Hi I am just wondering if anybody knows what may be the cause of a problem I have. When I got the Lekmek, the candohm was burnt out between each section and the cathode resistor on the 1st "detector" valve was completely open. two other 500k resistors were high one was 540k and the other 570k. I don't have the 425 ohm one but made up 3x 1 watt ones to get 400 ohms for the detector cathode resistor and I am waiting for a 425 ohm 5 watt one to come from the US. Now the radio works fine, but the + side of the candohm is sitting at 285 volts, but the schematic says it should only be 225 volts. The other two voltages at the candohm taps are correct at 100V and 50V as I simply moved the taps to obtain the correct voltage. As I said, it works fine for about ten minutes then starts to sound "Sketchy" if I turn the volume down to a whisper of up until it is blaring, it sounds good, but if I turn it down to where it is a nice listening volume it sounds sketchy.

Now I have swapped out the two high 500k resistors replacing them with 470k ones as I couldn't 500k ones. Expecting this to sort out the sketchy performance, I was dismayed to find that in less than ten minutes it started sounding thin and sketchy again.

I cannot find any reason why the HT voltage is too high, it comes through the speaker field coil and that's what I have 285 volts. All I can guess is that someone before me may have replaced the transformer, as it was loose when I got it. The new candohm does appear to get hot but I thought that would be normal and if it were drawing too much current, I would expect a voltage drop, not increase? Maybe the candohm getting hot is upping it's resistance perhaps? Not sure... Is there a simple way of losing 60 volts, maybe using an output between two beefy resistors, one to chassis and the other to HT? Any thoughts?

I can think of a couple of reasons. But the model & circuit is always useful.

If it has a tapped primary, and many did, what is it on? I am getting 245V on the mains. So 240-260 is appropriate. Check it.

HV is often caused by a valve not drawing current & with that much voltage that's on the cards and likely the output. The voltage across the cathode resistor should indicate that.

Did you up the cap values of the filters especially the first one. Apart from straining friendship with the cathode / filament of the rectifier, that will produce extra volts. What is the AC Plate to CT volts at the rectifier? Not P to P you need a 1kV+ meter for that one.

Is the field coil actually in series with the B+?. 310V DC at rectifier cathode, is not unusual for 250V on the load side of a choke / field coil /resistance.

Marc

Sorry it's a Model 540 1938 Console. I've tried swapping 42 valves but no change, same with 80's. I've used 10μF 600 volt filter and reservoir caps, originals were 8μF. I'll measure those voltages again next time I pull the chassis out. Once that 5W 425 ohm resistor arrives, if I get a chance before that. Field coil is in series, but there's three wires coming from the speaker going nowhere. One from each side of the voice coil and one from the output transformer, the other output transformer wire goes straight to one side of the voice coil. To get it to work I've joined the one from the output transformer to the one going to the side of the voice coil not connected to the transformer. These wires are ancient enough to be original, but what for? Another speaker? Or resistor perhaps...

If its electrodynamic the VC side has normally 3 nodes and 2 windings secondary side, that is normally shown on the circuit: Thats the Hum buck winding that goes in series with the VC. Often these carry an earth back to chassis.

Set is back biased, is that wired correctly? CT does not go to ground, back bias resistor looks like 300 Ohm? Should be 16- 20V across it: Is it missing? (Chassis positive)

400VAC into an #80 with capacitor input yields around 390VDC.

Check the heater volts if they are high Primary tap should be on black & green for an original transformer.

Looking at the speaker plug, if rewired It can be messed up. #42 is drawn as a Tetrode & its not, its got a suppressor grid, so its a Pentode.

Is the 80 really an 80? With directly-heated filaments that look like toaster element wires? Or is it an indirectly-heated rectifier, something like a 5Z3 plugged into an adaptor? These are a bit more efficient and so will deliver more volts into the load.

What is the DC resistance of the speaker field? 1500 ohms or 2500 ohms? Maybe the speaker has been changed.

If you have the wrong speaker, it's possible to insert some resistance to knock the volts back. The "kindest" place to put it is from the 80 cathode (i.e. 5v winding on transformer) to the input filter capacitor, which then goes on to the speaker field coil. Start with 47 ohms 10 watts.

