Ekco A274 - Project No 2

[Barnjet]

Managed some time back on the radios and pleased to report 2 new developments with this project.

Firstly, I decided to sand off all the polyurethane varnish that had been applied sometime during its history. No power tools as this would have risked completely ripping off the remaining veneer trim. As it was, some trim was lost even using a hand sanding block very gingerly!

I used my band saw with a sharp 1/4 inch blade to cut some paper thin pieces of walnut veneer, which I could then shape very accurately with scissors. I also removed the glass dial so that I could get tons of old varnish crud out of the corners. While the dial was out, I touched up some of the missing paint. Finally, I used a light oak wood stain to bring up the grain colour and finished the cabinet off with 3-4 costs of clear Briwax wood polish. I'm pleased with the result, though the photo doesn't do justice to its final appearance.

The second development is the discovery that the Tuning Indicator does indeed work! I just couldn't see it in bright daylight.

[Sideband]

Originally Posted by Barnjet

The second development is the discovery that the Tuning Indicator does indeed work! I just couldn't see it in bright daylight.

It is very common for the high value resistors that feed the anode and target pins of the TI go even higher in value so you might like to just check those and replace if they are way out (typically a 470K feeding the target might be found to read several megohms). If indeed the resistors have gone high in value, the simple act of replacing them might bring a worthwhile improvement to the TI making it brighter.

[paulsherwin]

Dimming is the standard magic eye failure mode though.

You should be able to clean the dirt from the speaker apertures by removing the speakers and spraying the fabric with some foaming cleaner. Agitate the foam with a paintbtush, then gently wipe it off from both sides at once using two soft rags. I'm pretty sure the fabric is synthetic in the A274 so won't shrink (it is in the A277).

[paulsherwin]

Reopened by request.

[Barnjet]

During some unrelated maintenance on this set, I noticed a red glow from the top of the V5 output valve (EL86 vice OEM EL84 - see photo).

I suspect this is a fault condition and the anode is glowing red due to over-current. My limited understanding of valve functioning is that this may be caused by a leaky capacitor putting the incorrect control grid bias on the g1 pin. That points the finger at the AF coupling cap, C31(see extract of cct diagram). However, this cap was replaced with a new component. So I have taken some pin voltages - see photo Trader sheet values are in brackets. All measurements wrt earth.

First, can members confirm this is a fault condition; secondly, is my initial diagnosis correct (i.e C31leaking) and thirdly do the pin voltages reveal anything?

Thanks

[Nuvistor]

Yes a faulty grid coupling capacitor can cause excess current to flow through the valve. Have you checked the cathode bias components, resistor and capacitor also the 220pf capacitor that connect from anode cct to G1 cct?
If the cathode components are correct the 10volts indicates around 70 ma flowing through the valve, max is around 53 and in this circuit should be around 40 ma.
If the grid coupler, cathode components and 220pf are good you could have a faulty valve, the valve base appears to be ceramic so I would not expect that to have a leakage.

[paulsherwin]

If you replaced C31 with a modern plastic film type then it's very unlikely to be bad, and the grid volts are roughly where they should be.

The valve is certainly overcurrent though. Did you change C35, the cathode electrolytic? A leaky cap there might cause these symptoms.

[Sideband]

From that photo, I don't see the anode glowing. I see the cathode glowing, looks normal, centre of the cathode is brighter due to the heater, normal and I see a reflection of these glows on the getter ring. If the anode IS glowing a feint cherry red, best capture this by taking the photo without flash and the lights off. At present, that photo doesn't show up anything abnormal.

It does appear from your voltage reading though that the valve is drawing excess current assuming the 10.4V cathode reading is correct. A very quick check is to short g1 (pin 2) to chassis. This won't hurt the valve. If the cathode volts reduce down to nearer the published figure, then it's a sure indication that the coupling capacitor is faulty. If you used a good replacement, this is unlikely. It could be the valve going in to 'grid current' due to an internal leak. Only a replacement valve will cure this problem. I seem to recall that EL86's can suffer from this problem. Best to use the correct EL84 if you have to replace it.

If you can't find a new one at the right price, I can probably find you a good one from my 'pulls'.

[paulsherwin]

The anode voltage does suggest the valve is overcurrent though, even if it's not enough to make the anode glow.

[Sideband]

Yes indeed. To help Barnjet, two voltages give the clue, high cathode volts indicate excessive current through the cathode resistor combined with the low anode volts. If the cathode resistor is 150 ohms then 10V across it is (V/R) 10/150 = .069A (69mA) so you have 69mA flowing through the valve. If the correct cathode voltage is supposed to be 6V then 6/150 = .04A (40mA) which is within the rating of the valve.

[Leon Crampin]

The EL86 is not a suitable substitute for an EL84. The anode current is much higher with the same bias and screen voltage, and the anode load will be mismatched. The EL86 is more like a 6.3V version of the UL84.

A much better substitute is the 6CH6. Rewire the base and shunt the cathode bias resistor to give a total value of 100 Ohms. The anode current and load matching will then be correct.

The 6CH6 is a far more reliable device than the EL84 and is much cheaper.

Leon.

