Beam Tetrodes vs Pe...
 
Notifications
Clear all

Beam Tetrodes vs Pentodes

Page 1 / 2
 
Dr Wobble
(@dr-wobble)
Member Deactivated Account
Posts: 702

Am I right in thinking that beam tetrodes are essentially the same beast? But for reasons of copyright etc, RCA had to call their pentodes-beam tetrodes?

If not not how do they differ in operation?

Lastly,I,ve seen 807 beam tetrodes used as replacements for EL34,s with a little tweeking, does it also follow that a EL34 could replace a KT66?

Thanks,Andy.

ReplyQuote
Topic starter Posted : 15/06/2012 3:57 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

They weren't the same, as the circuit symbols will prove. The Beam Tetrode was developed by manufacturers who wanted to make output valves without infringing the Mullard/Philips patent for the pentode.

You will see the differences between the two and see how a Beam Tetrode works here:

http://www.r-type.org/static/grid14.htm

Cheers,

Steve P.

ReplyQuote
Posted : 15/06/2012 4:12 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

There are three devices that perform like Pentodes:

1) The triple helical grid tube, the original Pentode. Control, Screen and Suppressor grids. g2 increases "partition noise" compared to a Triode and g3 makes it a little worse.

2) The Kinkless (Hence KT) or Beam Tetrode uses electrostatic lens (beam forming plates) instead of grid 3, the Suppressor grid. It's lower noise and more efficient, but not usually able to be designed for low power RF/IF. So many Philips power o/p types are actually Beam Tetrodes (DL96, ECL86 etc) but signal types tend to be "true" pentodes

3) The logical development of combining a Heptode that uses beam forming plates for g2 and a Beam Tetrode g3 Beam forming plates is the Rod Pentode. No grid wires at all. It was tried using indirect heater in the late 1930s in the West but a failure. The Russian guy in early 1950 realised it would only work with a thin filament and grid plates either side of the filament at right angles to the beam. It has thus no partition noise at all. An Indirect heater Rod pentode isn't possible. Also there is no sensible way of creating a true Triode, Tetrode, Hexode or Octode with a Rod Pentode (a Heptode or Nonode might be possible)

If the Cathode is not at "AC" ground the g3 (grid or beam forming plates or rods) has to be earthed separately. If the Cathode is at AC ground then g3 (grid or beam forming plates or rods) can be connected internally to Cathode. It's only on common grid RF amp, some kinds of oscillators and Cathode Followers that the g3 can't be internally connected to cathode.

Beam Tetrodes (aka Kinkless Tetrode) do really have five electrodes, but it's rare for the "beam forming" plates (equivalent to g3) to be brought out to a separate connection.

ReplyQuote
Posted : 15/06/2012 5:23 pm
Dr Wobble
(@dr-wobble)
Member Deactivated Account
Posts: 702

Thanks for that link Steve. So the beam reflecting plate in a BT is essentialy the same as g3 in a pentode,but differs in how it does the same function. If I also have understood correctly,the beam tetrode is superior in its ability to reduce harmonic distortion?

Re harmonic distortion,are we talking about the same thing -3rd,s 5th,s etc as in music theory? So the note A in the chromatic scale the third will be B,the 5th C# ? Therefor the BT is better at elimenating #,s or "Jazz" notes

Lastly,although I have read several documents,I,m still uncertain about the term "knee" . Is this the same as the saturation point of a valve.

Thanks for your input chaps,Andy.

Edit Ive just re-read your post Michael-Re ECL86,I thought the ECL 82-86 were a triode and a pentode all in one envelope,are you refering to the "pentode" section only?

ReplyQuote
Topic starter Posted : 15/06/2012 6:01 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

There isn't any real difference between a Beam Tetrode and Pentode in normal operation for harmonic distortion. Triodes are different.

Yes the "Pentode" section in most Triode + Pentode tubes (PCL, UCL, ECL series) is actually a Beam Tetrode. The triode part is a triode. There are some tubes made in "Pentode" and "Beam Tetrode" versions.

ReplyQuote
Posted : 15/06/2012 7:23 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

The "knee" doesn't apply to Triodes

Above a certain Anode voltage a Pentode (or Beam Tetrode) will stop increasing the current very much with increased voltage. This "knee" is between 3V and 20V depending on g2 and model of tube. It's NOT saturation! The actual nearly constant current above the "knee" voltage up to the "breakdown" or power limit of the tube is set by the g1 control grid. Saturation is where increasing g1 (closer to cathode) gives no more current.

