Forums :: NEW! Web Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Oct Tue 20, 2020 3:38 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 34 posts ]  Go to page Previous  1, 2
Author Message
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Jul Sun 26, 2020 4:31 am 
Moderator

Joined: Jan Thu 01, 1970 1:00 am
Posts: 30662
Location: Livermore, CA
That is a lot of work and looks great.

_________________
Norm


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Jul Sun 26, 2020 5:15 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 36177
Location: SoCal, 91387
Incredible workmanship!

_________________
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\He Who Dies With The Most Radios Wins/////////////////////////


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Jul Sun 26, 2020 8:13 pm 
Member
User avatar

Joined: Aug Wed 26, 2009 1:50 am
Posts: 1509
Location: Mesquite NV 89027 (from Coventry, UK)
Very impressive!


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Jul Fri 31, 2020 7:05 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 768
Lovely job on the variable capacitor, I really admire your work.


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Sun 02, 2020 10:28 pm 
Member

Joined: May Sun 18, 2008 3:18 pm
Posts: 84
Location: Sioux Falls, SD
I was struck by Steve's description of what period correct was during the audion era:
Quote:
At this time the "receivers" were just that and the "detectors" were separate devices. So I guess a likely true representation would be a loose coupler for your typical DIY station with your external detector of choice. (audion or crystal based). There is a combination unit in the QST article above but I don't think that would be typical of the time.

While I think Steve is correct for this project my goal is for a complete radio in a single cabinet and the article in the Dec 1916 QST tells me that there were at least some amateur builds like this. Also interesting are the comments written in over the next several months either boasting about the great performance or asking build questions.

All said I thought I better try building the QST radio. For this prototype I used my air variable and fixed mica capacitors. The 4.5" coil is wound on a PVC DWV coupler as finding a (free) cardboard tube this size was not looking good. The 4" coils came from a carpet cardboard tube, all soaked in paraffin wax. I believe I have copied the schematic correctly with two deviations 1) I added a grid leak resistor into the grid of the type 30 tube, and 2) I did not use a potentiometer to control the grid voltage, no difference was heard at 27 vs 36 volts, I suspect better regen control could be had from some plate voltage control but the variable capacitor did fine. See below for a marked up schematic and the current radio.

Performance was not as good as the earlier design. I could pull in 3 local stations loud and clear, but nothing else. It does make me wonder what a better antenna could do. I currently have a long wire that run up inside my wall to the attic and then spans across the attic for 40'. Grounding is via a copper post driven into the ground.
1000 - Sioux Falls (local) 10kw
1140 - Sioux Falls (local) 10kw
1320 - Sioux Falls (local) 5kw

From here I would invite any suggestions for component value changes as the article only described the coils and left the remaining parts up to the builder to determine. Also I would take any feed back on a final design. I like the Dec 1916 QST radio as it matches to what an amateur builder could have constructed and experienced, but I also like the "Navy" regen design found in the May 1917 Everyday Engineering article as there seems to be a lot of adjustments allowing for improved dxing.

Image
Image

_________________
Brad KE0MZY


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Sun 02, 2020 11:22 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 14860
Location: S. Dartmouth MA 02748-1225 USA
There is no return circuit for the grid. This can be via connection the grid leak resistor to the filament +...

The early tubes leaked enough that this could happen, more or less on its own. Components such as mud sockets leaking coil forms and a vulcanite panel all contributed to the leakage. The "modern" 30 is not going to have grid leakage as well as using modern materials. What happens passing electrons will continue to build the grid more and more negative until all plate current comes to a stop. In the mean time the characteristic curve is pulled and the detecting action becomes very weak or stops all together. In some instances, a very high value of grid leak and circuit losses will cause the grid to become negative very abruptly. Repeats rapidly and makes a Putt-putt sound in the phones.

With all the losses gone in a new construction grid leak mast be permanently fixed but easily replaceable so a value can be selected for sensitive operation.

Once the leak is "fixed" the the polarity of the plate coil has to in correct phase to add to the input coil. If it is not, the circuit will never regenerate.

A correctly functioning regenerative grid leak detector will consume very small amount plate current. A series "B" variable resistor will be ineffective in controlling the B+. A better solution is to use a tapped plate battery bank... Once a B+ voltage is found, that setting is usually permanent.

