Home  Resources  References  Tubes  Forums  Links  Support 
Delmonico T-11 FM Problem
11/22/2005 2:44:31 PMRick Skowronek
Am repairing a '50s Pseudo Ac/DC 3 band (AM/SW/FM) radio. Have all necessary test equipment including FM gen., Tek scope and Digital Multimeter. Have completely recapped and checked voltages and resistors. Have Sam's for the unit. FM is distorted. Uses a Ratio Detector and joint FM/AM IF transformers.
It was obvious someone previously had attempted to eliminate the distortion by tweaking all FM IFs, Det. transformer, etc. Have gone through a full alignment per Sam's instructions but it is erratic on how it responds to tweaking. Adjusting for peak IF voltage at the ratio detector just increases distortion. Sounds like overload. Limiter seems to be at correct voltages. Balance on ratio detector is right on at nearly 0 volts. Tube complement is 17EW8-FM RF and LO/Mixer, 12BA6 FM/AM IF Amp., 12BA^ FM Limiter, 12AL5 - Ratio Detector, 5 other tubes for rectifier, audio pre and final and AM/SW RF Amp/Mixer and Tuning Eye tube.

This circuit has no AFT. I have been through this thing for MANY hours and am not much further than when I started in cleaning up the sound. AM & SW are perfect soundwise. I have over 40 years in electronics both in design and troubleshooting. Any guidance would be appreciated.

11/22/2005 5:24:59 PMRadiodoc
Rick,

I don't have a schematic for the Delmonico, but have you checked the small electrolytic cap that should be connected to one of the 12AL5 plates to ground, could be pin 2 or 7. If your radio has FM agc from the ratio detector to the frontend or IFs, could be a defective component in that circuit that could cause overload.

Radiodoc

:Am repairing a '50s Pseudo Ac/DC 3 band (AM/SW/FM) radio. Have all necessary test equipment including FM gen., Tek scope and Digital Multimeter. Have completely recapped and checked voltages and resistors. Have Sam's for the unit. FM is distorted. Uses a Ratio Detector and joint FM/AM IF transformers.
:It was obvious someone previously had attempted to eliminate the distortion by tweaking all FM IFs, Det. transformer, etc. Have gone through a full alignment per Sam's instructions but it is erratic on how it responds to tweaking. Adjusting for peak IF voltage at the ratio detector just increases distortion. Sounds like overload. Limiter seems to be at correct voltages. Balance on ratio detector is right on at nearly 0 volts. Tube complement is 17EW8-FM RF and LO/Mixer, 12BA6 FM/AM IF Amp., 12BA^ FM Limiter, 12AL5 - Ratio Detector, 5 other tubes for rectifier, audio pre and final and AM/SW RF Amp/Mixer and Tuning Eye tube.
:
:This circuit has no AFT. I have been through this thing for MANY hours and am not much further than when I started in cleaning up the sound. AM & SW are perfect soundwise. I have over 40 years in electronics both in design and troubleshooting. Any guidance would be appreciated.

11/22/2005 6:53:17 PMRick Skowronek
Thanks guy for your kind reply. Yes, that was the first suspect. I went through the schematic looking for the usual suspects. That was sure one of them as were all the electrolytics and paper caps. Went even so far as to replace the pico caps associated with the circuits. Amazingly made absolutely no difference in the performance. The only real suspects left are the IF transformers or something really weird in the RF side of the fence which no replacements occured. Been feeding an absolutely dead on 10.7 MHz pure RF through the RF converter (which is a combo Local Oscillator and Mixer) and the system goes through the process while tweaking the IFs and ultimately the Ratio Detector. Sadly, it doesn't help much. Low level FM radio station signals come in fairly clean. Strong signals go into severe distortion and the only way to minimize the distortion is to off tune. I know this is one of the sadest examples of FM circuitry but it must have worked fairly well sometime. Tubes are good (and replaced). The only thing basically left are the IFs. If that's the case, time for a deep six.

Just looking for anyone that may have seen this before and may have a slightly different tuning procedure or checkpoints.

Thanks again.

:Rick,
:
:I don't have a schematic for the Delmonico, but have you checked the small electrolytic cap that should be connected to one of the 12AL5 plates to ground, could be pin 2 or 7. If your radio has FM agc from the ratio detector to the frontend or IFs, could be a defective component in that circuit that could cause overload.
:
:Radiodoc
:
::Am repairing a '50s Pseudo Ac/DC 3 band (AM/SW/FM) radio. Have all necessary test equipment including FM gen., Tek scope and Digital Multimeter. Have completely recapped and checked voltages and resistors. Have Sam's for the unit. FM is distorted. Uses a Ratio Detector and joint FM/AM IF transformers.
::It was obvious someone previously had attempted to eliminate the distortion by tweaking all FM IFs, Det. transformer, etc. Have gone through a full alignment per Sam's instructions but it is erratic on how it responds to tweaking. Adjusting for peak IF voltage at the ratio detector just increases distortion. Sounds like overload. Limiter seems to be at correct voltages. Balance on ratio detector is right on at nearly 0 volts. Tube complement is 17EW8-FM RF and LO/Mixer, 12BA6 FM/AM IF Amp., 12BA^ FM Limiter, 12AL5 - Ratio Detector, 5 other tubes for rectifier, audio pre and final and AM/SW RF Amp/Mixer and Tuning Eye tube.
::
::This circuit has no AFT. I have been through this thing for MANY hours and am not much further than when I started in cleaning up the sound. AM & SW are perfect soundwise. I have over 40 years in electronics both in design and troubleshooting. Any guidance would be appreciated.

