Philips 2514

[boiss]

Ok fitted the O/P bottle after having attached high Z headphones,paralleled a 5K res just in case across the O/P terminals. The O/P stage appears to be working OK although the HT is a bit low and G1 a bit high at -22V, Still have the 1K5 dropper resistors on the 1N4007s and another 12K load on the HT. Will start playing with the LND150 very soon, 15V zeners arrived, will start with the source R.

[boiss]

Mounted the LND150 on a bit of terminal block in place of the E415, back to back 15v zeners gate to deck, source has 10k with o.1 uF and 10uF decoupling. Put a a little resistor box across the 10K set to 10k. DSv about 15V, G and S about the same at 13V, Reduced SR and DSv went down to 6V. Increased SR and things started to change. With SR at 10K, Sv was 8V, GV 7V and Dv had gone up to 137V DS current slightly lower than 1mA, not far off the valve value. Don't understand where the + gate voltage is coming from as in front of the 1M grid stopper there is -2V. However the reaction control is working fine and it really howls. Certainly covers the mw range. I heard a station at one point, without an antenna or front stage valve etc. Starting to look hopeful !

[boiss]

Just a couple of pics, Quite rudimentary with electrical connectors for the LND150/E415. I will use some old veroboard to make the eventual circuit bd.
I will do the same for the B442, note the side contact on the base, this was not used in the radio as I can see that the chassis connection was never used. It could have been pin 5 in the base, but this was not mounted either. Obviously the valve that had been used originally had had directly heated cathode and was connected via the 2 120r resistors across the filament supply which are connected to the electrical -Ve (ground) not quite the same thing. I will connect the side contact to the nearby terminal on the panel which is connected to electrical -Ve.

[PJL]

My original valve version needs careful tuning as the RF gain is only good when close to oscillation, it also distorts quite badly. I have not checked the valves so maybe one of them is the cause. The wavechange switching is also odd as the aerial coil has 2 positions but the anode coil has 3 so the mid-position does not always require the aerial tuning capacitor to be reset to the beginning.

The original E415 (154V in UK version) are indirectly heated. Maybe yours had been modified to use a 2V directly heated valve?

[boiss]

My valves are or should I say would be on 4V AC, The 4V heater wiring has 2 120 Ohm resistor one from each side and they go to the "Chassis" so it is like a centre tap. As I said the B442 equivalent that the radio originally had was directly heated.

OK now I have added the other LND150 in place of the B442. After a few tries I found that a 470R resistor in the source to chassis gave the best results without a decoupling capacitor as the gain was too high and caused instability. Volume control works better too. Ids is around the 2.5/3 mA mark like the valve.

The radio now works although the regen is a bit violent, sensitivity is good but I think I need to play with the decoupling on the Regen Mosfet to calm it down a bit. Instability here is also a function of the long wiring etc I reckon. Only other problem is to reduce the hum, this is not from the HT or -Ve smoothing, increasing the Cs makes little difference, I think it is due to the switching of the 1N4007s so I need to look for some 10nF 1kv caps.
I will tidy up the Mosfet mountings etc with small bits of veroboard.
Anyway these are the pics of how it is now.

[boiss]

I made up a piece of veroboard with the components from yesterday (B442) switched on no sound HT pulled down to less than 20V. ARGH ! Disconned the drain Lead. HT returned. Burning smell from inside side panel. Mosfet is fried. Checked cct all OK Zeners and R OK.
No shorts on board. Voltages on Radio OK. Swapped Mosfet and tried again, Smoke from Mosfet and volume control very stiff. Found that if there is a short on the HT the -Ve goes way high and overheats the Volume pot. Obviously why the resistive part was offered as a replacement. I think I will isolate the HT su;pply going to the B442 and fit a 100K resistorand decouple it. This is because the HT comes up before the -Ve and overloads the Mosfet, Max dissipation 0.7W. Without bias and only the tuning coils as a load the I max of the Mosfet is 30 mA and with 150V on it. I would have thought it would self bias off with the 470R in the source. But I have 2 mosfets more so I will play it safe. I will just have to trace out this nightmare wiring. Sad to have to modify something on the radio. Odd that it was fine all the time yesterday!!

