Thursday, November 29, 2018

dual output high gain amp - continued

I posted this a couple post back. I had a chance to build one today. It is as good as it gets. I connected it to my power supply and listened with no input. Nothing, not even a hum. I put a test lead on the input and when I touched it I had hum. When I put the lead on the bench, nothing. I connected the antenna and had some hum. I put a coil across the input and nothing. When I get some more shop time it will go in a box with some jacks for the I/O.
I put my scope on the I/O but it was to low a signal to read on the Input when it started distorting the Output. Very sensitive and very stable makes this one a keeper.

Monday, November 26, 2018

filter response - with 4 position selector and tuning capacitor.

There was a question asked and this is a start point towards the answer.The question is what is this filters response at the different ranges.

So we have 4 bands and 4 responses to look at. The cap is variable and should cover the ranges shown on the drawing. Starting in position 4 we have 2 coils and one cap in the circuit.

This is the circuit with the cap a max.
This is the frequency response with the cap at max.
Reducing the cap to 200pf gives this response.  The schematic note says it tunes 16 - 54 MHz. A 300nh coil needs 329pf to tune 16 MHz and 29 pf to tune 54 MHz. And so it goes....
To see the response for position 3 move the switches to 3 and add the coils. You feed L4 to ground and L1,L2 and L3 in series to the capacitor..............

Dual output High Gain AF Amp

1 1 micro volt gives 3 mv to the ear bud or 30 mv to the head phone.
This will be tomorrows build if I get shop time. The sim was a little unstable. when I added the divider - filter on the input it became stable (in sim).
The 300 - 3k hertz frequency response is good. It rolls off well before the MW band so should be immune to the local boomer bleed thru.

Wednesday, November 14, 2018

DBM testing - observations

I put some jacks on the DBM I made a couple of days ago. I used shielded cable to feed the audio amp.
This was the audio amp I used for the test.
With the signal generator connected to the LO port and a 6' lead clipped to the AC register for antenna I turned it on and it hurt my ear. My generator was still set for the test I ran on the DBM. Even with no lead on the RF signal port it sang. I turned the RF signal off and heard voices. The 75 meter ham net was taking roll call. A little down the band I found a round table rag chew. The were discussing the parts and suppliers for a final amp build one was working on. And so went the rest of the evening. I did make a few adjustments and observations. The DBM sings with an LO signal of  0.1 volt to about 0.25 volt. Mini Circuits recommends 0.7 volt drive for their DBM which is the one we copy. I guess the core material makes the difference. Their goes up to 1 Ghz. My core is rated to 50 Mhz.
Anyway it is looking good and working well. The interference from the second generator makes me wonder how many times some one built a perfectly good DBM or NE602 circuit and then it failed because of stray signal bouncing around the room. I put RCA jacks on the DBM and used shielded cables with good results. So as a test I used jumpers to connect a DBM using the same core, AF amp and generator. It would buzz but no reception. I turned on the second signal with a 1Khz offset and it sang. The conclusion I drew was the DBM requires shielding between the ports to prevent overloading. In other words the LO signal is connected to one port by wire and the other port by radiated signal.
The NE602 having a amplification factor would make this more difficult to overcome. So shield those cables and keep the lead short. A good ground plane would help too.
I guess I will try a NE602 test next.

Tuesday, November 13, 2018

Z matching a crystal set

The short answer is the headphone Z needs to be equal to the diode Z plus the generator Z. You have little control over the headphone Z. (I have a high Z and a low Z headphone and high Z and low Z earbud.) Chances are the circuit will need to be adjusted to the phones. The diode Z can vary with diodes but here again choices are limited. Your main control in the situation is the coil taps. Tapping up or down to adjust the generator Z will give results.

 * After making this post I looked at the datasheet and found the Rds is 104KOhms. The the original poster told me the selectivity is very poor with the diodes. This tells me I went at it completely backwards. The diodes are loading the tank dropping its Q and spreading it bandwidth. The MOSFET reduces the loading and allows the tank to boost the signal. In the final analysis the tank load needs to be light to get the Q needed to boost the signal and this allows the headphone to draw more power without killing the signal. It seems a bit contradictory light load means more output. 100% of zero is zero but 10% of 1 micro watt is 100 pico watt. The old headsets worked quite well with less than 1 micro watt.

Monday, November 12, 2018

making a DBM

I put a DBM together last night and did some testing. The core is as the previous test indicated. 10 turns = 100 uh. My test indicated a 10 to one VFO to signal ratio to produce the best results. example 500 mv LO and 50 mv RF produces 10 mv IF. I plan to do another with more inductance and see how it effects the operation. I'm posting the construction.
I cut 3 wires to the same length and wrap them together.