11/22/2005 8:19:53 PMThomas Dermody
Aren't FM IF transformers supposed to be staggered somewhat? I know that television IFs are supposed to be staggered, but not sure about radio FM. I adjust mine right on with fair results, though I've heard of people using sweep generators so that the IF transformers pass a wider bandwidth. Here may be your trouble: strong stations distort more than weak ones. Well, the stronger the sound, >>>>>the farther the frequency shift<<<<<. If your IF transformers are not tuned broadly enough, then you won't be able to pass these wide frequency shifts. Distortion will result. I can stand by this sort of because my Crosley 1117 picks up random FM stations on its short wave band due to some harmonic problem within the receiver. They are distorted, obviously, though, because when a station is tuned in, only its center is picked out by the tuner (or edges if it is off-tuned). My AM RF section isn't designed to pass wide band FM stations. Try stagger tuning your IF transformers manually if you do not own a sweep generator. I forgot what the exact bandwidth of FM signals is, but it's fairly wide. Tune one section of each IF transformer so that it'll pass a signal well at a frequency slightly below 10.7 MC. Then tune the other section so that it passes fairly well with a signal slightly above 10.7. Go for say 10.6 and 10.8. Continue with this practice with the other IF transformers. If this improves louder stations, but still doesn't improve the loudest stations, then go for an even broader spread. Ideally, when looking at the signal passage on an oscilloscope, there's a certain frequency coverage that you're looking for. With the two halves of each IF transformer peaked just above and below 10.7 MC, you'll see the two peaks almost combine so that it becomes almost flat with a slight dip in the middle. The width of the flat will be that of the two peaks plus the slight dipping in the middle. I've never done this, but I've seen articles on this.

Thomas

11/22/2005 8:20:40 PMThomas Dermody
Note: you don't want the spread to be too wide, but fidelity and clarity depend upon this spread.

Thomas

11/23/2005 1:37:53 PMRick Skowronek
Thomas,

I believe you're right on on the bandwidth of the IFs. The alignment instructions for the unit show two ways to align it. One with a sweep generator and scope and the other with a sig geneneraroe set at 10.7 mHz and a VTVM on the out put of one leg of the ratio detector. With that you align for max output. With the sweep/scope combo you adjust for the mesa shaped waveform you talked about. The sweep frequency they suggest using (and thus the bandwidth) is 450 KHz which would be roughly 225 KHz on either side of the 10.7 MHz IF center.

The generator I have is the big HP 8640B which is a beautiful highly accurate generator capable of both AM and FM output. Instead of a sweep though it outputs a user selectable deviation of so many hertz. Not really great for setting up wide bandwidth IFs.

Will try your stagger tuning idea. Great thought. As I mentioned I can set this generator accurate to 1 Hz at 10.7 Mhz so stagger at 10.475 on the bottom coil and 10.925 on the top of each IF can. I'm sure, to add to my misery, these are dual IF cans with both the 455 KHz AM IF and 10.7 hooked in series with each other. Through the Band Select, various components and circuits are activated and deactivated. On a positive note, no one has adjusted the 455s.

I guess if this doesn't work correctly I probably have a bad 10.7 IF transformer section which is screwing up the overall bandwidth of the FM portion. Now I wish I had a full-fledged FM signal/sweep generator.

I would appreciate any other input from you or other forum mambers as I continue to tackle this monstrosity.

Rick

11/23/2005 6:13:15 PMThomas Dermody
Well, FM alignment is quite possible without a signal generator. Results may not be as faithful, and you may do a little more "hit-and-miss."

Thomas

11/23/2005 6:25:10 PMRick Skowronek
Thanks Thomas,

Do have an excellent generator just not the right exact generator for this situation. Something to be said for those older FM sweep gens. Newer FM equipment uses PLLs and synthesized oscillators,Varactors and crystal IF filters. Sure a big difference from our day, huh?

Just hoping some of the other forum gurus have also tackled this situation with some of the same limited equipment or situations. The more input the better. Whatever the outcome, I really appreciate your and radiodoc's help.

