29th Dec 2011, 4:26 am | #141 | |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
Quote:
The AGC circuit looks very strange, and it took me a while to figure it out (if indeed I have done so.) It consists of a grounded grid EF80 pentode, whose cathode is fed with low-level video from the PL83 video amplifier cathode, in series with a triode (half ECC82) with grounded anode. In fact the triode is functioning as a diode, with the grid as “anode”, fed with bidirectional line flyback pulses. From the grid is also taken the AGC negative DC voltage. It seems to me that the EF80/½ECC82 combination functions as a gated, grid-controlled shunt rectifier. The input “AC” to be rectified consists of bidirectional line flyback pulses, and these also provide the gating function, which, in view of the positive modulation case, would be timed to capture the back porch period. The rectifier output is controlled by EF80 current, in turn determined by its cathode voltage relative to the grid. And it is arranged so that without any switching, the gain is higher for the positive modulation case, where the signal is sampled at around its 25% amplitude point, as compared with around 75% for the negative modulation case. The OA70 demodulator diode is DC-coupled to the PL83 video amplifier, whose cathode is shown as being at +4.7 V DC. The video signals at this point are shown as 2.1 V peak-to-peak, negative-going for the positive modulation case, and positive-going for the negative modulation case. The diode demodulator (OA70) orientation is such that the video signals would be negative of the datum point, that is in the range 2.6 to 4.7 V. The respective back porch (close to black) levels would therefore be around 3.1 V for the positive modulation case, and 4.2 V for negative. Normalized to grid bias numbers at zero cathode volts, these would be --3.1 V for positive modulation case, and -4.2 V for negative. Although the EF80 is sharp cutoff, these numbers probably put it on the curved part of its characteristic, and the gain at -4.2 V grid bias is thus somewhat less than that at -3.1 V. I don’t think that the use of half an ECC82 strapped as a diode is so unusual. The common domestic receiving triodes mostly come in pairs, so the chances are that a spare triode was more easily found that a spare diode in a multivalve receiver. But I thought it was normal, when using triode as a diode, to strap the grid to the anode, whereas in this case the anode is grounded and the grid is used as an anode. Would there be a technical reason for doing it this way? Cheers, |
|
29th Dec 2011, 1:35 pm | #142 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Valve Questions
Less capacitance, and a screen?
|
30th Dec 2011, 2:13 am | #143 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
Thanks very much, Dave. Yes, that makes sense. Since the other half of the ECC82 was a video amplifier for the positive modulation feed to the sync separator, you probably wouldn’t want any significant transfer of the 190 V peak-to-peak line pulses, hence the need for a good screen around the agc section, which a grounded anode should provide. And the ECC82 data show the anode-to-alternate grid capacitances as being about an order of magnitude lower than the anode-to-anode capacitance.
The ECC82 wasn’t in the original Philips/Mullard World Series of TV valves, although the ECC81 was. But later it became an important TV valve, and Mullard highlighted the fact that it was manufactured and tested with special regard to its pulse applications. Cheers, Steve P. |
30th Dec 2011, 7:45 pm | #144 |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,081
|
Re: Valve Questions
It is astonishing what uses valves were put to. I'd also be certain that the connection under discussion was for lower capacitance and screening reasons.
What I'm now thinking of is, using the grid as an anode, how much can it take? First, there's the current - we don't want to melt the fine grid wires. Mullard's ECC82 data has no figures for grid dissipation! No that it would be particularly easy to measure in a pulse application, without voltage and current probes and a digital scope with maths functions. Secondly, how much reverse voltage can you put on the grid? Again, bearing in mind that it's close to the cathode, 190V seems a lot without the risk of flashover! No doubt Mullard didn't envisage this when the valve was released - and again, possibly their applications guys would have been available to advise. The again, maybe the circuit designers stressed a few ECC82's to destruction, measured the conditions they failed at, and if their circuit was well within these conditions, went home and slept happy. Not the sort of evaluation you could get away with using tr*ns*st*rs of course... |
30th Dec 2011, 8:14 pm | #145 |
Dekatron
Join Date: Sep 2006
Location: Kilmarnock, Ayrshire, UK.
