UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Other Vintage Household Electrical or Electromechanical Items

Notices

Other Vintage Household Electrical or Electromechanical Items For discussions about other vintage (over 25 years old) electrical and electromechanical household items. See the sticky thread for details.

Closed Thread
 
Thread Tools
Old 9th Dec 2015, 12:48 am   #61
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

It's now the end of day 3 (Sunday), progress is just about on target and the tone generator is starting to sound good, albeit still played one or two notes at a time using clipleads. Oscillators, filters and preamp all working OK, not exactly 100% up to spec but eminently musical for our purposes. A production discussion and some planning with the organist confirm that we will not be using a Leslie speaker, and that the organ will be heavily amplified by the venue PA system. The original 807-based power amplifier was not expected to be required for this one-off performance and is now officially removed from the critical path. It can be overhauled when the pressure is off and it has a suitable speaker to drive.

The matrix relays were next in line for attention, but a chance discovery while investigating a few items on the console uncovered a new problem: sticky keys. Three or four of the Great keys were tending to stick down and others were marginal. The order of play for Day 4 was thus revised to:

*Service key action and anything else important on console
*Inspect / clean / adjust matrix relays
*Tidy up power supply & fix tremulant
*Start on Main Cable

At the left of the preamp chassis is the light gate for the swell pedal. Inside this light-tight housing an incandescent lamp is controlled in brightness by progressive contacts on the pedal shorting sections of a resistor ladder. This illuminates an LDR that is part of a divider, controlling the volume in as smooth a manner as could at that time be achieved without using a pot, with all its attendant risk of crackles, noise and short useful life.
Attached Thumbnails
Click image for larger version

Name:	33 Filter panels.jpg
Views:	474
Size:	198.7 KB
ID:	116649   Click image for larger version

Name:	34 Filter recovery amp.jpg
Views:	464
Size:	155.7 KB
ID:	116650   Click image for larger version

Name:	35 Preamp.jpg
Views:	466
Size:	193.1 KB
ID:	116651   Click image for larger version

Name:	36 Swell cell.jpg
Views:	430
Size:	195.1 KB
ID:	116652  

Last edited by Lucien Nunes; 9th Dec 2015 at 12:55 am.
Lucien Nunes is offline  
Old 9th Dec 2015, 1:06 am   #62
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Power amp and speaker, neither of which are needed for the initial performance, although if the speaker can be recovered intact in the future it would be good to re-unite with the organ. In the meantime a conventional cabinet will have to be substituted, once the amp is overhauled.
Attached Thumbnails
Click image for larger version

Name:	37 Power amp.jpg
Views:	506
Size:	197.3 KB
ID:	116653   Click image for larger version

Name:	38 Power amp open.jpg
Views:	499
Size:	69.9 KB
ID:	116654   Click image for larger version

Name:	39 Speaker.jpg
Views:	522
Size:	123.2 KB
ID:	116655   Click image for larger version

Name:	40 Speaker mouth.jpg
Views:	518
Size:	134.0 KB
ID:	116656  
Lucien Nunes is offline  
Old 15th Dec 2015, 8:46 am   #63
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,636
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

The speaker is something else! Full range Tannoys at the back? They look like Goodmans drivers on the left but where are they ported? Or are they acting as a LF vibrator? Its a hugh speaker for what must have been 70w total OP max. I wonder what it sounded like? Not Hifi but it must have had some throw to be able to be heard everywhere in the church.

Interesting thread Lucien, how did it go on the day?

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 15th Dec 2015, 1:32 pm   #64
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

The throat of the horn is on the left with the two Goodmans drivers mounted in it. It descends to floor level and doubles back on itself to form the visible mouth. So the drivers' rear radiation is into the mouth of the horn, the effect of which I am not sure about. It seems to have done a reasonable job by way of efficiency, although I never heard its full potential as the organ (and especially the audio side of it) was well below par when we tested it in situ. If the speaker remains in position and operable in the future, I would be interested to make a direct recording of it, take it to the church with the amplifier and replay it through the speaker as though the organ were still there.

