Our "Second Dose" of "Fine Tuning" - A TOTAL Dynaco FM-3 reconstruction!

0 Members and 1 Guest are viewing this topic. Read 42104 times.

Pathologymd

  • Jr. Member
  • Posts: 3
Hi Bill,

Absolutely phenomenal work!  Starting an FM-3 project with the "motherboard" soon.

You have probably stated this before, but I see that you are using solid wire.  What gauge do you recommend--20 or 22 for general purpose chassis wiring??

Looks like PVC insulation.

Thanks and keep up the good work!

William

Bill Thomas

  • Full Member
  • Posts: 424
Hi, William!

     You are correct.  I am using wire with PVC insulation.  I tend to prefer 20 ga. wire for several reasons.  The filament wires *do* carry enough current that it *might* make a difference.  But the *main* reason I like 20 ga. has to do with those pesky eyelets.  The larger wires "fill" the eyelet hole a little more.  There is a larger "contact area" due to the larger circumference of the wire, as well as less solder between the surface of the wire and the inside of the eyelet hole.

     In "plain-speak", it's just a little easier to work with the larger wire, but it's not really a "critical" concern.

     I'll be posting another "update" shortly.  This tuner looks like it's going to be purty swell!  (Considering what it has taken to get to this point, it had BETTER be "purty swell" or *better!*)  We'll find out together soon.

Sincerely,

Bill Thomas

Bill Thomas

  • Full Member
  • Posts: 424
September 2nd, 2009 - Update:  "At Long Last..."

     The "Construction Phase" of our tuner rebuild is now complete!  Let's take a look at *most* of the finished unit.

     Let's start off with a shot of the front panel on our finished tuner.

     

     Since we're using the same aluminum face plate we started with, this shot looks a LOT like the original "face plate shot" posted earlier.  The parts have all been cleaned thoroughly so my grimy fingerprints are a thing of the past.  I *did* wind up replacing the Volume Control knob to get rid of a "divot" in the original part.

     Here's a picture of the completed tuner "from the top":

             

     Looks a bit different than a "stock" FM-3, doesn't it?  That "cleaner look" is mainly because these new circuit boards don't have the silk-screened legends on top of the boards.  This is probably the only FM-3 with stock-style printed circuit boards that are now made from FR4 fiberglass.

     Here's how things look from underneath the chassis:

     

     Looks pretty much like a "standard" FM-3 tuner with the exception of the added terminal strip and the lack of a quad capacitor.

     Let's look at the underside of the PC-12 multiplex board:

     

     Other than the color of the printed circuit board substrate, this too looks quite "normal."  Here's a picture of the multiplex board "from the top":

                   

     You'll notice a "mix" of metal-film and carbon-composition resistors.  The metal-films are Dale/Vishay and Corning resistors.  The carbon-comps are Allen-Bradleys.  The metal films are used where carbon-comps might introduce a bit of added noise.  The carbon-comps are used in the audio signal path for their "sonic signatures."

     I would LOVE to tell you that things are all SWELL.  Unfortunately, there appears to be a problem in the IF board.  The tuner IS "making music", but attempts to properly align the IF strip have been less than satisfactory.  In fact, this "problem" is the main reason it has taken so long to post pictures of the completed tuner.

     In order to troubleshoot the IF board, I retrieved an FM alignment generator from storage.  It too is a "Vintage Piece."  Unfortunately, all these years of simply "sitting" in storage haven't been too kind to the generator either.  I basically had to rebuild IT as well before I could use it to find our "alignment problem."  Hoping to speed things up a bit, I "borrowed" a Vintage Heathkit FM generator from a local repair guy who was EXTREMELY kind in loaning it to me.  Unfortunately, it ALSO has some "age issues."  (SOME days, you just can't win!)  So where does that leave us?  Well, I am *almost* finished rebuilding *my* FM alignment generator and will be using it during the next week or so to FINALLY get to the bottom of the problem on the IF board.

     That brings us up to date on the project.  We have a completed tuner that plays music, but lacks the sensitivity a "normal" FM-3 should have.  That too will be resolved VERY soon and I'll report on the eventual cause of the "difficulties."  At least we know that the RF board and the Multiplex board are working properly.  As soon as we can certify the performance of the IF board, we'll be ready to call this one "A Wrap!"