[paulsherwin]

If there isn't a fault present, then you could simply increase the value of the cathode resistor. This is often a good idea anyway, as it reduces stress on the output valve and related components, and only has a marginal effect on perceived maximum output volume.

Leon is right that there will be an output impedance mismatch with an EL86, but that shouldn't be significant in practice.

[Sideband]

Barnjet, I have some EL84's. Let me know if you want one. I've never had any problems with EL84's other than normal replacement particularly after a fault.

[Barnjet]

Thanks - lots of advice to digest and things to check. In order of simplicity

Originally Posted by Sideband

A very quick check is to short g1 (pin 2) to chassis. This won't hurt the valve. If the cathode volts reduce down to nearer the published figure, then it's a sure indication that the coupling capacitor is faulty. If you used a good replacement, this is unlikely. .

Good tip, thanks. Check conducted and no reduction in cathode voltage confirming that the new coupling cap C31 is good

Originally Posted by paulsherwin

Did you change C35, the cathode electrolytic? A leaky cap there might cause these symptoms.

No, I did not, so C35 was my next focus

Originally Posted by Nuvistor

Have you checked the cathode bias components, resistor and capacitor also the 220pf capacitor that connect from anode cct to G1 cct?

So I checked C35 (V5 cathode by-pass, 50 microF) by lifting the cathode end from pin 3. See photo. Using a DMM cap tester, it showed 16-19 microF, unstable resistance, but more importantly zero current when I put a 12V battery across it, with the DMM in series.

R23 (V5 grid bias, 150 ohm) measured 176 ohm with one end lifted

C33(part tone control, 220pf) lifted one end of this unusual looking (to me) capacitor and tested at 20nf but was OC on resistance and showed zero current with 12V DC. See photo.

For good measure, I also tested R22 (V5 anode stopper, 27 ohm) which showed 29.2 ohm

Originally Posted by Leon Crampin

The EL86 is not a suitable substitute for an EL84. The anode current is much higher with the same bias and screen voltage, and the anode load will be mismatched. The EL86 is more like a 6.3V version of the UL84.
.

Noted - replacing the EL86 is probably the best course of action and seems to be the majority position. But before I do that, are there any other checks I should perform to ensure I have identified the correct fault condition, namely the EL86 valve.

Originally Posted by paulsherwin

If there isn't a fault present, then you could simply increase the value of the cathode resistor. This is often a good idea anyway, as it reduces stress on the output valve and related components, and only has a marginal effect on perceived maximum output volume.

Leon is right that there will be an output impedance mismatch with an EL86, but that shouldn't be significant in practice.

Thank you, this may well be a work-around solution. What sort of increase in R23 would be sufficient?

Originally Posted by Sideband

Barnjet, I have some EL84's. Let me know if you want one. I've never had any problems with EL84's other than normal replacement particularly after a fault.

Great thanks - I'm thinking this may be the simplest path to take.

thanks

[Nuvistor]

I thought 220pfd would be 0.22nfd, 12 volts is too low a voltage to test the capacitor with, it could well be ok but it appears to be a wax paper dielectric type, not the most reliable after 60 years.

[Barnjet]

Originally Posted by Nuvistor

I thought 220pfd would be 0.22nfd, 12 volts is too low a voltage to test the capacitor with, it could well be ok but it appears to be a wax paper dielectric type, not the most reliable after 60 years.

Yes that’s what I expected, but the cap tested at 20nfd, about 100 times higher than spec. Maybe it’s sensible to replace it anyway.

[paulsherwin]

I'm sure people who know more about valve operating conditions than I do could tell you exactly what cathode resistor value to use, but I would just use trial and error. So, if the original value is 150 ohms, try increasing it to 200 or 220 - whatever you have to hand. If it's still overcurrent, try 270 or 330. The objective is to get the cathode current down to about 40mA. You can reduce it even below this if you don't mind some reduction in maximum volume.

[Nuvistor]

While you are substituting the bias resistor I would replace the 50ufd cap as well, also use an EL84.
A slightly higher bias resistor would be no harm to an EL84 either.

[Barnjet]

Originally Posted by paulsherwin

I'm sure people who know more about valve operating conditions than I do could tell you exactly what cathode resistor value to use, but I would just use trial and error. So, if the original value is 150 ohms, try increasing it to 200 or 220 - whatever you have to hand. If it's still overcurrent, try 270 or 330. The objective is to get the cathode current down to about 40mA. You can reduce it even below this if you don't mind some reduction in maximum volume.

Thanks - a general question about purchasing resistors, as I don’t hold any. Is it better to buy a “selection” of resistors in a mixed bag, so to speak. That seems cheaper and more expedient in the long run.

[paulsherwin]

Absolutely, you should keep a range in stock or you'll constantly be ordering single items, which is both expensive and inconvenient. This applies to all commodity parts, not just resistors. Forum members often dispose of 'goody bags' here, and there are bargains to be had on eBay.

Scrap electronic equipment is also a cheap source of spares - a 70s or 80s TV or cassette deck will be full of useful stuff to get you started.

This is OT for this thread though - please start a new thread if you want to discuss sourcing components.