The wrapped cores with 3 coils .
Separate the wires and prepare to clean the varnish from them.
The wire can be soldered without stripping the varnish. The varnish will burn off IF you get the wire hot. I put a drop of solder on the bare end and work down it. You can also use a small flame to burn the varnish off. An alcohol lamp is ideal for this but a torch work too. After heating removing the soot and ash with an alcohol wipe works well. I heard of a guy using a small cup of alcohol to dip the wire in. After setting it on fire he turned it over then things went bad. The wipes are safer.
Use the ohmmeter to find a coil, this will be the primary twist it to ID it. Find another coil. Connect one end of it to the other end of the remaining coil.
Make a pin board with the layout for a diode diamond and the I/O. A piece of perf board might be better if you have some.Run a ground lead from LO and RF inputs. Connect the primaries and center taps of the secondary.
The remaining wires go to opposite sides of the diamond. Now all that's left is mounting the diodes.
I use clip leads as heat sinks. Pair the head and tail of a pair of diodes and connect them to a pin. Pair the other diodes head to tail and connect them to the opposite corner.
Take a head and tail and attach them to a corner.
connect the remaining ends to the last pin.
The finished product. This works well for a test bed to try the cores. The final product will probably be put in a metal box. You can attach the cores to a ground plane by soldering the ground connections to it. Then use the 4 leads that connect to the diodes as stand offs to solder the diodes to.
I'm feeding it a 4001KHz @ 500mv and 4000KHz @ 54mv  signal and getting a 1KHz @ 5mv output. (best I can remember. I should have wrote it down. I took pictures of the setup but can't read them.) Anyway it works it is not sensitive enough. I will try more turns on the coils.

The 40/80 Dual gate first mixer - crystal filter - DBM detector

Dual Gate mixer - Xtal filter - Dual Gate IF
Dual Gate IF - Logic Gate Xtal controlled oscillator - Dual gate oscillator buffer. DBM detector.

The antenna fed a hi/lo bandpass filter. The VFO fed the first mixer which fed the Xtal oscillator. By switching the antenna selector the same VFO worked with 40 and 80 meters.

We used both the Dual Gate mixer and the DBM. The DBM requires a large signal so we used it after getting so signal gain in the amps.

NE-602A examples and a little data

 I like the DBM because I make the transformers and basically feel like it is a home brew. Take a look at the circuit.

Sunday, November 11, 2018

Another MOSFET amp

With a supply of 3 - 9 volts it will drive and earbud or headphone.  It could be a good starting point. I think a 2N7002 would work without any circuit changes?
I'm using the K669 because that is what I have on hand in leaded MOSFETs. My 7002's are surface mount, I may try one with them. A match box amp would be good to have around.

* you can connect the earbud / headphone directly to the source. I don't like putting DC thru mine and I have a bin full of the little transformers so I use them. You can wind your own. A 1:1 or 1:2 will work.

Saturday, November 10, 2018

NE602 update

This one has 2.1 MHz in and 100 KHz out.

DBM mixer

I'll let the sims do the talking. Here's the schematic. So shall it be a DBM or a JFET mixer?

JFET mixer

First a look at the schematic.
It is straight forward. Feed a signal to the gate another to the source and draw the output from the drain. Seems simple enough. The circuit is a bit to simple but we will see.
The display is the top of L1, L4 and L6. The inputs are 2.0 MHz and 2.1 MHz One of which is modulated with a 1 KHz tone.The green is the 100 KHz output. You can see the inputs are a higher frequency but the 1 KHz tone is lost because the sweep is to fast.
The input transformer is step up. I moved the probe to L2 and the input is a phasing is a little easier to see.As the red and blue go in phase the read peaks.
I changed the sweep to show the 1 KHz modulation better. As you can see the orange signal is constant in amplitude. The green is the modulated signal and the pink (difference signal) is carrying the signal too.
Next step will be to wind 3 cores and gather some test equipment and probes. First I should point out the mods I made to the original circuit. I did not put the resistor in the drain circuit. In the build I may use it. I put the bypass below the tank in the drain circuit. I wanted to bypass the tank with the high frequency signals. I used a different JFET simple because I don't have any MPF102.
Anyhoo, time for a build.

toroid frequency response and number of turns

In the short wave frequency range 1 turn can made a big difference.
The core material can make a big difference too.
The first choice to make is the core material?
The calculator give the frequency range for the material. The circuit under development is operating at 2 Mhz so the 77 material could be a good choice.
The next question is what will be the input Z. The rule of thumb is the primary should be 4 times the input Z. So with a 50 Ohm system we need a 200 Ohm primary. The next question is will we use a diode ring or feed a Jfet mixer?

Toroid core transformers - part 1

This calculator will provide the winding data for our transformers. It has the core data built in for many cores but in case your core is unknown it has a function for that too.

The calculator

The first step is to download and install the calculator. You can do the math if you like but I will use the calculator.

Since we don't always know the core type I will go thru the steps to determine the core parameter first.
(this requires an inductance meter or test procedure to determine the inductance.) The calculator also ID's cores by color code.

run the calculator and click tools.

Select the first option.
The program ask for turns, inductance and dimensions.
I made 4 turns and measured 17uh. The core is 0.5" so I look at T50 for the dimensions.
I transfer the data and the program gives me the AL and ui values.
The FT50-77 give me a fair match. The program displays 3 decimal places which could imply it is very accurate but a single turn will give a large step in inductance with this core. As you can see when I ask about 4 turns it gives 17.6uh. When I ask for a 17uh coil it answers 4 turns. There has to be some round off in the process. Wire spacing will cause some deviation in our results. Inter-winding capacitance will cause some variation.
Try inserting a frequency and you will get the reactance. In my case the coil has a reactance of 220 Ohms at 2 Mhz. This can come in handy as we look at the transformers.