     It's been a LONG time coming, but I feel quite certain this tuner will be "in the pink" within the next few days.  Wish me luck!

Sincerely,

Bill Thomas

Cure

  • Jr. Member
  • Posts: 6
So how is the IF board? Are we "in the pink" yet?

Bill Thomas

  • Full Member
  • Posts: 424
Friday, January 29th, 2010 - Update:  "It's Alive...  It's Alive... It's ALIVE!!!"

     There has been a LOT going on, personally speaking, so the final chapter on our "Second Dose of Fine Tuning" has been "on hold" - until NOW!

     The truly AMAZING thing about Dynaco's FM-3 tuner is that you can properly align it without any expensive test equipment (unless you "tweak" the Primary of the first IF transformer like *I* did while trying to "troubleshoot" the IF problem earlier.)  But if the Primary of the first IF transformer is set properly, everything else can be aligned using only the EMM801 tuning indicator.  And this isn't a "compromise" situation either.  You can get OUTSTANDING results - easily as good as using a whole cadre of test equipment.  This also means you can easily KEEP the tuner in optimum alignment later, in case you have to change tubes or other parts.  Pretty darned clever!

     When we last left our FM-3 with the all new Printed Circuit Boards, we were having a problem with alignment.  Between then and now, a bad capacitor was discovered *inside* one of the IF transformers.  This rendered the tuner "un-alignable" for some time.  The capacitor tested ok using my trusty, old Eico 950B R/C bridge UNTIL the Voltage on the part under test was raised above 200 Volts!  Then it showed leakage.  This one *teensy* little part had me "chasing my tail" for a LONG time, but the GOOD news is that the new IF circuit board was NOT the source of the problem!  (It also meant that my initial fears of installing one of the IF transformers backwards was NOT the culprit.)  But during the attempts at alignment, I managed to "tweak" the IF transformers WAY off the proper frequency of 10.7 MHz.  It took a crystal-controlled signal source to get us back "into the ballpark" and properly functional again.  Once the Primary of the first IF transformer was back "on-frequency", the rest of the IF alignment fell right into place.

     IF alignment isn't rocket science.  Simply follow the instructions in the FM-3 manual for aligning the IF section.  As with most analog gear, it pays to go through the alignment several times in order to INSURE that you have "tweaked" every last bit of possible signal.  You won't notice any LARGE changes the second or third time around, but you will probably find a *tiny* bit of improvement.  After the third pass, leave it alone.  Since these transformer adjustments are all "in-line," you don't really have a problem with adjustments interacting.  It's simply a matter of peaking each transformer for maximum signal as shown on the lower portion of the EMM801.  Remember to make these adjustments with the tuner set completely "between stations" so the only thing you hear is inter-station "hiss."

     Once the IF transformers are properly "peaked" for maximum signal, it's time to move on to the discriminator adjustment.  This is an absolutely CRITICAL adjustment and the tuner must be tuned "dead-center" into an FM station.  This setting is best accomplished using the EMM801 tuning indicator.  With a strong FM signal, the tuning indicator will close completely, but by "rocking" the tuning knob back and forth, you will be able to determine the EXACT center-point for the FM station.  make SURE you don't allow this setting to change while making the discriminator adjustments.  If it *does* move, you will have to start the discriminator alignment over.

     The FM-3 manual describes two methods of aligning the discriminator.  The "normal" method involves using the EMM801 as a "Voltmeter."  This requires un-soldering and re-soldering several connections.  The second method involves the use of a "VTVM" to set the discriminator alignment correctly.  Well, things have changed since the days when the FM-3 was originally produced.  Modern multimeters generally have a fairly high Input Impedance and will work just fine, but in order to keep the test equipment leads from de-tuning the discriminator due to lead capacitance, connect a 1 Meg resistor in series with the Positive lead of the multimeter.  Use the other end of the resistor to connect to the tuner.  Now you can back out the tuning slug on the top of the discriminator transformer and peak the bottom slug.  Don't be too concerned if you can't get around 8 Volts at the prescribed measuring point - just adjust the lower slug for maximum Voltage.  Then, adjust the top slug until the meter indicates as close to Zero as possible when the meter is connected to the other measuring point.  That's it!  Once this is done, don't forget to install the "gimmick capacitor" to eyelet #29 on the IF board.