The advantage is that the output current is controlled by g1 (and actually g2, but it is usually constant high voltage) and NOT the voltage or load on the Anode. This means more consistent performance on Batteries and some hum rejection on Mains sets. It also means for IF and RF tuned circuits that above the knee the output impedance is much higher than a Triode. This allows better selectivity (higher Q) than Triodes for RF amp or IF amp. It's essentially a Constant Current source (sink) controlled by g1 and g2.

With a Triode the current will vary with grid volts AND the anode voltage.

A bipolar transistor (PNP or NPN) has a similar knee to Pentode (or Beam tetrode), hence also above a minimum collector voltage (no matter what the current is) acts also like a constant current source.

ReplyQuote
Posted : 15/06/2012 7:39 pm
Kalee20
(@kalee20)
Member Deactivated Account
Posts: 256

Re harmonic distortion,are we talking about the same thing -3rd,s 5th,s etc as in music theory? So the note A in the chromatic scale the third will be B,the 5th C# ?

No it's not.

If you put a pure sine-wave input into an amplifier, second harmonic distortion causes an extra signal to appear in the output at double the input frequency. Third harmonic, at 3 times input frequency, etc.

On the other hand, in music theory thirds, fourths, etc refer to steps in the musical scale.

It turns out that two notes an octave apart have a frequency ratio of 2:1. So second harmonic distortion always sounds OK-ish.

A fifth (from musical A to E) is a frequency ratio of 3:2. So, third harmonic distortion actually creates a frequency of a musical octave and a fifth above the fundamental (ie 2 x 3:2 = 3:1). This sounds OK, too.

Fourth harmonic distortion creates a tone two octaves above, which is OK.

Fifth harmonic distortion will give rise to a tone two octaves and a major third (a musical third being a frequency ratio of 5:4, plus two doublings of frequency gives 5:1). This is not unpleasant.

Sixth harmonic distortion will be like third, with the extra octave, (ie 2 x 2 x 3:2 = 6:1), so OK.

Seventh harmonic distortion gives a note which is not part of the musical scale, and this starts to sound nasty. Eighth is of course just 3 octaves. Ninth happens to be a major tone plus 3 octaves, which is a dissonant combination, and most of the higher harmonics are dissonant.

ReplyQuote
Posted : 15/06/2012 8:02 pm
Kalee20
(@kalee20)
Member Deactivated Account
Posts: 256

So many Philips power o/p types are actually Beam Tetrodes (DL96, ECL86 etc)

The DL96 is actually a pentode - I've opened one! On the other hand, the DL92, although described by Mullard as a pentode, really is a beam tetrode (I've opened one of these, branded Mullard, too).

So the beam reflecting plate in a BT is essentialy the same as g3 in a pentode,but differs in how it does the same function.

It does give rise to the same favourable characteristics, yes, but it differs in how it works.

The beam tetrode works because it turns out that with a certain, critical electrode geometry, an electric field is created which itself repels electrons knocked out of the anode, back to the anode. In a pentode, an extra grid is inserted in the gap to do this repulsion back to the anode.

However, the spacing in the tetrode really is critical, and features like grid support wires upset the field. So, plates are put in which just stop any electron flow at all in these regions. Strictly, they should be called 'beam confining plates', but the accepted name has stuck.

I believe there are some high-power tetrodes, with chunky self-supporting grids, which don't need any beam plates at all. But don't quote me on this - I may be wrong (and please shoot me down if so)!

ReplyQuote
Posted : 15/06/2012 8:12 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

So many Philips power o/p types are actually Beam Tetrodes (DL96, ECL86 etc)

The DL96 is actually a pentode - I've opened one! On the other hand, the DL92, although described by Mullard as a pentode, really is a beam tetrode (I've opened one of these, branded Mullard, too)

You mean the one you opened was a "real" pentode. 

Also I have tested a "Rod" Pentode in a DL92, DL94 and DL96 socket and on the AVO VCM162 as those. 

I have no idea why some are "really" Pentodes and some "really" Beam Tetrodes. Mullard of course made the Philips recipe tubes. The DL92 (Philips or Mullard) is "really" a 3S4 RCA 1940 design (hence Beam Tetrode) and the DL96 no matter who made it a Philips 1953 design. Except of course for DL96 valves that are not really DL96.

You really can't tell (Beam Tetrode vs "real" Pentode) from schematic or datasheet always as sometimes the makers are a bit misleading.

