|
Vintage Radio (domestic) Domestic vintage radio (wireless) receivers only. |
|
Thread Tools |
11th Feb 2017, 4:23 pm | #1 |
Triode
Join Date: Jun 2014
Location: Reading, Berkshire
Posts: 12
|
Hum problem on Ekco U159
I'm restoring an Ekco U159 radio, and am having problems trying to eliminate a rather loud mains hum.
The circuit is a conventional AC/DC superhet with a UCH42/UF41/UBC41/UL41/UY41 line-up; there is a hum-cancellation winding on the output transformer, and all valve cathodes are connected directly to chassis, with bias for the output stage provided by a resistor in the HT negative line. I replaced the main 50+50uF electrolytics and all paper capacitors (apart from the one connected to the volume control); the 2uF electrolytic decoupling the valve screen grids was checked and OK so I left it in place. I didn't have a replacement for the 2nF "negative feedback" capacitor so left it out. A check of voltages and (where possible) currents against the service sheet shows a good match, and the waveform on the UY41 cathode is pretty similar to that given by a SPICE simulation, so I believe that the valves are generally OK, and indeed the radio works pretty well in terms of receiving stations. However, there is a very loud 50Hz hum from the speaker, independent of the volume setting, although it becomes less noticeable when receiving a signal. The two attachments are as follows: IMG_1522 - signal on UY41 cathode (sawtooth as expected, 5V/div) IMG_1532 - signal on loudspeaker (0.1V/div) Plainly, the hum-cancellation winding is doing a good job, since the waveform on the loudspeaker shows no significant sawtooth component. I tried various things to see what effect they had: 1. Replace UL41 or UBC41: no effect 2. Connect UL41 grid directly to its bias supply - reduced amplitude but more high frequency components, subjectively worse 3. Connect UL41 grid to chassis via 1000uF - large improvement 4. Connect UL41 bias supply to chassis via 1000uF - amplitude increased but more sinusoidal 5. Connect 47uF across (in turn) the HT filter capacitor, the 100n capacitor filtering the UBC41 HT and the 2uF screen grid filter capacitor - little difference 6. Restore negative feedback - amplitude increased but more sinusoidal 7. Connect UBC41 grid to chassis - no effect 8. Increase separation between UL41 grid/UBC41 anode wiring and heater connections - no effect Having read other comments here, I had already removed any connections to the unused pin 4 of the UL41 (it had been used to support the negative feedback components). Since all valve cathodes are directly grounded, heater-cathode leakage seems an unlikely explanation. I checked the UL41 grid resistor and it is reasonably close to its nominal value of 1M, as indeed are all other resistors in that area. I'm running out of ideas for what might be the cause of the problem, and would appreciate any suggestions. |
11th Feb 2017, 4:40 pm | #2 |
Moderator
Join Date: Jun 2003
Location: Oxford, UK
Posts: 27,787
|
Re: Hum problem on Ekco U159
I would normally suspect the UBC41 and UL41, but you say you've checked them. Can you see any contamination or heat damage on or around the valveholders?
|
11th Feb 2017, 11:22 pm | #3 |
Triode
Join Date: Jun 2014
Location: Reading, Berkshire
Posts: 12
|
Re: Hum problem on Ekco U159
Thanks for the suggestion. I had a look at the valveholder and it seems pretty clean and free from heat damage (though as it's ceramic I wouldn't expect this anyway). However, I also looked at the waveform on the valve pins. G1 had about 2V p-p of 50Hz which is consistent with what I'm getting on the anode (and explains why decoupling to chassis stopped the hum), G2 also had a significant sine wave but I think that was "effect" rather than "cause" since, as already mentioned, decoupling further had little impact. The UBC41 anode, however, had a much smaller amplitude, which suggests to me that the problem must be on the UL41 side, and so the valve may indeed be in some way to blame. I shall try to get it tested on a friend's valve tester.
|
12th Feb 2017, 12:39 am | #4 |
Nonode
Join Date: Jun 2007
Location: Coulsdon, London, UK.
Posts: 2,152
|
Re: Hum problem on Ekco U159
Double check the negative connections of the power supply electrolytic capacitors.
With this type of valve biasing, the capacitors are NOT connected to the chassis. Sometimes the 'cans' should be insulated from their mounting brackets. |
12th Feb 2017, 12:54 am | #5 |
Dekatron
Join Date: Oct 2003
Location: Croydon, Surrey, UK.
Posts: 7,549
|
Re: Hum problem on Ekco U159
Yes as silicon says make sure that there IS negative bias on the UL41. At least one of the electrolytics is probably isolated from chassis. If there is no isolation, the bias resistor is effectively shorted out......
Basically just carefully check the earthing arrangements in the power supply
__________________
There are lots of brilliant keyboard players and then there is Rick Wakeman..... |
13th Feb 2017, 1:19 am | #6 |
Hexode
Join Date: Sep 2010
Location: Stourbridge, West Midlands, UK.