You'll have to wait to find out how it went on the day, we haven't got to the exciting bit yet...
Lucien Nunes is offline  
Old 26th Dec 2015, 5:21 pm   #65
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Day 4 AM

It's Monday and we have to resolve the sticky keys issue. A change in environment often reveals itself in wood components changing shape slightly, and nowhere is this more important than in key action. The console is of decent construction, not quite as traditional as say a Compton, but still with wood-core keys and craftmanship in its execution. What could have been a can of worms turned out not to be, because the manuals did not carry on twisting and bending and forever throwing new faults as I had feared. One of the main problems was, just like in an autochanger, stiff grease. A traditional organ manual does not have points requiring grease, but through the latter half of the 20th century an increasing number of electronics used manuals with metal components in which any metal-metal contact requires greasing. In this case it was the touch springs, aided by sticky and dislodged bushings, that were dragging. To fully restore the touch of these manuals the bushings in at least the middle 3rd of the compass need doing, however with a bit of cleaning, greasing and tweaking all 122 keys were back in commission.

The remainder of the console was sorted out there and then as it was already open and powered up. A few dirty / damaged contacts to fix, loose bits and bobs, a high-resistance at a thermal circuit breaker, nothing very serious. One surprise was that all the indicator lamps for the voicing selection had a sound filament. Many look original and somewhat blackened but not one needed replacing. To recap how this works, you have a stopkey marked not 'Flute 8' but simply '8', over which there are two backlit legend indicators, 'Flute' and 'Clarinet'. If you press the 'Flutes' selection piston, a selection relay latches on and lights the 'Flute' indicators above all tabs that can be Flutes. It also operates the voicing relay in the generator cabinet that selects the Flute filters in the voicing panel for the Flute / Clarinet rank. Pressing the mutually exclusive 'Clarinets' piston releases the relay and selects the Clarinet indicators and filters. Operating any of the associated stopkeys will engage a contact bar of the right footage in the matrix relay, controlling generators of that rank at the appropriate pitch, but it is the piston that selects which of the voices those stopkeys will sound.

And then the cable arrived. 100m of 15-pair 24AWG flexible, Belden 95xx style, to be cut into 8 pieces to replace the main cable left behind at the church. This was put aside while the matrix relays were tackled.

The matrix relay is a kind of logical multiplier or coincidence detector. It consists of an X/Y matrix of contacts that close only at the intersection of an energised magnet in column X operated by a depressed key, with one in row Y operated by a selected stop. Matrix relays have been in use in extension pipe-organs since the 1920s as an alternative to the stop-switch relay. The latter uses common-contact relays for the keys, with as many isolated fixed contacts as there are stops in the division. These are then gated by multi-pole relays for the stops with as many completely independent poles as keys (usually 32 for pedals and 61 for manuals). Between the two stages, a very large amount of interwiring is required along with double the theoretical minimum number of contacts. The matrix relay does away with all the surplus, by detecting the coincidences mechanically.

One popular method, typified by Compton, was to make the key contact wires fall, and the stop contact bars rise, each by around 2/3 of the separation at rest. In the Miller, the stop contacts take the form of rotating bars, shrouded over 1/2 the circumference by a moulded insulator. An energised key magnet pulls down its paxolin trace rod against a spring, freeing the contact wires to press down on the stop bars. For stops that are off, the bar lies with the insulator uppermost, so that the wire cannot contact it. Putting on a stop causes the stop magnet to rotate the insulator to the side, out of the path of the wire so that it makes contact when keyed. In the Miller, the output from each contact is an HT feed to energise a specific oscillator, ballasted by a 10K resistor to form an LPF to minimise keying thumps. Because all the contacts are mutually independent, multiple contacts can simply be ganged together without erroneous 'backfeeding' occurring.

Day 4 PM.

The matrix relays had been somewhat shaken up in the move and dust that had been laying harmlessly in the bottom of the felt-lined casings had been scattered all over the contacts. A few parts had also dismantled themselves, as loose screws and register pins that had been slackening over the past half century of use and vibration had finally been dislodged. A few contacts deranged in the process needed some straigthening and tensioning but overall the three relays went back together easily and after a good clean, seemed to key quite reliably without extensive adjustment. As with most of these things there are usually one or two troublemakers that don't quite come up to spec like the others, often for no apparent reason. Over the next two days there were occasions to pop the fronts of the relays open to tickle them up a little but after a programme of exercise (hitting every key 500 times with all the stops on) missing notes were few and far between.

