|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
5th Sep 2019, 2:14 pm | #1 |
Hexode
Join Date: Jun 2012
Location: Chester, Cheshire, UK.
Posts: 421
|
ECL86 issues
Afternoon all, whilst servicing one of the hacker cavaliers I have, I noticed severe distortion after a while. Checking valve voltages lead me to the grid of the pentode section of one of the push pull pair, it was reading several volts. (I had already replaced the grid coupling caps) swapping the suspect valve with its partner made the voltage follow the valve so I obtained another (used but tested good). This one while better still has a volt and a half on the grid after about 5 mins of operation. Are ECL86s prone to internal leakage like the good old UL41?
Thanks Rob |
5th Sep 2019, 3:18 pm | #2 |
Dekatron
Join Date: Nov 2009
Location: Lincolnshire, UK.
Posts: 5,000
|
Re: ECL86 issues
Absolutely! This is very common in these and many other valves - EL84 etc. etc.
This is why valve testers are a complete waste of time for this type of fault in my opinion and are really just a 'nice to have toy' to play around with. As is always repeatedly said, the best test for any valve is in the circuit it's intended to work in. That valve would have been plonked in a tester for a couple of minutes and tested good with no other faults showing up. You take a chance with a used tested valve and the only real answer is to buy unused old stock. I don't have a valve tester. I used to have two of those one arm bandit Mullard high speed devices. They were completely useless for any sensible testing and I very regretfully dumped both of them when I move house a couple of decades ago. I really regret dumping them (well I didn't actually dump them, but left them in the garage of the old house when I moved), because I'm now much more serious about interesting old vintage items and would have enjoyed 'playing' with them now, but you can't turn back time and I can't say that I'd have had any real use for them in the last 20 odd years and hadn't used them for a good decade or more before that. If I really need to test a valve, I rig up a test circuit with power supplies, Avo meters and a load of 4mm plug and croc clip leads and take readings and sometimes plot a graph. Having said all that, if a valve tester came my way at a sensible price, then I certainly wouldn't turn it down, as it would be a nice vintage item to play with and restore in the workshop - I wonder what happened to my two old Mullards? |
5th Sep 2019, 4:19 pm | #3 |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,088
|
Re: ECL86 issues
As Techman says, really.
I do not have any experience of ECL86's (I have NOS one which I have never used), it is an impressively spec'd valve, high power and with a decent amplifier triode thrown in as well. But in a B9A bottle, it must get stinking hot, and suffer from the troubles that arise after prolonged operation at high temperatures. |
5th Sep 2019, 4:59 pm | #4 |
Octode
Join Date: May 2009
Location: Stevenage, Herts. UK.
Posts: 1,518
|
Re: ECL86 issues
Yes ECL86s are prone to this. Even Mullard seemed to struggle so it's no surprise that the Russians and Chinese don't seem to make modern versions. Tech man is a little hard on valve testers, put one of these in my AVO VCM and anode current will magically rise after a minute or two.
|
5th Sep 2019, 5:26 pm | #5 |
Dekatron
Join Date: Aug 2013
Location: Wigan, Greater Manchester, UK.
Posts: 9,433
|
Re: ECL86 issues
Have checked the grid leak resistor, if it’s gone high in value it will not help.
__________________
Frank |
5th Sep 2019, 7:31 pm | #6 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: ECL86 issues
The GP42 says the cathode resistor is 220 which seems low before making allowances for mains voltage/component tolerances. Push-Pull ECL86 will deliver enough power to propel the speaker cone across the room so I would reduce the quiescent anode current.
|
5th Sep 2019, 8:54 pm | #7 |
Nonode
Join Date: Mar 2008
Location: Aberdeen, UK.
Posts: 2,858
|
Re: ECL86 issues
Back at the weekend, a VMARS chum phoned me on the scrounge for two ECL82's & an ECL86.I found all three, boxed but untested. Popped them in my AVO MK3 - after warming up, both 82's tripped the testers alarm after about 10 seconds of drawing Triode Ia. But the pentode halves were OK. The same thing happened with the 86, but its pentode half failed. All three exhibited a "Gas" reading. The 82's were a Mullard & a Brimar, and the 86 was a "Ds BANKS"(Never heard of them ?).
