[TTech]

Came across An Atwater Kent model 20 big box and brought it home a couple weeks ago. Will be oldest radio I've worked on yet. First battery powered radio also. It has all it's tubes, all the filaments check good. One audio transformer missing, other audio transformer, open primary. Replaced one open grid resistor and one condenser looks bad. Going to take it slow and work on it when I get tired of other projects. But I do have a couple of questions to anyone that has experience on this type of radio.

1. When the variable condenser / tuning gang knobs are set at zero, what should be the position of the plates? Number 1 is approximately 80% closed. While number 2 is approximately 35% open, number 3 knob 45% open. I was expecting all three to be the same. I took #1 knob off and could find no alignment mark, all 3 knobs can be secured any where on tuning gang shaft. 

In the future when I fire it up and want to see if I can tune the radio to my sig gen putting out 454.3 wave lengths. According to the dial setting chart my 1st dial should be set at 62, #2 dial at 63, and #3 dial at 63. If the dials are not in the correct position in relationship with the plates, my dial position becomes unknown. Would like to solve this if I can. Unknown radio operation status plus unknown radio dial position could lead to a frustrating afternoon.

2. Speaker that came with with the radio has the name Perfectone on it. Horn has no curves and stands straight in the base. Measures 23" from bottom of base to top of the horn. Audio coils are good and can hear static when leads are tapped across a 9 volt battery. Have been able to find no information on this name of speaker. Any info or reference to a web site would be appreciated.   


     Thanks

     Tim

[TTech]

Sorry for the gaps and spaces in the thread wording. None of this is what I typed. When I saw it in the preview I didn't know how to fix it.

[Chas]

When each tuning condenser is fully meshed, the cursor on the panel should point to 100...

So expect that the radio will not tune all of the BC band and may only reach 1350Khz and no higher.

Although the speaker clicks the diaphragm may rattle..

The type of driver in the speaker will determine if polarity of the speaker leads ar to be observed. If there are (+) and (-) markings on the driver, connect the (+) wire to the B+ and the (-) to the plate circuit of the last audio tube...

If the driver is the bi-polar pin driven diaphragm motor the polarity is not important.

Be best to find used A-K transformers for the set. The audio transformers can be rebuilt by using modern units that have a high resistance primary. Some modern make-do replacements lack the primary resistance the net result is low volume and/or distortion...

If this is your first 20's battery set, be aware that scrimping on the power sources can also introduce problems like oscillation, poor sensitivity and poor volume. An absolute must is the "A" supply of at least 6 to 6.6 volts, even though the tubes are 5 volts. The reason is the rheostats lower that voltage and the difference is part of the tube bias. The rheostats also control the volume and detector sensitivity. The "B" battery must also have a low impedance using 9 volt transistor battery's will work but the set can oscillate. There is a chassis mounted cap in a metal can the should either be "stuffed" or replaced. This cap helps prevent oscillation.

http://www.atwaterkent.info/TechData/Dra...640sch.pdf

The url will help with the tubes:

http://www.nostalgiaair.org/References/A...lash08.htm

[TTech]

Thanks Chas, per your dial and tuning gang alignment instructions all three of mine are way out of line. But not difficult to correct. Have been looking at the ARBE-3 battery eliminator. Since I am fairly new at old radio repair I don't have a supply of bits and pieces to build one and would have to buy every thing new or travel a lot of thrift store miles looking for a panel or box to build one on. The new schematic with info is very handy and you have given me a good amount of reading to do on old speaker horns. I don't see any polarity markings on the speaker coil assembly or leads so I'll have to see what I think after I read the material you were kind enough to supply. The assembly that holds the diaphragm and provides a means to adjust the audio coil assembly towards or away from the diaphragm was in broken and missing form. Super glue to hold the pieces together then JB weld seems to have done the trick. The diaphragm had only 1 gasket, but it seemed one on each side would be better, so I made two and will see how it works. It will sound laughable but I made the adjustment assembly out of the cut off threaded  end of a wide mouth canning jar and it's mating metal ring. More JB weld and some minor successful testing using said 9 volt battery. I can adjust the audio coils onto the diaphragm so static from tapping the battery post is very weak, adjust the coil away from diaphragm for stronger sound, then away from diaphragm till static is weak again. The test may be meaningless, but keeps me hopeful. From the price I'm seeing there's not much hope of picking another one up cheaply for parts. Thanks again.