     Once the discriminator is correctly aligned, you *should* be able to hear local FM stations, but they will *probably* be in Mono.  We'll get to the Stereo Multiplex board soon, but first, let's go "by the book" and align the RF board.  This will make SURE the FM stations show up on the dial where they are supposed to be.  It will also insure that the RF section of the tuner is operating at maximum sensitivity.

     Unlike the IF adjustments, the RF adjustments tend to interact - a LOT!  You will go back and forth MANY times while setting the oscillator adjustments for proper dial tracking.  You will also spend a good bit of time going back and forth between the Antenna Coil and the Mixer Coil.  The dial tracking adjustments of the oscillator are pretty sensitive.  Small adjustments can make BIG changes in the dial tracking, but the Antenna Coil and Mixer Coil adjustments tend to be VERY broad and are not as "touchy."

     Once the RF board has been aligned properly, there's only one section left:  The Stereo Multiplex Decoder.  Since we're measuring the signal strength of the 38 kHz sub-carrier, we need a Voltmeter capable of responding to MUCH higher frequencies than a typical multimeter can see.  Fortunately, our trusty EMM801 can EASILY react to the 38 kHz sub-carrier level.  Just follow the manual and make SURE the tuner is tuned EXACTLY to the center of a Stereo FM station's frequency.  Temporarily connect the 1.2 Meg resistor to each Test Point and re-connect the EMM801 to the proper test points.  When adjusting T71 and T72, you will find that the adjustments interact a bit so go back and forth several times until no further improvement can be made.  Next, move the EMM801 connection to the second test point and adjust T73.  These particular adjustments are absolutely CRITICAL for best Stereo separation and lowest distortion, so make them VERY carefully.

     Guess what?  You have just aligned a Dynaco FM-3 tuner!  Once you've done it a few times, you'll find it pretty easy to do a full alignment in less than 30 minutes!  Honest!

     And THAT brings us (finally) to the completion of our "Second Dose of Fine Tuning."  It's been *quite* a journey getting here.  But the final result is a tuner that is sensitive, stable and sounds REALLY nice!  Had I known that the original printed circuit boards were in such unreliable condition, I *might* have made some different decisions, but in a way, this has been a VERY positive outcome.  Now, we have the "master artwork" available to produce NEW printed circuit boards that appear to be MUCH more reliable than the original phenolic boards and since those original phenolic boards are now approaching their 50 year anniversary, this is a VERY good thing!

     The tuner is currently being "burned-in" for the next day or so to check for any instability or other "glitches," but it *appears* that everything is working just like a NEW tuner!

     My SINCERE thanks to this tuner's owner.  He has shown INCREDIBLE patience in waiting for his tuner to return from this VERY long odyssey!  It will be shipped back home on Monday!  I can hardly wait to hear what he thinks about his "new" tuner.

     Now that the tuner has been completed, we can head on to our NEXT project!  It will be a companion piece to one of our previous "projects" and it has been "waiting in the wings" for quite some time.  THIS one will be really SUPER!  I'll introduce you to it with a brand NEW thread, coming soon!  Until then, keep those old Dynaco units working.  If I can help answer your questions, just let me know.  I don't have ALL the answers, but with each "project" our "collective knowledge base" increases markedly.  Only GOOD can come from THAT!

Sincerely,

Bill Thomas

Bill Thomas

  • Full Member
  • Posts: 424
Monday, February 1st, 2010 - Update:  "One Last Thing..."

     I thought it would be good to close our "Second Dose of Fine Tuning" report with the results of the "burn-in" process.

     The completed and aligned tuner was "benched" and tested for the past few days.  It proved to be totally stable within about two minutes of turn-on.  (Dynaco suggests a 20 minute period to achieve full stability.)  After numerous "heat-up/cool-down cycles," it appears to be an EXTREMELY stable unit.  There is no sign of drift whatsoever.  Both sections of the EMM801 tuning and stereo indicator close fully.  This is not *always* the case with an FM-3 tuner, but THIS one appears to be operating at the "top of its class."

