[R-390] R-390A Cosmos PTO has erratic linearity

Mon Jun 16 00:58:14 EDT 2014

Probably the best written article ever on the Cosmos PTO.  Bravo, Tom Marcotte!

vy 73 de Norman, KG4SWM

On Sunday, June 15, 2014 7:15 AM, Tom M. via R-390 <r-390 at mailman.qth.net> wrote:

>
>
>Bill,
>
>This is an article I wrote about how to calibrate the Cosmos PTO.
>
>Take a look and see if you find any nuggets.
>
>Tom
>
>-------------------------------
>
>Calibration of The Cosmos Permeability Tuned Oscillator for the R-390A
>
>by Thomas Marcotte, P.E.
>N5OFF      
>This article will focus upon the calibration of the Cosmos permeability tuned oscillator (PTO) for both linearity and endpoint adjustments.   Few devices have been so maligned or poorly understood than the Cosmos PTO employed in the R-390A HF receiver.  This device has been mentioned in articles about the R-390A many times, as well as on several Internet mailing lists, often negatively.    The features of endpoint and linearity adjustments for this PTO have often been misunderstood, leading one to suspect it may be a poor device.  This article will attempt to convincethat the Cosmos PTO is worthy of service in the R-390A when adjusted properly. 
>Collins designed and built the first R-390A receivers utilizing its own 70H-12 PTO.  This PTO covers a frequency range of 3.455 Kcs to 2.455 Kcs in ten clockwise turns.  It is a robust device constructed of the highest quality materials, including an encapsulated main coil, sturdy compensating stack, and a temperature controlled oven. 
>The R-390A receiver was produced by many manufacturers other than Collins. These manufacturers were at liberty to employ the PTO's of suppliers other than Collins, as long as their performance met the specification MIL-R-13947.  Consequently, PTO's employed in the R-390A were supplied by many other manufacturers including Motorola, Progresstron, Dubrow, Cosmos and Raytheon.  Of all PTO types, the Cosmos seems to be the most common.  Electronics Assistance Corporation used these PTO's exclusively in its production run of 11,000 R-390A's around 1967.  
>Cosmos Industries was a New York manufacturer of radio equipment located at 31-28 Queens Boulevard, Long Island.  Among its products was the well known  Cosmophone HF-SSB transceiver.  The improvements made in the Cosmos PTO over the original Collins device were designed by Lewis Metzger and Harold Goodman, both instrumental in the development of the Cosmophone. They received US patent number 3,098,989 in July, 1963 for their idea and working model of a PTO that could be linearized with external adjustments, unlike the Collins PTO. Anyone who has ever worked on a R-390A PTO soon finds out that there are two important adjustments that can be made.  The first is that the PTO must be adjusted such that its output is exactly 1000 Kcs in ten turns. This is commonly referred to as the endpoint adjustment, and is as far as most users will go in PTO calibration.   Endpoint adjustment is important as it not only affects dial calibration, but can also
 compromise
>front end track tuning on the lower bands. 
>The second important adjustment is linearity.  If one were to graph the ideal PTO output frequency versus the shaft position in number of turns, the graph would make a straight line.  This was the goal of Collins Radio in all of its designs of the era, including the 75A, 32V and all of the later models.  Art Collins wanted a linear output, and that was that.  The Collins PTO was famous for being able to achieve this goal, as well it had to, because the inherently linear Veeder-Root counter of the R-390A exposed nonlinear PTOs with errors as low as 0.1 Kcs. 
>All PTO's employed in R-390A service have an endpoint adjustment screw. This screw is typically located underneath a sealed screw hole behind the PTO' s transformer can.  As a PTO ages, its output will typically decrease, i.e., it will have an output of less than 1000 Kcs in ten turns.  To remedy this, the endpoint adjustment screw must be turned (usually clockwise) until the PTO's output is increased to exactly 1000 Kcs in ten turns.  The turning of this screw moves a tuning core into a second coil in the PTO to make the adjustment.  PTO's of the Collins design will have two inductor coils, a main coil and an endpoint adjustment coil. 