I seem to remember, some time back, one or two Forum folk mentioned that these types of valves were unreliable & short lived. Most VCM's & Testers have the advantage of indicating a Gas problem, as well as low inter-electrode resistance. Some can give a reasonable indication of Ia & subsequent mA/V. Having recently had hundreds & hundreds of old ex "silent Key" valves dumped on me, my MK3 has done sterling service weeding out all the clapped out rubbish. To do this by "substitution method" - I would need a hundred different radios - with all their different valve usage & valve holders - plus working eight hours a day for a month. Right enough, don't go to the expense of buying a decent valve tester if you only need to occasionally change a suspect duff one. I would recommend Forum folk who live near to others in a locality - set up a pool of spare valves &/or chat up an obliging someone nearby who has a tester. Regards, David Last edited by David Simpson; 5th Sep 2019 at 8:56 pm. Reason: alteration |
5th Sep 2019, 11:33 pm | #8 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: ECL86 issues
Probably the only likely way the grid electrode could acquire a positive charge is if it was being bombarded positive ions. Normally there should be very few of those in the valve, provided that is, that the metallurgy of all the electrodes is high purity and specifically that metalwork is not raised to a temperature above that which was present when the valve was evacuated at the factory. This of course is the factor that sets the published specification on a valve's maximum dissipation. If the temperature is raised above that manufacturing value, contaminants exit the metalwork and "poison" the valve.
Probably a lot of new manufacture clone valves have issues with the metal purity and might not have been out-gassed at the correct electrode temperatures, so in use the problem crops up, but it might not in a lower power dissipation application. |
6th Sep 2019, 5:37 am | #9 |
Nonode
Join Date: Aug 2007
Location: Walsall Wood, Aldridge, Walsall, UK.
Posts: 2,873
|
Re: ECL86 issues
Hi!
I have to confess I've always been lucky with ECL86s – all those I've come into contact with have been absolutely fine with no noticeable grid–emission or leakage! I will add, however, that Mullard's "Low Consumption EL41" operating conditions, (Va = 250V, Vg2 = 210V, Rk = 220 Ω) , which is also applicable to lower–power use of the EL84 as well, does go a long way towards prolonging the life of these valves – the Marconi 4238 reel to reel I sold has it's ECL86 used under these conditions, and it's o/p valve was still in first class condition with plenty of volume and almost total absence of hum and noise! Chris Williams
__________________
It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed! |
6th Sep 2019, 7:00 am | #10 |
Hexode
Join Date: Jun 2012
Location: Chester, Cheshire, UK.
Posts: 421
|
Re: ECL86 issues
Hi Nuvistor, All resistors have been checked and any out of tolerance were replaced . As the fault follows the valve I'm going to have to bite the bullet and buy a New one. At least I should be able to get a refund on the one I have.
|
6th Sep 2019, 11:33 am | #11 |
Octode
Join Date: Nov 2005
Location: Surrey, UK.
Posts: 1,874
|
Re: ECL86 issues
This is a common problem with high dissipation output valves mounted in small glass envelopes - B8A, B9A and B7G.
