Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: Prev 1 2 3 [4] 5 6 7 8 9 ... 18 Next   Go Down
  Print  
Author Topic: Designing & Building a High-Peformance Subminiature-Tube Regenerative Receiver  (Read 177103 times)
KB1WSY
Member

Posts: 1309




Ignore
« Reply #45 on: May 08, 2015, 06:28:48 PM »

Adding a variable-capacitor regeneration throttle

I added a "variable capacitor throttle" for extra control over regeneration. The capacitor is basically a variable RF bypass between the choke-end of the tickler and ground:



(I have also updated the main schematic that appeared a few pages back -- http://tinyurl.com/lva95cm --  to add this new capacitor.)

I used a 140pF variable because that's what I had on hand ... but it seems to be about the right value in this application.

It provides precise additional control over regeneration. When used in combination with the screen-voltage control on the detector tube, it makes a big difference to ease of reception and particularly, adds considerable control over (1) the signal-to-noise ratio and (2) controlling overload. When copying a weak signal, I keep the screen regeneration control low and close the capacitor plates for maximum "threshold of oscillation" sensitivity. With a stronger signal, I increase the screen regeneration control and open up the capacitor. This is excellent for avoiding "pulling" of CW signals, and for clean demodulation of strong SSB signals.

The presence of the capacitor also provides additional, and welcome, overall gain (G3RZP had already suggested adding at least a fixed capacitor at that point in the circuit).

Edited to add: So far the only downside of the throttle capacitor is that adjusting it has a noticeable "detuning" effect. I am not sure whether anything can be done about that -- which means that calibrating the main tuning dial will be difficult (I think there will have to be a "bandset" or "calibration" control on the front panel because the tuning points will change when regeneration is adjusted with the capacitor). I believe this is a known issue with regenerative sets. Perhaps adding the extra isolation of an RF stage will help???

(On the other hand, the screen regeneration potentiometer doesn't seem to have any detuning effect.)

After using the new "capacitor throttle" configuration for a little while, I changed the value of the grid leak bias resistor back from its "experimental" value of 6.8MΩ back to its original value of 2.2MΩ. I was finding that the larger-value grid-leak resistor was now a source of too much noise.

Today's other major improvement was replacement of the 20pF antenna series capacitor with a small-value "gimmick" cap (two wires twisted together), which had the following effects:
(1) Eliminated almost all of the problems with stability and hand capacity effects. It is now possible, for example, to tune in an SSB signal and hear a normal voice without "Donald Duck" making an appearance every time I move my hands toward or away from the dial, and I can listen to a long QSO without needing to "touch up" the tuning. Also, the cords on my high-Z headphones are no longer affecting the tuning or the regeneration when the cords are moved or when I grasp them with my hands.
(2) Enabled grounding of the set, which had originally been impossible because it choked off oscillation. Grounding has caused a drop in sensitivity, but also led to an obvious increase in stability and additional reduction of hand-capacity effects. Grounding also eliminated a "microphonics" audio feedback problem I was having in the audio stage (this problem appeared when I switched from resistor load to a 7-Henry choke load), and solved several other problems with "touchy" components.
(3) Eliminated the problem I was having with restricted tuning range.

I'm very excited by the progress so far. Now it's time to work on the inductors and build the RF stage.

73 de Martin, KB1WSY
« Last Edit: May 08, 2015, 08:30:17 PM by KB1WSY » Logged
KB1WSY
Member

Posts: 1309




Ignore
« Reply #46 on: May 09, 2015, 04:35:50 AM »

Some questions about the 1AD4 tube and about the B+ supply:

(I have edited this post to fix the link to the 1AD4 data sheet. The link was correct, but the host website now seems to be having issues, so I've now linked to a different but equivalent version on a different site.)