[Chas]

The correct polarity of a diaphragm speaker when no (+) sign is noted and the speaker is used in the plate circuit of the last audio tube is such when the plate current is flowing, the diaphragm is drawn inward. This polarity re-reinforces the magnets and increases the sensitivity of the speaker. Time, drains away the strength of the magnets that even the correct polarity cannot restore. In such a speaker it is possible to re-magnetize the magnets by the use of a neodymium magnet. Careful observation using a compass will determine the the N & S poles of the internal magnet. By shunting the OEM magnet with the neodymium  the raping the OEM magnet with a small tack hammer then a stronger magnetic field will be set.

If the original tinsel cord is still in place for the speaker it may have a red thread in the cord, this is the positive thread and should connect to the tip of a phone plug if it is used. When inserted into a radio that has a speaker jack once the connection is made the frame of the jack becomes connected to B+ via the speaker but the shell of the plug makes shock contact difficult.

This same polarity should be applied to high impedance magnetic headphones as well.

The Antique Wireless Association has recently made available a circuit board and a parts list for a power supply that could power this radio. A part list and sources are included, cost estimate should not exceed $100, fabricate ones own cabinet.

https://www.antiquewireless.org/homepage/pcb-item-5/

You may find more battery set parts available at the Antique Radio Forum Classified ads... There have been recent posts in regard to A-K 20. Google with Antique Radio Forum in the search terms.

The ARBE-III folks have suitable high resistance audio transformers that would be more appropriate for this radio. The Hammond Radiola III replacement transformer is also O.K.

When the radio is completely restored a short and direct earth ground in place and an antenna of 35 to 50ft. used, if, the reception is still weak, consider lowering the value of the grid resistors from the 600 ohms to 500 or even lower. That will increase the gain of the RF stages. This value can only be lower until the radio oscillates uncontrollably. These resistors can be 1/4 watt carbon and do not have to be wire wound.

A major issue with 20's era radios are the connections both soldered and mechanical. They must all be re-done even if they check O.K. Do keep the wire size at least 18ga for the "A" battery for minimum voltage drop...

Chas

[TTech]

Once again appreciate the much needed info. Won't be able to digest all the info quickly due to several other large projects in the queue. Thinking of lowering the grid resistor values. Have many A.M. stations in the area but they are all fairly small output stations, one-K to 5-K outputs. average distance from home, twenty to thirty miles. I'll need all the help I can get. Some drop to 10 watts at night.

Since I read your last reply have found several video's on re-magnetizing the permamagnet. But all leave out the info I'm in need of. Guess it's common knowledge that I slept through during a long ago class. Does {north equal +} and {south equal -}.
This speaker has sharp pins at the ends of the wires. Looks like the + wire connects to the radios 90 volt B+ and the negative speaker lead to the last amplifier tubes plate ?

I know this is very basic. The radios I've worked on from the 40's and 50's though similar to the solid state then digital radios I've worked on, are worlds apart in components and how those components function. This radio and speaker takes me back even further, once again similar but a bit of a different world again.  Just trying to avoid a little mistake that causes catastrophic damage. Thats a cause and effect that covers everything.

[Chas]

Thinking of lowering the grid resistor values.

I should not have mentioned that But I will repeat, that is done only when the radio is used in a remote location. If the procedure is done it may obscure other problems or make them worse. Keep the value at 600 ohms for now.

Look around the Atwater-Kent web site. I believe there are many model handbooks reproduced that explain how to set up the radio, antenna, ground, batteries and speaker.

Begin with a triage, take ohmmeter (I highly recommend the use of an analog meter) readings of all the connections, but plan on re-soldering all of them. Be sure the rheostats are working and not worn or burned out.

Be sure all the tube pins are cleaned.

Determine exactly what is to be done about the audio transformers...
A-K recommends the use of NuJol a pharmaceutical mineral oil to lubricate any moving parts. If the rheostat compartment is opened petroleum jelly is used to lube the slider and finger contact.

The "A" or filament circuit is critical in respect that getting a crossed connection to B+ will destroy the tubes.

This may help:

https://www.arborsci.com/blogs/cool/thre...omagnetism

However, the only need is to see which way the diaphragm pulls in or out with the polarity. Tracers on wires are for identification for the layman who would not have an ohmmeter.

If this radio uses a battery cable be aware that the color coding will have faded. Confirm where all the cable wires are going. Re-color the wires with a Sharpie or make tape flags with the proper identifications. The ends of the wires can be cleaned up and tinned, a little sealant applied to keep the textile from fraying.

Chas

[Eliot Ness]

The Antique Wireless Association circuit board and parts list for a power supply recommended by Chas is a fine choice if you want to build a power supply.  There will still be some items you'll need to source on your own such as an enclosure, and while none of that is overly difficult (especially with on-line sources), it is something you'll have to do.

