UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

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.

Closed Thread
 
Thread Tools
Old 9th Jun 2008, 10:18 am   #21
GMB
Dekatron
 
GMB's Avatar
 
Join Date: Aug 2003
Location: near Reading (and sometimes Torquay)
Posts: 3,086
Default Re: Another (possibly silly) valve question

When I was a child I was told that the thing that poisoned cathodes was having electrodes disconnected (floating), especially the cathode itself.

This advice could be taken to mean that having the cathode more positive than other electrodes is the bad situation.

Floating the cathode could also result is h-k insulation breakdown I suppose (?).

By the way, over-running a valve can also be done by normal running but with reduced heater power.
GMB is offline  
Old 10th Jun 2008, 4:11 pm   #22
Steve Smith
Retired Dormant Member
 
Join Date: May 2008
Location: Isle of Wight, UK.
Posts: 24
Default Another (possibly silly) valve answer

Quote:
Originally Posted by majoconz View Post
they just wander around aimlessly. Perhaps they all pile up in the bottom of the valve and eventually short all the pins out........perhaps I'll think this through again.....
This may be close to the truth.

A few years ago I made a valve pre-amplifier for a friend for his double bass. Usually when I make something like this I use a 9 volt AC power supply which drives a reverse connected mains transformer for the HT and the incoming 9 volts AC is also rectified to give 12 volts LT for ECC type valves with heaters in series.

This pre-amp came back to me for investigation as it worked fine for a while and then the gain reduced. Usually after about an hour.

Eventually I found that I had connected the heaters in parallel instead of series so I was putting 12 volts into 6 volt heaters. Due to the extra current load, the 12 volts was actually pulled down to 9 volts. Still 50% more than expected.

After correcting this, the unit worked perfectly.

The only explanation I could think of was that after a period of excessive heating, there were too many 'free electrons' in the valve which somehow impeded the flow of 'legitimate' electrons.


Steve.
Steve Smith is offline  
Old 11th Jun 2008, 8:34 am   #23
Studio263
Octode
 
Studio263's Avatar
 
Join Date: Apr 2007
Location: Hampshire, UK.
Posts: 1,574
Default Re: Another (possibly silly) valve question

Not silly at all! Here's a little bit of text taken from Morgan Jones's excellect book "valve amplifiers" (get a copy, it's really interesting). The text refers to oxide coated cathodes, that covers most of the small valves we are likely to encounter.

"Another problem with oxide cathodes is cathode poisoning. If the cathode is kept at full operating temperature, but little or no current is drawn, a high resistance layer of bariun orthosilicate is formed at the interface between the barium oxide emissive surface and the nickel cathode structure. This layer not only reduces emission, but more significantly it increases the noise resistance of the valve.

Poisoned cathodes may be gradually recovered by operating the valve at a high current..."
Studio263 is offline  
Old 11th Jun 2008, 8:39 am   #24
ppppenguin
Retired Dormant Member
 
ppppenguin's Avatar
 
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
Default Re: Another (possibly silly) valve question

The Wikipedia entry on Hot Cathode suggest that silicon contamination is a significant problem in the reliability of valves. See Failure modes at bottom of page.

http://en.wikipedia.org/wiki/Hot_cathode
ppppenguin is offline  
Old 11th Jun 2008, 12:53 pm   #25
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Cathode poisoning in valves.

This is an interesting thread! It seems to me that if the cathode is made without damaging imputities (such as silicon mentioned above) then running the heater with no anode current will not cause poisoning. But if there is silicon present, then there could be poisoning. This could account for differing anecdotal evidence from different people.

Myself, I have always believed it to be good practice to not have zero cathode current for very long periods. But the mechanism of poisoning has not been clear to me (it still isn't why having a current flowing could prevent the barium orthosilicate layer building up).

I am aware that a 'poisoned' cathode can have strange high frequency performance, and that is because the layer of orthosilicate is only a few atoms thick so, although it adds significant resistance, the conductive materials either side give a reasonable shunt capacitance. Thus the valve has a drooping response to square pulses.