At this point the maintenance to the cabinet appeared to be nearing completion, with just the power supplies and tremulant left to service. With the exception of the rectifiers, all heaters are supplied at 12.6V from one large transformer with four isolated secondaries. Each rank of oscillators complete with its filter preamp is thus electrically independent, with its own humbucking pot, the 7.2A rating of each secondary being just sufficient for the largest rank with 48 12AU7s. The four ranks also have independent HT feeds; two HT transformers and GZ32s each provide two 250V HT rails Via C-L-C-L-C smoothing (serving the preamps / filters), from which four VR150s ballasted by 3k provide up to 30mA stabilised 150V HT for the oscillators via the matrix relays. Of the 10 electrolytics, four had additional caps paralleled up and there were signs of seal failure, so although there was little HT ripple evident it was thought wise to just temporarily byapss the lot with wire-ended caps pending finding suitable sized dual-section cans to make a permanent job of it. One of the 3W carbon stabiliser load resistors had gone very low, under 2k, so this was replaced.

The tremulant is another Hartley oscillator with a large iron-cored coil that runs at a few Hz. Or should, as in this case it was non-oscillating despite evidence of all the passive components having been changed, not very tidily. Unsurprisingly, one half of the coil was O/C, but due to the large number of turns and limited available time it was not though wise to attempt to rewind yet. All the trem oscillator has to do is drive the grids of two ECC82 shunt modulator stages which, when their 2k anode loads are inserted into the 150V HT lines, tug gently at two of the four oscillator HT rails. With the coil disconnected, the trem oscillator valve grid was driven by a function generator but no tremulation resulted until the other ECC82 was replaced with one that actually had some emission. With the component values chosen by Miller, the trem was found to have only a very mild effect, and in the event it was not needed for the performance.

The end of the Monday consisted of tidying up loose ends in the generator cabinet. With all component replacement completed and the preamp substitute 6SN7s running on their temporary external stabilised DC supply, the oscillator heater humbucking was adjusted for each rank in turn. The pots were found to have significant effect, and no position was quite satisfactory on two of the ranks, although with the total amount of heater wiring and the large, low-impedance pickup circuits located nearby it was thought that a very faint background hum was probably unavoidable. Nonetheless it seemed worth trying to pinpoint any valves with poor H-K insulation in the suspect ranks. With the wipers of the pots lifted from earth, the aggregate figure of all ranks was very poor, as low as 5k from heater to ground. Before trying to find the culprits, a few valves were tested at random and an interesting fact emerged - they were all rather low, 500k being about the best. Whether this relates to having been run with no HT for most of their lives (and they are all similarly down to about 50% emission), is an interesting question that I have not had time to investigate. After pulling a couple of serious offenders and re-trimming the pots, it was found possible to get the whole thing acceptably quiet, so the generator rack was declared ready-for-service at about 11pm. The workshop was tidied ready for the wiring marathon to remake the main cable, starting the following morning.
Attached Thumbnails
Click image for larger version

Name:	41 Heater transformer.jpg
Views:	416
Size:	197.3 KB
ID:	117463   Click image for larger version

Name:	42 Power supply.jpg
Views:	398
Size:	190.5 KB
ID:	117464   Click image for larger version

Name:	43 Tone setting pistons.jpg
Views:	415
Size:	195.6 KB
ID:	117465   Click image for larger version

Name:	44 Pedal matrix relay.jpg
Views:	403
Size:	114.4 KB
ID:	117466   Click image for larger version

Name:	45 Pedal relay hinged down.jpg
Views:	389
Size:	141.0 KB
ID:	117467  

Lucien Nunes is offline  
Old 26th Dec 2015, 5:27 pm   #66
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

A close-up of the contact arrangement in the pedal matrix relay. This one of the three was chosen as the busbars are visible between the groups of traces. Having only 32 keys instead of the 61 required for the manuals, but being made to a standardised design with all octaves of each pitch class adjacent for ease of bussing contact outputs together, the pedal relay has these 12 gaps for notes above middle G where the pedalboard stops.
Attached Thumbnails
Click image for larger version

