Philips ECL86 book

[Terry_VK5TM]

Worldradiohistory has just added the 'Philips Circuits with the ECL86' book

https://www.worldradiohistory.com/BOOKSHELF-ARH/Technology/T...20Circuits%20with%20the%20ECL86.pdf

[ajgriff]

Fascinating book Terry. Thank you for posting the link.

Alan

[vinrads]

Nice one Terry, good to see there are no fancy tapings on the o/p tx's should be an easy build . Mick.

[wd40addict]

Interesting! In practice I think, despite the soothing words, 9W was pushing it a bit too much. The 86 seems much more prone to grid leakage and runaway than its 82 cousin.

No modern maker attempts an ECL86, probably for this reason.

Some of the designs it was used in didn't help. in the S-33H, for instance, Heathkit used a 100R cathode resistor rather than the Philips specified 170R! After an ECL86 failed in mine I've upped them to 150R with hardly any effect on power or distortion, but a noticeable reduction in standing bias.

[G6Tanuki]

Yes the ECL86 dissipation does seem to be optimistic, they developed a reputation as a valve that got tired quickly, though not as quickly as that awful ELL800 that had a brief popularity in the early sixties.

I remember also the ECL86 being used in very small two valve HF amateur transmitters for mobile use, one ECL86 used as a crystal oscillator and PA stage, the other as the modulator, driven from a crystal microphone and coupled to the PA using a small push-pull output transformer in the same style as a Codar AT5. wire both valves' heaters in series across the 12V supply, and get the HT from a couple of OC28 transistors in a simple inverter, and you have a very compact top-band transmitter with low power consumption.

[Boulevardier]

Very interesting and useful, Terry. Thanks for posting. My ECL86s have 150 ohm cathode Rs; wondering if I should up them to 170 ohms as per Philips - those ECL86s aren't cheap to replace...

Mike

[Aub]

Yes very interesting. Thanks for posting. I made a stereo 3w per channel amp using ECL86s on their sides. At the time, I never thought about the orientation of the valves but it's been working fine since 2007. Perhaps I got lucky.

Cheers

Aub

[GrimJosef]

Thanks for this Terry.

It includes a pretty detailed analysis of the conditions for stability with such a high-gain valve. Philips point out that every effort has to be made to keep the ratio of CaPgT (the capacitance between the pentode's anode and the triode's grid) to Cag (the anode-grid capacitance for the triode) as low as possible. To this end metal screening is placed inside the valve itself, and the use of a valveholder with an external skirt at the bottom of the valve is also strongly recommended.

Obviously all of this effort can be wasted if an amp designer doesn't separate the external wiring for the triode grid from that for the pentode anode as completely as possible. I've seen a couple of examples where more recent designers haven't done this (nor have they used a skirted valve socket) and the amp can then be very prone to becoming an inadvertent high-power HF oscillator.

Mullard (part of Philips, of course) make these points again in their instructions for the 3W single-ended amp here http://www.r-type.org/articles/art-003g.htm and the 10W push-pull one here http://www.r-type.org/articles/art-003f.htm. They do allow for tagboard construction of the 10W amp but they recommend using a pcb to 'force' the wiring into the best places, and they actually specify that for the 3W amp.

In each case the component list just says 'Technograph' for the pcb with no part number that I can see. I'd be interested to see how the copper tracks should be laid out for an ECL86. Does anyone have one of the specified Technograph pcb's that they'd be prepared to photograph, or at least a reference for a clear picture of one ?

Cheers,

GJ

[Valvepower]

Good call…

Ah, ECL86 stability… there is that 6.8pF capacitor (which isn’t on a lot of the official diagrams) that pops-up in the output stage of the Mk3 Rogers Cadet…

I picked up a 'chassis' single ended ECL86 amp for a fiver at this year’s Audio Jumble… it’s even simpler than the Economic Mullard ECL86 3W circuit.

Terry.

[kalee20]

What a super book! Starting with the design aims of the ECL86, and how they achieved it - even down to the type of glass used, to minimise secondary emission and thus minimise DC feedback.

It's always seemed to me to be a quart-in-a-pint-pot sort of valve - definitely one to derate a bit, and to keep cool, for best life.