Posts: 5,422
|
Re: Valve Questions
Hi.
I am not too sure of all that switching on the high definition 819 line system its bound to damp out some high frequencies and reduce resolution.
__________________
Cheers, Trevor. MM0KJJ. RSGB, GQRP, WACRAL, K&LARC. Member |
30th Dec 2011, 11:43 pm | #146 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Valve Questions
The ECC82 was derived from the EC90, which was regarded as a power triode so had a fairly robust grid. In Class C it could cope with some mA of grid current, so OK as a signal diode.
|
31st Dec 2011, 2:12 am | #147 | |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
As far as I know the 6C4 (later EC90) was an early B7G valve, and the 12AU7 (later ECC82) was an early B9A valve. It is interesting how the original design requirements for RF power amplification facilitated its later use in quite different applications.
The 6J6 (later ECC91) was also used strapped as a diode, although whether with grid as anode I don’t know. But per the attached scan from the Stokes book, the 6AL5 (later EB91) was developed to avoid the necessity for this practice. Quote:
It’s curious that the IF strip consisted completely of sharp cutoff pentodes, with AGC applied to the 1st and 3rd stages, as well as to the PCC84 RF amplifier. Certainly the EF85 would have been available by the time Belgian TV started in later 1953. The Mullard papers on sync-cancelled AGC give some clues as to this practice. In that on the negative modulation AGC system it is stated: “With a negative modulation system, using frequency modulation sound, the problem of cross modulation in the receiver is not so serious as for systems with a.m. sound. Sharp cut-off pentode valves (EF80) are normally used in the i.f. amplifier.” There is the caveat though that to avoid cross-modulation, the RF and IF agc voltages must reach their respective maxima at the same, notwithstanding the delay. There is a worked example that shows that to achieve this condition with -5 V agc on two (of three) EF80 IF stages and -5 V agc on a PCC88 rf stage, after 3 V delay, an agc voltage of -12 V must be available. The cross-modulation issue referred to with a.m. sound systems presumably applied to combined vision and sound 1st IF stages. In the circuit at interest, the sound is split off after the tuner, which one would infer obviated the problem. Given that the original Philips/Mullard World Series of TV valves did not include a remote cut-off pentode, one would assume that the design plan included using the sharp cut-off EF80 in agc stages, and that such use was not simply an afterthought. As discussed earlier in this thread, the EF85 was added later, although from the Mullard materials that I have seen, was initially featured more as a member of the new FM/AM radio receiver valve series, and so perhaps just quietly slipped into the TV series. Cheers, |
|
31st Dec 2011, 12:11 pm | #148 |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,081
|
Re: Valve Questions
|
2nd Jan 2012, 4:46 am | #149 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
But were there any new ED valves in the modern era? Philips/Mullard does seem to have avoided classifying any of its modern receiving triodes as “D”. For example the ECC91 was described as being suitable for RF power amplifier use. And the ECC40 data, although mostly covering AF voltage amplifier applications, shows an application as a push-pull power amplifier. With pentodes, there was a grey area between “EF” and “EL”. Whilst the L part of the ECL80 could be used for AF and TV field outputs where low power was sufficient, in many cases it seemed to be used more in an “F” role than an “L” role. But then as has come up quite often in this thread, valves often found uses well beyond those applications originally included in the datasheets, and as well, their makers sometimes “slanted” what they said according to the subset of applications they were writing about. For example, the thread “EF91 in B9A” at: https://www.vintage-radio.net/forum/...ad.php?t=77974 prompted a look at the Mullard 1949 “Valves for Industry and Commerce” book. That book predates the EF80, but the EF91 is listed and described therein as an “H.F. pentode primarily intended for use as an R.F. amplifier or mixer valves in television receivers.” The EF42 and UF42 are also both listed, but described a bit more differently than their actual relativity might lead one to expect. Thus, for the UF42: “R.F. pentode having high mutual conductance, primarily intended for use in television receivers in which the cathode heaters are series connected.” But, for the EF42: “High slope pentode, intended for use as R.F. or I.F. amplifier in A.C mains operated equipment.” There was no mention of specific television