Name:	46 Pedal relay detail.jpg
Views:	346
Size:	191.7 KB
ID:	117468  
Lucien Nunes is offline  
Old 26th Dec 2015, 10:22 pm   #67
Courtney Louise
Octode
 
Courtney Louise's Avatar
 
Join Date: Sep 2005
Location: Welshpool, Powys, Wales
Posts: 1,325
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Wow!
__________________
33, 45, 78, around and around they go...
Courtney Louise is online now  
Old 26th Dec 2015, 11:42 pm   #68
Phil G4SPZ
Dekatron
 
Phil G4SPZ's Avatar
 
Join Date: Apr 2005
Location: Bewdley, Worcestershire, UK.
Posts: 4,735
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Although by no means knowledgeable on the detailed architecture of electronic organs, and being a mere Grade 1 player many years ago, I'm finding this thread absolutely fascinating and have been following it avidly for several reasons. Firstly, the heroic rescue and restoration (and I'm eagerly awaiting the concluding instalments) but secondly for the extremely readable and understandable description not only of the work undertaken but also the underlying theory. I can only hope that this thread can be elevated to 'sticky' status, and/or that Lucien uses it as the basis of a book!
__________________
Phil

Optimist [n]: One who is not in possession of the full facts
Phil G4SPZ is offline  
Old 27th Dec 2015, 10:19 am   #69
Bobdger
Hexode
 
Join Date: Jan 2015
Location: Rustington, West Sussex, UK.
Posts: 381
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

I to would love a book and to hear the organ played. Lets hope that Lucien uses it as the basis of a book.

Bob.
Bobdger is offline  
Old 27th Dec 2015, 11:52 am   #70
Tractorfan
Dekatron
 
Tractorfan's Avatar
 
Join Date: Mar 2008
Location: St. Frajou, l'Isle en Dodon, Haute Garonne, France.(Previously: Ellesmere Port, Cheshire, UK.)
Posts: 3,177
Smile Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Hi,
This just an absolutely fascinating thread. Love reading about the progress and it will be amazing to hear the Miller when the time comes.
Forgive me if I've missed mention of it earlier, but what are the two electric lamps for in the base of the cabinet? I'm assuming that they're anti-condensation heaters? When I worked in the power generation industry, the switchgear cabinets all had heaters which resembled ceramic dropper resitors inside a little cage which were energised when the motor contactor dropped out.
Cheers, Pete.
__________________
"Hello?, Yes, I'm on the train, I might lose the signal soon as we're just going into a tunn..."
Tractorfan is offline  
Old 27th Dec 2015, 2:01 pm   #71
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Thanks for all the positive comments, I wasn''t sure whether the machine itself would be of much interest but the story of its rapid resurrection seemed to be unusual enough to be worth telling. As for a book, this project and many more like it will appear in 'The Making of Electrokinetica' as and when that is standing on its own legs and something to brag about. We're still at the beginning of that story, although there have been thrills and spills aplenty on the way to where we are now. 2015 has been a very slow year for EK due to work commitments, but the CP TX 'Eureka! moment' made up for it. As Bruce Galvin and I examined the old klystron left from the previous TX that we had been offered (the official reason for my visit to CP) the penny dropped, I turned abruptly, gestured at the Sigma TX and said 'But this is actually what we ought to be saving,' and the rest is history.

With the less accessible aspects of 20th century technology it is very important to seize the moment and make something of them, thrust them into the limelight at least once to be recorded in the collective consciousness while people can still relate to them. If you're digging a hole and you find a T.Rex, everything stops, the palaeontologists and reporters are called in and everything proceeds in a very controlled and conservative manner, to make the best of the find. If instead you find an Otis 6850 selector, well, that might just be chucked in the scrap man's van before anyone can say 'Going Up'. But the chances are that almost everyone over 30 reading this has unknowingly interacted with a 6850 which (thankfully) you can't say about the dinosaur.