Posts: 434
|
Re: Hum problem on Ekco U159
The can on my U215 has thick brown paper wrapped around it's body where the chassis holds it in place.
|
13th Feb 2017, 8:18 am | #7 |
Hexode
Join Date: Sep 2010
Location: Stourbridge, West Midlands, UK.
Posts: 434
|
Re: Hum problem on Ekco U159
Are you sure the hum is really a problem?my U215 hums at low volume but it doesn't really bother me,it's had new caps and out of tolerance resistors changed,my UL41 has got plenty of life still left in it,as i have noted elsewhere on the forums,AC/DC sets are prone to more hum than an AC only set.
|
14th Feb 2017, 11:40 am | #8 |
Triode
Join Date: Jun 2014
Location: Reading, Berkshire
Posts: 12
|
Re: Hum problem on Ekco U159
Thanks for the comments. The smoothing capacitor can is indeed insulated from chassis, by both its own plastic sleeve and the cardboard one that surrounded the original metal can (if not I think I'd have worse problems than mains hum!)
I tested the valves and there was no significant leakage or other problems. I also tried removing the UL41 (patching its heater into the circuit using flying leads and grounding its cathode); there was still significant 50Hz on the grid pin of the valveholder, although I think the amplitude was smaller. Interestingly, there was significant 50Hz on the valve grid pin, and an even greater amplitude on the screen grid pin, but as this were measured with all other pins open-circuit it's probably not fair to read much into it. I tried a couple of other things too. With the set cold, I looked at the waveform on the UL41 grid pin as it warmed up. The 50Hz came on immediately, then increased in amplitude slowly as the heaters warmed up (there's a thermistor that gives a slow warm-up), and fell back a little as the set started to produce output. To me, this is a clear indication that the coupling is somehow from the heater circuit rather than directly from the mains. I also looked at the amplitude with my scope probe on x1 and x10, which changed the load impedance. The change in amplitude was broadly consistent with what I'd expect from a high impedance (leakage or capacitance) feeding the UL41 grid (as opposed to the UBC41 anode, or bias supply). I calculate that this would have to be of the order of 30M or 100pF, so resistive leakage seems much more likely - and probably wouldn't have shown up on the valve tester. I will try cleaning the valve and the holder more thoroughly to see if that solves the problem. |
14th Feb 2017, 12:11 pm | #9 |
Nonode
Join Date: Jun 2007
Location: Coulsdon, London, UK.
Posts: 2,152
|
Re: Hum problem on Ekco U159
If you are using an oscilloscope on an AC/DC set I assume you are using an isolation transformer or a battery powered oscilloscope.
Volume controls with mains switches are sometimes a source of mains hum. You could temporarily remove the mains from the switch. I reduced the hum in a valve radio by fitting a screened cable to the grid of V3 in place of the original plain wire. Would it be worthwhile powering the set or just the heater chain from a DC supply? |
14th Feb 2017, 12:12 pm | #10 |
Banned
Join Date: Dec 2013
Location: Middlewich, Cheshire, UK. & Winter in the Philippines.
Posts: 3,897
|
Re: Hum problem on Ekco U159
As the capacitors were replaced, presumably one at a time, at what stage did the hum start?
Or if they were changed all in one go without testing, are you absolutely sure a man made error has not occurred? |
23rd Feb 2017, 4:05 pm | #11 |
Triode
Join Date: Jun 2014
Location: Reading, Berkshire
Posts: 12
|
Re: Hum problem on Ekco U159
Thanks for the various suggestions; I think I've finally solved the problem. As often with this sort of thing, it was a combination of factors.
I'd taken care to connect the capacitors I'd replaced correctly, but not so a previous repairer: the 50nF capacitor that should have gone from mains live to mains neutral went instead to chassis. Correcting this made only a slight improvement in the 50Hz amplitude, but cleaned up the waveform noticeably. Having done that, restoring the negative feedback again only slightly reduced the amplitude of the 50Hz but had more impact on its harmonics (not easily visible on the scope), giving a great subjective improvement to the point where the hum was acceptable in the light of what people have said. I guess the response of both the speaker and the human ear at 50Hz is not all that great, so while it was 50Hz that I was seeing on the scope, it was the harmonics that I was hearing. Moral of the story is not to assume that previous repairs/restoration were done correctly! |
23rd Feb 2017, 4:33 pm | #12 |
Moderator
Join Date: Jun 2003
Location: Oxford, UK
Posts: 27,787
|
Re: Hum problem on Ekco U159
Good stuff.
Pretty much all domestic mains valve radios have some level of background hum. This was seen as normal and acceptable at the time, and some people saw it as a welcome indicator that the radio was switched on. Some radios hum even if you take all the valves out, because the mains and output transformers interact, so hum elimination can't have been a design priority. |