2018-01-15 2nd FM Microphone Kit Assembled

I ordered some other FM Microphone kits, and assembled one last week. This is a better one, labeled EEQKIT RF-02FM VER170525. It has three transistors, one of which is a microphone preamp and it should perform much better. The kit also has a jack that will allow you to plug in a stereo signal from a CD or other audio source and broadcast a mono signal. The link and schematic are below.

I assembled it with no problems, the parts were labeled on the PC board. The “7 turn” is the coil with the plastic and the tuning slug. One thing I like is they used a 10k for the LED current limiting resistor so the LED draws only a fraction of a milliamp, which saves the battery. The red LED is still bright enough.

I powered it with a 6 volt Ni-MH battery pack, the board is marked DC 3 – 6V. I monitored its output with an FM radio. I tried to find my ‘station’ on the FM band but I didn’t hear anything. I got my Gooit frequency counter and put it next to the antenna and it was saying 320 MHz! That’s way too high! So I found a tiny screwdriver and gently and slowly turned the black tuning slug in until I got the frequency down to the bottom of the FM band around 90 MHz.

I tuned in the radio and blew into the microphone but the audio was very weak. That should not happen because the kit has a microphone preamp. So I measured the collector of Q1 and it was less than 0.8 volts. It should be about half the supply voltage or about 3 volts. I checked the 9014 datasheet and saw that it is a high gain preamplifier transistor. The schematic shows that its base is biased by R4, which was 100k. Most low level preamps use a resistor much higher, from 470k to 1 megohm or more. So I replaced the 100k with a 470k and the collector measured about 1.8 volts, which is much closer to half the supply voltage. I powered it up and the microphone was much more sensitive, I don’t have to yell in it to be heard on the radio. So the conclusion is that the kit should have a much higher value for R4, 470k should be okay.

A link to the kit instructions (.PDF)


2018-01-11 The 1N60P “Germanium” Diode Issue

From FB group

Markku Vuorensivu
I bought some 1N60P “germanium” diodes and I found they are not germanium, they are Schottky silicon diodes. They do not work like germanium diodes in circuits that depend on the leakage that germanium diodes have. The Schottky or germanium diodes can be checked for reverse leakage and germanium will have less than 1 hundredth of the microamps of leakage that a germanium diode has. I suggest that anyone who recommends these also reminds readers about this issue.

Here is a video on YouTube about how to tell one from the other.


2019-01-10 The Ninth Anniversary Of The Philips LED Candelabra Lamp

On Jan 11, 2010 I connected a Philips candelabra lamp to the AC line and turned it on, and for the last nine years it has been running continually, except for a few days time when the power was out, which might add up to a few days, probably less than a week. I did this test because there were several companies giving LED lights a bad name by selling lights that would go dim after less than a thousand hours. I was sure that Philips was not one of those companies so I used a Philips light for the test. Philips LED lights, and any LED lights were expensive in 2010. The prices for 60 watt equivalent LED lights were $20 or more. The prices have come down dramatically since then, and the LED lights have contributed substantially to the reduction of electricity usage.

So 24 hours times 365 times 9 years gives 78,840 hours, and to cover the days of power outages I’ll round it down to 78,600 hours. Most LED lights say they will last 25 thousand hours but some LED makers said the LEDs themselves should last for 100 thousand hours. My LED light is still bright after 3 times its 25,000 hour lifetime. The light has gone more than 3/4 of the way to the hundred thousand hour mark.

In 2010 I didn’t have a luxmeter to measure the light’s output, so I tried to measure it with an old GE light meter for photography. But it didn’t give an accurate reading and it doesn’t work anymore. I can only give a guesstimate by eyeball. There has been some drop in the light output but it still looks bright.


2018-01-09 Bootstrap Principle In Audio Amps

FB group Building transistor radios 2018-01-09

Willie Barnett
That is one deficiency of your circuit: the output transistors do not have the ‘bootstrapped’ bias resistor.

I’ll explain why it’s needed. In your circuit, the signal turns the driver transistor as close to off as possible. The current through the base bias resistor causes the top output transistor to turn on. But the transistor’s base and emitter voltages go higher, and as those voltages approach the + supply V, the base bias resistor has less and less voltage, so it can’t supply the needed current. The transistor becomes current starved. So in your circuit the upper output transistor can’t drive the speaker close to the + supply V.

To solve this problem the spkr is connected to positive and as the voltage across the spkr goes positive it adds to the + supply V. The base bias resistor gets its current from the capacitor end of the spkr, and the bias resistor always has voltage across it, so there is no problem with current starvation.

Another way to solve this problem is to split the base bias resistor and connect the split point through a capacitor to the output. In this case the spkr can remain connected to ground.

The 4.7k can be lowered or made zero. But the circuit should have a small capacitor from collector to base of the driver transistor to prevent it from oscillating. But then maybe you might want it to oscillate, so it can drive a 50 ohm antenna at a ham frequency??


