193 valves in 5 days - rescuing the Miller Classic IV organ

[Diabolical Artificer]

I take my hat off to you Lucien, this sort of restoration is usually done over a several years; five days!! Impossible.

I doubt you have time to film the restoration, but it would make a great one off documentary.

If you need any parts, I'd love to help.

Good luck, Andy.

[SiriusHardware]

Is this an Organ Transplant? Seriously though, well done.

[PaulR]

Originally Posted by Dr Wobble

I doubt you have time to film the restoration, but it would make a great one off documentary.

I bet the BBC would have been interested if there had been more time.

[Hartley118]

I'm uncomfortable with this prejudice against Hunts Moldeal capacitors. In my youth, for those of us who struggled with the inherent leakage of wax paper capacitors - even in NOS ex-WD stock, Hunts Moldseal were a revelation - a quality leak-free component.

Sure, after more than half a century, they've typically developed some leakage, especially if the Mold is no longer Sealed due to cracks. However, unless connecting anode to grid, most applications can cope with a few megohms leakage. The 'replace on sight' policy that we so often hear really can benefit from a bit of technical thought.

Martin

[Lucien Nunes]

There was nothing foolish about Millers' choice of caps - there are no waxies, they are all either Moldseal (in non-critical positions) or Dubilier metal cased (in critical positions). Most of the caps in the oscillators can be allowed a little leakage, so while both types are inherently suspect, there was no cause to assume that the organ would not function until they were all replaced. Had this been necessary, a Herculean effort would have seen it done but I did not take that route. There were capacitor problems for sure, but I will return to these later in the story. FWIW, I had virtually no trouble with the Moldseals at all, most of my woes were with the expensive Dubiliers.

I paid little attention to the valves, experience being that church organs usually arrive with all valves life-expired, thus it was sensible to anticipate changing the relatively small number outside of the oscillators. Most of the latter appeared to be original, but they have lived strange lives devoid of anode current because the organ uses HT keying. Although the heaters might have seen 5000+ hours use, the top notes of one or two of the ranks probably haven't had more than a few minutes of emission in the last 50 years. What effect this would have had on the cathodes was yet to be discovered, however with 72 tested spares in stock and a 3-tier emergency strategy to deal with ECC82 problems, they were of minimal concern.

Of maximal concern was the removal of the generator cabinet to ground level. It had stood near the console at the back of the balcony / choir loft since the church was built, there being no clear indication how it got there. The narrow, winding stairwell did not look like an easy ride for an 8 cwt box the size of a phone kiosk, yet there were also problems with heaving it off the side of the balcony onto a hoist. We measured and re-measured the stairwell and tested it with a cardboard template, coming to the conclusion that while awkward, it would make it down. In places it would be such a tight fit that it would have to be lifted over the skirting, a small groove would have to be made in the plaster, and we would have to rely on interlocking some airspace inside the cabinet with the projecting corner at the first wind.

There was no video of the process but by good fortune the wife of the organist who first contacted us about the organ is a photographer, so we have excellent stills from start to finish. Stand by for day 1.

[Lucien Nunes]

Day 1 am. The easy bit.

[Lucien Nunes]

Day 1 am cont'd.

[Herald1360]

Were any Totectors involved in that process?

[Radio Wrangler]

It looks like both the console and the decor came through unscathed - as well as all the toes.

What was the organ replaced with?

David

[Lucien Nunes]

Were any Totectors involved in that process?

Yes, and at one stage the generator cabinet was standing on mine for a couple of minutes. I take safety on vintage recovery projects seriously - there are always written risk assessments, task-specific method statements, crews are briefed and operations are supervised. I have insurance, lifting gear is tested, PPE and other resources are provided. It gets tricky, though, when there are volunteers who are not actually part of our team - sometimes they like to turn up and lend a hand, without being fully equipped. Hard hats, gloves, safety specs and hi-viz I can lend them when needed, footwear I cannot, so it is occasionally necessary to ask people to stand well clear.

In any case, that was only the warm-up exercise not involving any heavy lifting. That is yet to come.

What was the organ replaced with?

When the works and re-ordering are completed, they plan to install a digital. The Miller had fallen out of use some time previously, so its removal did not leave a musical vacuum.

A technical interlude: Stop controls, extension and unification.

Before we delve into the control system of the Miller, it is important to understand the principles of stop-control in a pipe organ. In Bach's day, things were fairly formulaic. Each manual, and the pedalboard, controlled a functionally separate instrument, complete within itself, that we would now call a division. Every division possessed a number of pipe ranks, each comprising as many pipes as there were keys in the division (typically 58-61 for manuals and 30-32 for pedals) and controlled by a drawstop. Drawing N stops and pressing M keys would cause N x M pipes to speak. E.g. a triad played with Diapason and Flute drawn together will sound three Diapason pipes and three Flute pipes. As with instruments of the orchestra, a pipe organ usually contains ranks of different pitch ranges, often one or two octaves above or below unison (the pitch at which a piano sounds). Drawing an 8' Flute and a 4' flute (one octave above unison) and playing the same triad would cause three pipes of the 8' Flute and three pipes of the 4' flute to speak.