Try that stuff, see what happens.

Oh, 100pf is a grid cap value used for waves shorter the 160 meters, for the BC wavelengths 250 pf is used...

Post the next drawing with "A" battery polarity...

If you cannot get the receiver to squeal from excessive regeneration then it is not working at least properly.

As much of a pain it is using a bank of AAA or AA cells in a group of 3 or 4 four cell holders would be a better "B" battery until the correct "B" voltage is found for your experiments.

For a detector much over 24 volts rings a bell something may not be right...

GL

Chas

_________________
List' & I will Enchant Thine Ear


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Mon 03, 2020 1:29 am 
Member

Joined: May Sun 18, 2008 3:18 pm
Posts: 84
Location: Sioux Falls, SD
Chas, thank you for the suggestions. See below for an updated schematic. After looking I was using a 250pf capacitor but labeled the schematic wrong, I did make your suggested changes and tied the grid leak resistor to A+. This increased the volume but also seemed to reduce the selectivity as the few local stations really "bled" across the band. Depending on the selected taps when adjusting the capacitor the volume would go from nothing to loud with lower capacity yielding higher volume. I could not get the radio to sequel which has me questioning the wiring. I tried swapping the coil polarity but still no squealing, just the same three stations, maybe a volume difference.

I plan to double check my wiring and then will run the radio some more to better evaluate. Thanks, Brad.

Image

_________________
Brad KE0MZY


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Mon 03, 2020 4:56 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 14860
Location: S. Dartmouth MA 02748-1225 USA
This is an opinion.

The tuning coil, marked loading coil, it is not common practice to shunt the turns by shorting, that "kills" Q regeneration could be absorbed in that short, try removing the the wire from the start of finish that is creating the short after the first tap...
Then increase the value of the grid leak in steps 2.2, 3.3, 4.7, 5.2 and up. Beyond 8.2 megs the bias developed by the leak will probably become too negative and cut the tube off.

A third observation is the use of series tuning and using such a feedback connection that introduces a number of variables. Anything in the three circuit tuner will have a large effect on regeneration and the resonant frequency. I am aware there are alternate ways to achieve regeneration, however, some of these, sets, lets call them "early" circuits had these interaction defects. That is unless you are attempting to do just that...

I am suggestion setting up the tuned circuit so it is parallel tuned, moved the feedback coil to the plate circuit. The feedback coil if it slides can control the regeneration or placing the 50k rheostat in shunt with the feedback coil can control regeneration.

Granted changing the plate or filament voltage would also control the feed back, but there is more interaction and loss of sensitivity in that method.

The sensitivity will come up greatly once feedback is achieved. Broadband response is from the loss of "Q". Reducing the antenna coil position or taping down should increase the "Q"

Too long an antenna places a load on the regenerating circuit and can also inhibit regeneration.

It was common in the early 20's to use a flat top antenna these had a lot of capacity and sets in order to provide for both capacitive and inductive antenna loads using series/parallel switching of the tuning condenser. Once the "lashup" is set for the antenna type it is left that way.

BTW, if the coil positions are optimized that is one variable fixed, that will allow the tuning condenser dial to be logged or "calibrated"...

Fun eh :D

_________________
List' & I will Enchant Thine Ear


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Tue 04, 2020 3:51 am 
Member

Joined: May Sun 18, 2008 3:18 pm
Posts: 84
Location: Sioux Falls, SD
In the QST article the directions for the loading coil note "A connection is made between the last tap and the switch which acts as a reducer for the dead-end effect since it short-wave circuits the unused turns."

Out of curiosity for coils with taps is it best to just leave the unused turns or is there a better way to handle this?

_________________
Brad KE0MZY


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Fri 07, 2020 8:12 pm 
Member

Joined: Nov Mon 06, 2017 2:35 pm
Posts: 126
Location: Texas, U.S.A.
Your construction of this receiver is inspirational. Thank you for sharing your project.