     So how does it sound?  Revealing!  For those LOUSY FM stations that are compressed so badly they average over 90% modulation at all times, this tuner reveals just how DREADFUL this practice can be!  But there is a local station that uses VERY little compression and in THIS case, the sound quality is simply magnificent!  My brother decided to take a listen on headphones.  He noticed a bit of a "rumbling" sound.  I took a listen and realized it was rumble from the station's turntable!  Another source eliminated that problem with the next selection.  It was a well-recorded choir that was *probably* recorded using an M-S recording technique because every voice was positioned in a unique location in the Stereo sound field!

     And WHAT a sound field!  Separation is OUTSTANDING.  Frequency response seems extended too; serving up the BEST that FM broadcasting has to offer.  One of the GREATEST assets of the FM-3 tuner has to do with its ability to achieve full quieting with a poorer signal than other tuners.  But THIS unit seems even better than most other FM-3's I have dealt with.  Also, there is absolutely NO tendency for this tuner to be microphonic.  Even relatively hard "taps" to every area of the chassis seem to be ignored.

     So, we'll call this one DONE!  The tuner is now on its way back to the original owner.  I will relay any further information I receive, when it becomes available.

Sincerely,

Bill Thomas

baldrick

Great work, Bill.  Nice that you have re-created the PCBs as well.  This is something that I have been working on with one of my own FM-3s.  The IF strip is a bit more challenging to recreate because if it's length, but it should be done shortly.

I have a question about the gimmick "capacitor". 

The alignment instructions tell us to install this gimmick partway through the alignment process but doesn't explain why and doesn't tell us to remove it.
I've heard various responses on whether it is supposed to stay in or be removed after alignment.

What's you take on the situation?

Bill Thomas

  • Full Member
  • Posts: 424
     Actually, Dynaco *does* explain the reason for the "gimmick" capacitor. (sort of).  According to Dynaco, "It corrects for the effects of interaction between the adjustments of the two slugs in the discriminator transformer."   It is added after the discriminator transformer alignment has been completed.  When re-aligning the discriminator transformer, the "gimmick" should be removed.  After alignment, it should be replaced.  It has no effect on the IF alignment or the RF alignment, but it really MUST be added after properly aligning the discriminator transformer in order to achieve the lowest distortion.

I hope that clears things up.

Sincerely,

Bill

rlee8394

Bill,

The "gimmick" capacitor is a piece of wire extending 11/16" above PC-8. Is the value of this gimmick cap determined by both the length of the wire as well as the insulation? If so, would a gimmick based on a wire with Teflon insulation be any different than one with PVC insulation as when the tuner was designed? Basically, does the type of insulation matter?

Ron

baldrick

I would think technically yes.  Most won't dispute that mylar is different from polypropylene which is different from paper/oil and different from polystyrene.
But the value of its capacitance must be infinitesimally small since it's basically sticking out in space and whatever capacitance it does generate would be to ground.  A few femtofarads?



rlee8394

I think that most of the confusion is based on Dynaco's own statement for the use of the gimmick cap. Obviously, the adjustment of the discriminator transformer occurs first. After that process is complete, you install the gimmick cap. Yet, Dynaco states the rationale for the gimmick as,

 "... a small value capacitor which corrects for the effects of interaction between the adjustments of the two slugs in the discriminator transformer. At such times as realignment is performed, this "gimmick' should be removed before aligning the discriminator."

These two sentences seem to contradict each other as it would seem to imply that the "gimmick" needs to be present during the adjustment of T5. This of course is not the case, as the "gimmick" is installed after the adjustment process. Maybe Dynaco should have stated that it corrects for interaction of the slugs after adjustment is complete. That seems more accurate.

Ron

gmcjetpilot

  • Newbie
  • Posts: 1
I am thinking of buying a FM-3.
What should I look for. I assume the tuning eye is critical.
These were Kits so I assume quality varies.

Also is there a list of mods with instructions.
I guess the two power resistors and the "Berning Mod" PEC module. Is there a schematic for those modules.