>Adjustment of the PTO's linearity is another matter entirely.  In the Collins PTO, the company employed an internal corrector stack which accomplished this task.  This stack is constructed from a set of adjustable shims upon which a cam follower rides.  As the PTO shaft is turned, the main tuning slug is advanced axially on a lead screw, and the cam follower is simultaneously advanced on the corrector stack. Naturally, advancement of the main slug on the lead screw is exactly linear, but unfortunately the coil and powdered iron tuning slug do not usually cooperate in linear fashion.  The cam follower mechanism allows the tuning slug to either increase or decrease the overall rate of advancement on the lead screw by imposing a slight twist of the tuning slug.  This increase or decrease in tuning rate is determined by the shape of the corrector stack and is what accomplishes the linearity correction.  If a graph were made of the nonlinear response
 of the
>PTO, it might look very much like the profile of the Collins corrector stack when properly adjusted.  
>Linearity correction is necessary because it is virtually impossible to  wind the main tuning coil and match it with an iron slug for an overall linear output.  The only problem with the Collins design is that the corrector stack is inside the PTO cover.  Naturally, adjustments must be done with the cover off, but unfortunately, replacement of the cover influences the adjustment requiring multiple attempts and no shortage of frustration.   Metzger and Goodman solved this problem by improving the Collins PTO design.  The Cosmos endpoint adjustment is similar to the Collins adjustment, however the real innovation is found in their external adjustment for linearity.  This is the most misunderstood feature of the Cosmos PTO. 
>Upon careful examination of the Cosmos PTO, one will find the endpoint adjustment in its usual location as described above.  To the left of the endpoint screw hole will be found an additional screw cover.  Underneath this second screw cover will be found a series of very small screws.  During clockwise rotation of the PTO shaft, this series of screws passes underneath this window from left to right.  One screw will pass with every 90 degree turn (25 Kcs) of the PTO shaft.  It is this series of screws that are used to adjust linearity of the Cosmos PTO.  
>The Cosmos PTO has three inductor coils instead of  two employed by Collins.   There is the familiar main tuning coil, an endpoint adjustment coil, and a third coil that is part of the linearizing device.  Like the endpoint coil, the linearizing coil is in series with the others.  The inductance change of the corrector coil is controlled by a core which goes in or out as the PTO is rotated through its ten turns. 
>To help visualize the corrector mechanism, picture if you will a Teflon disk laid onto the front face of the PTO.  This disk acts as a cam.   A cam follower contacts the slug of the third coil.  The corrector disk rotates through a reduction drive as the PTO is advanced its ten turns. Under this disk is a series of screws which are laid out in a staggered fashion all along its circumference, forming a circle.  The flat heads of the screws, which are adjusted to varying heights in close proximity to their neighboring screw, make up a (typically uneven) surface upon which  the flexible Teflon disk rests.  As the screws are adjusted in or out, the Teflon disk (formerly and initially flat) is shaped into a wavy surface (a cam) upon which a cam follower rides.  You've seen the kiddy motorcycle rides at the carnival where the miniature vehicles go round and round, and also make a gentle up/down motion on the wavy surface of the merry-go-round floor. 
 It is
>this slight vertical displacement of the vehicles that is analogous to the in/out motion of the core in the compensating coil.  Note: One should refer to the R-390A service manual before attempting to service the radio.  
>The two things needed to calibrate the PTO at 25 Kcs points are 1) an accurate means of measuring frequency, and 2) an accurate means of turning the PTO shaft exactly 90 degrees at a time.  The author uses a frequency counter to measure the PTO output, and the Veeder-Root counter in the radio to measure the turns of the PTO.  When using the radio's counter to measure turns, take the time to tape exposed 120 VAC power and fuse connections to prevent electrical shock while performing the more than 40 PTO remove/install maneuvers.  The frequency counter is connected directly to the output cable of the PTO.  A handy crossover connector from the PTO fitting to BNC can be found on the back bulkhead of the radio.  Simply borrow the adapter that is present at the IF output jack.  There are many alternatives to using the radio frame to measure turns, including using a sacrificial R-392 frame (credit Wally Chambers, K5OP), and various calibrated knobs. 