Referring to the QST December 1916 circuit, short-circuiting of the unused loading coil turns, as shown in the schematic, does reduce the possibility of unwanted absorption of energy from the tuned circuit (dead-end effect referred to in the article).
It looks like the tuned circuit for the tube consists of the loading coil and secondary coil in series, in parallel with the series combination of the 27 - 301 pF variable and the .0026 uF. The .0026 is much more capacitance than the 301 pF variable, so the first approximation is that the tuning capacitance is the 301 pF variable alone. Without shorting the unused turns of the loading coil, it is possible that the inductance of the unused turns in combination with the capacitance of the coil and stray capacitance could create a series resonant circuit which absorbs energy which may be near the desired reception frequency, if I'm not mistaken.

Again referring to the QST December 1916 circuit, I would be inclined to put an RF choke between the battery positive and the plate circuit, because the RF impedance of the battery and headphone is unpredictable but is likely to be low. For medium wave broadcast (say 400 kHz to 1.8 MHz) I'd try 20 to 50 millihenrys; the self resonant frequency of the choke would need to be above 2 MHz, if I have not made a mistake in haste.

The '30 tube as a grid leak detector can be run at 3 mA of plate current, which occurs with right around 45 volts on the plate. This gives much louder audio.

The circuit referred to in your opening post of this topic (shown here), from Everyday Engineering, May 1917, page 117 is the commonly constructed Armstrong circuit. It does usually give good performance, as you observed.

Attachment:
EVERYDAY_ENGINEERING_CIRCUIT_02.jpg
EVERYDAY_ENGINEERING_CIRCUIT_02.jpg [ 82.52 KiB | Viewed 682 times ]


An example of amateur use of the circuit was published in the Sept. 1917 issue of QST:
Attachment:
QST_Sept_1917__Kruse_Armstrong.jpg
QST_Sept_1917__Kruse_Armstrong.jpg [ 70.59 KiB | Viewed 682 times ]

C2 needs to be placed fairly near to coil "C", say within around one to three coil "C" diameters.
I would put an RF choke between the headphone and the C2 - coil "C" junction, placing the choke within a few inches of C2 for medium waves.
.......
WB5HDF

.......


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Sun 09, 2020 4:24 am 
Member

Joined: May Sun 18, 2008 3:18 pm
Posts: 84
Location: Sioux Falls, SD
More progress today. It would have been interesting back then with the amount of time, experimenting, and troubleshooting needed to build a really good performing radio. I have flipped through all of the above noted magazines 1916 through 1917 and while there are some examples and dimensions a lot is left up to the builder to figure out. Here in the present day I have the convenience of a digital meter to check capacitance; For the grid capacitor from reading/estimating (and present day knowledge) I first came in at 150pf, added one more plate and then cut it back until I got to exactly 250pf. For the bypass capacitor as earlier noted it measured 0.0026uf, I now have removed one plate of foil to bring it to 0.002uf. Will these values operate better, maybe?

Image

infzqi, thanks for the information and thoughts on design. One limiting factor may be using a design where I only need to build one air variable capacitor. In the Sept 1917 Everyday Engineering there is a good article on radio design with discussion along with values given for a radio receiver of 200 to 2500 meters (same "Navy" circuit as listed in the May 1917 issue). Of interest was a suggestion to use a fixed capacitor that has 2 or 3 values that can be cut in for the primary circuit series capacitor in conjunction with the use of coil taps.

Values for the primary series capacitor were recommended at no less than half of the antenna capacitance and up to 1.5 times the capacitance. Suggested values range from 100pf to maybe 800pf depending on the antenna. 0.002uf was recommended for the "telephone condenser", what I am assuming is the bypass capacitor.

My next idea is to prototype the secondary coil with a top variocoupler for the tickler and a bottom variocoupler for the primary antenna coupling. This would eliminate using a variable capacitor for regen control and would also eliminate coil taps. Additionally tuning from say 180 to 580 meters can be done reasonably with a single untapped secondary coil.

_________________
Brad KE0MZY


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Sun 09, 2020 6:32 pm 
Member

Joined: Nov Mon 06, 2017 2:35 pm
Posts: 126
Location: Texas, U.S.A.
Very nice capacitors you have made there.

Good idea on the variocoupler coils.