Question 1:

According to the data sheet (http://tinyurl.com/pm4pn56 -- this link has now been fixed), this tube's suppressor grid is composed of "two deflector plates," one of them connected to the negative side of the filament supply, the other one connected to the positive side. Der leitende Überzug ist mit dem Anschlussdraht 3 verbunden. g3 besteht aus zwei Platten, die mit je einem Heizfadenende verbunden sind (adding random German because it sounds so nicely technical). For simplicity, I have schematically drawn these tubes as if they are traditional-suppressor types, internally connected to only one side of the filament. Can someone explain the advantage of the 1AD4's suppressor-"grid" configuration? Tnx.

Question 2:

Concerning the B+ power supply. For the "desktop" (non-portable) version of this receiver, I'll be switching from miniature low-capacity (55mA/h) 23A 12V disposable batteries to high-capacity (600mA/h) rechargeable 9V Lithium-Ion batteries. Currently there are four 23As wired in series. With new 23A batteries, this yields a B+ voltage under load of about 51V.

What I have found is that the receiver's performance deteriorates sharply once the batteries have discharged down to about 42V. This could be a voltage issue; or I guess it could be that the discharged batteries are struggling to supply a steady current??

A 9V LiOn battery has a nominal voltage of about 7.4V and an "almost discharged" voltage of about 5V. Presumably however, when fully charged it's more like (say) 8.5V, at least for a short time. I'm thinking of using seven of these cells. This would yield nearly 60V at full charge, about 52V at the "nominal" 7.4V and about 35V when nearly discharged. The alternative would be to use only six batteries, which would yield nearly 45V "nominal" but probably wouldn't take all that long to dip below the 42V below which my set's performance deteriorates.

So, looking again at the 1AD4's data sheet (http://tinyurl.com/pm4pn56 -- this link has now been fixed). It specifies a "typical plate voltage" in Class A amplifier operation of 45V. Depending on the manufacturer, the "absolute maximum" plate voltage is either 90V or 100V.

Am I right that, therefore, putting voltages as high as about 60V on the plate (at the top of the LiOn battery's charge cycle) would be absolutely fine? Safe for the tube, and not likely to deteriorate the receiver's performance (in fact, quite the opposite)?

Question 3:

With the B+ supply that I have now (maximum of 51V, using low-capacity miniature batteries) I have never been "zapped" even though I've made no special effort to avoid putting my fingers across the supply while the set is on -- the B+ has its own on-off switch which is a cheap miniature knife switch with exposed metal lever and contacts. I wonder whether increasing to a maximum of 60V (and using batteries with 10 times more capacity) will make any difference. Am I courting electrocution by being so cavalier? I promise I won't sue you, whatever your answer is!

Why batteries?

Edited to add: Perhaps you are wondering why I'm using a battery supply for the shack-bound receiver, rather than just building an AC-powered PSU. Well, several reasons. (1) Later on, a portable version of the set will be built, so keeping the battery supply is a useful "proof of concept." (2) Batteries should help eliminate hum issues, especially since the 1AD4 is a directly heated tube. (3) There's something cool about a battery-powered tube set and besides, it will be "portable enough" to be hauled to a picnic table on Field Day or whatever, and I can monitor the bands during power cuts! (4) If I started using an AC PSU, there would be no reason not to use standard indirectly heated tubes, which have higher power consumption but also higher performance and are easier to design circuits for!

73 de Martin, KB1WSY
« Last Edit: May 09, 2015, 05:30:27 AM by KB1WSY » Logged
G3RZP
Member

Posts: 1315




Ignore
« Reply #47 on: May 09, 2015, 05:06:44 AM »

The usual requirement for marine radios was that anything over 50 volts had to be accessible only by the use of a tool such as screwdriver.
Logged
W1BR
Member

Posts: 4195




Ignore
« Reply #48 on: May 09, 2015, 06:34:49 AM »

I've had good luck with battery voltages as low as 9 volts (for the detector) on some of my sets.  "Smallest" tubes I've used are type 1G4, 1H4 and some '30's.  Battery powered regens are cool.