On which subject (I hope Paul doesn't mind me putting an inline plug here) I'll be resurrecting our forum at EK as of today (it's been dormant, with no news to report) but it will carry new stories of jousting with electromechanical monsters that are just too far off topic to fit in here. Next year will be the Year of the Lift for us - as well as the Year of the First Open Day - so expect to see our own Brontosaurus craning its neck over a skyscraper and gently picking up the express advancer panel in its teeth.

Back to the Miller, the light bulbs are indeed anti-condensation heaters, energised by N/C contacts on the main generator power relay. In truth it doesn't need them half as much as the electrostatic generators that might have been used had a competitor's product been selected. Those work with such high impedances that the tiniest trace of moisure absorbed into surface dust can cause all manner of weird effects. Similar bulb-shaped measures were taken by Compton, until superseded by the proprietary 'Dampp-Chaser' (sic) tubular heaters that were popular in pianos and organs alike.
Lucien Nunes is offline  
Old 28th Dec 2015, 2:52 pm   #72
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

You will recall from the removal episode that the original wiring from console to generator cabinet was found to be permanently trapped in the fabric of the church building and had to be abandoned. It consisted of eight 15-pair solid-core telephone cables, in which most the 240 cores carry 24V 'action current' switching signals of around 100mA, derived from the power supply in the console, to the coils of the matrix and auxiliary relays in the cabinet. These comprise 61+61+32 = 154 keys, 30 stops and a dozen or so miscellaneous functions. A dozen cores carry the variable AC lamp current for the swell light gate via the expression pedal contacts, and the remainder serve as the 24V return.

A type of cable I have always found useful is the Belden-equivalent 95xx series of screened 24AWG multipair, of which CPC stock most sizes in 100m rolls at a very advantageous price. 9515 seemed an ideal replacement for this project; with 7/0.2 stranded cores it is more flexible than the original telephone cable and less likely to break at the terminations after repeated handling. For economy and speed of wiring, 37-pin D connectors were chosen, one per cable with 7 contacts unused. It was thought that connectors at both ends would be a good idea, but under pressure of time the console end would be wired directly onto the tagboard for now and connectors inserted in-line later. The stubs of cable at the generator rack had deliberately been left as long as possible, with the intention of putting connectors at that end from the outset.


Day 5 AM.

The console tagboard has no labels whatsoever. The cable stubs were therefore stripped back, traced, and found to follow the old PO 'BOGBSWRBYV' colour code in a convoluted but consistent way. Key contact tags are arranged not in chromatic order (as they usually are in a pipe organ, following notes from 1-61 up the keyboard) but grouped in a kind of mimic of the matrix relay magnet layout, which follows a pitch-class grouping so that octaves of the same note are adjacent. These account for five of the eight cables, the remaining three serving the stops and other functions. Due to its US origins, the 9515 cable follows the ICEA / ANSI colour code which bears no relation to the PO code, although thankfully they are two of the three codes that I know well enough to avoid having to look them up as I worked. They are both pairwise codes of single-colour cores, i.e. each core is only identifiable in conjunction with its partner, making it necessary to strip back enough sheath to unambiguously identify the pairs. When preparing a new end, it is helpful to lay the pair up more tightly for a few turns, so that the cores don't drift apart as wiring proceeds.

The new cables were cut and stripped, a small flight case placed behind the console as a stool, the soldering station parked on another to the side, and with the aid of a cup of coffee, wiring began.

Black/Blue > pair 1 > Black/Red
Black/Orange > pair 2 > Black/White
.
.
Blue/Slate > pair 10 > Red/Blue
Orange/White > pair 11 > Red/Yellow
.
.
Orange/Slate > pair 14 > Green/White
Green/White > pair 15 > Green/Blue

When all 120 pairs were done, I got up and wandered outside to stretch my legs, reflecting on the likelihood of a mistake amongst the blacks or reds. Where seven cores of the same colour are clustered together rather than alternating along a row of tags, with only their partner for id, it's not difficult for two to become transposed. If they were both keys or stops, the effect of a split pair would be obvious and easily traced to its cause. Other functions, or a return crossed with a feed, and anything could happen including burnt-out contacts.

Day 5 PM.

