware) as said before ?

[murphyv310]

Hi David.
I can edit the description or remove the video, it's your call.
I'd not think there would be an issue with a 50v rated cap in the board with an AC DC set.
Look at it this way, the board is grounded on the chassis, the supply is from an isolated mains TX rectified and smoothed to the board, the audio output is fed to the top end of the volume control normally a 1m or 2.2m value pot the slider goes to the grid. There is no high voltages on the control grid. Even if the mains live is connected to chassis I don't see how it will damage the board unless someone is poking about and being careless. I'd like to think all here will doubly check that the chassis is at mains neutral.
Incidentally I did use a 0.01uf 630v cap in this set.

PS I'd Probably just get another one, saves faffing around with post etc. I'll let you know in due course.