2019-01-08 99 Cent Wireless Microphone Kit Assembly

2018-01-08 also FB group Building Transistor Radios

Before the Xmas holidays I ordered a few of these at 99 cents apiece, and they gave a delivery date of late Jan to early Feb, so I was pleasantly surprised to get them early.

There are many of these single transistor microphones around, either as schematics or as kits. They have low audio sensitivity because the microphone puts out only millivolts and that’s not enough to fully modulate the RF oscillator. Blowing in the microphone is easily heard, but for your voice to be heard you have to talk loudly very close to the microphone.

This kit comes with a coin cell holder that will take a CR2032 3 volt lithium cell – it’s very compact, but the lifetime is not very long. I didn’t have a coin cell so I soldered a two AAA cell holder to the kit to try it out. All of the resistors are 1 percent, which is better, but the color bands are harder to read than 5% resistors. All of the capacitors are very low cost ceramic disks, and do not give the circuit good frequency stability.

This kit came with the hole in the PC board for a short antenna, but there was no wire for it, they assume you have a short piece of wire available – short meaning a foot or 300 mm. But when the antenna is used the frequency will change when the antenna is brought near other objects. Without the antenna the signal can be picked up a few tens of meters.

The coil comes already made, and measured 100 nanohenrys, which is typical for the low end of the FM band. The turns of the coil can be spread to increase the frequency, because the 30 pF tuning capacitor is fixed and can’t change the frequency. But this 30 pF value is on the very high end of the typical values found for the low end of the FM broadcast band. Also C5, the capacitor connected across the emitter and collector, is 10 pF, which is double the values typically used. In addition there is a capacitor C6 across the emitter resistor is usually not used, but in this case is 30 pF. All this adds up to more capacitance for this circuit and a lower frequency.

So after assembling the circuit I powered it up and measured the frequency with my Gooit frequency counter, and I got 78 MHz, more than 10 MHz too low. So I removed the 10 pF C5 and put a 5.6 pF in its place. I checked the frequency and it was still below 80 MHz. So I removed the 30 pF C6 and put the 10 pF in its place. The frequency then measured 81 MHz, still way too low. So I had to lower the 30 pF tuning capacitor C4 and put a 20 pF capacitor in its place. Now I measured 91 MHz, just about right. I can put a 0.5 to 5 pF variable capacitor across this 20 pF and move the frequency down to the 88 to 90 MHz part of the band.

The circuit needs another transistor to amplify the microphone so the listener can hear what is being said in the room. I will have to see if I can add a transistor and a few parts to make it more sensitive. As it is now it’s more of a kid’s toy, not really useful for listening.


2019-01-07 TRF Receiver Experiment

2018-01-04 also FB group Building Transistor Radios

I thought that the ‘tinkerers’ who were tinkering with the loop antennas may have been getting a stronger signal because the loop antennas have a large area to capture the RF signals, so I decided to make a single channel AM Band receiver, and I chose KNX 1070 all news all the time as the station.

I wound sixteen turns of 24 AWG PVC insulated telephone wire around a small cardboard box, and taped the wire in place. I measured the inductance at 128 microhenrys, and used the resonance tool in the free app ElectroDroid to calculate how much capacitance the coil would need. I connected four 470 pF 1% silver mica capacitors in series- parallel to get that required capacitance and I connected the capacitors to the loop.

I added another four turns to the coil to tap off the signal to go to the base of the RF preamplifier. The schematic is in the photo. The supply is two 1.25V Ni-MH cells connected in series. I plan to connect a 4V PV solar cell to charge them.

The preamp is okay, and so is the audio amp stage after the detector diode. But I was getting poor sensitivity using a “1N34” diode from Asia, which tests as a Schottky diode, not germanium – it doesn’t have the leakage a germanium diode has. The volume was low, so I experimented with that detector. When I soldered a real germanium diode across that diode, the volume got much louder. It is performing up to my expectations now. I get a bit of harshness in the audio so I may add more capacitance after the diodes to filter out more of the high frequencies and not sound so harsh. The schematic shows the 5.1k connected to +2.5V but it is actually connected to ground.


2019-01-05 Substituting Chokes For A Coil In A Reflex Circuit

From FB group Building Transistor Radios 2018-01-15

Nick wants to use two 10 mH chokes close to each other in place of a coil that’s used in the typical reflex circuit. I’ll include a schematic of the circuit (from somewhere on the net). This is labeled L2.

(begin my comment)

Nick Bastian
The coupling is not as good as if they were wound on the same form. What I see is the secondary is putting out RF for the detector diode to rectify and turn into audio. The secondary has to put out enough voltage to forward bias the diode and give a reasonably strong audio signal. Hopefully two chokes close together will give a signal that’s strong enough.

But this coil is needed because it’s a reflex circuit and the RF has to be separated from the audio. If the circuit is changed so it’s not a reflex, then the coil would not be needed. The coupling could be done with resistors and capacitors. 👍
The cost of an additional transistor and a few resistors and capacitors is like ten cents or so. And it saves more than a dollar for the coil. 👍👍

Another way to do this is to use a ‘link coupling.’ You wind a few turns of wire around each inductor and connetct them with a twisted pair. Use the same number of turns for both windings.

