[462ron]

Thanks guys for your input! @Vinzer, yes the 37 is an AC power set but my parts chassis didn’t come with one. That and the fact that I already have the tubes needed for my 35, I’m going to keep with my original plan of getting it up and running.
@Chas, lots of good info there you supplied. I had already ohmed out the audio transformers before deciding on buying the parts chassis and they seem fine. I did have the less expensive option of just buying the brass pulleys and bands but decided on getting the whole chassis as it would have other parts I might need down the road. I did notice the tuning caps are a bit stiff and I believe it’s what caused the original bands to break. One good thing with this set is it only has a 0-100 tuning scale instead of actual frequencies so there is no real alignment to be done other than the crude procedure of getting all the tuning caps synchronized! I have added a drop of 3 in 1 oil to the bearings and it made a world of difference. Where it says to use petroleum jelly on the rheostat, I’m assuming that’s regular Vic’s petroleum jelly straight out of the medicine cabinet.
Also, as I stated before, I mainly resurrected this radio for the the sole purpose of experimenting with the grid leak detector so for now, it will be an abomination on my bench with having a computer power supply for the “A” and “C” supply and a bunch of 9V batteries for the “B” supply. Not very pretty and certainly not what the Mrs. wants in her living room   Someday I may want it for display but then I will get a more streamlined power supply for it like an ARBEIII and a period correct speaker!
@MrFixr, not sure why having a 27 for the detector would affect the 3rd RF coil to be 20uh more than having the 200A. Lots of trial and error experimenting ahead.
For anyone needing good info on Atwater Kents, I found this website to be a decent source…http://www.atwaterkent.info/

Ron

[Ron Ramirez]

Vaseline, not Vicks. Vaseline is a lubricant; I doubt your AK will be getting a chest cold.

[462ron]

 Oops, thanks for the heads up Ron. I ran over to my medicine cabinet and saw that I have Vaseline, although I do rather like the smell of the Vic’s!

Ron

[MrFixr55]

I like the Vics too. I wondered (did not check) if the 27 had different interelectrode capacitance than an 00A or 01A does. They may have had a different inductance for the 27 than the 01As.

I have a working 40. never noticed whether there was sufficient slack in the bands to "bump" the other caps to optimize the tuning. The radio is plenty sensitive (when I turn off the computer, LCDs and other noise pollution). My Crosley 601 has 2 vernier adjustments but like most bandboxes, the tuning caps are totally disintegrated pot metal.

[462ron]

MrFixr, I did a quick look at the interelectrode capacitances between a 200A and a 27 and it looks like there is substantially less pfs in the 27 so that would probably account for needing more uh in the resonating coil. Well if I have problems getting the 3rd RF stage synchronized I will swap in the coil from the parts set. I’m still a bit of a ways off yet to firing this set up. 

Ron

[Chas]

FYI, the '00 gas detector is for DX at the sacrifice in gain over the 01a. Usually the 1st AF transformer is 5:1 some of the voltage loss made up by the 1st AF transformer with a sacrifice in fidelity. Same '00 is fussy about its filament voltage and grid leak resistance to achieve full sensitivity. Many 20's radio fans found the '00 difficult to keep in good, sensitive adjustment.

I suggest lighting up the converted 35 with a known '01a with good emission, even if it has to be rejuvenated.

The change from '27 to an '01a in the detector will not be noticed as any shift in tuning will go away with capacitor alignment. The 1st AF transformer is likely different plate load.

BTW, the info for A-K lube came from some original service info I found at a local A-K dealer estate.

FWIR the coils in the doner radio are binocular. These should be fine if mounted as they were in the original set. No need for the orthogonal mounting as was the 35...

The info about changing the value of the grid damping resistors in the RF can be found in some repair histories for TRF radios in Ghirardis Radio Troubleshooters Handbook.

A follow up on changing the tuning condenser array alignment frequency. "Move the frequency from 1mhz to 950 khz" . Do the opposite if the radio is more sensitive in the other direction, ie: 1050khz.

Sand paper the tips of the pins on the tubes the oxidized solder can give poor connection for RF. Re-Solder everything...

GL

Be interesting how this transplant works out.

Chas

[462ron]

Ok, got to spend some time on the AK and had my first snag. The 200A detector was dead, filament out. I figured I had nothing to lose so I tried heating and resoldering the filament pins and now it lights. Not having ever worked on a TRF set before, I had no idea what to expect from it as far as performance goes but I was rather disappointed in it. The one station that I get here from 50 miles away usually comes in very strong but I couldn’t find it. I did get one station though. I had a nagging feeling that the tuning caps I transplanted were the problem as they had 2 more fins than the originals even though the RF transformers measure the same uh from my set and the parts set. I swapped the tuning caps back and it works better. There is a lot of range in trying to set the outer caps to synchronize to the main tuning cap. I will need to try with my signal generator and see if I can dial this in better. It’s hard to know exactly where a station should be on the dial or a signal from the generator as this only has a 0-100 dial scale! If I get this set tuned up better then the next step will be to experiment with changing the detector from a grid leak to a diode detector using a 1N34 or some such diode!