Conversely some differently marked valves from SAME maker are actually the IDENTICAL tube not even selected on test (ECH81 for Mains Radio and ECH83 has 12.6V data sheet for car radio and they are the identical tube). But some tubes from different makers or the same maker at different time have different "insides" but same part number.

ReplyQuote
Posted : 15/06/2012 8:26 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

They were two equivalent ways of doing the same thing, avoiding the tetrode kink caused by secondary emission from the anode. Philips/Mullard held the pentode patent and I think RCA developed the beam tetrode and held the patent.

I believe there was also a lot of work in treating the anode surface to reduce secondary emission.

As for which is the best, well, they'll both have advantages and disadvantages. BTs were generally held to be better for power valves with a high current density and pentodes for signal valves. There were certainly a few signal valves such as W63 and KTW63 which seemed to be pentode and BT versions of the same thing. I believe triode-hexodes such as ECH42 are usually beam tubes.

I think that when you look back at the jockeying, the technical merits are just one consideration and it's difficult to know the full history. There's also cross patenting agreements, avoiding paying licences, Not Invented Here, the question of things such as having a competitive advantage in being able to cope with close manufacturing tolerances or needing a design which was tolerant, having expertise in developing one or the other, selling the product you've got and not the product you'd like to have, etc..

Pete.

ReplyQuote
Posted : 20/06/2012 2:23 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

Good summary I think. ? 

ReplyQuote
Posted : 20/06/2012 4:14 pm
Synchrodyne
(@synchrodyne)
Reputable V-Ratter Registered
Posts: 421

I believe triode-hexodes such as ECH42 are usually beam tubes.

Now that’s interesting. I must say I have wondered why Philips/Mullard chose a hexode rather than a heptode for the mixer part of the ECH41 and ECH42. On the face of it leaving out the suppressor grid seemed a bit odd.

The ECH42 successor, the ECH81, did have a heptode mixer, though. Here there was deliberate intent that the heptode section be used as a 10.7 MHz FM IF (1st IF) amplifier in FM-AM receivers, as well as the AM mixer, so maybe for this application the “pentode” form was seen as preferable to the “beam tetrode” form? The ECH42 was also so used, but perhaps as an afterthought rather than by design. It was also intended that the heptode section of the ECH81 be used as an externally fed mixer in say communications receivers, with the triode either unused or as an oscillator buffer, and in this role it seems to have been preferred to the EK90.

Cheers,

Steve

ReplyQuote
Posted : 29/06/2012 12:13 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

On reflection and having looked it up, I think to call triode-hexodes "beam tubes" is a bit misleading. It wasn't to do with suppressing secondary emission from the anode.

In the late 20s and early 30s there was a lot of R&D on frequency converters. They had a number of conflicting requirements, simplicity and low cost, good isolation between the LO and signal grids and so no pulling, good conversion transconductance, they were trying to work with higher IFs, they wanted a design which was vari-µ. They had all sorts of problems to solve and tradeoffs to make.

Beam forming, so far as it was done, was to do with tighter control of the electron paths in the valve and deal with transit time problems.

There's a section on frequency changers in Langford-Smith.

A lot of comms receivers by around 1940 used a separate oscillator valve, often with a neon regulated supply and with special attention given to thermal stability. It wasn't necessary and was much too expensive for domestic receivers.

According to this

http://www.r-type.org/static/grid04.htm

The tetrode kink wasn't really a problem with RF mixers because they always had a resonant load.

Philips had a go at producing a beam octode.

http://www.r-type.org/static/grid06.htm .

Pete.

ReplyQuote
Posted : 29/06/2012 5:23 pm
Kalee20
(@kalee20)
Member Deactivated Account
Posts: 256

The tetrode kink wasn't really a problem with RF mixers because they always had a resonant load.

I'd guess also it wasn't a problem because amplitudes were lowish here, so the anode voltage never dropped down to the kinky region.

ReplyQuote
Posted : 29/06/2012 11:32 pm
Synchrodyne
(@synchrodyne)
Reputable V-Ratter Registered
Posts: 421

That might have been more of the reason than the resonant load. After all, RF and IF amplifiers usually have resonant loads, but pentodes are preferred over tetrodes for these.