Has anyone experienced this, and what is the order of the effective time constant? Milliseconds perhaps?
kalee20 is online now  
Old 11th Jun 2008, 2:02 pm   #26
Neil Breward
Retired Dormant Member
 
Neil Breward's Avatar
 
Join Date: Aug 2005
Location: Keyworth nr. Nottingham, UK
Posts: 140
Default Re: Cathode poisoning in valves.

Interesting theme this one! If Barium orthosilicate (Ba2SiO4) or even barium metasilicate (BaSiO3) is forming, this implies a source of silicon dioxide. Is this an impurity in the cathode metal, in the original Ba-Sr carbonates used for making the cathode coating, or could it even be derived from the glass of the envelope (unlikely as that may be)? I'm intrigued!

And why would an electric current break it down? Barium oxide is stable to nearly 2000 degrees C, but maybe the silicates will dissociate into BaO and SiO2 again in the environment of an actively-emitting cathode surface?

I'll try to look up some dissociation energy values . . . . . .

Cheers,
Neil
Neil Breward is offline  
Old 11th Jun 2008, 2:24 pm   #27
Leon Crampin
Octode
 
Join Date: Nov 2005
Location: Surrey, UK.
Posts: 1,869
Default Re: Cathode poisoning in valves.

I suppose one possible source of silicates could be the refractory coating applied to the heater wire before its fitment into the cathode tube. This is usually done by dipping the heater filament into a slurry in order to uniformly coat the (usually) wound wire. As the heater is always hotter than the cathode - by definition - unwanted emissive effects are always possible in the absence of a large electric field from the anode.

Normal operation of an indirectly heated valve would make the heater and control grid most negative, followed by the cathode screen and anode becoming sucessively more positive. The positive potential on the cathode tube relative to the heater would normally ensure that stray electrons from the (very hot) heater are collected on the inner surface of the cathode.

It's possible that operating a valve at zero anode volts and consequently zero cathode volts due to there being no current in the biasing resistor gives rise to the emission of stray electrons from the heater.

I wonder if directly heated valves such as large triodes and rectifiers suffer from this effect. If they don't, the heater or its insulation may possibly play a part in this.

Leon.
Leon Crampin is offline  
Old 11th Jun 2008, 3:44 pm   #28
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Cathode poisoning in valves.

I remembered seeing something in the radio designers handbook and think I have found it and it goes along the lines of:

There are always some positive ions floating around either from gas, from emssions from the cathode itself etc. Under normal operation these ions are drawn to the grid(s) where -ve current is flowing. Without this -ve current flow the ions bombard the cathode causing damage to the cathode.

I presume the grid draws the electron cloud away from the cathode surface attracting all those nasty ions...someone elses theory but it does seem to make sense.
PJL is offline  
Old 11th Jun 2008, 5:17 pm   #29
Leon Crampin
Octode
 
Join Date: Nov 2005
Location: Surrey, UK.
Posts: 1,869
Default Re: Cathode poisoning in valves.

Thinking about this a bit more, I think you're right. Middle period television CRTs had an ion trap magnet at the gun end of the neck, the theory being that the electrons would be deflected by the magnet along their correct path through the gun structure, but the ions would not. The reason was always given as positive ions damaging the cathode.

CRTs of this type and application differ from receiving valves in that all of the gun electrodes downstream of the cathode are positive in relation to it.This might explain why these tubes needed trap magnets wheras the valves of the same period, which no doubt used the same cathode material technology didn't.

Comments from anyone with detailed knowledge of these types of tube would be interesting.

Leon.
Leon Crampin is offline  
Old 11th Jun 2008, 5:41 pm   #30
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Cathode poisoning in valves.

I've just tried to find anything in my copy of the RDH, to find the paragraph PJL refers to. The nearest I can find is in Section 1.1:

"If a slight amount of gas is present some of the electrons will collide with atoms of the gas and may knock off one or more electrons, which will serve to increase theanode current, leaving atoms deficient in electrons. These are known as positive ions since they carry a positive charge (brought about by the loss of electrons), and the process is known as ionisation. The positive ions are attracted by the negative cathode, and being comparatively massive, they tend to bombard the cathode coating in spite of the protection offered by the space charge."

This is a different mechanism to the sub-surface insulating layer referred to Posts 23-27 above. However, assuming that positive ion bombardment is the cause of cathode poisoning, I can see the following flaws:

1: It will only take place in a significantly gassy valve.

