Components that underachieved.

[G6Tanuki]

A while back I was sorting out some old valves and came across a seemingly unused Cossor 42OT, which I put on one side intending to do something with.

Now, I get to study the spec, and realise that this valve, though looking nice, is a spectacular underachiever!

Details here: http://www.r-type.org/exhib/abb0009.htm

Heater is 4V at 2A - that's Eight Watts! And looking in another reference, it only gave 3 Watts of audio output.

It was introduced in 1937, must have been one of the last introductions to use the old "British" base.

Compare it to the first 6V6 - also introduced in 1937 - which could produce 5.5 Watts of audio, needed only 2.8 Watts of 6.3V heater-energy, and came with the much more-modern octal base.

6V6 shows up the 42OT as a spectacular underachiever; so why was it introduced? And did anyone ever use it? I can only think Cossor wanted to be able to say they had a valve for all positions in a radio.

What other components can you think of which have similar performance-deficits-from-new?

[RobRusbridge]

Don't be too hard on 42OT. It's much the same as AC2/Pen and KT41. It's the way things were back then. And don't forget the 42OT is a high-slope output pentode - sensitive enough to be output bottle in a short superhet (hence the AC2/PenDD). You can't say that of a 6V6G. If you want to do a short super with the old American octals you'd be looking at a 6AG6G (EL33) or in Europe an EBL1, EBL21 or EBL31. I suppose you could do a short super with a 6V6G if you have a 6B8G for the IF amplifier, use its diodes as detector and AGC and then reflex some audio gain which you could pick up from the screen grid. But a reflex super isn't really the same as a short super.

If you want an under-achiever, compare SP41 or SP61 with EF50. The EF50 is much more frugal with heater current and comparable in other ratings.

[G6Tanuki]

Yes I can see that the 42OT could have a place in a 'short' Superhet, but there were US valves in the same era whose gm was as high and only took half an amp at 6.3V; the massive heater-power of the 42OT looks equally crazy when you realise that a 6L6 only took 0.9A @6.3V !

The SP41/61 were equally heater-intensive, but apocraphylly they were preferred in some WWII RADAR gear over the EF50 because they were more reliable. The 6AC7

http://www.r-type.org/exhib/aaa0053.htm

did quite well, and served the same niche as EF50/SP41/SP61.

Is there some sort of intrinsic rule that high-slope valves generally have high heater-dissipation? If so, why? Later TV video-output valves often had impressively high gm but didn't need disproportionate amounts of heater power.

Another 'series' of valves that underachieved were the Mullard/Philips secondary-emission series like the EE50

http://www.r-type.org/articles/art-112.htm

with its crazy electron-path, which in the lab promised spectacular gain but they never seemed capable of getting it to deliver this reliably.


I've got a couple of SP61 and some Mazda Octal bases as well as the aforementioned OT42 - part of me is minded to build something with them [either a classic 1-V-1 shortwave regenerative receiver or a similar-lineup VHF superregen - in the second case the 'RF amp' would be there mainly to stop the superregen stage spewing 'sharsh' back through the antenna and annoying other listeners]. Biggest problem will be finding a 4V heater-transformer though - I could always stick a 1Ohm resistor in series with the 42OT heater to lose a couple of volts.

[turretslug]

The 6TH8 triode-hexode seems to have been rather greedy in terms of heating power with 6.3V at 0.7A, more than double than that of several contemporaries. I wondered if it was simply an earlier 4V 1A valve re-heatered- the change from 4V, often B7 based to 6.3V, often IO based types seems to have been accompanied by a reduction of around a third in heating power with a number of types- perhaps there had been genuine progress in cathode coating materials with most but the 6TH8 was a "minimal development" upgrade from an earlier valve. Post-war, the rimlock Mazda 6C9 had double the heater power requirement of its Mullard counterpart the ECH42.

I'd tried to rationalise the seeming correlation between high gm and high heater power but could only come up with the thought that closer-wound grids rob the nearby cathode of more heat, which seems kinda tenuous....

Assuming this thread is open to electronic under-achievers generally, I'm sure I recall a class D field output IC that was meant to be a Wonderful Thing, replacing the TDA1170/1270 type devices but whose main achievement was charred PCBs (so much for that switch-mode efficiency!) and early ritual suicide. The TV folk here will surely recall it, perhaps with some pain....

[Cobaltblue]

Originally Posted by G6Tanuki

It was introduced in 1937, must have been one of the last introductions to use the old "British" base.

6V6 shows up the 42OT as a spectacular underachiever; so why was it introduced? And did anyone ever use it? I can only think Cossor wanted to be able to say they had a valve for all positions in a radio.

I don't know Cossor's thinking behind the 42OT but it was used in a lot of the 1938 model year sets (so released in 1937).

This included the "48" and "58" series with their large over size dials such as the 484 and 584

The alternative was the 42MP/PEN

Shortly afterwards Cossor reverted to using triodes 4XP, 2XP and 2P in 1939 and 1940 model years sets.

Some 1938 model year sets used 41MP which is an indirectly heated triode.

I am guessing that filament power consumption was not a big factor for them in large cabinet-ed mains powered sets, just conjecture on my part I have not seen any documentary evidence to support this view to date.

If I can locate it I have a large advertising piece extolling the virtues of the 48 series I will see if it mentions the use of an output pentode.

Cheers

Mike T

[paulsherwin]

Heater power doesn't seem to have been an important design consideration for British indirectly heated valve manufacturers in the 30s. I guess high power mains transformers didn't cost much more than low power ones then. I wonder if the Americans were keen to minimise heater current because they had one eye on automotive battery supplies.