I looked back at the Philips data (Book IIIA) for the ECH41 and ECH42. Therein might be the answer. The screen voltage needed to be fed from a potential divider in order to stop it rising when the gain was reduced by agc bias. Screen voltage rise would cause secondary emission, reducing gain (not in and of itself a problem since that was the objective) but also decreasing valve resistance, and so adversely affecting the loaded Q of the tuned circuit load. The net of this is that the hexode was operated in a manner that kept it away from the kink. That it had a kink indicates that the hexode was not actually of the beam tetrode type.

Presumably then the heptode configuration was chosen for the ECH81 to allow greater freedom in setting its operating conditions, as suited its wider range of intended roles.

The hexode would appear to belong to the same group as what one might call the first kind of heptode or pentagrid, that is one intended for use as a self-oscillating mixer, with a dedicated oscillator anode “grid”, and without a suppressor, and so belonging to the tetrode family. The octode was similar but added a suppressor grid. The second kind of heptode had a suppressor grid but no dedicated oscillator anode grid, and could be configured as an externally fed mixer (e.g. 6L7) or a self-oscillating mixer (e.g. EK90).

Although nominally superseded by the ECH81/UCH81, the ECH42/UCH42 was used until quite late in the valve era. Eddystone used it in new (e.g. 820) and revamped (e.g. 670C, 840C) models after, sometimes long after the ECH81 was established. Thus one might say that the tetrode type survived in MF/HF receiver applications, in disguise as a hexode, long after the conventional wisdom had it superseded by the pentode.

Cheers,

Steve

ReplyQuote
Posted : 30/06/2012 2:12 am
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

Steve,

A.T. Witts in the Sixth Edition of "The Superheterodyne Receiver" describes several FC valves including the triode-hexode and the beam triode-hexode, the example of which he gives is the 6K8, which was a very popular valve. In this case the beam did suppress secondary emission from the anode.

The second edition is availble free on the web, but doesn't discuss the 6K8.

I can scan the pages if you want.

http://www.r-type.org/exhib/aaa0058.htm

Shows the electrode structure.

I don't know if there was a B7G or B9A version of the 6K8 as there there was with many other valves.

I've no idea why the latest designs were triode-heptodes rather than triode-hexodes, although in the case of the ECH81, it was so that it could serve as an IF amplifier as you point out.

Self oscillating mixers and double valves were always something of a compromise and the better and more expensive solution was to have a separate oscillator valve.

Pete.

ReplyQuote
Posted : 30/06/2012 5:00 pm
Synchrodyne
(@synchrodyne)
Reputable V-Ratter Registered
Posts: 421

Thanks very much for that and for the offer to scan. That’s probably not necessary, as I do have a late edition of Witts, but along with all my other books, is packed in a 40 ft container in transit somewhere between Sydney and Auckland.

Anyway, your comment about the 6K8 reminded me that Langford-Smith had something to say about this, and that it was definitely “different”. The ECH35 is probably covered as well, but maybe not the later Philips valves. So I’ll be doing some research when we have our books again. As far as I know the 6K8 was not replicated in the B7G/B9A range, and neither was the 6L7 heptode mixer. For MF/HF frequency conversion the 6BE6 (EK90) seemed to reign supreme in American practice in the B7G/B9A era, even in the AM sections of high quality tuners such as those from Scott. The 6BE6 was also used as a standalone mixer, in which role it might have been seen as at least an indirect replacement for the 6L7. I should say though that the best successor to the 6L7 was probably the heptode section of the ECH81, driven by an external mixer.

There was a 12AH8 triode heptode, B9A, but I do not know whether that was an American development, possibly as a successor to the 6J8G, or from Brimar, who may have found that the UK (and European) preference for triode hexode/triode heptode types necessitated adding one to its standard radio receiver range, which was otherwise based upon the 6BE6 as frequency changer.

Regarding the frequency changer hierarchy, given that an ECH42 or ECH81 is better than a 6BE6, particularly at HF, one wonders why the latter was used at all for (UK/European) mains-powered receivers at least. I do not imagine that there would have been a material cost difference between the ECH42/ECH81 and the 6BE6, nor in their surrounding circuits.

But then so ingrained was the pentagrid in American practice that the 6BE6 was said to be suitable for use in FM receivers, and the 6BA7 (B9A base) was developed as an improved pentagrid for FM use. That practice did not last very long, though.

As you mentioned in an earlier posting, the use of a separate oscillator was reserved more for communications receivers. The only UK domestic example that comes to mind (in the miniature valve era, anyway) is the Dynatron T139. (I don’t know if its predecessor T99 was the same, nor if the practice was carried over to the later T10.) One assumes that the market for a standalone heptode mixer, say the heptode part of the ECH81, for use in communications and upmarket domestic receivers, was not large enough to justify its production – or looked at another way, the best economics and spares availability was provided by applying the ECH81 with the triode unused or in a buffer role; Eddystone did both for example.