Ron

[MrFixr55]

Hi 462 Ron.

Other than the detector, TRF sets are actually easier to figure out than Superhets.

The 4 big issues with TRF sets of this vintage (assuming you have it powered right):

First issue, Yes, very few TRFs had calibrated scales. I have a 1027 Federal 10 "Orthosonic" (Wonder if Victor smelled trademark infringement here) The scale was calibrated but was calibrated in meters. Back in the day, people kept logs as to station callsigns, frequencies and what number on the scale that the station appeared on. the "1 dial" sets started to appear in earnest in 1925 - 1926. This is why you have 3 separate caps ganged by pulleys and belts. A-K did not have time to arrange for a 3 section single frame capacitor; neither did anyone else except possibly RCA and Philco (?).

Second is selectivity. The A-K 35 was introduced somewhere between1926 and 1927, By then, there were at least 3 clear channel stations in NYC, WEAF (Later WNBC, now WFAN) at 660KHZ, WOR (710KHZ) and WJZ (WABC) at 770KHZ The TRF set of this type with 3 tuned circuits had reasonable selectivity but not as good as a superhet. There should be a reasonable amount of RF Gain from this system . Note that the 1st RF stage in this set is not tuned. The RCA Radiola 16 (battery) and 17 (AC) were very similar to this set. However, the A-K 35 has separate bias connections for the 1st AF vs the AF Output, and has separate B+
BTW, make sure that the 1K Ohm wire-wound "grid resistors" on the 2nd RF and 3rd RF (first and second tuned circuits) are intact. They almost always are open. If open, bridge with a 1K resistor, but leave the windings there. These resistors were there to help prevent oscillations in these stages (These sets were not neutralized (Neutrodyne) sets.



Third, the detector (the reason for the experiments) is certainly not HiFi. in addition, a 200 is much more distorted than a 201A tube, but more sensitive. You can use either.

Finally, the available tubes did not have earth shattering power in the Audio Output stage. A 01A at 90V and -4.5V Bias but out less than 20 mW. a 112 in the same condition put out 30 mW A 71A could put out 700mW but at 180 Deg and -40V grid bias (C-)

f you are using an ARBE III, their table assumes 01A tubes in all sockets, and does not follow Atwater Kent instructions.

Make sure that you are set up per the following chart:
For 01A and 112A, for 90VB+, Bias (C-)should be -4.5V. For 135V B+, bias should be -9V.

For the 71A (UX171 or 171A), there must be much greater bias. Per the RCA Receiving Tube Manual RC10 (Ca 1930), for the '71A, for 90V B+, bias is -16.5V. For 135V B+, bias is -27V. For 180V B+, bias is -40V.

Connections to power cable are as follows:
Red- A+, B-
Black- A-, C+
Yellow- B+ for detector, usually 22.5V but also try 45V
White- B+ for all tubes except Output, should be 67V
Green / Yellow- C- for output stage, set per above table.
Brown- B+ for Output stage. Use above table.

If using a horn or magnetic speaker, polarity of the speaker wiring is important. I lik to use an 8" dynamic speaker with an output transformer primary impedance of about 2500Ohm. One from an AA5 works fine.

Make sure that you have a long, high antenna and no CFLs, LEDs computers, etc.

Attached below are links for instructions from the Atwater Kent website:

https://www.atwaterkentradio.com/ak_booklets.htm

Download both booklets, they should be useful and are interesting.

Hope this makes sense (I am hopped up on cold remedies!)

Best Regards,

John, MrFixdr55

[Ron Ramirez]

A-K did not have time to arrange for a 3 section single frame capacitor; neither did anyone else except possibly RCA and Philco (?).

Philco did not enter the radio manufacturing business until June 1928; consequently, all their sets had ganged tuning condensers.

[MrFixr55]

When you get around to converting to a diode detector, I think that you can get away with just changing the cap from 0.00025 to 0.01. The Grid leak resistor can likely remein the same. You can crank the voltage up to 67V and apply the 4.5VC- to the leg of the grid leak that does not connect to the grid (usually grounded, lift the ground and put 3 1.6V batteries in place.

I have been meaning to test the quality of sound from 01As used as a preamp. Haven't gotten around to it.