The next job was to put 37-pin females on the other 9515 ends; this could be done at comfort of the bench, so with a new cup of coffee round 2 began. No code translation this time, just wire pairs in sequence along the connector pins, so it went ahead quickly and with good confidence in the accuracy. It was also easier to check visually than the tagboard, albeit with the same risk of split pairs. Once completed, it was time for a late lunch before starting on the relay cables. Round 3 began at about 4pm, sitting on the flight case by the side of the cabinet. This was the easiest of all as the PO code is the more familiar, and the solid cores quicker to lay up as pairs, so it was finished and inspected in under 3 hours.

TESTING TIME AT LAST!

At just after 7pm on the evening before the show, I was finally in a position to plug the console into the generator cabinet and try playing a tune. Would the two ends talk nicely to each other? Would there be smoke, jumbled-up notes or just a blown fuse. Find out what happened in the next episode...
Attached Thumbnails
Click image for larger version

Name:	47 Console tagboard.jpg
Views:	342
Size:	155.9 KB
ID:	117494   Click image for larger version

Name:	48 9515 cable end.jpg
Views:	337
Size:	124.8 KB
ID:	117495   Click image for larger version

Name:	49 PO cable end.jpg
Views:	331
Size:	160.1 KB
ID:	117496   Click image for larger version

Name:	50 Main cables.jpg
Views:	320
Size:	150.8 KB
ID:	117497  
Lucien Nunes is offline  
Old 28th Dec 2015, 4:36 pm   #73
Guest
Guest
 
Posts: n/a
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Arghhh... the suspense!
 
Old 28th Dec 2015, 8:20 pm   #74
mole42uk
Nonode
 
mole42uk's Avatar
 
Join Date: Aug 2010
Location: Resolfen, Wales; and Bristol, England
Posts: 2,588
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

I'm on tenterhooks...
__________________
Richard

Index:
recursive loop: see recursive loop
mole42uk is offline  
Old 29th Dec 2015, 7:41 am   #75
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,636
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

That's a lot of fiddly soldering on those plugs, the wires inevitably shift halfway through a joint.

Ditto above,

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 29th Dec 2015, 9:05 am   #76
Hunts smoothing bomb
Octode
 
Hunts smoothing bomb's Avatar
 
Join Date: Oct 2005
Location: Wimborne, Dorset, UK.
Posts: 1,407
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

This really is an awsome beast,
I'm absolutely loving the read so far, this really would make a great book!

I can't wait for the next installment, it makes restoring my early 60's Bird twin manual organ look like childs play!

Cheers
__________________
Lee
Hunts smoothing bomb is offline  
Old 29th Dec 2015, 9:37 am   #77
Phil G4SPZ
Dekatron
 
Phil G4SPZ's Avatar
 
Join Date: Apr 2005
Location: Bewdley, Worcestershire, UK.
Posts: 4,735
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

My money's on it playing like new...
__________________
Phil

Optimist [n]: One who is not in possession of the full facts
Phil G4SPZ is offline  
Old 29th Dec 2015, 8:01 pm   #78
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Switch on console power... generator power relay operates and valves warm up. No breakers trip, no ciphers. I wanted to try the stops and controls first as miswiring of the three miscellaneous cables was more likely to cause damage than the key cables.
Hold a pedal down, try each Pedal stop, all the correct stops sound.
Hold a chord on the Great, try each Great stop, all the correct stops sound.
Hold a chord on the Swell, try each Swell stop, all the correct stops sound.
Try the six voice selection pistons, all the correct voicings are selected.
Rock the swell pedal - volume varies smoothly.
Put on a couple of 8' pedal stops, play each pedal in turn, all notes sound in correct order (rough tuning notwithstanding)
Repeat on Great, all 61 notes sound in correct order.
Repeat on Swell, do, re, eek, so, fa...