>Setting the endpoint on a Cosmos PTO is similar to the Collins PTOs. Test the PTO to see how many turns it takes to achieve an output starting at 3455 Kcs and ending at 2455 Kcs.  Most aged PTO's that have not been recalibrated will typically require an additional 1-15 Kcs past the ten turn mark to achieve the proper range.   To bring it back to 1000 Kcs output in ten turns, adjust the endpoint screw (usually clockwise).  If the proper output cannot be achieved within the adjustment range of the endpoint screw, one must open the PTO and remove one turn from the endpoint adjusting coil.  This change in overall inductance of the coil will bring the endpoint back 7 Kcs or so.  Once this is accomplished, setting the endpoint can usually be achieved. While the PTO is open, the lead screw and bearing can be lubricated. The author usually takes this opportunity to bake an open PTO in an oven at 150F for several hours to dry it out and regenerate the
 desiccant. 
>Perform the calibration only after baking, and keep the PTO in a zip lock bag if you plan to leave the adjustment window screws open for an extended length of time. 
>When setting the 25 Kcs calibration points, it is extremely important that the screw being adjusted is directly lined up behind the adjustment window.  This may be accomplished by simply rotating the PTO shaft until the nearest screw is lined up exactly with the window.  The reason this is so important is that the cam follower which rides on the compensating cam is directly behind this window, and will thus be directly under this screw at this point.  If this caveat is not followed, adjustment can still be made, but there will be interaction between the point that one is attempting to adjust, and its neighboring points.  This will yield poor results and will be very frustrating.  Be advised that if the screw being adjusted is directly behind the window, the PTO position may not yield an exact 25 Kcs calibration point on the dial, e.g. 000, 025, 050, etc.  This is not a real problem, but it may be necessary to shift the starting point of the 25 Kcs
 check
>points by 5, 10, or 15 Kcs up or down.  Simply remember that the PTO has an output of 3455 Kcs at dial indication 000, and 2455 Kcs at dial indication +000.  Armed with this information, one can make a spreadsheet calibration chart for all 40 calibration points in
>25 Kcs increments.  An example chart might look like this: 
>FrequencyVeeder-Root Error  ( Kcs)Counter(Kcs)
>3465-010
>3440-015
>3415-040
>3390-065
>And so on . . .  for 40 to 42 calibration points 
>The above example is for a PTO that has the nearest linearity alignment screw falling at a frequency of 3465 Kcs on the frequency counter.  This corresponds to a starting point on the Veeder-Root counter of -010. Note that this starting point is not on an even 25 Kcs point, but that is OK as it is most important to start with the nearest screw directly lined up with the adjustment slug at the center of the window.  To complete the chart, subtract 25 Kcs from the frequency and add 25 units to the Veeder-Root counter for each point out to 2455 Kcs. 
>Each screw will have an adjustment range of about 5 Kcs +/-.  Clockwise rotation of the screws will reduce the PTO's frequency output at a given checkpoint.  One should avoid adjusting the screws to near their full clockwise position as this will cause excessive drag in the PTO. 
>Advanced Procedure for the Stout Hearted  
>
>To remedy problems with excessive drag due to adjustment screws reaching their clockwise limits, back all 40+ screws out to their counterclockwise stops, and then turn each screw one turn clockwise. This will give the cam an initially flat position.  Start PTO calibration at the point nearest 3455 Kcs moving up the dial toward 2455 Kcs (+000 on the Veeder-Root counter).  
>Remember to check each calibration point against the frequencies on the calibration check chart (it is not hard to get 5 Kcs off).  Some check points may require a touch-up calibration after completion of the first pass.  This will be especially true if the PTO calibration was re-started with the cam in the initial flat position as described above.  Extreme
>adjustment changes (5 Kcs) at any given point may cause output changes at neighboring points.  These can be worked out with multiple checks and adjustments at problem checkpoints. 