[462ron]

MrFixr, I double checked and have all the proper voltages between the computer supply and a stack of 9V batteries. The original grid resistors mounted on the tuning caps are open, hair fine wire, so I transferred the good ones from the parts set that measure around 350 ohms. We’ve done away with our landline phone and I have made my old phone line out to the pole a long wire antenna, guessing about 60’ between what runs up the side of the house and then out to the pole, plus whatever additional footage inside the house. It works quite well with the added benefit of now I can plug any radio in any room I want and have access to that antenna from the nearest phone Jack!
As far as volume, I am using an output transformer and speaker and feel that the volume would be sufficient if I had better reception. I’m going to switch over to the Dell computer supply as it seems to have less noise and hash. Appears like getting the 3 tuning caps aligned/synchronized is being a problem, sensitivity drops way off to the point that there is hardly any background noise at one end of the dial compared to the other. Using my signal generator, synchronization is iffy as I can hear it over a wide range of rotation, no real peak. I have seen a video on YouTube of a 35 and the gentleman’s set receives pretty well. AM in my area isn’t great but I’m in PA and received a station in NC one night although weak, yet a station about 50 miles from me is almost nonexistent even though it comes in loud and clear on other radios. I thought about buying a set of 01s for it even though the ones I have test good. Getting to the point of putting crazy money into it! I’ll see after switching out power supplies.

Ron

[MrFixr55]

Hi 462. My post just disappeared! A couple of questions:
Your Long Line Land Line, do u usually use this with 1930s vintage superhets or do you use it with TRFs also.

Is this your first TRF?
First TRF that uses 01As?
Are you using a 00 or a 01A in the detector stage? If using a 00, try a 01A. A 00 is more sensitive, but 01A is louder.

An interesting point that you made was that you had difficulty peaking and syncing the tuning caps.

Try the following.

Replace the Detector grid leak with a 3 MOhm Resistor. Replace the Detector grid Capacitor with a 250 pF (0.00025 uF) Cap. replace the Detector "Phone" cap with a 100pF (0.0001) uF Cap. Use an 01A in the detector.

If you have a '112A tube, us that in the output stage.

Per previous comments from a Phorum expert, if you have a good 6VDC Power supply of about 2A or greater, substitute the 5VPS with that. If you have 8 D Cells, hok them up in series parallel. They won't last long but should last a couple of hours.

Line up the tuning caps so that at 100 on the scale, all 3 tuning caps are fully meshed.
Connect your signal generator to the junction of the detector tuning cap and the detector grid cap. Set the frequency to a clear spot around 1200KHz, turn the modulator on and adjust the attenuator for minimum but hearable signal. "Rock" the dial till maximum signal is heard.

Adjust the "detector rheostat" (the one on the right) by lowering it till no signal is heard, then raising it till there is no improvement in volume. Retain all these settings and document scale reading on tuning dial.

Connect the signal generator to the junction of the tuning cap stator and the grid resistor of the rf RF Amp. without disturbing any settings. There should be a significant increase in volume of the signal. Lower attenuate the signal till the volume is low but can be heard. Lower the RF Amp filament rheostat (the one on the left) till the sound is minimized, then increase filament voltage to the RF amps till no further increase in volume is heard. "Rock" the dial as before. Attempt to sync the 3rd RF tuning cap (the one with the knob) with the detector tuning cap.

When peaked successfully, move the signal generator connection to the 2nd RF Tuning cap at the junction of the tuning cap stator and grid resistor. There should be significant increase in volume of the tone.

Repeat the "Rocking" procedure.

Leaving all settings the same, change the signal generator frequency to approximately 600KHz. Tune that frequency on the radio. If significantly louder, there may be an issue with one or more tubes in the RF Section.


Finally, connect to the antenna input. there should be increase in volume.

These radios are not loud by any stretch of the imagination. A 01A in the output stage at 90V produces only about 10mW. a 112A under the same conditions produces about 30 MW. a 112 at 135V B+ and -9V C- will produce 100 mW. For comparison, a pocket transistor radio produces 100 - 200 mW and an AA5 table radio produces 1000 -1500 mW.

I am under the weather today and have to finish projects for my job, as I retire this Friday. However, ASAp, I will pull out my AK35 and play with it for comparison.

Best Regards,
John, MrFixr55

[Chas]

I concur with all the operational procedures recommend and add.

Use a ten turn link to get the RF signal into the radio do not directly connect to any tuning circuit.
Modulate at 30% and keep the level low. If the upper part of the band is dead move the align frequency by 50khz in the weak direction.

Do keep the "A" supply at the radio at 6 to 6.6 volts. starving the radio will create a bias error. The rheostats provide an initial bias so result is 5 volts on filaments.
Reduced "A" will also paralyze the tubes over long term operation by not completing the thorium cycle, filaments not get hot enough so thorium is stripped and not renewed from the core of the filament wire.

Grid leak detectors must not be overloaded with an RF signal, so excessive signal from either the generator or the antenna must be avoided nor can they deal with 100% modulation without noticeable distortion.