Grrr. Two notes crossed, something like tenor E and top D#. A quick count along the tags revealed that these are nearby black wires on the console tagboard (although the notes aren't adjacent, their magnets are). I actually had a small wager on that I could do all 720 connections without a mistake and I was kicking myself for such a boring and obvious slip. I swapped the Swell and Great cables, localising the fault to the generator side which was a surprise. A closer inspection of the 37-pin revealed that the disposition of the two split pairs in the cable made it unlikely for them to be confused, as they were lying well apart where they emerged from the cable sheath. After stripping off a further inch of sheath it became quite clear that the pairs were not split at that point. Nonetheless both manuals played correctly all the way through when connected to the Great relay, so the fault had to be at the Swell relay itself. And sure enough, the original wiring that I had not touched had a visibly split pair, that must have been 'corrected' during installation by making the same substitution at the console tagboard. Thinking back, having determined the tag sequence I had stripped all the original wiring off without proving every conductor, the two crossed blacks would have been inconspicuous and the correction thus overlooked and discarded. It had been done that way because the wires in the relay were tied out in different bundles and difficult to correct. For the same reason I re-instated a corrective uncrossing at the relay side of the 37-pin which took only seconds, to be rewarded with a fully functioning action and the satisfaction of having made a better job of my 720 connections that the original wireman did of his 480, and my bet won.

It was something after 8pm, I played Vater Unser, had a coffee, played Alain's Postlude pour Complies, listening to the tone of each stop behind the slowly moving harmony. Not too bad, a couple of notes were a bit intermittent in the relay again but the oscillators were all there, albeit in some cases horribly out of tune. I phoned Eloise; she didn't answer so I left a message saying things were generally coming together OK, then plonked the phone in front of the speaker and added the last few bars of Widor's Toccata from the 5th Symphony.

Two major jobs left to do.

a) Sort and pack all the tools and spares to take with us
b) Do a trial tuning of one complete rank to time how long I would need to allocate for that at the venue.

My list (abridged)

DI box (to get preamp Hi-Z signal into PA)
Phones amp
Closed-back phones
Tone tracer
Variable PSU
Function gen
Various audio & power cables
Stock of spare valves
Boxes of passive components
Crate of handy bits - wire, connectors, sleeving etc
Kit of organ-action spares (contact wire, leather buttons, felt etc)
Normal toolkit
Electronic service aids kit
Inspection lamps
Slings
Trolley
Console skid
Sheet of 1" ply for laying cabinet on its side
Barrier tape & danger notices

10pm. all packed and lined up for loading. 96 oscillators to tune, maybe a minute each as some will need regulation too. Done by midnight then.

Tuning proceeds in much the same way as on a pipe organ. Starting in the middle of the compass, the scale is set on 12 oscillators of the Diapason, which has a clear and distinctive tone, by counting beats or using an electronic tuning meter (effectively a frequency meter calibrated in semitones and cents). The oscillators are energised in turn using a flylead carrying HT derived from the appropriate regulator, and a 10k series resistor. The next octave is tuned to the first by energising each oscillator with the already-tuned one in the octave below, the following octave is tuned to that one, etc. Then the process repeats in the opposite direction until the rank is completed. Other ranks begin by having their middle octave tuned to the first rank's middle octave, then as above.

I laid the scale on the Suflex-equipped middle octave of the Diapason at A=430, believing this to be a good compromise for the paper-equipped oscillators that won't tune comfortably around A=440. At first things were OK but as I went down, I started finding oscillators that would not quite get there. A=420 seemed OK, so I tried an octave of Suflex at that pitch and just about made it. So back to square one with A=420. I should mention that these pitches are as I remember them and the principle is correct, although the actual numbers used on the day could have been different as it all became a bit of a blur later. A few oscillators looked as though they might need their caps changed, or we'd have to start again at A=410 or lower and I didn't think that was going to happen. As I worked, a little mental alarm bell started to ring. Why had early experiments with tuning not led me to consider such a low pitch? I recalled that was the Flute rank, so tried A=420 on the Flute and the Suflex oscillators definitely weren't having any of it. *expletive deleted!*. By not previously comparing the tuning on multiple ranks simultaneously, I had never proven that all oscillators would tune within a common range with their existing tank capacitors. *another expletive deleted*.

11pm. Ready to bail out and invoke the backup plan. Phone Eloise? She has a very tough and critical day ahead and probably doesn't want to be told at this point that the star of her valve-themed show is now a bog-standard tonewheel Hammond. Leave it until the morning, there's nothing to gain from disturbing her.

