Hi all, I have a Tasma Baby 180 radio. I have changed every capacitor checked all the resistors and all are within 20%, replaced every valve, aligned it and it works great BUT...

On voice it sounds good but when music is playing the sound is not sharp, the singer is OK but the music behind is distorted, like the speaker cone is ripped. It sounds maybe like only the higher frequency sound is being affected which would account for the voice spectrum being OK.

I temporarily removed the electro speaker and replaced it with a resistor, output transformer and permag speaker but the result is the same.

Edit: If I disconnect the cap from the second detector grid (valve 57) and inject a music track from a phone into the grid the sound is clear, so the issue is before that stage.

Any ideas? Thanks, David

https://www.radiomuseum.org/r/thomand_baby_180.html

Sounds like a grid bias issue?
Open circuit grid resistor or return path?
Assuming that when you hook the phone up, the return path is established
Fred.

Normally I use an Oscilloscope to sort issues like that. Plate detectors, like that has, are notorious for distortion, especially when their current is wrong. they have to work at a specific current that is why the cathode resistor is very high.

Those things rarely have any form of AGC / AVC so a strong signal can swamp them, normally there is a Local / DX switch to cut back the strong signals. 58 is a 6U7 and 6D6; heater & bases the only difference. Not shielded they can be a problem by causing oscillation.

If its a strong station try 68pF in series with the antenna.. In one set I needed to up the bias on the last IF tube to stop it overloading the Plate detector.

Thanks Fred and Marc, that gives me something to work on. This my first foray into a set so old tech.

Fred, what do you mean "grid return path"?

Marc, I did connect my oscilloscope and the signal looked quite clean up to the second detector.

For info: The grid bias on the 58 is 39 Volts at min volume and about 2 at max volume. The 57 is 2.5V and the output valve (59) is 15V.

Interestingly, if I put my finger on any of the grid caps it cuts all sound, I would expect it should hum. If I remove the grid cap and put my finger on it it hums very loudly. I don't know if that's a clue.

I will work through your suggestions, Thanks

If you remove the grid cap you remove the bias from the tube. that can in some tubes do damage as it becomes a diode. Beware some top caps are the plate 6CM5 is one 807 in a HMV Radiogram another.

Grid return is the resistor on grid 1 and the fly lead to the grid cap: Grid caps can come loose: Some fall off. Fixed a few of those.

#59 Grid 1 bias is quoted -18V 250V plate (Philips) #58 is normally -3V as an IF amp. Volume control is in the divider train, if that is one of the long ww resitors with tabs on it? They can corrode.

# 57 aka 6J7 again heater change. Whist it may vary with the manufacturer. Cathode current to shut the tube off -7V. (relative to cathode) that would be 7V across the cathode resistor. Plate current around 0.1mA Varies with plate & screen volts. The parameters are different for an AF amp / If amp or a detector.

General (Philips) Grid leak 1Meg Plate resistor 500K (470); Screen resistor 2.9Meg; Cathode resistor, 2200 Ohms. (-6V)

The info you want is "Tung Sol" & their saying more like 25K on the cathode. (measure across resistor (Digital meter). I would be putting a signal in via the antenna 5-12microvolts or 50 micro volts Signal grid of mixer, prefer knowing the quality of the signal. I prefer antenna for this one.

Thanks Marc for your concise reply.

I found another schematic with voltages on it, it's from the console model of the 180.

https://www.radiomuseum.org/r/thomand_180.html?language_id=2

I surveyed the voltages on mine and all voltages are low. B+ is 199V, should be 255V. Plate on the output (59) valve is 193V, should be 240V. The grid bias is correct on all valves except the #59 which is 14, should be 20V, all grids are at 0V. There is no voltage stated for rectifier output but mine is 345V. I replaced the #80 with another #80 but it made no difference.

The voltage divider seems to be OK, the voltages from it are linear and what I would expect. I will try and clean it anyway.

Is the low voltage enough to cause problems and why is it so low?

I am using a DMM, will that underead that much?

I concur with what Marcc said about anode bend detectors. The other problem with them is the AM modulation index on today's stations is much higher than was possible with most transmitters of the 1930s. What you are hearing is intermodulation distortion and it gets much worse with high modulation indices. There is a good reason the anode bend detector was consigned to the dustbin of history.

If you actually want to listen to this radio I'd seriously consider a small modification - a diode detector. You can hide one very effectively in the 2nd IF can and no-one will ever need to know! Then add a 1m resistor from the output plate to the 57's plate to introduce a little negative feedback and reduce the distortion still further. You should be able to hide this resistor somewhere.

Radio and Hobbies once did an article on updating old radios.

Reason for low HT might be a wrong value field coil on the speaker or maybe the mains transformer was changed at some time in its life. Everything looks workable to me, I doubt it's your problem. Radio will run a little cooler and live longer.

I agree with Ian, I have a 57 working as an anode bend detector in a set and in "phono" mode with normal bias the record player sound is dead clean.
When working as a detector, the sound is pretty clean on talk stations but distressing on music from 2CH and ABC stations.I did not understand why and left it to Ron :"Lateron". I guess back then a bit of distortion was not an issue in a plain Jane home set and probably not so evident in the program material of the time.
The anode bend detector is inherently non linear and just cant cope with the range required as Ian describes hence more distortion.
Ians suggestion of a diode detector is a good particularly if the operating voltage is high that is in the 10's of volts to get clear of the knee chracteristic. I'll modify my set the next time I pull it apart because music is what I want to listen to, not flapping parrots.
Fred.