2: If there is no anode current flowing, there will be no positive ions created because there will be no fast-moving electrons to energise the gas.

If I'm wrong about this (for example, in a multi-element valve where one section is passing current and creating ions which could migrate to another cathode), the cure would be to arrange for another electrode to be more negative than the inactive hot cathode, and thereby trap the ions.

I also note in the RDH, section 3.1 (h), a paragraph about stand-by operation. The advice given is, if the heaters are powered, that "in general, better life will be obtained when the equipment is designed so that some, rather than no, cathode current flows during such periods. It is also essential that the heater or filament voltage be maintained as close to the rated value as possible.... In equipments which are designed so that no cathode current flows during stand-by periods for periods of less than 15 minutes, the filament voltage of quick-heating valves should be reduced to 80 per cent of normal, while the heater voltage of indirectly heating cathode types should bemaintained at normal rated value. For longer periods, both filament and heater power should be turned off."

Again, it would be nice to know what the degradation mechanism is. I'm sure F Langford-Smith of the RDH is right here, but I do like to know why!
kalee20 is online now  
Old 11th Jun 2008, 7:27 pm   #31
BGmidsUK
Retired Dormant Member
 
Join Date: Oct 2007
Location: Leicester, UK.
Posts: 809
Default Re: Cathode poisoning in valves.

I thought this was to prevent "ion burn" on the screen .. for example the first section of the gun in the AW43-80 is angled upwards, with a corrective magnet fitted to the tube neck. However, this magnet is also fitted above the gun, with its retaining clip angled upwards (which may contribute to the magnetic field?), so I'm not sure how this works! I will investigate further. However, the magnet is adjusted to give the brightest picture, meaning it is used to align the electrons WRT the rest of the gun assembly, suggesting the ions are trapped or dispersed by the first section.


Brian
BGmidsUK is offline  
Old 11th Jun 2008, 7:30 pm   #32
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: Cathode poisoning in valves.

Quote from "Electronic & Radio Engineering" (1955), page 207, by F.E. Terman.

"Gas molecules (within the valve) may damage the cathode by reacting chemically with the cathode material in such a manner as to destroy the (cathode's) emitting properties. This result only occurs for certain combinations of gases and emitting materials; however, it only takes a small trace of an undesirable gas to have a serious adverse effect on the emission. Oxide-coated cathodes are particularly susceptible to 'poisoning' of this type."

The rest of the text makes it quite clear that the 'gas molecules' undergo ionisation, and that ionisation within a valve - irrespective of the nature & mechanism that causes it - and there are more than one - will damage the cathode of the valve. F.E. Terman goes on to say "with oxide-coated cathodes, ion bombardment causes mechanical disintegration of the emitting surface. The effects begin to appear when the velocity of the ions exceeds about 20 volts, and increase with increased voltages and number of ions."

There follows a section about the reasons for gettering - which I haven't quoted, since we are all aware of the significance & importance of this - and it is moving Off Topic.

Al / Skywave.
Skywave is offline  
Old 11th Jun 2008, 7:59 pm   #33
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Cathode poisoning in valves.

That's a good quote, Al Skywave.

It's conceivable that ions could be created in a valve with no anode current flowing - possibly if it was in bright light (photo-ionisation). But, having generated positive ions, unless there are significant voltages of the right polarity applied to electrodes, I can't see the ions being accelerated to a velocity equivalent to 20 volts.

Tractorfan's original post asked about faster valve degradation with zero anode current flowing - all other things being equal, it seems to me that poisoning by positive ion bombardment would increase not decrease with current flow.
kalee20 is online now  
Old 12th Jun 2008, 12:25 pm   #34
Neil Breward
Retired Dormant Member
 
Neil Breward's Avatar
 
Join Date: Aug 2005
Location: Keyworth nr. Nottingham, UK
Posts: 140
Default Re: Cathode poisoning in valves.

I don't think there's any doubt that bombardment by positive gas ions would damage a cathode. There's always some residual gas in a valve - there's no such thing as a perfect vacuum! - but the formation of such ions depends on the energy of the electrons being sufficiently high to remove another electron(s) from a gas molecule on collision. The energy of the electrons depends on the accelerating voltage, so this should be a minimum when the anode voltage is zero! Furthermore, a dense space charge of low-energy thermal electrons is more likely to 'mop-up' any loose positive ions rather than form more of them.

