K8TND’s Take on the Kitchin Regen
Some of the most popular feedback I have had on this blog revolves around the Kitchen regenerative receiver project that I took on initially back in Feburary of 2012. From time to time I still receive interesting e-mail messages and comments about other people’s designs and this one is of no exception!
Cliff Donley K8TND sent me a very interesting e-mail message about his take on the Kitchin regenerative receiver. He has made several modifications and also has some pointers to share with us. First up the schematic that Cliff sent me is the latest schematic of the Kitchin receiver that he could find: (you can click on the image below to open just the schematic in another window for printing)
Cliff told me that he has some pointers on how to eliminate the hand capacitance issues that plague this type of receiver:
“Regarding hand capacitance issues, with the tank coil being wound on a
toroidal core, that helps tremendously on my version and the
varactor/diode and potentiometers totally eliminate it.”
And here is where he gave me a tip about grounding the rotor of the capacitor to reduce the hand capacitance:
“On the older regens I’ve built, using traditional variable
capacitors, I have eliminated MOST of the hand capacitance problems by
grounding the rotor of the capacitor and connecting the stator to the
hot side of the tank circuit and then of course enclosing the circuit
board in a metal case.”
I am amazed how far this circuit design has been modified and taken to yet the next level. This is truly “open source” before that term was original coined, and it is still going strong today!
Here is a little slide show that shows Cliff’s regen that he build with the modifications: (Nice blue LED indicator light Cliff! – You can click on the slideshow to stop it to examine the photos in closer detail if you wish)
And one last tidbit of wisdom from K8TND about air variable capacitors:
“BTW Nick, I was also going to mention an old trick from way back in the
60s. If you have a variable capacitor that’s too high in capacitance and
you want to reduce it, simply put the capacitor in a vise, take a sturdy
pair of long nose pliers, grasp ONE plate at a time, starting from the
outside and YANK it out of the ROTOR. Do NOT try to remove plates from
This is a great little trick as finding the proper air variable capacitors is getting increasingly difficult and when you do find them they are relatively expensive.
A huge thanks to Cliff Donley K8TND, and thanks for sharing! I know that others will benefit from your radio wisdom.
73! de Nick N9SJA
Thanks Nick. Here’s some additional info that I think is important regarding this regen I built. This circuit was originally designed to run at 6 volts DC. Seeing this circuit is now running on 9 volts, I have been experimenting with a few component changes to match the new voltage. All changes have been reflected in the current schematic except ONE, and that’s the Zener diode. I have been experimenting with using a 1N4737 which is a 7.5 volt Zener and I have found it definitely provides clearer CW and SSB, in that it eliminates a tiny bit of warble. I have made this change permanent in my regen now. Remove 1N4736, insert 1N4737.
Another important thing I want to comment on it the RF choke. RF chokes are VERY expensive now and that’s if you can even find one. I tried to find some info on the web for winding my OWN RF chokes on toroidal cores. I couldn’t find a thing on it, so I had to do my own research. I have been using cheap green Chinese cores I get off Ebay. You can try other cores, but most of them won’t provide the amount of inductance you need to reach 2.5 mH.
I have used a gray core and a natural ferrite colored core, but I run a freq test on them before I call them finished. To test a core out, I quickly wind 34 turns of #26 wire on it and then measure it with my inductance meter. It SHOULD read somewhere near 2.5 mH. If it reads in the uH range, then it’s not the right kind of core for an RF choke. Once it checks out around 2.5 mH, I do a frequency run on it with my RF generator and my scope. Put the choke in series with one leg of the circuit to the scope probe and then sweep it with the RF generator through the range that you want to attenuate. You should see serious attenuation of this range compared to the raw output of the RF generator. If you DON’T see proper attenuation of the signal, then you either need a different core OR more or less wire on the core. It’s a learning curve, but you’ll get the hang of it and be able to make your own effective and cheap RF chokes.
A couple other things I’d like to mention. I’ve understand there’s some concern about the high cost of acquiring electronic parts these days. Personally, I use Tayda electronics. I have no stock or interest in their company, I just found that they have a pretty good stock of parts at an extremely good price. You can find their web site here:
They ship from Colorado, so you parts will arrive quite quickly.. !!