applications in this description, even though the EF42 was the 6.3 V heater counterpart to the 100 mA UF42. So one might infer that for AC TV receivers, Mullard was in 1949 offering the EF91 not the EF42 for RF and IF use, whereas for AC/DC receivers, it offering the UF42. But all three could be used in essentially the same range of applications, which went well beyond TV receivers. Further to the recently described unusual ECC82 application as a TV agc rectifier with grid as anode and anode as a screen, there is an unusual EF80 application described in Kerkhof and Werner, in an early CCIR (system B) receiver. Here the video amplifier was two-stage, an EF80 DC-coupled from the demodulator (configured for a negative-going output) followed by a PL83. The EF80 had a second function as the agc valve. Positive-going line flyback pulses were applied to the EF80 suppressor grid, from where the agc negative voltage was also taken. During the pulse interval, the valve acted as a controlled rectifier with its suppressor grid as anode, its output proportional to the input sync pulse amplitude. One might say that the EF80 concerned was time-multiplexed, for most of the time acting as a voltage amplifier, but periodically acting as a controlled rectifier. Suppressor grid curves for pentodes don’t seem to be part of the usually supplied data, but where they are available, as with the dual-control type like the 6AS6, it looks as if the anode current curve flattens out by about +20 V on the suppressor. Presumably though, at high enough g3 voltages, such as the line flyback pulse would supply, significant current is diverted from the anode to g3. Cheers, |
2nd Jan 2012, 3:54 pm | #150 |
Octode
Join Date: Nov 2005
Location: Surrey, UK.
Posts: 1,870
|
Re: Valve Questions
There certainly were "D" triodes in the modern era - the PD500/510 shunt stabiliser valves were examples, as were their "E" counterparts, although seldom seen in the UK.
Philips shot themselves in the foot with the EF42 as its heater current was 330 mA, precluding series operation in TVs. Suppressor grid control is an interesting topic, and I suspect it was not liked by the valve makers as it is subject to large production spreads. It was first tried in a different form with the E/UAF42 where the suppressor was used as an AGC delay diode with somewhat variable results. The EBF80 rather rapidly followed as a more sensible way of achieving the required result, albeit on a 9 pin base. As an observed result, a standard late '50s early '60s valved AM/FM receiver with the ratio detector DC output tied to the suppressor grid of the the final IF amplifier will self-limit with a maximum DC control voltage of about 33V - however strong the signal. Valves such as the E/UF89 and 6F18 clearly operate under suppressor bias controlled gain under these conditions. Leon. |
2nd Jan 2012, 9:40 pm | #151 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Valve Questions
I'm not sure what are the criteria for distinguishing between C and D (triode) or F and L (pentode). Max cathode current? Max anode dissipation?
EC90/6C4 was intended (among other things) for Class C RF amplifier and oscillator service, so it was fairly rugged. The classic octal low-mu dual triode was 6SN7, with 6FQ7/6CG7 as the B9A version. This needed 600mA at 6.3V; the ECC82 (two EC90 in one bulb) could do similar things in a TV set with only 300mA heater current. |
2nd Jan 2012, 10:49 pm | #152 |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,081
|
Re: Valve Questions
Me neither - the ECL80 was definitely used with the 'L' section more as an 'F', as Synchrodyne observes.
It does seem rather arbitrary, use of 'C' or 'D', and 'F' or 'L'. I reckon it was down to the initial, target application. Thus the shunt stabiliser triode PD500 is definitely a power application, this is clear cut, disregarding the fact that the max cathode current is only 1.6mA, in sharp contrast with the PC97 which is 20mA. Similarly, the DL96 has max cathode current 6mA and max anode disipation 0.6W, whereas the EF80 greatly exceeds both these (15mA and 2.5W respectively). Having decided a target application, and numbered the valve appropriately, it seems that an appropriate set of data was given. So, having decided that the ECC82 (for example) would be directed at amplifier triode applications, figures for µ, ra, gm, Ikmax etc are stated (all relevant to amplifier circuits), but not PIV (which is relevant in a rectifier application), despite the fact that some designers used it as such, as noted by Synchrodyne. |
2nd Jan 2012, 11:23 pm | #153 |
Dekatron
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,395
|
Re: Valve Questions
Even the hairy-chested ECC230 (aka 6080) doesn't appear to have justified "EDD..".