Rick

:Well, FM alignment is quite possible without a signal generator. Results may not be as faithful, and you may do a little more "hit-and-miss."
:
:Thomas

11/24/2005 2:23:33 AMThomas Dermody
Well, in my "day," which is actually not my day, because I was just born ('79), bell bottoms were very in.

Still, just about everything my parents owned back then was tube type because it was hand-me-downs....so I guess in my day things were like that.

Thomas

11/24/2005 1:16:45 PMRick skowronek
Sorry for aging you guy. Usually see tubeniks more my age. I used to actually build and design with them long before the advent of transistors. But have also stayed up through full blown digital logic and micros. Fun watching it grow though miss the ability to work on stuff anymore. Most are literally impossible to repair down to the component level.

Oh well.

Thanks again for the feedback.

11/24/2005 2:37:56 PMThomas Dermody
Yeah, the new stuff is hard to work on, which is why I don't usually work on it. It isn't beautiful to look at, either. It is very functional. About the only stylized logo you'll find on anything anymore is the cool "N" for National Semiconductor components. I love all of the old components with their awesome logos, which is why I balk and and freak out at people who gut the insides of sets and replace all old components with new ones (gutting the insides of the components and replacing them with new is of course quite acceptable as long as the shell and logo of the original component are preserved).

Still, the new stuff is amazing, and it functions quite well. I like how new televisions never need adjustment. They always hold their picture, and color is fairly stable. I never really cared for readjusting the picture constantly, but I live with it and enjoy it when working with a cool old television. I guess it's kind of part of the experience. If it hadn't been for progress in technology innovations, though, we wouldn't all be chatting here on a web site. Perhaps we'd never meet, and so many more people who wanted to restore old radios would not have the resources or information they need, and they'd probably feel that they're the only ones on the planet who enjoy restoring radios. That's how I felt years ago, though there were people much older than me who lived on my block who used to be radio repairmen. ...Really cool people. Also, who ever thought that you could purchase an antique waffle iron of your choice just about any day of the week? eBay makes things so much more available. One thing that'll be kind of weird, though, is in the future things won't be located in their "proper" area anymore. People in California will be playing records that were originally bought in New Jersey. ...But then, that's kind of the way it always has been. You see things on Antiques Road Show all the time about a piece of furniture that has moved all over the place, and not necessarily because of eBay.

Thomas

11/24/2005 2:28:32 PMMarv Nuce
Rick,
My many years of electronics are laced with FM horror stories like yours, but don't recall ever seeing a properly tuned FM if system having too narrow bandwidth. Properly tuned means the steps you have already pursued. As a matter of fact, the industry struggled for years to get steepest skirts and the narrowest usable bandwidth for FM broadcast, without real success. As I recall, it was and still is +/- 75kc max deviation, for a total of 150kc per channel, and each channel has a 25kc guard band either side to adjacent channels. So a nice shaped passband of 150 kc @ the -3db point is adequate. Few stations ever exceed these limits, and only rarely come close. I suggest you take a closer look at the ratio detector xformer, and its associated "S" curve. Actually it only slightly resembles an "S" leaning over severely. If the straight portion of this curve is too narrow or not straight at all from misalignment, severe distortion will result. The center of the straight portion should sit at "O" volts DC, with equal lengths both above and below this crossover. If your sweep gen has correct markers, you can set the straight portion at its required (+/-75kc) length and linearity by tuning the detector xformer. A simple resistive pad can be used to add a second RF gen in the signal stream, if the appropriate markers are not installed, or not adjustable on the sweep gen.
marv

:Am repairing a '50s Pseudo Ac/DC 3 band (AM/SW/FM) radio. Have all necessary test equipment including FM gen., Tek scope and Digital Multimeter. Have completely recapped and checked voltages and resistors. Have Sam's for the unit. FM is distorted. Uses a Ratio Detector and joint FM/AM IF transformers.
:It was obvious someone previously had attempted to eliminate the distortion by tweaking all FM IFs, Det. transformer, etc. Have gone through a full alignment per Sam's instructions but it is erratic on how it responds to tweaking. Adjusting for peak IF voltage at the ratio detector just increases distortion. Sounds like overload. Limiter seems to be at correct voltages. Balance on ratio detector is right on at nearly 0 volts. Tube complement is 17EW8-FM RF and LO/Mixer, 12BA6 FM/AM IF Amp., 12BA^ FM Limiter, 12AL5 - Ratio Detector, 5 other tubes for rectifier, audio pre and final and AM/SW RF Amp/Mixer and Tuning Eye tube.
:
:This circuit has no AFT. I have been through this thing for MANY hours and am not much further than when I started in cleaning up the sound. AM & SW are perfect soundwise. I have over 40 years in electronics both in design and troubleshooting. Any guidance would be appreciated.

11/28/2005 1:38:16 PMRick Skowronek
Hi Marv,

Thanks for your kind reply. My problem is that I do not have a proper FM sweep generator, one especially with marker capabilities. I do have an HP 8640B which is a beautiful gen with full FM and AM capabilities with a settable deviation frequency. Perhaps if I was a bit more familiar with FM alignment, it would be ideal. I also have a TEK 100 Mhz scope.