Now, another thing I’d like to share with you fellow Regen builders is some of my findings on Varactors.
After doing some research on Varactor diodes, I found out you don’t really even NEED a Varactor diode. In one of my Regen receivers, I’m using a RED LED as my fine tuning Varactor. Red LEDs, large LEDs, Schottky diodes, rectifier diodes and even FETs, will serve well as a Varactor IF you reverse bias them.
I’m not going to go into all the technical details on these devices, but if you do a Google search on using them as Varactors, you will find LOTS of reading material and charts. After looking at the charts, I decided that the 1N5822 would be a good substitute for a Varactor diode. I ordered some off Ebay and they have worked out very well for me. If you look on the internet, there are a number of sites that describe how to employ Varactor diodes in your circuits. There are also testing circuits you can use in your shack to test out various LEDs and diodes in your junk box to see if any of them will work as Varactors. I usually just do a quick hookup across my capacitance meter to see how much capacitance they have. If it’s over 70 pf, then I put it in a little zip-lock bag for future Varactor use. Oh, by the way, Radio Shack has an AM radio Varactor they sell for 3 bucks, but you must order it from their web site, get them on Ebay.
Note: When you are building your Regen, it’s not necessary to be neat. Most of my stuff I build “ugly” construction and it generally works just fine. I’ve tried both “Manhattan” and “Ugly” styles of construction and also using 10 Megohm resistors or capacitors for standoffs. All work just fine, so it’s up to your personal preference.
BTW, when you order copper clad board, ask the seller if they will include some SCRAPS for FREE in the package. You can take small scrap strips of clad and cut it up with side-cutters to make little solder islands that you can glue on your chassis with Super Glue.
One last thought for this post: A good way to cut your copper clad board is to use your wife’s paper cutter. Just don’t tell her about it..!! 🙂
I’ll have to employ these ideas on my regen…whenever I get to that project. Good stuff.
In case you wonder why my receiver looks so hacked up and tacked together is because it’s an ongoing experiment. I’m continuously trying new things to make it better. I have been recently trying to improve the stability of it and have found that the .001 capacitor that feeds the tuning network to the hot side of the main tank coil, is EXTREMELY sensitive to temperature changes. I use a SODA STRAW to blow air on different components and check their ability to change frequency of the receiver. I have tried several different types of capacitors in place of that .001 and the most stable I have found so far is a tiny multilayer ceramic cap. They are yellow and about as big as a small pea. It still drifts when I blow on it, but WAY less than anything else. Stay tuned, I’m going to try the Radio Shack Varactor diode today.
I got the new Varactor installed in the regen and I must say it’s AMAZING. I’m getting coverage from 4.5 to 17 Mhz now. If you want to order these, the cheapest place to get them is Radio Shack and they sell them on Ebay, NOT in their stores. The part number is NTE618. When I saw that these were made for AM radios, I figured they’d have tons of capacitance and a low minimum capacitance.
This Varactor is a keeper and is staying in the radio. I’m going to add a couple turns to the main tank coil to bring the range down to 3.5 Mhz. Remove !N5822, install NTE618 in it’s place.
Happy Thanksgiving to all. Here’s a couple of my most recent changes to this regen machine.
The capacitor that links the tuning network to the top of the tank coil is shown as a .001.
Change this component to a 620pf silver mica. The capacitor that links the regen potentiometer
to the tickler coil is shown as 430pf. Change this capacitor to a 470pf silver mica. These two capacitors are extremely important to the stability and smoothness of operation of this radio. I changed these values
to accommodate the new Radio Shack Varactor diode. Everything is working VERY well now, I only hope
someone will build one of these fantastic radios.
Very nice Cliff. I have updated the schematic on the main page to be the latest revision (11/9/2012). I just removed the other schematic and put the new one there. I will have to find time to build one of these radios using the Varactor diode. I just have so much going on at the moment, it is hard for me to find time to build a project. Anyone else out there, if you build one, let me know. I would love to post your build of this radio also!
73! de Nick N9SJA
PLEASE READ before construction of this receiver.