We may have drifted away from the original topic, although I think RF applications are in scope and it we have found mixer valves that belong in the tetrode, pentode and beam families.

Back to output valves, the conventional wisdom is that for a given anode dissipation, beam tetrodes typically have lower slope than pentodes, lower screen current relative to anode current, and lower heater dissipation. This certainly applies when one compares a 6BW6 with an EL84. But then I wonder if these differences more-or-less disappeared with later beam tetrode models. The 6BW6 was after all the late 1930s 6V6 in a B9A bottle, whereas the EL84 was an early 1950s development.

At first glance the ECL86 looks as if it was the combination of an updated EL41 (a pentode) with half of an ECC83. But if in fact the “L” part was a beam tetrode, then one could say that by the time it was released, beam tetrodes could mirror pentodes in their key properties, albeit with pentode-like screen current relative to anode current. But the ECL86 heater, 6.3 V, 0.7 A, suggested beam tetrode efficiency, particularly if one allows 0.15 A for the triode. The EL41 was 6.3 V, 0.71 A, and the EL84 6.3 V, 0.76A, the last-mentioned showing an efficiency gain along the pentode development vector, as well.

Cheers,

Steve

ReplyQuote
Posted : 01/07/2012 2:42 am
Synchrodyne
(@synchrodyne)
Reputable V-Ratter Registered
Posts: 421

By way of an afterthought to my above post, as I recall the GEC KT77 was a beam tetrode (in the tradition of the KT66 and KT88 before it) that was designed to match and be a direct replacement for the EL34 pentode. That would indicate that by the time the KT77 was released (mid-1960s?), beam tetrode technology was capable of matching the pentode in terms of slope. But its heater power, at 6.3 V, 1.4, was only marginally lower than that of the EL34, at 6.3 V, 1.5A.

Cheers,

Steve

ReplyQuote
Posted : 01/07/2012 5:11 am
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

12AH8 was only made by Brimar. I've never seen any reference to any other maker in connection with it..

Brimar developed a number of valves in the late 40s/early 50s intended to take advantage of the new B9A format, 6BW6, 6BS7, 6BR7, which achieved some success but which were overtaken by the Mullard/Philips equvalents, EL84 and EF86. 6BW6 had a reputation for being somewhat troublesome and EL84 for putting up with anything. Brimar also positioned the 6BW6 as a VHF output valve and there was an SQ version, the 6061, which they also offered as a solder in version. 6BW6 was a 6V6 in B9A.

They must have registered them with the US naming convention authority, but they seem to be completely unknown in the USA.

Come to that, the ECH42 and ECH81 have US type names as well, 6CU7 and 6AJ8, but as far as I can seem they didn't sell in the US.

I don't think it's as simple as saying the ECH81 was the best mixer valve and there was a love affair with the 6BE6. Eddystone used 6BE6 in some of their sets and ECH42 or ECH81 in others. A lot of respected comms receivers used 6BE6 or 6BA7. They weren't doing that because the ECH81 was a foreign thing.

I haven't looked closely at post war US domestic receivers, but I gather there was a big market for cheap kitchen and bedroom radios, MW only, transformerles and running off 110V mains and for listening to local stations based in the city you were in. That could be why the self-oscillating pentagrid was popular and the triode-hexode and triode-heptode were ignored. There were different market conditions in Europe.

In any case, if you were designing a receiver, especially a comms receiver, I think there'd be a lot more to choosing the mixer(s) than looking at the datasheets.

BTs and pentodes were two different ways of solving the same problem with inherent strengths and weaknesses. As time went on and the technology matured, the weaknesses could be minimised, however, there was a tendency for customers to stick to things that were tried and tested, like the 807.

In some cases it's hard to know what the truth is or whether a valve existed in two physically different forms with the same electrical characteristics. E.g. 6688/E180F/CV3998. Some say it's an RF BT and others a pentode.

Pete.

ReplyQuote
Posted : 01/07/2012 4:18 pm
Anonymous
(@anonymous)
Group Deactivated Account
Posts: 16765

There are definitely tubes with SAME model but some are "real" Pentodes and some Beam Tetrodes. Often different makes though.

ReplyQuote
Posted : 01/07/2012 5:59 pm
Page 1 / 2