[boiss]

Originally Posted by PJL

My original valve version needs careful tuning as the RF gain is only good when close to oscillation, it also distorts quite badly. I have not checked the valves so maybe one of them is the cause. The wavechange switching is also odd as the aerial coil has 2 positions but the anode coil has 3 so the mid-position does not always require the aerial tuning capacitor to be reset to the beginning.

The original E415 (154V in UK version) are indirectly heated. Maybe yours had been modified to use a 2V directly heated valve?

Does yours hum?

[boiss]

I had a bit of a disaster and lost 2 LND150s with the new veroboard. Couldn't work out what was going wrong then I saw it arcing across the tracks (quite wide spacing 1.5mm) Cleaned the tracks but lost a second LND. Leakage between the tracks. put it down to the old bakelite veroboard.
Made a new cct on some fibreglass PCB with the dremel. ( No ferric Cloride)
After having tinned the copper I made some measurements between tracks, around 5Ms. Cleaned with surgical spirit. OK.
Mounted everything and cleaned again, Ok. Fitted the board and it is working Ok.
Limited the Ht current on the RF section with a 39K and decoupled to chassis.
Working fine now, albeit a bit of squegging and howling from the regen. Biggest problem is the hum which is quite noticeable.
2 0pics of the first failed board and the second OK one.

[boiss]

Just Noticed the transformer voltage is also visible on the Transformer cover, See the RH pic.
I used flux meant for surface mounted component soldering, never imagining any probs with it. I put the testmeter probes in the flux and got a couple of Megs resistance. This is no good for high impedance devices. It needed to be cleaned off twice.
The old Veroboard I used was a different pitch and larger spacing compared to the newer stuff which incidentally is also quite bad if you measure between tracks without doing anything with it. OK for low voltage low impedance stuff.
I tried a 1925 BTH horn speaker, it was rather lousy, OK for voice, but music really LoFi.
Tomorrow E415 onto a piece of PCB, Fingers crossed.

[boiss]

Mounted the LND and all components on the little pcb, Negative, gain has gone, regen is all to pot. Seems to like long wiring and instability. Think I'll leave it for a day or two.

[boiss]

I think the LND150 got a bit stuffed while putting it on the little PCB. Trying to mess with it as little as possible. Removed the 10uF cap and put a 47nF in it's place so I have 50nF across the 10K Source resistor. Seems to be back where it was albeit the PCB is a bit messy. Ids is about 0.6 mA so there should not be a problem with core saturation of the interstage transformer. Definitely a bit of HF instability, will try some ferrite beads in a few strategic places in the next day or 2. Wondering about the hum, the O/P bottle has a centre tapped directly heated cathode. I have tried an old 2K Horn, sound wasn't very loud and hum was noticeable. Maybe the valve is a bit tired. The only mod to the radio so far has been to add a 39K resistor going to the reaction tuning coils with a 2uF cap for decoupling. This limits the possible current to the (B442) Mosfet to about 3mA at 150V well within the max dissipation limit. After the ferrites, next step is to attach the phones to the O/P of the interstage transformer without the valve and have a listen, If it is good then a FETand BJT combo will go on the base which I made up already.

[PJL]

As I recall, hum was amongst the problems with my set but as it still has its original valves I put it down to a likely faulty valve and did not try to mess around with it.

This was also using an old horn speaker so would not expect much 50/100Hz out of it but I will try with something a bit more modern! and something better for an aerial too which was just a short wire...

[boiss]

Had a bit of try with ferrites on the HT line, RF stage drain and Det drain.
regen is much smoother now and hum is maybe a bit better. My house is incredibly bad for RF pollution and the wide bandwith really makes it worse.
I tried the phones on the Interstage tranny and there was hum there too. The quality of the audio was better though. The Gain of the O/P bottle I would guess at about 4. Very dodgy running this without the O/P bottle HT goes from 155V to 243V. Something a bit weird? The 2.7Ks after the 1N4007s drop 15V, About 5.5mA so the total HT I is about 11 mA. Need to check it on Duncan's PSU calc.
O/P stage I is about right as my 2 MOSFET valves only take about 1mA.
I will try and knock up an O/P valve with a 2n3819 and a BF47*. I tried a horn speaker, volume was a bit low and it was pretty poor quality. Another possibility is to use a small 220/12V toroid and an 8R speaker. I have some excellent sounding 3" Grundig units. I looked at the O/P transformer, It doesn't appear to be gapped and the distortion is LF, maybe the 10mA is saturating the core.
So 3 things to sort, Volume, Quality and Hum.
Here is a pic E415 PCB is a bit messy, will tidy later

[boiss]

Originally Posted by PJL

As I recall, hum was amongst the problems with my set but as it still has its original valves I put it down to a likely faulty valve and did not try to mess around with it.