With the coming of pneumatic, electropneumatic and electric action, logic functions could be implemented and various possibilities emerged. Suppose the 4' Flute pipes were identical in tone character to the 8' of the same pitch, just displaced by one octave along the keyboard. This is not necessarily the case but an organ can be tonally designed so that it is. One rank of pipes could be made to serve both stops, by selecting pipes in the correct octave(s) for the stop(s) drawn. All that is necessary is to extend the 8' rank of 61 pipes by one octave upwards (another 12 pipes) and the separate 4' rank is redundant. The only compromise is that when doubled chords are played with adjacent footages drawn, causing multiple requests for a pipe of the same pitch, only one pipe of that pitch sounds instead of one per rank. This is noticeable to the classical organ scholar but not so much to the man in the pew, and with the refinement of the principle that took place during the 20th century a considerable number of ranks could be saved without interfering with the overall musicality of the organ.

Extension can cover a number of octaves in each direction. A Diapason might be extended down to 16' and up to 2'. Four stops from one rank, using 61+(3x12) = 97 pipes instead of 61x4 = 244. Those savings could be directed to adding more variety of tone colours, e.g. by adding a different stop, possibly itself extended. This was exploited heavily in the Cinema Organ, where variety was essential and space and budget limited. Many cinema organs contained just one rank of each tone colour, extended as far as required. This was arranged to serve all the divisions, a principle termed unification, to get the maximum bang for the buck. Instead of a Trumpet for the Solo manual and another for the Great, one extended Trumpet would do for every pitch on every division.

In case this sounds like penny-pinching in the extreme, look at it another way. If your organ contains 10 ranks it is unlikely to contain more than one Trumpet. In a classical 'straight' organ, it would be available only on one division and at one footage. On a cinema organ, you can use it at three footages, on any division, all at once if you like. I.e. much more flexibility for minimum extra cost. Most electric organs employ some aspects of extension and unification, and when we examine the oscillator arrangement and matrix relays of the Miller it will be apparent how closely it follows the principles outlined above for the pipe organ.

The matrix relays are pertinent because they contribute a fair chunk of the weight of the generator cabinet, which we are about to carry downstairs...

[Herald1360]

Organists had to be quite fit in the early days to physically operate the keys on a large tracker with several ranks in use simultaneously. The keyboard also had to be part of the organ itself. Pneumatic action took the load off fingers, but was too slow responding to allow the keyboard much freedom either. It took electrickery, cunning Victorians and some serious multicore cables to set the keyboard free.......

Looking forward to more of this

[Radio Wrangler]

There are some good videos on Youtube of the grand organ at Saint Sulpice which illustrate how involved drawstop operation got before the thumb-piston and stop motor were invented to add a degree of automation.

The matrix relays amount to a huge mixer with hundreds of inputs. Relays have their problems, but with electronic switching, the addition of more and more channels of noise floor becomes a limitation.

David

[peter_scott]

If you are ever out of work Lucien then I'm sure there would be a job for you in Atlantic City where they are restoring this monster with its 32,000 pipes.

Peter

[Lucien Nunes]

In this case the matrix relays handle only DC keying, therefore they have no impact on the audio. They generate a logical sum of products, where n = relative octave number:

NOTE(n) = KEY(n).STOP(n) + KEY(n-1).STOP(n+1) + KEY(n-2).STOP(n+2)...

For a unified rank there may be multiple terms at each footage, from the various divisions that can access it. In pipe organ action, matrix relays receive action voltage (usually 12-24V DC) and output likewise. In the Miller they receive action voltage (24V DC) but output regulated HT (150v) to key the oscillators. They are the three grey metal boxes in the base of the generator cabinet - one per division - they are very heavy and they are on the move:

[Lucien Nunes]

More

[Lucien Nunes]

And more

[Lucien Nunes]

Removal completed

[Radio Wrangler]

Impressive work!

I take it you moved it straight to the venue for the hoped for performance and fettled it there?

David

[Lucien Nunes]

No, it could only move into the venue on the day of the event, during which time there would be tech checks and rehearsals going on. We took it back to the workshop with the intention of getting it as ready as possible, preferably completely tuned and tested. The most we could reserve was 2 hours of production time for tuning, testing and repairs to any faults that had occurred in the move, prior to the organist starting rehearsals.

Day 2 am.

Testing in priority order:

1. Test the oscillators, get a rough indication of the number of each type of fault. Measure repair time for each type of fault, compute total time to repair oscillators.

2. Inspect matrix relays for damage / maladjustment due to moving, quick test of contact operation, decide whether cleaning / realignment would be needed.

3. Check console operation likewise.

4. Test ancillary circuits, power supplies, preamps, filters etc.

[wave solder]

What happened with the speakers and their enclosure(s)?