(End my comment)


2019-01-01 Loopstick Of Ferrite Suppressor Sleeves

Also to FB groups Building Transistor Radios and Ferrite Rod Antenna Experimenters

I got a bunch of ferrite RFI suppressor sleeves from I found a wood dowel the same size as the hole in the sleeves, and I cut it slightly shorter than the length of 5 sleeves. I drilled 1/8 inch holes in the ends and screwed #6 sheet metal screws into each end. After getting the threads formed, I screwed a rubber faucet washer onto one end, slid all five of the sleeves on, then screwed another faucet washer on the remaining end to keep all 5 snug and tight. I’m not concerned about the steel screws, I think they have very little influence on the magnetic field. But if you think it might make a difference, the sleeves could be glued on instead.

I cut off 8 feet (2.4 meters) of 24 AWG solid conductor PVC insulated telephone wire and wound it onto the sleeves, and taped it with electrical tape. I measured the inductance with my LC meter, and it measured 172 microhenrys. I wanted about 230 microhenrys, so I think that if the wire was longer, about 10 feet or three meters long, it would have been about that much inductance.

I plan on using a two section air variable capacitor with one section of 365 pF. The other section could be connected in parallel and with them in parallel with the 172 uH coil it would cover the whole AM band. I realize I could have chosen better wire, Litz wire would be best. But I don’t have any that’s heavy enough for this coil. I suppose I could use 24 AWG enameled magnet wire, and the turns would be closer together giving more inductance, but I don’t think it would be much better Q than the telephone wire. I have also thought about making some wire similar to Litz wire. I would cut four lengths of thin wire, such as 32 AWG, and twist them together enough to wind onto the core. The wire wouldn’t be as many stranded as Litz wire but it would be better than solid wire.

Update – I connected in parallel a dual section variable capacitor, both 365 pF and the oscillator sections in parallel. I can hold the coil on the AM radio next to the internal ferrite loopstick, and when I tune the variable capacitor connected to it, I can receive the station louder, with better signal strength.


2018-12-31 Measure Negative Voltage With Arduino

from FB group Arduino Projects 2018-12-31

Visualize this. You have a resistor, say 10k, connected to your 10 volt negative source. You have an opamp which has an output that can swing up to positive 10 volts. You connect this output to another 10k resistor, and connect the remaining end of that resistor to the first resistor. The opamp’s output is at +10V and the -10V cancel out so the connection of the two resistors is at 0 volts. You connect the opamp’s + input to 0 volt ground, and you connect the opamp’s – input through a resistor to the connection of the two resistors.

If the negative voltage changes, the output of the opamp will change to keep the connection of the two resistors at zero volts. Then you can connect the opamp output to your analog input through a 2 to 1 voltage divider to keep the analog input voltage from 0 to 5 volts.


2018-12-31 Zenith Royal 500 Unique Issue

Also to FB group Building Transistor Radios 2018-12-31

I have several Zenith Royal 500 radios with the ‘long range’ eighth transistor added as a RF preamplifier (chassis number begins with 8). See the attached photo. This does not require a third section on the tuning capacitor, as some ‘TRF’ radios have. Zenith added a tuned circuit between this and the next mixer stage. But this tuned circuit was made very broadband so it covered the whole AM band. The broadbanding was done by putting a low value resistor across the capacitor and inductor and that changed the normal peak of the tuned circuit into a flat table with sloping ends. In this schematic it is R1, in parallel with L2.

The Zenith and most other radios of the ’50s and ’60s used carbon composition resistors, which are prone to drifting out of tolerance, typically going higher in value by more than 20 percent. As the value goes up the circuit becomes less functional, until it doesn’t work properly or not at all. The resistor across the tuned circuit might be 2200 ohms, but it drifts higher so it might be 3300 ohms or even higher. The tuned circuit flatness changes to a more rounded peak, giving good reception in the middle of the band but poorer reception at the ends of the band. The resistor can’t be measured in circuit because the coil is a very low resistance in parallel. It must be removed to measure its resistance. If it has been removed, it is very easy to replace it with a carbon film resistor which will not have the drift problem.

An alternate way of checking the performance of the tuned circuit is to check it with a sweep generator, but I don’t know if there are sweep generators broad enough to cover the whole AM broadcast band.

If there are a lot of radio stations in the area and they are strong in the middle but weak at the ends of the band, it might give an indication that this is the problem. But it isn’t certain because other factors could cause a similar problem. In fact, the tech may be trying to align the radio but have the problem with weak performance at the ends of the band and believe it is being caused by poor tracking of the two sections of the tuning capacitor. This may cause the tech to spend a lot of time trying to get the two sections to track, when the problem isn’t there but instead the RF preamplifier’s broadband tuned circuit. Let’s hope this information saves the tech some time in the future.

© RustyBolt.Info/wordpress