[stevehertz]

There was a secondary emission pre war TV valve, the TSE4 that never caught on. It was used in one of my then TVs and I also had a spare once in a while. It was very rare and I recall some ridiculous price that Tudor Reese was asking for them. Apparently it was eclipsed within months of its introduction by the EE50. More info here, it was basically a research avenue in an attempt to overcome a construction problem in valves: http://www.r-type.org/exhib/aaj0185.htm

A large cathode (more heat) needs less precision in the grid/cathode spacing. Easier to make for a high mu valve. As things got better and smaller, the cathode isn't needed to be so big.

[Hartley118]

Originally Posted by paulsherwin

........... I wonder if the Americans were keen to minimise heater current because they had one eye on automotive battery supplies.

I'm sure that was a factor Paul - after all the fact that 6.3V was chosen for valve heater voltage in the US was geared to the then standard 6V car battery.

I fear that British valve manufacturers, heavily protected by the BVA cartel for many years, had little commercial incentive to innovate until the late 1930s when American competition began to bite.

Martin

[RobRusbridge]

It came to my attention this morning that some of those very old one-pin one-socket coils underachieve. Others are much the same to look at but give better results. It seems two thirds of mine underachieve and a few of the better ones cover wavelengths which struggle to be relevant in the modern world.

[Top Cap]

I remember my old boss George Hampstead telling me that the Mazda U801 was a most terribly unreliable valve. I had one once in my collection but was lost somewhere in my house move. It was a pull anyway and I don't think I would have trusted it

[Refugee]

I have never seen one of those 4 anoded beasts but have heard stories about them.

[G6Tanuki]

I've only ever seen one U801, thankfully.

The first production of them had issues with the emissive coating flaking off the filament and causing flashovers as the flakes fell to the bottom. Often they were used with pairs of anodes strapped - with current-balancing resistors (50-100 ohms) in series with each anode - problem was, in at least one TV these resistors were run very close to their rating. When one went O/C the current it would have passed to its U801 anode was forced to the other anode of the pair - whose future failure was guaranteed - usually in a bright blue glow as the prelude to a fireworks-display!

Another component that had promise but turned out to be a real problem: tantalum capacitors. Whether the 'wet' type in sealed metal cans as found in things like Clansman radios and 1970s oscilloscopes, or the blue resin-encapsulated 'bead' types, they always seem to be the cause of issues a few years down the line. I guess you could consider them the 'waxies' of the late 20-th-century.

[SiriusHardware]

If we are not limited to valves I'd like to nominate the SN76013 Audio power amplifier IC, almost doomed to fail at some point in their Fidelity record player hosts. There is some suggestion, though, that Fidelity ran them at or beyond their specified supply voltage, so it may not have been entirely their fault.

[G6Tanuki]

Those TI SN76xxx amplifiers were a bit of a liability, yes - I replaced quite a few in 1970s TVs, usually after the kids had turned Top of the Pops up to full-volume.

[Wasn't it also the same series of TI chips whose production-rejects were rebadged by Uncle Clive as his "IC-12" ?? http://rk.nvg.ntnu.no/sinclair/audio/super_ic-12.htm ]

[Radio Wrangler]

Magic eyes...... life-expectancy thereof

David

[turretslug]

Side-contact valves?- I assume that they were actually intended to give a good contact by the action of scraping the contact surfaces on insertion, but gained a reputation for precisely the opposite result.

Amongst the above, the EFM1 seemed designed to be side-lined as an oddity- a combination of magic eye and variable-mu AF pentode. Someone should have said, "Very good- but why?" at the development stage.

Loktal valves were pioneers of short, low-inductance electrode-to-base connections, but had the same by-then large footprint as the established International Octal types and the short, stiff pins often gave contact problems by contrast with the robust and unambiguously seating IO. It wasn't long before they were usurped by miniature glass-based types post-war. Rimlock valves performed well (in general), particularly when TV and VHF radio development demanded valves with good HF capability and they were popular with European manufacturers for a few years post-war- but resistance from US manufacturers meant that, like loktal, they weren't much more than a side-story in overall valve history.

[Julesomega]

If we are not limited to devices, I'd like to rant against a series of equipment enclosures which were used for Lab-assembled and small production-run test equipment. Front and rear panels were joined together by lengths of extrusion, and the case sides, top and bottom were attached to the extrusions by M3 screws fitted into square nuts which floated up and down channels in the extrusions. All very nice if there were any provision for fixing the locations of the nuts, but there was none and they would all relocate to the ends of the channels while you try to screw the panels on. As the panels were pre-drilled, you'd think they could have drilled the extrusions or solid bars to suit.

have added a blob of glue to the nuts here in the hope they will retain the nuts next time I open up the case.

[G6Tanuki]

Turretslug's comments on valves remind me of the "Catkin" range that were introduced between the Wars; the envelopes being slim metal tubes that also served as the anodes. Of course with the anode being at HT potential a clear danger presented itself, so "Catkin" valves were sold with perforated metal screens fitted over them - which increased the outer diameter to the same as a normal glass-bottled valve so any packaging-benefit was lost.

Also: Mazda-octal based valves. Invented as a way to lock-out US 'international Octal' imports from the UK market, but failed for obvious reasons.

[samjmann]

My nomination would be the 'printed resistor'.

I remember 'printed' resistors that were used in some Sharp ghetto blasters. You couldn't solder to the end pads and ended up point to point wiring a replacement. I've not seen them since in any consumer equipment since, but perhaps they're used elsewhere. SJM.