I'll sure be looking at it when I get it close per Edd's procedure below.

Thanks,

Rick

:My many years of electronics are laced with FM horror stories like yours, but don't recall ever seeing a properly tuned FM if system having too narrow bandwidth. Properly tuned means the steps you have already pursued. As a matter of fact, the industry struggled for years to get steepest skirts and the narrowest usable bandwidth for FM broadcast, without real success. As I recall, it was and still is +/- 75kc max deviation, for a total of 150kc per channel, and each channel has a 25kc guard band either side to adjacent channels. So a nice shaped passband of 150 kc @ the -3db point is adequate. Few stations ever exceed these limits, and only rarely come close. I suggest you take a closer look at the ratio detector xformer, and its associated "S" curve. Actually it only slightly resembles an "S" leaning over severely. If the straight portion of this curve is too narrow or not straight at all from misalignment, severe distortion will result. The center of the straight portion should sit at "O" volts DC, with equal lengths both above and below this crossover. If your sweep gen has correct markers, you can set the straight portion at its required (+/-75kc) length and linearity by tuning the detector xformer. A simple resistive pad can be used to add a second RF gen in the signal stream, if the appropriate markers are not installed, or not adjustable on the sweep gen.
:marv

11/27/2005 6:19:28 PMEdd
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rick:
You can tune a tuna if you can tune a Delmonico.
I have no reference to a T-11 model of Delmonico but referring to e-bay there is a Delmonico that looks like a larger Grundig-Telefunken knockoff type of cabinet with the "glass" type of wood finish. And then, there is a common plastic cased cabinet, which seems like I might remember seeing long ago as a Japoni imported Delmonico.I would guess the latter to be your type of cabinetry.
Initially, I also am in question of the Ratio-det circuitry as to the balance of the like resistive elements on its secondary and if the low value secondary electrolytic filter is a new one. Old original units were required to DC filter as well as present low ESR to the RF present, with time their low ESR capabilities lacked fiercely ( unless it was designed incorporating an additional .1 ufd disc ceramic shunting it). Also, an AC coupled- scoping of the B+ side connection of the Ratio-det transformer to see if there is only pure Dc there with no unbypassed RF there due to decreased power supply filtering capability over that long time span of its age.

As to the electronics aspect on the IF's I was also a little fuzzy as to their construction. As per the AM and FM portions possibly being in a unitized / shared housing, or were they using the more common utilization of a series mechanically separate pair with the 10.7 being at the tube end of the series connection and the 455 being a the B+ end with there being such a mammoth frequency / Z differential such that at FM frequency use that the 455 unit is merely an insignificant little bit of transparent series inductance at the B+ end . Equally, in the 455 mode of operation, the top series FM IF winding is of no consequence at the much lower 455 kc frequency scheme of its operation..
It would be nice to know, informationally, the manner of construction of the 10.7 transformers. Specifically, the Ratio det xformer. I am thinking of possibly either a "K-tran" design of transformer, with a top coil on a form and then another coil spaced away from it at the bottom, with their respective companion encircling top and bottom powdered iron tuning cups.
The second design scheme in mind would be spaced out top and bottom coils on a common coil form with merely respective top and bottom powdered iron cores with their imbedded threaded brass screws that are used to run the cores up and down within the respective coil forms.
And lastly, possibly, ones that use the same coil / form configuration, but with internal coarse threaded powdered iron cores that have an internal hex /or/ square hole that accepts a hex / sq alignment tool…..but being Japoni, they were much more into using a recessed slot into the end of a slug, for acceptance of a flat screwdriver type of insulated tool. What I am leading into, is if you used the outermost resonance point of the slugs positioning in the coils in setting your tuning of the ratio det coil and even the three other coils in the FM coil lineup.
Further clarifying that:
Visualize the two separated coils on the common coil form with one coil being the common single winding coil of the final limiters plate circuitry. The other coil, being the special secondary winding with its adjunct top counterwound phasing winding. Its the one that you phase null to get the zero voltage output from its secondary detected / developed voltage.
Lets go back to the limiter plate coil of the Ratio-det coil and visualize that if that the slug was far out of the coil, that its L/C resonant freq might be 12 Mhz. then you start running the slug into the coil and maybe ¼ of the way into the coil, you find that 10.7 resonance is acquired, but you keep entering the slug until it centers within the coil for max obtainable L value with possibly, say, a 9.6 Mhz resonance point. However, you then continue turning, and then its decreasing L value results in the running up the freq spectrum and hitting 10.7 resonance point again, but with the slug deeper into the coil. BUT this time the positioning being within the intercoupling path between the adjunct other coil, which has somewhat of a coincidental effect on upsetting the other coils tuning as well as the degree of inter coil coupling.
The identical situation holds true for the transformers other coil in its slugs positioning. There are three probabilities of misalignment out of the four combinational possibilities in the scenario. Sooooo you want coil 10.7 resonance acquired on both coils with their respective slugs at their outermost mechanical positioning within the coils at their 10.7 resonance tuning..