I would agree on checking the entire PSU for correct parts. A resistor ahead of the field with no input cap will tend to behave as a choke input filter.

One can never predict what bizarre mods & mistakes are inflicted on some sets in their lifetime. I have one on the bench that has never been touched, as opposed to the one next to it that had wrong parts & wiring errors. For info neither were safe to plug in. Weeping can caps say no way, for the untouched one.

Well I've got the voltages better. There was another tap on the transformer with slightly higher resistance than the one the active mains was connected to so I connected the mains to it and got 248 volts after the field coil, then I adjusted the candohm taps to get the required 105V and 155V and was hoping that I'd have it sorted, all put together and the 425 ohm cathode resistor in place, but no... After 15 minutes it starts sounding like a kazoo until I juice up the volume? I'm thinking of putting a thick speaker grille cloth in so I can have the volume up louder to get a decent sound without the overall sound being too loud as I just can't get it to sound nice and clear at lower volume? I'm planning on putting a 47 ohm 10 watt resistor between the rectifier and smoothing capacitor to see if that drops the volts down a bit more and gets a nicer sound. Supposed to be 225V, in any case 248V is better than 285V.

I think I'm best just replacing all of the 500k resistors, there must be one going high when powered and dropping back after. That's all I can think of as it sounds great when I first plug it in and I can't think of anything else which can deteriorate over 15 minutes except a resistor going high?

A resistor can go open and that can be a cathode one where fitted. Observe the voltages, Normally if B voltages go high, the set is drawing too little current. A leaking cap can change the grid bias. An overheating resistor can cause B voltage issues.

You need to measure voltages before & after the incident. A valve is not out of the equation. AGC where fitted cannot be ruled out.

Marc

Test and measure! Don't guess!!!

If the voltages don't change more than 10% from cold to fault condition, the problem lies elsewhere.

For a start, I'd suggest you couple another audio signal into the audio path, say, from your phone, and see if the same thing happens. This so you can be sure the fault is in the audio stages or later.

If so, the timing and nature of the problem sounds to me like it could be the speaker poling as the field winding heats up. Try disconnecting the output of the speaker transformer from the speaker and connect it instead to an external speaker.

Just re-reading this, your set has an anode bend detector. These are notorious for distortion. They are horrible things!

The key is the cathode bias. The valve characteristics will change with warmup. You may need to adjust the cathode resistor to get a clean sound, or swap the 58 (I assume) with the one in the first (autodyne) stage.

Well I've gone over all voltages and they are as specified on the 6A7 and 6D6 screens and plate, namely 100V and 150V only the 75 has 169V and the 42 now has 246V with a NOS valve, they are all now NOS. So I'm only 20V too high, but that is straight through the output transformer. I've gone over all of the resistors and they are all within 10%, no change after running half hour on the bench, if anything voltages dropped 1 or two volts over this time. I've fitted the grille cloth and overall I think it's sounding better. I just need to let it clock up some hours running. I'm waiting for a 47 ohm, 10 watt resistor which I might put between the rectifier anode and field coil / filter cap to see if this washes off some HT. Overall though maybe I'm over pedantic, the voltages maybe 20V high in a couple of spots baut well within the handling parameters of the valves, caps, coils and resistors, there's no evidence of excessive current draw and I'm questioning if maybe the xchematic is a bit howdy doody, as there is one anomaly I've already spotted as one valve, the oscillator has two plate voltages listed, 115 and 150 volts?? So I'm not going to stress overly, just let it run and keep an eye on it for the time being. Thank you Ian and Marc for your help. Cheers guys.

Be aware that modern multimeters present next to no load to the circuit, compared to a VTVM that would have been used to take the factory readings
20V high would be pretty close to spec IMO.

OK, so it's newer than I thought. No anode bend detector.

Try injecting an audio signal (e.g. from your phone's earphone jack) to the cap on the 75.. See if that still sounds distorted.
That will be a start and will tell me where you should go next.

You should be able to use a 3.5mm to RCA audio cable. Put the inner pin of the RCA on the grid cap (in place of the existing connection) and ground the outer of the RCA to chassis. Play some music on your phone.

Ok Ian, that's worth a try even though I am starting to think some crapola, dust and what not, has built up between the voice coil and magnet, I've dislodged a fair bit of fluffy dusty crap from around the inside underneath of the cone.