Here is an article about making an RF choke, from Popular Mechanics, June 1929, p. 1012
The windings occupy the slots between the discs.
Attachment:
RF_CHOKE_01.jpg
RF_CHOKE_01.jpg [ 300.76 KiB | Viewed 639 times ]

URL for the June 1929 issue: https://books.google.com/books?id=w94DA ... e&q&f=true
.....
WB5HDF


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Thu 13, 2020 3:12 am 
Member

Joined: Nov Mon 06, 2017 2:35 pm
Posts: 126
Location: Texas, U.S.A.
Of course the Everyday Engineering "Navy" circuit shouldn't need an RF choke; due to the fixed .002 uF capacitor between the cathode and B+ end of the tickler, which (hopefully) should keep the RF current out of the 'phones and battery.

-------
WB5HDF


Top
 Profile  
 
 Post subject: Re: 1917 Regenerative Receiver - Build Project
PostPosted: Aug Sun 16, 2020 10:35 pm 
Member

Joined: Nov Mon 06, 2017 2:35 pm
Posts: 126
Location: Texas, U.S.A.
The 1917 article "Design of Radio Receivers" by M.B. Sleeper is very informative, thank you for referencing it!

Here is an example of calculated primary inductances (no series primary capacitor) for resonance with an example antenna of 100 pF and 31 uH (microhenry) or 31,000 cms (centimeters). Of course a series primary capacitance will increase the inductance needed for a given frequency (8/18/2020 correction). First the total inductance is found, then the effective inductance of the antenna [ref. W. L. Everitt, Communication Engineering, 2nd ed., 1937, p. 505] is subtracted to arrive at the following inductances: (revised 8/23/2020)
1.75 MHz (wavelength 171.43 meters), primary inductance around 73 uH (73 x 1000 = 73,000 cms)
540 kHz (wavelength 555.5 meters), primary inductance around 859 uH (859 x 1000 = 859,000 cms)

To eliminate primary coil taps, a "high impedance", slot wound, untuned primary could be used - of inductance to make the resonant frequency of the antenna and primary just below 540 kHz, and loosely coupling it to the secondary. [ref. A.A. Ghirardi, Radio Physics Course, 2nd ed., p. 551] Although this type of antenna transformer primary would probably be anachronistic for a 1917 set. The coil construction would be similar to the RF choke in the previous post. The self resonant frequency of the primary coil alone would need to be at least 1.75 MHz.
So for example: primary inductance for resonance, with the example antenna given above, at 530 kHz = 869 uH or 869,000 cms.

Looks like some sort of slider contact variable inductance would be another example of technology used around 1917.

Looks like the secondary coil would be around 290 uH, if I have not made a mistake.

8/23/2020 addendum:
After some more study, I would be inclined to try a variometer as the primary for the antenna transformer. The variometer was definitely in use prior to 1917 [A. P. Morgan, Wireless Telegraph Construction for Amateurs, 3rd ed., 1914, p. 196]. A random wire antenna (such as the single wire strung up the wall and across the attic type for example) would be connected to one end of the rotating center coil; the far end of the variometer fixed coil would go to ground. The secondary (coil that connects to grid circuit) (appx 290 uH) would be wound on the same form as the fixed coil, near the ground end of the fixed coil (could include provision to slide secondary closer or farther). The tickler could be on a separate coil form arranged to be movable to vary the coupling to the secondary. The maximum inductance of the variometer for the example 100 pF, 31 uH antenna would need to be around 859 uH. The variable inductance of the variometer will accommodate antennas with some variation of capacitance and inductance from the example.

A series capacitor won't be essential in the primary circuit as long as the frequency to be received is lower than the fundamental frequency of the antenna (wavelength greater than fundamental wavelength of antenna). Table following is for single wire antennas. [ref. H. P. Manley, Drake's Radio Cyclopedia, 1927]
Attachment:
DRAKE_1927_RADIO_CYCLOPEDIA_012.jpg
DRAKE_1927_RADIO_CYCLOPEDIA_012.jpg [ 87.22 KiB | Viewed 339 times ]

Attachment:
DRAKE_1927_RADIO_CYCLOPEDIA_011.jpg
DRAKE_1927_RADIO_CYCLOPEDIA_011.jpg [ 192 KiB | Viewed 339 times ]

-------
WB5HDF


Top
 Profile  
 
Post New Topic Post Reply  [ 34 posts ]  Go to page Previous  1, 2

All times are UTC [ DST ]


Who is online

Users browsing this forum: mblack and 10 guests



Search for:
Jump to:  


































-->


Privacy Policy :: Powered by phpBB