11.30pm. Actually, what have I got to lose? Only the bottom few octaves of each rank will need their caps changed. Maybe 100 caps. Perhaps some kind of system might work where they are all moved along one semitone if the drift was fairly constant (it wasn't). Start again at A=430, prove it works on all or most Suflex and mica oscillators, then tune the whole rank, downwards first, changing caps wherever necessary. Time it, do a new feasibility study.

2am. Two octaves into the paper oscillators and I have a kind of a routine going. Set the tuning coil midway, try the note, guess the ratio between current and target values, or read off the marked values if the labels are all still attached. Choose between replacing the caps outright (maybe with the ones from the note above), just one of them if multiple in parallel, or putting something large in series. If it lands outside middle 2/3 of the tuning range, try again. If within, tune and move on.

5am. I have tuned the Diapason throughout and tweaked the regulation of the anode transformers to a nicety. There are capacitors everywhere, much of what I had packed is now scattered around the workshop, but the Diapason sounds fantastic. It has to be possible, surely? That took about 5 hours. Let's see, doors at 7pm, ready by 6, rehearsals 4, troubleshooting 3, get-in 2, travel 12, load 11, If I ring Dan and tell him to come at 11.30 I have 6 hours available. I've got faster with practice, could do the Diapason in 3 hours so we have time for 16 octaves, that's the whole Flute and the middle of the Trumpet, the 12th misses the two lowest octaves so that will be quick, perhaps it can be left until we arrive.

6am. Another coffee, another capacitor. What a jumble - recycled papers, stripies, chunky polypropylenes, anything with two foils and something between them, never mind the tempco. Series, parallel, series parallel. Bits of card wedged in the tuning coils to open the gap, if it's not quite there on the tuning coil, loosening or tightening the bolts of the tank / sine takeoff transformer can affect the inductance just enough to make the difference. Cheating? Me?

12ish. Dan arrives, I'm still in full swing. We have two organs in one - a two-ranker in good tune at A=430 and another two-ranker all over the shop nearer A=410. Pull the cables, tip the cabinet on its side with the forklift (grams: sound of lots of capacitors falling out), repack all the chutney, chuck the whole lot in the van and follow it down to the venue in Soho. I rang David the organist. 'It's all working,' I said in a moderately upbeat tone, 'Tuning's a bit weird though, might have some quirks until I've tinkered with it once we're there and set up.'
'Don't worry' said David, 'I'll see you at 4, tell me what's OK to play and what not and I'll work around it.'

We reached Soho in very good time. After some negotiation with the parking people Dan managed to get the Luton van (which was too tall to get in) off the road and in a safe place to unload. We called some of the show crew down to help get everything up the 200 foot ramp to the venue above the car park.

'You can't park that there whatever it is.'
'It won't go under the barrier.'
'Just take a ticket and it'll open. You won't get charged unless it's got number plates* on it!'

And in we went, rack, console, bucket'o'caps and all...

*Apparently the registration ORG4N does exist, it was up for sale a few years ago for £35k.
Attached Thumbnails
Click image for larger version

Name:	51 Generator rack at car park.jpg
Views:	392
Size:	129.9 KB
ID:	117539   Click image for larger version

Name:	52 We've arrived.jpg
Views:	344
Size:	54.7 KB
ID:	117540  

Last edited by Lucien Nunes; 29th Dec 2015 at 8:06 pm.
Lucien Nunes is offline  
Old 29th Dec 2015, 9:27 pm   #79
sexton_mallard
Hexode
 
sexton_mallard's Avatar
 
Join Date: Sep 2013
Location: Ellesmere, Shropshire, UK & Co. Cork, Ireland.
Posts: 499
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

Fantastic stuff Any chance of a video posted somewhere soon with the beast in action?
sexton_mallard is online now  
Old 30th Dec 2015, 9:53 am   #80
Lucien Nunes
Rest in Peace
 
Join Date: Apr 2008
Location: London, UK.
Posts: 2,508
Default Re: 193 valves in 5 days - rescuing the Miller Classic IV organ

There's one more episode to write, then I will try and post some video. I only have rather scrappy mobile phone footage - the moody lighting with spots on the console was too contrasty for the phone camera and the audio tended to overload the mic as I was stationed right by a speaker stack, but you can get an impression.
Lucien Nunes is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 2:18 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.