On-the-other-hand, the ARBE III has been the gold standard for a pre-built power supply for several years with many satisfied customers, and if that is what you decide it will serve you well for many years.

[TTech]

Thanks again for the info Chas. I do have a copy of the Atwater Kent Radio instruction book, volume 2. This radio came with no cables. Thinking of selling off a few motorcycles Eliot. So with my retirement goal of becoming lazy, the ARBE III is pulling at my wallet fairly strong.

With compass in hand along with 9 volt battery, multimeter, and spring scale. Determined what was North and South, Positive and negative. 
Since I don't yet have a working radio tried just sending a 1k tone audio from my sig gen to see if speaker would respond in any way. I know I'm not the first one, but a new adventure for me. O-scope  said my 1000 hz was 20 vpp .5volt DC. Expected a very weak response from the speaker especially with my mason jar and metal band modification. Was surprised when I was rewarded with a loud very clear 1000 hz  tone I could hear in all rooms of the house. My mason jar glass threads allowed me to once again tune the speaker for loudness and clarity. With a little tuning it could also reproduce a clear 400 hz tone. At 1000 hz could still hear a tone even when I reduced the tone amplitude to .9 vpp. At this point was glad I could produce such a clear tone off the original  driver and metal diaphragm. Don't know how it will respond when connected to the real thing. But this also keeps me hopeful.

[Chas]

TTech,

Must post pix to the board of your work. I'm glad the speaker is working, I can't wrap my mind around a Mason jar and a horn speaker    

May want to "stuff" the .3mf RF bypass cap. The paper one will be leaking. FWIR it is in a metal can with a soldered bottom. A .33/630v may be too large a diameter, three .1/630 in parallel may fit better. The metal case should be the other terminal. GL  chas

[TTech]

I have not looked too closely at the caps yet. Will refer back to your reply when I do. My driver fix unfortunately does not include the entire mason jar, although I could have made it work using the entire jar. Just used a bottle cutter to remove the jars glass threads and metal ring you would tighten down on the threads if you were canning. My modification is not aesthetically pleasing but so far it seems to be working. I took some pic's as I was creating a mess making it. Will sort through and try to post some pic's tomorrow.

[TTech]

                    Final product installed and working in speaker horn base. Speaker is fully functional for tuning, just as original was. Based on testing so far.
The speaker wires don't allow the base to sit flat. My modification is a bit taller than original. Will need to make a wooden extension of about 1/4 inch and attach to base so speaker base sits flat.

[Chas]

Ah ha! I see... The entire driver housing was ruined (which often happens in other brands of speakers) so one was created out of "parts" that could be screwed together. The new diaphragm is the steel Mason jar cover.
Very good!

Chas

[TTech]

Very good, you described my fix better than I could. It's hard to tell from the second photo, that is the original diaphragm in the center. You did get me thinking of how the canning jar sealing lid might sound when used as a diaphragm.   

Showing my old time farm boy roots I'm afraid. Finding a fix with what I already have on hand. Only cost for me was the epoxy.
Keeping my fingers crossed.

[Chas]

Do note the "string" that is wound around the two connections made to the driver coils. That is the "tail" of the braiding over the tinsel connecting cord.

Most speakers use this "tail" to tie via a hole or ring next to the connections to take the strain off of the connections should there be an accident that pulls on the cord.

There may be a "tail" on the other end of the tinsel cord. When a phone plug is used to connect there will be a hole in one of the connection devices to tie this tail...

In the radio there might be a hole or not to tie the cord.

Re-connecting tinsel cord to the metal phone tips is a bit tricky. I'll post how to do that if need be.  Some tinsel cords it is possible to establish a new tail of braid so long as the braid is still in good condition. Simply start a hole in the braid and work the individual cords through the hole leaving the tail of braid.

Why not post an image of the set removed from the cabinet of the three tube area where the audio transformers go. Since FWIR one of the transformers are missing and one is bad. According to the new drawing the transformers are the same. If a dust footprint seems to indicate that. Post the dimensions of the existing round transformer and the spacing of the now empty mount holes for the missing transformer.

I suggest not cobbling an audio transformer, though "stuffing" an old round case with a correct transformer will suffice. Incorrect resistances or the wrong ratio will create distortion...

Want ads should not appear in a discussion form...

Post a WTB here for the transformer:  https://philcoradio.com/phorum/forumdisplay.php?fid=17

The grid leak resistor probably has increased value to infinity. A replacement can be wired in below deck or the original can be carefully dissembled, the glass removed and replaced with a larger glass tube and a 1/10w 2.2 meg. resistor hidden inside... If the grid leak resistor is higher than 8-10meg the detector stage will be cut off and not work...

Chas