This was also using an old horn speaker so would not expect much 50/100Hz out of it but I will try with something a bit more modern! and something better for an aerial too which was just a short wire...

Interesting you had the hum problem too and distortion. My antenna is about 10 ft of wire along the back of the sofa
Will be glad to hear of your results
M

[boiss]

Back on it this afternoon, Reduced the resistances in series with the 1n4007s to 110R Ht jumped to 195V from 155V. O/P bottle now gets a bit hot. before it was tepid. Audio much better quality. Music can be listened to. Changed Cs in PSU to 33uF and 20uF. Hum much reduced and HT the same. Regen stage draws 0.15 mA, LND a bit hot to touch 50C, I will make a heatsink from an Electrical connector. It is only dissipating 30mW. Tried increasing Smoothing Cs but it reduced the V, maybe due to higher leakage I through the caps. As it stands now it is quite good and useable. I would still like to make a SS O/P stage and not use the original valve.
So far only changes to original rado are a resistor of 39K and decoupling cap 2.2uF to deck and increasing smoothing caps. Duncans PSUD showed about 15V ripple with original values, now about 2V. No changes to original wiring. Noise floor here due to electro smog is very high unfortunately. Need to carry on with it as I have no base panel and just to get the perfectionists going, some LED scale illumination ! Aye AYe Aye!

[boiss]

Optimised the C13 and C14, now 47uF and 22uF, Hum just discernible, HT 199V. Sound quite reasonable now.
Tuning is a bit odd, correlation between tuning caps is a bit hit and miss. Just noticed I have a problem with CV2, It is shorting in 2 positions, Nothing to do with the correlation but annoying. Need a good look with a lens and plenty of light. hopefully a bit of corrosion, otherwise? I have some old very weak Yellow leds, In series with a 1k resistor across the heater supply maybe from the (E415/LND) base, they would be great for the tuning scales and modify nothing on the radio.

[Rich Woods]

I must admit the definition here of “non modification” is making me smile. Personally I would have converted to take more readily available contemporary valves as it would definitely have a more authentic feel to the restoration. Nevertheless an interesting project. Each to their own I guess.

[PJL]

Originally Posted by Rich Woods

I must admit the definition here of “non modification” is making me smile. Personally I would have converted to take more readily available contemporary valves as it would definitely have a more authentic feel to the restoration. Nevertheless an interesting project. Each to their own I guess.

Converting to modern valves is not easy on this set. The valve sockets are B4 with the 5th connection on a screw connector on the side of the valve base and there is insufficient headroom for adaptors. There is a picture of mine with the valves in place here https://www.vintage-radio.net/forum/...d.php?t=113519.

[Rich Woods]

Ah, ok I see where you are coming from. I did mean more common valves as contemporary to the originals as possible. I now get that the side connection would make things “interesting” and necessitate valve base changes. .

[boiss]

The biggest problem is the heater voltage ie 4V or actually just over 3.8V
Apart from the 2V battery valves and 1.5V DLs, Dfs etc there is not much choice and then there would be problems with the HT deing so high also, Too many changes. I Have not modified the radio apart from the 39K resistor and 2.2Uf which current limits the the Mosfet current at switch on. The main smoothing caps have been replaced with ones of larger values and higher WV. but It would have been a good idea anyway due to the hum. The caps in 1928 were very big and physically the biggest they probably could have used in those days. I have retained side contacts top caps etc. In reality I am running the original O/P valve and all 3 of the other positions could have the original valves inserted without any modification at all. The "Ia"currents drawn are approximately the same as with the valves within 0.5mA. The heater have been replaced with 100R Resistors all within the false bases. The only heat is from the O/P valve. I made heatsinks for the fets from Electrical connectors and small pieces of aluminium.which hold the TO92s very well with a bit of silicon grease they are not even warm now after several hours.
The radio is working very well and will pull in a few stations with just a couple of feet of wire for an antenna even with the local electrosmog.