Pete
Logged
KB1WSY
Member

Posts: 1309




Ignore
« Reply #49 on: May 09, 2015, 06:55:48 AM »

I've had good luck with battery voltages as low as 9 volts (for the detector) on some of my sets.  "Smallest" tubes I've used are type 1G4, 1H4 and some '30's.  Battery powered regens are cool.

Pete

Yes, I've heard several people mention their success with low plate voltages. So I'm a bit puzzled that I seem to need the full 45V to get decent performance. At one point I deliberately lowered it to 36V (by using three 23A batteries in series instead of 4) and the receiver stopped working altogether -- no oscillation, at any point on the screen regeneration control.

It's a bit frustrating, in the sense that several of the design guides I've consulted state that you'll get better performance overall if you drive the detector tube with a lower plate voltage -- this is supposed to have been one of the big regen-radio design tips in the 1920s. For a while I thought my failure with low plate voltages was because of the "modest performance" of the 1AD4 compared to cathode-type tubes, but then I realized that the tubes used in the 1920s were a lot wimpier than my 1AD4s.

73 de Martin, KB1WSY
Logged
KB1GMX
Member

Posts: 1824




Ignore
« Reply #50 on: May 09, 2015, 05:11:42 PM »

Q1, suppressor grid.  In the 1ad4 it was designed as a hearing aid tube and its a pentode.
The grid give more power gain, also better isolation from control grid (G1) to plate.

Q2 voltage.  Higher is not an issue the tube will accept it.

     Note if you having troubles at 42V and below that means you need a turn or two more on the tickler.

Q3 threshold for shock is in that region for most (varies with skin resistance and resulting current flow).
      above about 60V start to be careful.  Also with bigger batteries the total current possible can be
      serious.

NA4 (not asked) lower voltages are possible and some tubes work better than others down there.
         Obviously you need more feedback and many users use a large inductance audio choke for
         the plate resistor in your version as it has less DC resistance and high impedance at audio.
         A typical choke might be several dozen Henries.  You may try a tube audio output from a AA5
        (primary side is usually 5Kohms at audio), others have used ignition coils, pulls from old radios and
        small (physical) 220 to 6.3v transformers. I have a 50 henry choke from a seriously messed up
        BC221 and that works well.    In the old tube days this was called impedance coupling and it
        relied on the high AC impedance.


Allison

Logged
KB1GMX
Member

Posts: 1824




Ignore
« Reply #51 on: May 09, 2015, 05:17:03 PM »

Both lexan and mica filled are not lossy.  I've swept them on my 4191A to 1ghz
and they are pretty decent as everything is lossy there so less so is easy to see.

For HF even PVC (the white stuff) is fine and easily found at home despot.
The thinner wall CPVC is slightly lossier but the thinner wall makes up for that.
By lossier at UHF where I can see it to a fine degree. At HF it would be hard
to measure.

If you want to go to extremes more air the better but mechanical support is a must.
Look at spider wound and basket weave.

Allison
Logged
KB1WSY
Member

Posts: 1309




Ignore
« Reply #52 on: May 09, 2015, 08:50:09 PM »

NA4 (not asked) lower voltages are possible and some tubes work better than others down there.
Obviously you need more feedback and many users use a large inductance audio choke for the plate resistor in your version as it has less DC resistance and high impedance at audio. A typical choke might be several dozen Henries.  You may try a tube audio output from a AA5 (primary side is usually 5Kohms at audio), others have used ignition coils, pulls from old radios and small (physical) 220 to 6.3v transformers. I have a 50 henry choke from a seriously messed up BC221 and that works well. In the old tube days this was called impedance coupling and it relied on the high AC impedance.

Interesting, thanks ... I'll see what I can find in the junkbox. I know that I have an old filament transformer somewhere; it's not particularly small but it could be used experimentally. In my local electronics store today I saw a small Hammond tube output transformer with a 10KΩ primary and I almost bought it so I could play around with it as a plate load and/or use it as an output transformer for a future power audio stage. It was a low-current transformer, about 3/4" wide. However I choked on the price ($18).