This receiver design using the solid state tuning/Varactor diode has three minor issues. One is there’s a longer period to warm up and settle down to where it doesn’t drift. We’re talking 15 minutes or slightly longer. Then there’s a slight wandering back and forth in frequency when copying CW or SSB. Also you’ll notice a slight warble to the CW signals when you tune them to a low audio frequency. These issues are ALL caused by the Varactor diode tuning network. I have an identical receiver I built earlier that uses the same circuit only with conventional variable capacitors and it is drift free after 5 minutes and the CW tones are clear as a bell, even down to the lowest audio freq. The problem is NOT caused by the power supply as I’m using the same professional HP regulated supply on both of the receivers. I’m currently working full time to see if I can alleviate these problems. Any help or suggestions would be highly appreciated.
I have made a little progress on the stability and clarity of CW tone problem. Add a 1 uF multilayer ceramic capacitor between the wiper on the coarse tuning potentiometer and ground. Do the same thing with the fine tuning potentiometer. I’ve also noticed if I turn the RF gain down, it will stabilize the signal considerably..!!
I have tamed the stability problem by adding an additional Zener regulator pad close to the tuning pots. I also changed the value on the original Zener resistor on the regulator for the regen section. The receiver takes about 15 minutes to warm up and stop drifting, which is normal for these beasts. I’m very pleased with it’s performance and I think you will be too. I’m now going to build a finished “neat” version in a metal case. I’ve given the final schematic and photos to Nick. He will post them when he gets time.
I just had a minor revelation this morning. While listening to some CW, I noticed some instability after I had moved the main tuning pot a touch. Come to find out the 10K main tuning pot had some soft spots on it that would cause instability if you landed on one of them. I pulled it out and put a nice 10 turn Amphenol wire wound potentiometer in there and it makes a world of difference. It’s also very nice to have the main band tuning stretched out some more.
I would recommend either a good quality name brand OR wire wound pot for the main tuning component.
Through experimentation, I have found that the major cause of instability when listening to CW or SSB to be caused by drafts and air currents in the room. The Varactor diode is extremely temperature sensitive, so this receiver needs to be housed in a metal box of some sort. The metal will also shield the Varactor from being modulated by AC house current and other random RF radiation. The metal box should also have a good earth ground if possible. The receiver works very well and I sit down for a listening session with it every morning while I’m drinking my coffee.
I finally received the aluminum box for this Regen and got it installed yesterday. I must say it has made the receiver quite stable now. I sent a photo of the receiver in the new case to Nick and hopefully he will have time to post the updated schematic and final photos of the Regen soon..!!!
Pingback: K8TND’s Take on the Kitchin Regen: Part 2 « N9SJA
Thanks for your detailed reporting on using a varactor with the Kitchin regen design. It’s something I’ve been considering myself. I have a couple of comments and questions.
First, you mentioned that reducing RF gain increases stability with the varactor tuning. This is likely a sign that the signal voltage is large enough to modulate the varactor voltage, causing detuning. The solution is to use two back-to-back varactors, as described here: http://www.radio-electronics.com/info/data/semicond/varactor-varicap-diodes/circuits.php
Next, you can also use a varactor to control regeneration. Just use the varactor in place of the throttle cap and reverse bias it with a pot. This has the advantage (over you r scheme of a pot in series with the throttle capacitor) that the pot will carry no RF. In your scheme, your regeneration control pot will have RF on it – do you not notice any hand capacitance effects when adjusting regeneration?
Now, some questions:
Stability. Does your latest build, with all of the improvements, still take about 15 minutes to warm up? How much is the drift (and at what operating frequency) before and after warm-up?
Selectivity. I’ve read two separate articles that state that the low Q of varactors will degrade regenerative receiver selectivity, particularly at upper HF. Do you notice this effect?
Tuning resolution. I have a ten-turn wirewound pot connected to a poor man’s varactor (1N400x diode, as in the original Kitchin design). I notice that the pot does not tune continuously, but tunes in discrete steps, say 50Hz steps at 7 MHz. I can notice this when tuning slowly over CW signals. With your ten-turn pot controlling the very wide-range (4.5-17 MHz) main tuning, I imagine your main tuning is not continuous but jumps in steps of maybe a few hundred Hz. Is that right?
Thanks again for your detailed reporting.