Thanks George

baldrick

The PEC module mod is here as well as the power supply capacitor mod:http://quadesl.com/refurb/refurb_dynaFM3.html

The other changes are summed up here including the power resistor mod:http://www.avahifi.com/root/audio_basics/1990-11_fm3_keep_working.htm

If you are more concerned with functionality rather than aesthetics then you can use a single element EM84 tube instead of the dual element EMM801 but you would lose the stereocator indication.  The stereo multiplexer would still work fine, though.

Instructions for that surgery are here:http://www.the-planet.org/dynaco/Tuner/Replacing%20EMM801%20with%20EM84.pdf
PM me for other pitfalls.

handsome

excellent article. one question: the PEC modules have to be adjusted to suit country's where the deemphasis constant is 50uS. doe sthsi also apply to the PEC module on the IF strip or just the PECs on the multiplexer board?

Bill Thomas

  • Full Member
  • Posts: 424
Just the two PEC modules on the Multiplex board.

Bill

handsome

thank you!

Bill Thomas

  • Full Member
  • Posts: 424
     If you do the "Berning Mod" to replace the original PEC modules (Which can be seen at Sheldon Stokes' website:  http://www.quadesl.com/pdf/fm3_pec.pdf), Change the value of the 510 pF capacitors to 340 pF.  Since this is a non-standard value, you can either use 330 pF, or a parallel combination that equals 340 pF.  Either method will be fine (You'll *probably* never hear the difference between a 340 and a 330 pF capacitor in this application.)  This will be the correct value for the 50 microsecond de-emphasis necessary in Europe.  Use either mica or polystyrene capacitors here.  Works GREAT!

     Hope this helps.

Bill

avahifi

  • Industry Contributor
  • Posts: 4682
    • http://www.avahifi.com
Hi Bill,

Its great to see you back at AC again.  Hope you are feeling much better these days.

For your information, we recently posted a small upgrade to our Ultimate 70, Ultravalve, and even stock Dyna St-70 amplifiers.  All that is necessary is to install two 1N4007 diodes, one in series with each high voltage power transformer lead to the 5AR4 rectifier tube.  The diodes will mount on unused 5AR4 socket pins. The purpose is to prolong the life of the rectifier by reducing the impact of cold hard turn-on AC voltage to the tube.

We recommend this for any vacuum tube amp using a 5AR4 rectifier tube.

Regards,

Frank Van Alstine

Bill Thomas

  • Full Member
  • Posts: 424
Thanks, Frank!

     I feel like such a stranger nowadays.  I've been "tuned in" on the forum, but haven't felt much like any real meaningful *work* until this week.

     Needless to say, I'm still a bit tired, but I *am* back and will be finishing up my "lingering projects" very shortly.

     At least my "Life Force Meter" is getting LONGER these days.  For awhile, I was a bit concerned that it was running out.

     Getting old stinks, but there's only one thing worse - NOT getting old.

"Old" Bill

P.S. Thanks for the "mod" details.  Makes GOOD sense to me! Especially with the quality of current production 5AR4/GZ34's.

Bill

handsome

Hiya Bill,
thanks for the replies. I have also replaced the stereo PECs with the Berning mods adjusted to 50uS (340pF caps). I now find that there is too much treble (after two days of listening) and a fair amount of hiss. I also find i am getting very little variation when using the mono knob - that is very little stereo to mono change. I am using a strip antenna (not so great) so i wonder if a better antenna would help? Or do you think i should try align the Multiplexer board? I replaced the 47k resistors on the MPX board with 0.1% resistors but have had no change to stereo performance.

I was puzzling over the PEC boards and tried to simulate them in a SPICE program, unfortunately i really know nothing about RF or even complex electronics so could not simulate the input and output impedance's. nonetheless the graphs were quite interesting. i have attached a copy here. the Berning values give a dip at exactly 38kHz, our 50uS values do not dip at 38kHz (more 56kHz) and the dip is much shallower. I found that by adjusting the 100pF cap to 68pF the shape was corrected but i still suffer 8dBs less rejection of 38kHz.

Do you think this could be causing any of the problems i have outlined? Surely any extra bandwidth above 20kHz would not be audible? I would appreciate any input you can offer and any thoughts I am finding this a fascinating learning curve!