Just to be sure I understand, I believe you are talking about the plate load in the *detector* stage, right? I can see how that would increase the feedback. (As you know, I'm already using a 7H choke as the load in the *audio* stage.)

The receiver is working beautifully at the moment ... had a lot of fun monitoring the band tonight and heard lots of DX. So this idea of lowering the B+ is more in the nature of an experiment, rather than a "necessary fix." The builders from Days of Yore seemed to believe that running lower plate voltages (on the detector) was advantageous in a regen.

Both lexan and mica filled are not lossy.  I've swept them on my 4191A to 1ghz and they are pretty decent as everything is lossy there so less so is easy to see.

To start with I'll probably go with the 1" Lexan pipe.

Here's a view of the receiver and my patented "helter skelter" experimental construction technique. At the back is the "power supply": left to right, "C" batteries (about 5 volts negative bias from four NiMh AA-cells, which I will be using to bias the RF stage); "B" batteries (the empty space is where I'll put the 9V rechargeables, replacing the four tiny 23A batteries); and the "A" battery, which is a rechargeable NiMh D-cell. The power supply wouldn't be out of place in the 1920s, apart from the advances in battery technology and the ubiquitous plastic replacing bakelite and metal. The new regeneration "throttle" capacitor is just in front of the power supply. The screen regeneration pot is in the lower-center of the picture. The empty space to the left of the coil is where I'll put the RF stage.



73 de Martin, KB1WSY
« Last Edit: May 09, 2015, 09:00:07 PM by KB1WSY » Logged
KB1GMX
Member

Posts: 1824




Ignore
« Reply #53 on: May 10, 2015, 06:35:39 AM »

Is the large inductor for the plate load for the detector.  I have the latest schematic you posted.


Allison
Logged
N3QE
Member

Posts: 5593




Ignore
« Reply #54 on: May 10, 2015, 08:24:04 AM »

The suggested circuit is a no-no. It won't put bias on the tube because of the DC path to ground through the coil. It also shunts the signal path.

No, it's not obviously wrong. 1AD4 grid bias at 45V B+ is indeed often simply stated to be 0V.

In most every "real world" 1.5V filament tube battery radio, the grids are shown at DC ground in the schematic either by inductor or resistor. But usually neither or at most one of the filament ends is at ground. It often takes some inspection of the filament wiring to find out that the filaments form a circuit in themselves with voltage drops along the chain and sometimes resistor networks to get correct bias at each stage. Usually in multi-tube radios there are RF and AF bypasses for the filaments too (although in some cases they do not bypass the filament at some stages specifically to get some specific degeneration relative to the next stage - now that's complicated but they saved one capacitor in the design!)

This is a "feature" of many 1.5V directly heated battery-supply tubes. Correct grid bias relative to the heater on the spec sheet, is often 0V or or a multiple of -1.2/-1.5V for series connected filament chains.
Logged
KB1WSY
Member

Posts: 1309




Ignore
« Reply #55 on: May 10, 2015, 09:37:41 AM »

Is the large inductor for the plate load for the detector.  I have the latest schematic you posted.
Allison

Not sure whether you are asking a question ... the only large inductor in the current set is the 7H choke on the plate in the audio stage.

73 de Martin, KB1WSY
Logged
G3RZP
Member

Posts: 1315




Ignore
« Reply #56 on: May 10, 2015, 02:18:23 PM »

If you don't put a negative bias on the grid, then you have  very limited dynamic range on strong signals, because you really don't want the grid driven positive  - positive grid means grid current means distortion and likely gain compression - and you can't get an RF gain control function.. 

Getting round it with a grid leak leads to detection in the RF stage, lower selectivity and more cross-modulation.