Try: Reducing the RF grid squegging resistors, go as low as 200 ohms or until the stage "takes off" from self oscillation, then increase in 100 ohm steps. They do not have to be equal. Changing this resistor value is a recommendation for TRF's as found in A.H. Ghirardis Radio Troubleshooters Handbook.

https://en.wikipedia.org/wiki/Squegging

Get the radio working properly before experimenting with mods...

   

It has occurred to me that using a computer P.S. for "A" power not only is too low but the 5 volt return is both grounded to the case of the P.S. and to the "U" power line. That poses a problem if an external ground is used and could compromise the RF rheostat and the AF fixed resistor. As for the fixed resistor it is hidden in a length of varnished cambric as a strand of resistance wire. Can be overlooked as a length of rotted wire and inadvertently replaced with copper...

The common "A" source was a 90 ah L/A battery. This provided a low impedance and bulk of current drop occurs at the rheostats and not in any lightweight wire used for the "A" supply, generally 14 ga from the low impedance source.

Further, the grid return for the detector is via the junction of two bias resistors that are across the "A" supply, again, reduced "A" will effect the bias of the detector. Be sure the detector grid capacitor is not leaking and at the value given.

YMMV

Chas

[462ron]

Thanks John and Chas, a lot to digest here for sure! To answer your question John, yes this is the only TRF I’ve ever worked on or even listened to so I have no idea of what to expect as far as performance. My set uses all 01As except for the 200 detector. I have swapped in an 01A for the detector with not much difference. My phone line conversion to use as a long wire antenna works really well on my long wire superhet sets.
Chas, I didn’t realize these tubes/sets are that finicky as far as the “A” supply goes. I am using a computer ps that puts out 5V. I changed from a no name ps to a Dell with same voltages. The first supply I used turned out to be noisy, the Dell is much quieter. I’m going to see what I can rig up to get an honest 6-6.5V. That resistance wire in the AF filament supply was crumbling so I replaced it with a .47 ohm carbon resistor which measured close to what the wire does. I’ve tried using my signal generator by going stage by stage connecting it through a cap. Then I used a coil with several turns wound on a TP roll that fits quite well over these tubes. With the quieter ps now I can get hear 2 stations, probably 3 at night but it is alive mostly in the middle of the band, the upper and lower ends are dead. I’ve tried the AK synchronization method plus randomly experimenting.
John, I will experiment with that cap and grid resistor on the detector and see what changes occur. I have changed the B+ value up and down on the detector plate supply without much difference. So, I have lots of things to further look into and I appreciate you and Chas for all your input. OBTW, congrats John on your upcoming retirement. You obviously have antique radios as a hobby like me but in fair weather, I have my hiking and outdoor activities. Hobbies are a must otherwise you’ll be bored and go nuts, daytime TV is a wasteland! Now get yourself healed.

Ron

[Chas]

Can we see what the "new model" A-K 35 looks like and where it is located during testing?

FWIW although the binocular coils have a confined RF field any metal closer than an inch can effect them. It could be that a metal bench, insulated, yet metal could be effecting the radio. Yes, I know the 35 has a metal can cabinet. I have one in my collection too. Place a block of wood or panel under the radio.

Try: Removing the bands from the tuners and tune each independently, if there is a tracking problem independent tuning will solve it. If the bands are too tight, the rotor of the outside tuners can be pulled out of the cone bearings. Obviously if your model has two bearings in each tuner then that is moot.

Keep the B+ for the detector at the rated voltage, be sure the two bias resistors across the "A" supply that are used to obtain a bias for the detector grid are in the right order or the bias will be wrong. The detector works because the tube is set up to operate in a non-linear mode, it detects at the plate as well as amplifies. If the values of the external circuits for the detector are wrong detection will be weak or non-existent.

Be mindful of that ground "loop" caused by the A- and that the A- is not connected to ground via the power supply ALSO! that the antenna is NOT grounded at the far end too.

The 201a is a rugged tube and can recover if paralyzed by a rejuvenation process. They won't last forever in use as the thorium is eventually depleted from the interior of the tungsten wire. Using the 201a outside of its normal operation does deplete the surface thorium layer faster than it can be replenished at normal operating temperature (voltage).

Radio fans of the 20's would ever so often do a "low level" rejuvenation by removing the B & C batteries such that no potential is on the plate or the grids. Then, fully charge the 6 volt L/A battery, of of charge a L/A 6-volt battery is 6.6 volts. Then, turn on the radio and turn up the rheostats so all resistance is removed and allow the filaments to burn for 1/2 an hour. Thus process works so long as there are no fixed dropping resistors in the radio or automatic filament ballasts. For those radio fans with that type of radio one had to invest in a "Tube Rejuvenator" or combined rejuvenator/tester (plate current meter)...

I often wondered why the ends of battery radio power cables were so frayed

Keep trying all the suggestions made. The problems will soon be discovered...

GL

Chas