|
3rd Jan 2012, 10:19 am | #154 | |||||
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
Quote:
Quote:
Quote:
Quote:
Quote:
Cheers, |
|||||
6th Jan 2012, 12:29 am | #155 | |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
Quote:
The 12AT7/ECC81 was a pair of 6AB4/EC92 in one bulb. The 12AX7/ECC83 does not seem to have had a single-triode antecedent, but was effectively two of the triode from the 6AV6/EBC91. The EBC91 was an outlier amongst (radio receiver) double diode triodes (American TV DDTs varied somewhat), in that it had a 100 µ triode in place of the “normal” 70 µ type, as used in the EBC90/6AT6, EBC41, 6FM8 and EBC81. Cheers, |
|
6th Jan 2012, 11:48 am | #156 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Valve Questions
American DDTs tended to have lower mu, as they often had to work on lower HT voltage than in Europe.
The EBC90 triode was the same as half a 5751, but it originated way back before the 6Q7. Could it take the prize for the most different bases/versions of essentially the same valve? |
6th Jan 2012, 1:43 pm | #157 |
Pentode
Join Date: May 2008
Location: Leeds, West Yorkshire, UK.
Posts: 236
|
Re: Valve Questions
|
6th Jan 2012, 10:10 pm | #158 |
Hexode
Join Date: Mar 2005
Location: London, UK.
Posts: 318
|
Re: Valve Questions
And Marconi/GEC offered the electrically similar (identical?) A1834 which had the same electrode structure as the 6080 but fitted in a larger classic shaped envelope.
Last edited by Brased; 6th Jan 2012 at 10:26 pm. |
18th Jan 2012, 12:14 am | #159 | |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Valve Questions
Quote:
Technical advice on other applications was provided via other channels, such as the Mullard Valves, Tubes & Circuits (VT&C) series, and perhaps these were more persuasive, and probably used by many manufacturers as the starting point for their designs. Taking the PCF80 as an example, it was “announced”, along with the PCC84, in VT&C 15 of March, 1954 (“Valves for V.H.F. Television Reception), although it had been available for a while before then. VT&C 16 of April, 1954 (“PCF80: A Frequency Changer for Band I and Band III Television”) provided much more detail. But then diverse other applications were dealt with in later VT&C issues, namely #26 of February, 1955, “Additional Applications of the PCF80”, and #27 of March, 1955, “Recommended PCF80 Television Applications”. There was not similar treatment of the ECC82 in the VT&C series, but perhaps that series had closed before Mullard was ready to do this, and the diverse TV applications for this valve could well have been covered elsewhere in the Mullard technical literature. That the ECC81 had AF as well as its original TV applications was recognized in VT&C issues #10 of October, 1953, “Noval-Based Valves for Audio Amplifiers”, and #12 of December, 1953, “Applications of Double Triodes in Audio Amplifiers”. An interesting comment made therein was: “The audio applications of the low-µ ECC82 are more limited because of its low voltage gain, but it may be found useful in penultimate stages where a low impedance is desirable”. That at least implies that AF applications were secondary for this valve. And Mullard advertising of the period did cover other applications, such as that from WW October, 1953, where the trio of double triodes is presented as being part of the Mullard communications series. VT&C #10 was in the same issue. Cheers, |
|
18th Jan 2012, 2:01 pm | #160 |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,081
|
Re: Valve Questions
I've seen that issue of VT&C too, and I remember that phrase.
In which case, what were the 'target' applications of the ECC82? I can't help feeling that as a fairly rugged, general-purpose low-µ triode, it would have been helpful of Mullard to indicate max grid dissipation, and max negative anode (and grid) voltage, for those people who wanted to use it in weird pulse circuits. I've seen it as an AGC amplifier in a gated-AGC circuit, in which most of the time the anode is negative to cathode. |