There's something else going on that we're missing somehow. When a valve is 'conditioned' to optimise its performance, what is happening to the emitting surface? Does it become a metallic atomic monolayer? Can someone who understands 'work functions' give an opinion on this please? Can such a surface de-grade when heated, but not emitting?

Otherwise I think we're still on the 'contamination' and crystal chemistry path.

Cheers,
Neil
Neil Breward is offline  
Old 12th Jun 2008, 1:34 pm   #35
Tim
Dekatron
 
Tim's Avatar
 
Join Date: Jan 2003
Location: Bradford on Avon, Wiltshire, UK.
Posts: 3,301
Default Re: Cathode poisoning in valves.

Quote:
I think if cathode poisoning were a serious issue, many FM sets which had been operated for most of their lives on FM would show a poor E/UCH81 triode.
Ah yes but the E/UCH81 has a common cathode, so presumably the current flowing in the other section would prevent poisoning of the triode(one and the same) cathode.
I think Murphy also used to include high value resistors in anode circuits of "unused" valves. I have seen this done in Comms receivers too.
__________________
"Nothing is as dangerous as being too modern;one is apt to grow old fashioned quite suddenly."
Tim is offline  
Old 12th Jun 2008, 3:14 pm   #36
Leon Crampin
Octode
 
Join Date: Nov 2005
Location: Surrey, UK.
Posts: 1,869
Default Re: Cathode poisoning in valves.

A fair point, Tim but I suppose it all depends on how mobile these positive ions (if indeed they are the villains) are. In the E/UCH81 the triode and heptode are well separated, albeit being as you say on the same cathode.

If I can find a casualty, I'll break one open and see what isolation (mica spacer, screen perhaps) there is between the sections.

I guess the people who really knew are largely dead by now, but it's very interesting to speculate as to what's happening. Despite these theoretical "problems" valves are staggeringly reliable. I must change at least 10 capacitors for every valve.

Leon.
Leon Crampin is offline  
Old 12th Jun 2008, 9:39 pm   #37
Duke_Nukem
Octode
 
Duke_Nukem's Avatar
 
Join Date: Jan 2003
Location: Birmingham, West Midlands, UK.
Posts: 1,268
Default Re: Cathode poisoning in valves.

Oxide coatings are delicate little things. The thickness of the actual emissive surface is just one or two atoms.

During normal operation, the surface evaporates. Not very quickly, but evaporate it does, the resultant evaporate (?) landing on nearby electrodes such as the grid - so if the grid gets hot you get an emissive grid i.e. grid current (UL41's spring to mind at this point).

Anyhow, during operation, the surface is replenished by diffusion from the inner surface. Why should this diffusion go to the surface and not wander off anywhere else within the cathode (or indeed just stay where it is)? Answer : there is an electric field or current flowing.

Valves and CRT's are run with +ve grid and higher heater temperature specifically to activate the cathode during the manufacturing process, and indeed CRT's cathodes can often "rejuvenated" by this method.

So, with the heater running but no current flowing, the surface evaporates but isn't replenished, with adverse effects. But the surface may well be repaired by operating the valve.

When the heater is over-run, it is quite probable that the rate of evaporation exceeds the rate at which the surface is replenished, again detrimental to emission.

Cathode poisoning is different. One cause is electrons hit residual gas, knock an electron off a gas molecule making it +ve and hence the newly formed +ve ion scoots off towards the most negative point it can find - the cathode - smashing into and physically damaging the surface. Being damaged, it isn't so repairable. A good electron cloud loitering around the cathode, as is the norm for a valve under normal operating conditions, certainly helps protect the cathode but an ion with as little as 15eV energy smashing into the cathode will cause damage and its the main reason why oxide cathodes are not used in more serious big valves.

Electrode distances make little difference - the energy a charged particle obtains when accelerated by a given voltage is a function of the voltage only, electrode distances only affect how long the particle takes to travel the distance.