The second aspect to touch upon is the fact that at least you have a good stable Hoolett Plastered sig gen for stable freq referencing and possibly a good Tek scope for observation. I don't have the Sams to reference their alignment procedure for your particular Harmonica , however, I would suspicion that back in those days the FM alignment technique might have been nothing spectacular, just a stable 10.7 CW signal, being unswept. Marv was attesting to a sweep alignment, wherein I don't think that you would require such refinement. But, if using that technique, you could then observe the detected output for the overall peaking, bandwidth and both side skirt responses of the IF 10.7 aligned signal response. As well as the clipping on of a scope probe and then viewing the more important detector sections developed "s" curve to confirm its all important symmetry. The latter, I do believe, is potentially wherein your problem lies, specifically, the tandem core alignments are at such error that you have been left with an imbalanced output of the + and - nodes of the detectors signal.
Initial data:
The first relevant data that I might ask for in analysis might be for you to fire up the set and warm up, and then select your most problematic station in reception and use it to evaluate with. Hook in the DC metering to the detector stage in the manner prescribed in Sam's . Next you tune slightly off the station and watch either a + or - polarity of voltage start to increase, continue slowly tuning until that voltage peaks and then starts diminishing. Log in that peak DC voltage level. Next, tune back onto the station and then past it in the opposite direction. Expect an increasing voltage of the opposite polarity as well as its peaking and then diminishment. Log in that peak voltage value and compare the two for a sameness of values, if there is an imbalance you need to proceed as below:
How's about attempting a realignment in the manner in which I would perform it. If I remember correctly, your H/P, has a very good output attenuator...much in the order attainable with the old '603s and '608s.
How about initially putting a 10.7 CW signal into the unit in the Sams's suggested manner of injection. Since this units design has 2 stages leading up to and including limiting, it should quite easily accomplish hard limiting. However, this could be detrimental to accomplishing a waveform viewing in the making of a very precise 10.7 peaking of its transformers coils. Initially, see if you can scope probe into the plate coil circuit of the last limiter(Ratio-det primary) , using scopes AC-HF mode triggering on the incoming IF signal and portray ~ 10 wavelets across the scope face. The observation is to be able to then see if the attenuation capability is adequate to bring those wavelets down from hard flat topping/limiting action, down into discernable unclipped, clean sinewaves and then further attenuate the signal until the resultant display amplitude drops down to the "grass" level. If that analysis proves that to be attainable, with the combinational sig gen and scope capabilities, then bring up the sig input level back up just enough to discern a workable viewing amplitude of sines (well before any limiting action is onsetting). Then you start peaking the four 10.7 resonant coils ( Ratio-det coil secondary excluded naturally, since its on past the monitored circuitry, and remember, use outer positioning of the slugs potential two 10.7 resonance possibilities). While watching the displayed amplitude displayed on the scope, attenuate the RF input progressively to keep the scope display at the approximate same amplitude when peaking the coils. Thus, in no way letting them encroach into the threshold of a clipping / compression amplitude level. Run back through a second peaking of those coils to confirm no counter inter-reaction in earlier adjustments were experienced. Finally, then all coils will be peaked at optimum / peak gain. The final step of the series of adjustments will be the unclipping of the scope probe from the limiters plate circuitry and moving it to so as to being merely a clipping onto the plastic wiring of the plate lead going to the Ratio-det transformer. (Should it remotely be, that distance is so close, that a bare buss wire mikght have been used, instead, slip on a longitudinally slit piece of plastic insulation to let the scope probe clip around.) The idea here, being to compensate for any prior high Z capacitive loading imparted on the plate circuit during the previous tuning run. This time will require using a 1X probe with the vert gain set to its ~10mv decade and then observe the scope display and bring up the 10.7 RF CW input for a viewable/workable amplitude level and then tune and re-confirm that the prior peaking of the primary of the Ratio-det is still at max.