And incidentally, although described as a triode-pentode, it looks more like a beam tetrode output, with beam-forming plates rather than a helical-wound suppressor grid. But then Philips were always reluctant to acknowledge the BT invention!

[daviddeakin]

Ooh nice update!

[Leon Crampin]

The Philips ECL86 design study is inded a fasciating document. However, I think Lorenz in Germany should take some credit for innovation in this field. One of the ultimate audio devices of this type must be the ECLL800 which combines a twin tetrode or pentode for double single-ended or push-pull operation together with a unity gain triode for phase inversion. This mut have been a difficult design to get right.

Has anyone constructred an amplifier using this (now expensive) device?

Alternatively, experiences of servicing (mostly German) radios and other audio systems using this device would be very interesting.

I wonder how stable it was, and also its life span prospects.

Leon.

[kalee20]

I suspect the ECLL800 was less of a challenge to get stable than the ECL86.

The ECL86 was designed as the complete active component for following a pick-up or a detector stage, and had lots of gain thus facilitating negative feedback.

The ECLL800 was just an output valve that could replace (say) an EL84, but provide push-pull operation with all its advantages, and give a bit more power. As such, the input level was output-stage magnitude (volts, not tens of millivolts). The input pin fed the 'C' triode grid and one of the 'L' pentode grids direct.

The 'C' triode was very low gain, all it had to do was provide phase inversion at unity gain, to drive the other pentode grid.

Obviously, some care had to be taken, bit I suspect it was easier than the ECL86.

[Aub]

Slightly off topic, the QQV03-10 is a nice achievement. Two VHF tetrodes in a B9A envelope
I've used one for push pull audio in a modulator with good results.

Aub

[bobbyball]

A friend and I like to restore the "better" radiograms, usually Grundig ones.

One particular model my friend came across used ECLL800 valves for the output stages, however these had been replaced by a circuit using 4 x ECL86 valves, on a large strip of Veroboard. This then used B9A plugs to "plug into" the ECLL800 sockets. I presume this was done due to lack of availability of the ECLL800's. Depending on the circuitry the ECL86 in push pull is capable of higher output power than the ECLL800. We noted slight distortion with this circuit which disappeared when ECLL800 valves were fitted (we have a small stock!).

Back to the "humble" ECL86 and I am currently restoring a "Marconiphone" portable record player, more for it's slightly rarer BSR autochanger but also knowing how actually good these can sound, having serviced one for a friend/neighbour in the past. This is one of the very many examples, by numerous manufacturers, using a single ECL86 and sounding very good for it, the large gain of the valve and negative feedback circuitry working well. It is just a pity that none of the "modern day" valve producers can ably reproduce the ECL86 as stocks are starting to run low now, with prices to match!

[Marconi_MPT4]

ECLL800 was in the obsolete-obsolescent-maintenance only category by 1970. Mazda published substitution details using 2x ECL86 which stated no component changes were required. Since the original valve was typically operated in class B, probably the cathode bias resistor value(s) might have required a tweak depending on the actual configuration.

Round about 1962 Mullard developed a new cathode coating to reduce deposition of particles on the adjacent electrode structure for PCL86, and this might have been carried over to ECL86.

Thanks for posting link. Excellent book which gives an insight into valve design considerations. Back in the day Philips-Mullard were one of the best regarding supply of technical design data together with an occasional visit from the TLO.
Rich

[Richardgr]

Reading this discussion, one wonders why Europe didn't embrace the compactron, or even do more on the B9D base (look at the size of those EL519s!).

[bobbyball]

I understand radios were "taxed" in Germany based on the number of valves they contained, hence the drive to use as many multiple section types (eg ECLL800, ELL80 etc).

[Richardgr]

When reading the ECL86 document linked above, I was intrigued what the AD9057 OPT was like, and came across this interesting document describing Philips offerings at the time with some good details for transformers ...
Philips - Components and Materials part 3

OPT described on page 43

In the AD9057 the coils are wound around a C -type core of oriented laminated sheet.
The stray inductance and the winding capacitance are low and, consequently, the
resonance frequency is so high that heavy feedback can be applied without risking
instability.

[Steve_Bell]

The Mazda document referred to in post #16 is available on this forum. See post #4 at:

https://www.vintage-radio.net/forum/showthread.php?t=34989