>With this linearization procedure, it is possible to obtain checkpoint accuracies of +/- .100 Kcs with an initial dial calibration at 000 using the crystal calibrator.  This greatly exceeds the specifications listed in MIL-R-13947B which require a +/- .300 Kcs with a dial calibration at the nearest 100 Kcs check point using the crystal calibrator.  
>In conclusion, the reader should be left with these points. 
>- the Cosmos PTO can be calibrated for both endpoint and linearity.
>- calibration can be done at 25 Kcs points, typically exceeding military specifications. 
>73 N5OFF
>
>References:
>
>Military Specification  MIL-R-13947B, "RADIO RECEIVER R-390 ( )/URR", 26
>October 1960
>
>TM 11-5820-358-35, "Field and Depot Maintenance Manual, Radio Receiver
>R-390A/URR", pp. 107-108.
>
>"Serially Connected Course and Fine Inductors with Continuous
>Adjustment", U.S. Patent number 3,098,989 awarded to Metzger and
>Goodman,  23 July 1963
>
>Also credit: Discussions with Wally Chambers, K5OP, Memphis, TN.
>
>
>--------------------------------------------
>On Sat, 6/14/14, Bill Hawkins <bill at iaxs.net> wrote:
>
>Subject: Re: [R-390] R-390A Cosmos PTO has erratic linearity
>To: r-390 at mailman.qth.net
>Date: Saturday, June 14, 2014, 10:06 PM
>
>Well, 200 parts in 2
>million is 0.01% accuracy. I'd say that was amazing
>for a mechanical, analog device.
>
>The answer to your question
>may be in the Collins Engineering Report, if
>it states the accuracy for the PTO.
>I seem to remember +/- 500 cycles, but you
>should find that in print.
>
>The age of digital calculators does make one
>think that all of the
>digits ought to be
>accurate, with no regard for percent error.
>
>Kind regards,
>Bill Hawkins
>
>-----Original Message-----
>From: R-390 [mailto:r-390-bounces at mailman.qth.net]
>On Behalf Of Larry H
>Sent: Saturday, June
>14, 2014 4:16 AM
>
>I'm
>restoring a Stewart Warner R390A with a cosmos pto type
>136-1.
>Although I've had to fix a number
>of problems on the R390A, it looks and
>operates well now except for 1 issue (or I
>think it's an issue). I've
>graphed
>the linearity of the pto numerous times every 25 KC, and
>it
>tracks mostly +/- 100 - 200 cycles (with
>a few exceptions up to 300
>cycles), but
>after adjusting the linearity bad spots numerous times,
>different ones keep popping up.  IE, the
>measurement repeatability is
>poor, but still
>mostly with in +/- 300 cycles.  This is after letting it
>warm up for 2 hours.  The frequency stability
>is very good.  Also, if I
>graph the pto
>linearity, use it for a few minutes, and graph the
>linearity again (without making any adjustments
>to the pto) its
>different than the previous
>graph.  Everything inside the pto seems
>tight and the gear train to the readout is
>tight (IE: very little
>slack). I have
>cleaned the pto shaft grounding contact and checked  the
>contact pressure.  The endpoint also varies
>about +/- 100 cycles. I have
>the prescribed
>spacing in the oldham coupler (its in very good
>condition) and it is ligned up very closely
>with the KC shaft and the
>spring is
>installed correctly. The gear train and the pto operate
>very
>smoothly.
>
>Should this be better or is this normal?  Any
>suggestions would be most
>welcome.
> Thanks.
>
>Regards, Larry
>
>______________________________________________________________
>R-390 mailing list
>Home: http://mailman.qth.net/mailman/listinfo/r-390
>Help: http://mailman.qth.net/mmfaq.htm
>Post: mailto:R-390 at mailman.qth.net
>
>This list hosted by: http://www.qsl.net
>Please help support this email list: http://www.qsl.net/donate.html
>
>______________________________________________________________
>R-390 mailing list
>Home: http://mailman.qth.net/mailman/listinfo/r-390
>Help: http://mailman.qth.net/mmfaq.htm
>Post: mailto:R-390 at mailman.qth.net
>
>This list hosted by: http://www.qsl.net
>Please help support this email list: http://www.qsl.net/donate.html
>
>