We are to the point where I certainly hope that you are not saddled with only having your mentioned DVM, for here is where a VTVM and its zero center offsetting display capabilities, along with quick reading under a changing voltage polarity can really facilitate final tuning operations. I have found it to be quite disadvantageous in using a DVM with its inherent slow response time and bobbling digits (even those supplemented with linear LCD hash markings).If you don't have one of those VTVM's-TVM's-FETVM's with center dial scale zero center set up, a technique that I have done when not having one as a visitor to a bench was to take a higher Z (20K+Z) , like the Triplett '630 or Simpsons series, and set one lead in series with a "phantom" variable voltage source. E.G. a 9 V xstr type battery with a 1 K pots end terminals across it and taking off from that center tap as referenced to an adjunct lead off one end battery terminal in order to get a 0-9 VDC supply. (Or 0-18 VDC available, if two series batteries were utilized).That way you short the leads and adjust the pots injected voltage level to get center scale referencing on one of the top linear plotted voltage scales. Then, when voltage is tested it will be either additive or diminutive to that initial voltage and, there you have a 0 center metering with bipolar compatible displaying. That lower 20K Z does load down the circuitry just a bit, as compared to the prior mentioned almost infinite loading as per VTVM's or even DVM's.
We are now up to the precise zeroing of the Ratio-dets secondary tuning.
Hook op your DC metering to the Sam's prescribed monitoring test point(s) ,then bring up the 10.7 CW level and see which polarity of detected DC voltage starts developing. Irregardless, keep bringing the RF up slowly and do a slow continuous correction on the Ratio-det's secondary coil adjustment ( 0-ing) to make that voltage reading be corrected to 0 VDC. Once the RF is maxed there should be minimal variance to that O voltage reading. (Note that up to this point, we have never touched the frequency after its initial setting to 10.7) (Therefore, in the next step as we are slowly starting a small tuning above and below 10.7, so we are dependant on that final O VDC reference setting to refind our precise initial 10.7 freq setting that everything was peaked to. (Unless you have a good 8-10 dig freq counter to reset freq with)
Next you start tuning the sig generator frequency slowly off in one direction and watch an increasing either + or - voltage develop and continue until it peaks and then starts diminishing in value.Write down that peak voltage value. Then you slowly bring the sig gen freq back and pass 10.7 to go the opposite freq direction and watch the opposite polarity of detected DC level increase up to the point of its peak and then start diminishing in value. Log in that peak voltage and compare the two logged values and see what difference there is. ( is all of that procedure Deja vu ? )I would suspect them to be quite close. If you haven't figured it out by now, what we were doing ...two times now...was utilizing a simple metering technique to evaluate the symmetry of your detectors "s"curve response outwardly to its + and - node limits, without utilization of swept RF technology or use of a scope in its observation.
Overall now, if we were to plug in some hypothetical numerical value (since inline mixer-IF-and limiter design makes difference in overall system gains)...say like an easily referenced round figure of +2 VDC and -2VDC as being read out for the two nodes peak readings, some fine music should come out of that system. Conversely if we should come out with some imbalanced permutations , say like +1.5 VDC and -2.5 VDC, that should produce some ratty distorted performance.
Finally if you want to totally balance and precisely optimize you can go back to the sig gen still set to its freq and look up your first logged figures at the top of the page on that FM stations evaluation, and that would permit seeing the levels of peak voltages being produced at that particular RF input level being handled on that off air signal. Lets just say that they were in the order of the above secondly mentioned figures and we can see that their total differential voltage spread is 4 VDC so half of that value for an average is 2 ( should it instead, actually have read out as being ,balanced, at optimal + and- 2 VDC's levels.) So we would then want to tune to both sides of 10.7 and see if our + and - nodes peak voltages sum up to a combined total differential ~4VDC level ,and if not, use the attenuators RF level to permit the attaining of that figure. Then its just a matter of minute touch-up of the Ratio-det’s primary slug ( for the first time, since peaking it initially) , it is what will correct the signal imbalance, take note of its direction to be turned and move it ever so slightly. Go back and do offsets of > and < 10.7 tuning in order to take the nodes peak voltage readings again and then compare to see if there is a closer balancing or greater imbalance of the two readings and once the corrective tuning direction of the Ratio-det's primary is ascertained, is then just a matter of minute progressive correction of the primary until a perfectly equal balance is acquired on the detected output nodes + and - voltages. Also, you have then aligned using the same RF levels and limiting experienced when evaluating with that off air signal.That's it.
And, if that doesn't do it for you , I'm afraid that you've got trouble right there in River City .
73's de Edd