If the heater is under-run, then not only is the cathode not being replenished as described earlier, but the electron cloud will be significantly reduced (unless you suitably reduce the anode current) opening the cathode to more of the +ve ions.

So, how many atoms of residual gas are there ? Well, typically the vacuum might be one ten billionth of an atmosphere (or 0.01 atmospheres if it is Mazda ) and that equates to around one billion molecules per cubic centimetre.

Other reasons for poisoning include impurities present in, or leeching into, the cathode when it was manufactured.

TTFN,
The Anorak
Duke_Nukem is offline  
Old 13th Jun 2008, 9:55 am   #38
Neil Breward
Retired Dormant Member
 
Neil Breward's Avatar
 
Join Date: Aug 2005
Location: Keyworth nr. Nottingham, UK
Posts: 140
Default Re: Cathode poisoning in valves.

Good explanation, Duke!

Odd though, that the 'poisoning' cases seem to relate more to low-energy situations (ie zero cathode current), or are we confusing two different processes which both result in loss of emission?

By the way, has anyone every tried 'reactivating' a fading valve the way a CRT could sometimes be reactivated?

Cheers,
Neil
Neil Breward is offline  
Old 13th Jun 2008, 11:37 am   #39
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Cathode poisoning in valves.

I have read again the paragraph in RDH which is in the first few pages of chapter 3. As I understand it the grid suspends a space charge (electron cloud) around the cathode and presumably itself which prevents the electrons breaking away from the cathode and thus controlling the anode current. It is this space charge that prevents damage to the cathode from ions due to collisions with electrons.

It would be reasonable to assume that the distribution of the space charge would determine the movement of ions just as it determines that of electrons. It would also suggest that a near cut-off valve would suffer least and a valve with high anode current the most. But how would the space charge distribute in a valve with no anode/grid connections?
PJL is offline  
Old 13th Jun 2008, 12:59 pm   #40
Duke_Nukem
Octode
 
Duke_Nukem's Avatar
 
Join Date: Jan 2003
Location: Birmingham, West Midlands, UK.
Posts: 1,268
Default Re: Cathode poisoning in valves.

Quote:
Originally Posted by Neil Breward
Odd though, that the 'poisoning' cases seem to relate more to low-energy situations (ie zero cathode current), or are we confusing two different processes which both result in loss of emission?
Not quite. The zero current situation tends to be recoverable. Bombardment (and hence real poisoning) due to ions require a current to flow, else there'd be no electrons to knock the atoms to make the ions.

Quote:
Originally Posted by PJL
It would also suggest that a near cut-off valve would suffer least and a valve with high anode current the most.
Yes, but this would not be due to the electron cloud around the cathode, it'd simply be because there were no high energy electrons flying about to make the ions.

In normal operation, the cathode region is in equilibrium - the electrons in the cloud are in fact acting to stop / repel further electrons trying to leave the surface. Thus, as anode current starts to flow i.e. electrons leave the cloud and wander off to the anode, there are less left behind to repel those trying to leave the cathode surface, hence more electrons can get out and join the cloud. As a result, the cloud doesn't shrink quite as you'd expect as current starts to flow. The "protection" afforded by the cloud remains fairly consistent, but the number of ions it has to intercept increases as current starts to flow.

That is, until you reach saturation, where the cathode has reached max emission. The cloud rapidly disappears and so does the protection, and the high current makes lots of loverly ions at the same time. Thats one reason you should never run oxide cathode values into saturation.

Another factor is the cloud in effect distorts the electric field. If you had two plates 1cm apart with 100V between them you'd say the gradient were 100V/cm and thats what youd see at the surface of A. But a cloud of electrons around A will reduce the gradient at A (and increase at B). Applying this to a valve, the cloud acts to buffer the cathode from the full force of the anode. With the cloud gone, the cathode sees the full force which can be enough to literally strip the surface off the cathode - just as happens in a rectifier valve with an excessive reservoir cap or s/c HT.

Its a wonder they ever work at all !

Quote:
By the way, has anyone every tried 'reactivating' a fading valve
Nope, but always wondered. Most CRT's recover a bit after having been dormant for a long time and I dare say the same would apply equally to valves.

TTFN,
The Anorak
Duke_Nukem is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 9:33 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.