The cause is usually an excessive g1 temperature, giving rise to grid emission with the resulting rise in anode current exacerbated by gas - often released because the valve is just running too hot. Fortunately, there's a dead simple test for this for those without a valve tester. Assuming autobias with a cathode resistor, simply run the valve in its equipment with a DVM connected between g1 and earth - ie across the grid leak resistor. A good valve will probably show a small positive voltage, but this must remain practically stable as the valve heats. A poor valve will show an increasing voltage over several minutes of operation (I test for about 15 minutes, unless the valve is so poor that the test places other components at risk). For a 12W pentode, I would expect no more than 100mV positive grid voltage, stable after 3 minutes or so. I have found EL84s to be very poor in this respect, and ECL86s little better. You can't miniaturise the Watt. Beam tetrodes are noticeably better than pentodes in this respect - I suspect grid alignment gives higher efficiency and results in a lower g1 temperature. The 6BW6 and the 6CH6 are my devices of choice - unless there's room for an octal or loctal valve. The above test assumes that, of course, the grid coupling capacitor to the previous stage is above reproach. Leon. Last edited by Leon Crampin; 6th Sep 2019 at 11:43 am. |
6th Sep 2019, 12:12 pm | #12 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: ECL86 issues
We used to keep the "puckle" cards permanently out of the Mullard HST card box so they were ready to go to cut testing time down when doing rental recons.....no pressure
Lawrence. |
6th Sep 2019, 1:48 pm | #13 | |
Dekatron
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 14,007
|
Re: ECL86 issues
Quote:
Brimar seemed to get it right with the 6BW6 - Mullard/Philips never seemed to get the EL41/84 quite right even to the end. |
|
6th Sep 2019, 7:26 pm | #14 | |
Heptode
Join Date: Oct 2008
Location: Aalborg, Denmark
Posts: 903
|
Re: ECL86 issues
Quote:
Another way is to use a valve tester built for that same purpose. Have a look at: https://www.vintage-radio.net/forum/...d.php?t=148050 Rgds, /Torben |
|
6th Sep 2019, 8:53 pm | #15 |
Nonode
Join Date: Aug 2007
Location: Walsall Wood, Aldridge, Walsall, UK.
Posts: 2,873
|
Re: ECL86 issues
Hi!
Member Leon's explanation is exactly the same as I could have added, the only however I need to add is that the 6V6/6BW6 have only about half the slope of an ECL86 and about 45% of tbe slope of the EL84, so if you're replacing these valves you should use the high–slope 6CH6 – this was originally intended for video amplifier use, but with base rewiring and the correct bias resistor it should give virtually the same performance as the Mullard types! Using the 6BW6 might run into volume problems if the output voltage of the cartridge is on the low side – it's slope is only 4.5 mA/V! If you can get another (hopefully sound!) ECL86, try reducing the G2 volts on pin 3 to about 180–200 max and altering the cathode resistor on pin 7 to 220Ω if it isn't already! Chris Williams
__________________
It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed! |
6th Sep 2019, 11:00 pm | #16 |
Hexode
Join Date: Jun 2012
Location: Chester, Cheshire, UK.
Posts: 421
|
Re: ECL86 issues
Thanks chaps for all the really interesting info, I have learned a lot from this. Unfortunately, the seller won't refund me as he claims he doesn't machine test his valves, only tests them in equipment to see if they are working. No voltage tests etc, I'm tempted to go down the claim route but it hardly seems worth it for the bad feeling it will undoubtedly cause
|
7th Sep 2019, 10:50 am | #17 |
Nonode
Join Date: Mar 2008
Location: Aberdeen, UK.
Posts: 2,858
|
Re: ECL86 issues
Rob,
I would reiterate my suggestion, back in post 7, for vintage radio folk to get together with other enthusiasts nearby, & consider pooling surplus decent valves. If someone has a valve tester - all the better. Another option would be to build your own basic DC tester. Thoughts & designs have been talked about extensively in past Forum threads, over the years. If your field of vintage radio collecting/repairing uses mainly B7G, B9A, IO, & MO, then with simple wafer switch circuitry it should be a reasonably easy task. Especially if two or three heads get together. A fiver spent here & there at BVWS Swapmeets & other large sales/auctions, or at local ARS Junk sales should source 2nd hand switches & pots, meters, valve holders, LV & HV transformers, & general components. For measurement of Ia, then using an AVO8 on its trip protected DC mA Currents ranges is a good idea. Many of us valve & valve testing enthusiasts will be saddened to hear that your valve supplier doesn't test his valves properly, or offer a refund. In my experience, and hearing of other folk's experiences, there was a common practice in years past, for Radio/TV repair folk to put suspect "Pulls" into the boxes that new valves came in. The fact that many a valve has corrugated cardboard &/or old brown paper wrapped around it - is no guarantee that it is NOS. Heaps of these old "squirreled away" boxed valves still appear from time to time. Regards, David |