RF gain control by varying RF stage screen volts with zero bias still leaves the grid going positive with all that entails....
Logged
KB1WSY
Member

Posts: 1309




Ignore
« Reply #57 on: May 10, 2015, 03:18:31 PM »

If you don't put a negative bias on the grid, then you have  very limited dynamic range on strong signals, because you really don't want the grid driven positive  - positive grid means grid current means distortion and likely gain compression - and you can't get an RF gain control function..  

Getting round it with a grid leak leads to detection in the RF stage, lower selectivity and more cross-modulation.

RF gain control by varying RF stage screen volts with zero bias still leaves the grid going positive with all that entails....

That indeed is the message I got from the previous discussion about the RF stage (not in detail, but the outlines).

If you go back to the previously linked data sheet for the 1AD4 (http://tinyurl.com/pm4pn56):

Voltage on Grid 1 for S = 10 microAmps/Volt: Typical is -3.8V. Limit value, -7V.

Yes, it seems to be true that many of these "battery" tubes operate with zero-bias, or very low bias, in a number of situations.

However -- and maybe I got the wrong end of the stick -- but what Peter seemed to be saying back then is that if you want tube-level control over RF gain, you'll need to bias the tube with an explicit supply. I am guessing that, otherwise, you are limited to "passive" RF gain control methods such as resistors or capacitors in the antenna circuit. Plus, Peter's point about dynamic range (makes sense: restricted operating voltage = less dynamic range).

Meanwhile I have a question for Peter and others: Now that I have added a -5V battery source for bias, would anything be gained by applying that bias to other stages (not just the RF gain stage)? I assume we should leave the finely-balanced grid-leak detector alone, but what about the audio stage(s)? I assume adding additional bias would reduce the gain but increase the controllability of those stages -- and that might be a good idea in a regen, for stability, even if I have to add additional audio stages.

73 de Martin, KB1WSY
« Last Edit: May 10, 2015, 04:34:38 PM by KB1WSY » Logged
G3RZP
Member

Posts: 1315




Ignore
« Reply #58 on: May 11, 2015, 02:03:38 AM »

You may well find less overload problems with some bias on the AF amplifier grid. If you get distortion on strong signals, it may either be RF overload of the detector or overload of the AF amplifier if the grid starts going into grid current on positive peaks. A small amount of grid current can be acceptable, but it means that the input impedance starts to drop on signal peaks and that means the load on the previous stages is changing - which can reflect back there to give distortion. Which is why  (such as in modulators for AM transmitters) when Class AB2 and B audio amplifiers were used a lot, it was common to use a step down transformer to drive the grids and negative feedback on the driver stage to give a low output impedance to the push-pull stage.
Logged
KB1WSY
Member

Posts: 1309




Ignore
« Reply #59 on: May 11, 2015, 05:15:26 PM »

Incidentally, what material is the coil former? Even at 7MHz, some plastics aren't very good, especially coloured ones. A quick and nasty check is to put a glass of water and the coil former in the microwave oven and zap it for 30 seconds or a minute or so to see if it gets warm. If it does, that tells you that it's lossy at 2.4GHz: if it stays cold, you can be pretty certain it's not lossy at 7MHz.

So, today I finally found a second pill bottle with the same form factor as the first one. I was visiting some relatives in a retirement home and it only took 30 seconds to find one in a dustbin! (Removing the pharmacy instruction labels took a lot longer, but never mind.)

I stuck it in the microwave with a glass of water (pictured below) and ran it on High Power for one minute. The water was very hot, but the pill bottle was still room temperature, exactly the same as when I put it in.



So I'll be using the new pill bottle for the experimental RF stage. I like these particular pill bottles because they are exactly 1.25", the same as my collection of virgin Amphenol forms. I figure I'll use the pill bottles to mess around with, and then use the Amphenol forms for the "final version" of my receiver, transferred from the breadboard to "proper construction."

73 de Martin, KB1WSY
« Last Edit: May 11, 2015, 05:24:54 PM by KB1WSY » Logged
Pages: Prev 1 2 3 [4] 5 6 7 8 9 ... 18 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!