Hi Ian, thanks for that info.

I found the cathode resistor for the #57 frequency converter have been replaced with a 3k instead of a 4k and replacing it improved the sound quality no end.

How does one go about making a diode detector? A diode and a cap I assume but what values and how does it connect the the #57?

It's main problem now is it occasionally loses volume when turned on, not every time, maybe one in ten, it doesn't fade out. No amount of tapping and banging on components will bring it back so it's not a bad contact or valve sockets. It then comes back by itself sometimes by probing it with a meter, a bit of static and it's back. I have replaced every cap and changing the valves when the fault is active doesn't restore it. The oscillator is working when in fault mode and the stations can be heard when the volume is turned up full. I have taken oscilloscope measurements with it working and will compare when the fault occurs next. The voltages remain constant in fault mode or normal mode.

Maybe just throw it in the bin with the anode-beam valves

Hi Fred,

Yes, that describes mine exactly. It's at a stage now where it's OK and I'm happy to leave it as is, although Ian's suggestion of a diode detector is tantalising. It's wonderful that folks with years of experience are able to pass on their knowledge to us (me) retired hobby enthusiasts.

The lack of AVG in this set is a real eye opener too, it goes to show how effective it is in later sets.

After writing the last post I turned on the radio and the low volume problem was there again.

I checked all the voltages and although a bit lower than during the day were close to what I had measured previously. As I was checking the plate voltage on the #57 frequency converter when I touched the probe on the plate socket it crackled and the set went to normal volume.

That is the second time this has happened except the last time I probed the anode on the same valve. Thinking about it, another time it came good when I re-checked the aerial connection with it running and it crackled.

This thing a is a nightmare

As noted a Local DX switch can be an advantage & I would be amazed that one was not fitted. I have found holes where one should be on the back panel: That's known as its on the circuit diagram.

The wrong resistor, was on the bucket list under too much signal causing overloading.

Whist they are not common failures did you replace Mica caps? That's the sort of thing they do. Static they also do, but I would be looking for bad solder joints (probe with chopstick) Corrosion on valve pins that have not been out for a while. Do not pull them out via the envelope.

Diode goes in the grid one circuit. 1N914 small signal diode.

Hi Marc, no sign of a Dx switch or anywhere to mount one. Not a bad idea though.

I have replaced every cap including micas and that's including one in the oscillator can. The only ones left are the aligning micas in the IF cans but they have no affect on the problem if I twiddle them in fault mode.

I have pushed, prodded, twisted and banged every component in the set to no avail, it will neither induce or dissuade the fault, not even a crackle. It appears to be electronic, as I said, creating some interference on the frequency converter socket pins with a DMM probe will spring it back to life. I traced the signal with my scope to the grid of the IF amplifier and it seems strong but weaker at the #57 anode-bend grid which would indicate the second IF can but with the cover removed and gently manipulating the tiny coil wires makes no difference.

The fault doesn't appear while the radio is running, only from when it's turned on... i.e. if it works when turned on it will keep working. I also replaced each valve while the fault was there and it made no difference.

I'll keep plodding along.

I'll try the diode, so just in series with the grid to the #57 anode-bend. Seems too simple

Hi Tippy, try a diode preferably a germanium low forward voltage type, either in series like a crystal set or clamped across the gang.

The main point is the valve then needs to biased like an audio amplifier with something like 2 k on the cathode and bypassed with a 0.1uf.
The DC condition of the grid then needs to be satisfied as in a "grid leak" resistor around .5 meg so the grid can leak to earth.

With a series diode and some means for the grid to leak to ground, then to your amazement it may just work!
In theory there should be a resistive "load" for the diode to work into and the DC offset voltage generated (called the AVC voltage) may need blocking so it doesn't bias the 57 out of working. But just try it nice and simple and see what happens.
The unwanted RF signal may be sopped up by the circuit stray capacitances or some components all ready in there.

I will modify my TRF with the 57 anode bend detector to diode in a similar manner, but, at the same time I want to get rid of the cathode volume control in the 58 RF stages and implement a simple AVC which will use the diode offset rectified voltage. The diode load will then be a pot as a volume control just like a proper radio. There are some caps and an RF choke in it that I can swap out of the 57 plate ciruit and into the diode circuit to filter out the RF not wanted. Needs a bit of a re-design just like the old article I saw in a back issue of RH but I cant remember what issue it was, any body got a clue?????? I have the CD so can find it if I know which month/year????

I'm not allowed in to the workshop yet, still recouping from heart surgery so will not do this for some time maybe next week or the one after that...………………………bugger.
Fred.


I have not looked at the Tasma circuit I really should do that before issuing any sage advice!!!.

Just downloaded the circuit of the Tasma from Kevin Chants site.
Wow, the Tasma guys did not waste any parts in that baby!
Very classic circuit..
Yes, I reckon just put a diode in series with the grid lead from the IF (try one way then the other to reverse the Dc offset if it biases the valve off/on, clip a 1m resistor from grid to ground, bridge the 20k cathode resistor with a 2 or 4k and enjoy the music. The .001 cap and HC choke in the plate should top cut the RF as intended.
What could go wrong??????
Fred.