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
:Am repairing a '50s Pseudo Ac/DC 3 band (AM/SW/FM) radio. Have all necessary test equipment including FM gen., Tek scope and Digital Multimeter. Have completely recapped and checked voltages and resistors. Have Sam's for the unit. FM is distorted. Uses a Ratio Detector and joint FM/AM IF transformers.
:It was obvious someone previously had attempted to eliminate the distortion by tweaking all FM IFs, Det. transformer, etc. Have gone through a full alignment per Sam's instructions but it is erratic on how it responds to tweaking. Adjusting for peak IF voltage at the ratio detector just increases distortion. Sounds like overload. Limiter seems to be at correct voltages. Balance on ratio detector is right on at nearly 0 volts. Tube complement is 17EW8-FM RF and LO/Mixer, 12BA6 FM/AM IF Amp., 12BA^ FM Limiter, 12AL5 - Ratio Detector, 5 other tubes for rectifier, audio pre and final and AM/SW RF Amp/Mixer and Tuning Eye tube.
:
:This circuit has no AFT. I have been through this thing for MANY hours and am not much further than when I started in cleaning up the sound. AM & SW are perfect soundwise. I have over 40 years in electronics both in design and troubleshooting. Any guidance would be appreciated.

11/28/2005 1:21:16 PMRick Skowronek

Edd,

Here are the answers to your questions-

:Initially, I also am in question of the Ratio-det circuitry as to the balance of the like resistive elements on its secondary and if the low value secondary electrolytic filter is a new one.

- The receiver has been almost totally recapped. All electrolytics in the power supply (7 of them), all paper non-polarized and even the ratio detectors balance picofarad's as well as it's 10 ufd electrolytic across the output (believe they call it a stabilizing capacitor). All resistors have been checked for value in most of these circuits as have B+ voltages. All is now nearly new including the ratio detector and limiter tubes.
:
:As to the electronics aspect on the IF's I was also a little fuzzy as to their construction. As per the AM and FM portions possibly being in a unitized / shared housing, or were they using the more common utilization of a series mechanically separate pair with the 10.7 being at the tube end of the series connection and the 455 being a the B+ end...

Exactly right. dual units, same housing (except for limeter/ratio detector coil, hooked in series with 10.7 at tube plate and 455 khz at B+.

:It would be nice to know, informationally, the manner of construction of the 10.7 transformers. Specifically, the Ratio det xformer. I am thinking of possibly either a "K-tran" design of transformer, with a top coil on a form and then another coil spaced away from it at the bottom, with their respective companion encircling top and bottom powdered iron tuning cups.

Pretty much as you described with the brass slotted screw shaft adjusts out of top & bottom. Basic three coil design, that is one side on limiter, the other side on the two elements of the ratio detector (plate & cathode) with a third smaller coil going from the center tap of the detector side coil through a resistor to the audio out junction of the two picofarad sized caps on the ratio detector. Haven't torn out the coil itself to see the internal look.

Anyway, after reading your beautifully described directions for alignment with the tools I have on hand, I should be able to at least get this mama close. I do have an older Heathkit VTVM of the mechanical kind. I totally agree on the ratio detector secondary being much easier to tweak by a zero center mechanical meter movement.

And just for your own info, the HP 8640B is married to a beautiful 9 digit frequency counter. You can literally set this to parts of a hertz. It also has a frequency lock function that, once set, accounts for and corrects any minute drift in your gen's output freq. And, of course as you mentioned, a highly accurate and broad range attenuator system. Truly a masterpiece and well ahead of it's time.

Just from reading through your procedure, it really does sound like pretty much of a hosed alignment. As I said, someone in the past decided to tweak all FM circuits. The distortion is definitely worse on higher powered stations when center tuned and you can decrease it by tuning off station slightly.

I'll give it a good college try. Just may get it with your help. By the way, the 10.7 MHz signal input is fed directly into the grid of the local oscillator/mixer tube bypassing all of the FM RF circuits and the Sam's for this is Set 570, Folder 9.

Thanks,
Rick

11/28/2005 2:42:56 PMThomas Dermody
455 KC and 10.7 MC IF transformers wired in series are switched between the two by shorting out the 455 KC section. When both are in series, with the 10.7 MC section on the plate side, the 10.7 MC section acts simply as a wire connecting to the 455 KC section. It has no effect on the 455 KC section, and so the 455 KC section functions as though it were by itself. When switching to FM, the 455 KC section is shorted out, sending current directly to the 10.7 MC section. Never are the two left in series for FM. The choking action of the 455 KC section would ruin all FM operation. Either B+ is bypassed directly to the 10.7 MC section by shorting out the 455 KC section, or a condenser of large value (anything above .01 MFD) is connected from the junction between the two sections to B-. Obviously complete switching between the transformers could be done by an elaborate switch. This is an unnecessary expense and procedure. Wiring them in series and shorting out the 455 KC section for FM is ideal.

Just had to put that in there to rectify the explanation of piggy back IF transformers. Otherwise all procedures for tuning are accurate. The radio is distorting on loud stations either from poor limiting (unlikely) or from peaked IF transformers, as everyone seems to agree upon. Edd's tuning procedure should remedy this problem.

Thomas

11/28/2005 5:17:00 PMRick Skowronek
Hi Thomas,

Good to hear from you again. Well, you probably need to talk to the Japanese designers of this circuit.

First, just a bit of circuit clarification. There are two IF stages. Each has a dual transformer section. The first IFs indeed are bandswitch selected to operate individually. This is primarily done by by an interesting method. That is, in FM the mixer/Local Oscillator tube receives 90 volts B+ when in the FM mode. That B+ is turned off when in AM/SW by the bandswitch The AM/SW local oscillator is always powered up but it's section of the first IF is ungrounded on the secondary in FM mode, grounded in AM/SW. The common output of the dual IF is then sent to the first AM/FM IF amplifier. Since the FM LO is unpowered in AM and the 455 kHz IF secondary is ungrounded in FM, it is acting like two seperate transformers and interaction should be slight. However, in AM the two sections of the IF secondaries are pretty much in series. Probably not much of a problem in AM @ 455 kHz and solo transformer in FM.

Now, for the second IF Amp. stage. This is most definitely in series on the primaries. The tube is a 12BA6 labeled AM-FM IF Amp. The tube plate output feeds both the AM and FM IF transformer primaries in series and then off to the B+. Nothing is switched. As to the secondaries, the FM goes to the Limiter input and the AM IF goes to the AM detector which also doubles as the first AF preamp, 12AV6. So, there you go. That's the schematic and the radio in a nutshell.

Rick

:455 KC and 10.7 MC IF transformers wired in series are switched between the two by shorting out the 455 KC section. When both are in series, with the 10.7 MC section on the plate side, the 10.7 MC section acts simply as a wire connecting to the 455 KC section. It has no effect on the 455 KC section, and so the 455 KC section functions as though it were by itself. When switching to FM, the 455 KC section is shorted out, sending current directly to the 10.7 MC section. Never are the two left in series for FM. The choking action of the 455 KC section would ruin all FM operation. Either B+ is bypassed directly to the 10.7 MC section by shorting out the 455 KC section, or a condenser of large value (anything above .01 MFD) is connected from the junction between the two sections to B-. Obviously complete switching between the transformers could be done by an elaborate switch. This is an unnecessary expense and procedure. Wiring them in series and shorting out the 455 KC section for FM is ideal.
:
:Just had to put that in there to rectify the explanation of piggy back IF transformers. Otherwise all procedures for tuning are accurate. The radio is distorting on loud stations either from poor limiting (unlikely) or from peaked IF transformers, as everyone seems to agree upon. Edd's tuning procedure should remedy this problem.
:
:Thomas

11/29/2005 9:32:54 AMThomas Dermody
Interesting. I've never seen IF transformers left in series in a U.S. set when switched to FM. I suppose it could work, since the AM trimmer (whether fixed or variable) will pass the high frequencies that the AM coil wouldn't otherwise pass, permitting the FM coil to function. Odd. I can't say that this would be the most efficient set-up, but maybe it is. Typically, however, if the AM and FM IF transformers are wired in series, the AM is shorted out for FM reception.

T.

:Hi Thomas,
:
:Good to hear from you again. Well, you probably need to talk to the Japanese designers of this circuit.
:
:First, just a bit of circuit clarification. There are two IF stages. Each has a dual transformer section. The first IFs indeed are bandswitch selected to operate individually. This is primarily done by by an interesting method. That is, in FM the mixer/Local Oscillator tube receives 90 volts B+ when in the FM mode. That B+ is turned off when in AM/SW by the bandswitch The AM/SW local oscillator is always powered up but it's section of the first IF is ungrounded on the secondary in FM mode, grounded in AM/SW. The common output of the dual IF is then sent to the first AM/FM IF amplifier. Since the FM LO is unpowered in AM and the 455 kHz IF secondary is ungrounded in FM, it is acting like two seperate transformers and interaction should be slight. However, in AM the two sections of the IF secondaries are pretty much in series. Probably not much of a problem in AM @ 455 kHz and solo transformer in FM.
:
:Now, for the second IF Amp. stage. This is most definitely in series on the primaries. The tube is a 12BA6 labeled AM-FM IF Amp. The tube plate output feeds both the AM and FM IF transformer primaries in series and then off to the B+. Nothing is switched. As to the secondaries, the FM goes to the Limiter input and the AM IF goes to the AM detector which also doubles as the first AF preamp, 12AV6. So, there you go. That's the schematic and the radio in a nutshell.
:
:Rick
:
::455 KC and 10.7 MC IF transformers wired in series are switched between the two by shorting out the 455 KC section. When both are in series, with the 10.7 MC section on the plate side, the 10.7 MC section acts simply as a wire connecting to the 455 KC section. It has no effect on the 455 KC section, and so the 455 KC section functions as though it were by itself. When switching to FM, the 455 KC section is shorted out, sending current directly to the 10.7 MC section. Never are the two left in series for FM. The choking action of the 455 KC section would ruin all FM operation. Either B+ is bypassed directly to the 10.7 MC section by shorting out the 455 KC section, or a condenser of large value (anything above .01 MFD) is connected from the junction between the two sections to B-. Obviously complete switching between the transformers could be done by an elaborate switch. This is an unnecessary expense and procedure. Wiring them in series and shorting out the 455 KC section for FM is ideal.
::
::Just had to put that in there to rectify the explanation of piggy back IF transformers. Otherwise all procedures for tuning are accurate. The radio is distorting on loud stations either from poor limiting (unlikely) or from peaked IF transformers, as everyone seems to agree upon. Edd's tuning procedure should remedy this problem.
::
::Thomas



© 1989-2025, Nostalgia Air