Softrock Lite-type kit with no surface mount parts available from Japan

[AB2EZ]

Hi

I ordered one of these... for use with a 455 kHz input (i.f. output from receiver).

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&viewitem=&item=190239900546

After I receive it, I'll post a review. I don't expect any surprises... and this is just an alternate source of these I/Q dual down converters.

Note, the Ebay listing provides a link to this fellow's web site. He is apparently involved with a larger project that involves students around the world. His site doesn't include a schematic... but from the photo available on his web site

http://zao.jp/radio/sft66db2/images/P1010142.JPG

I could figure out what the circuit is. It is very similar to a Softrock Lite:

a. Crystal oscillator for basic local oscillator reference
b. Dual flip-flop chip => 4:1 frequency divider and 90 degree phase shift
c. Quad analog switch chip for dual mixer function
d. Dual op-amp at output of mixer to produce I and Q inputs for sound card
e. On-board reverse polarity protection diode and voltage regulator

Stu

[KF1Z]

He also make a dds type controller, that replaces the crystal,  that allows better use of an entire band...(I believe it was a few megahertz range)
Not just stuck to one frequency, if you care to use it as a reciever.

Complete schematic is here:

http://zao.jp/radio/sft66db2/index2.php

[AB2EZ]

Bruce

Thanks. I guess I did not look far enough down his home page to find the link to these details.

Looking at the schematic and the board layout... he has designed the oscillator layout on the pc board so that you can build it with a crystal and use a 14-pin dip that includes two inverters.... or, you can use a widely available 14-pin programmable oscillator by just substituting that DIP for the inverter DIP.

Standard 14-pin programmable oscillators (they actually have only 4 pins, but they fit in a 14 pin socket) are available from Digikey for around $5.00 each. You specify the frequency you want, and they program it in prior to shipment. For example, if you want a 444 kHz local oscillator for the mixer, then you specify the programmable crystal oscillator to be pre-programmed (burned in) at 4 x 444 kHz = 1.776 MHz.

As you point out, you can also use a DDS to produce the desired signal to drive the 4:1 divider/90 degree phase shift circuitry.

I'm also going to try one of these with my Drake R-4C...  using the R-4C's 50 kHz i.f. output. I'll use a (pre-programmed) 1952 kHz crystal oscillator, and divide it by 8 before sending it to the divide-by-4 circuitry on the board. [1952 kHz/32 = 61 kHz]. This will require me to cut just one trace on the board to divert the output of the crystal oscillator (mounted on the board) to the off-board divide-by-8 chip (I'll use 3 stages of a 74F161APC-ND 4-bit binary counter, in the form of a 16 pin dip) and back to the input of the on-board divide-by-4 circuitry.

Since the i.f. output from the receiver (whether 455 kHz or 50 kHz) is already reasonably well-filtered by the i.f. filter of the receiver... I won't implement the toroid-based bandpass filters at the input to the board. The mixer on this board is a single-ended design... but I'll still use a ferrite toroidal transformer (made from one of the 1-inch diameter cores we use in Class E transmitters)... with just a few turns on the primary and secondary... to provide isolation at the input.

[WD8BIL]

Hi Stu,
Let me know how the Drake setup works on the 50KC I.F..
I've built a SoftRock up for the 5645Khz I.F. in my R4A.

Should get some time to do a test drive with a week or 2.
It would be interesting to compare notes.

Buddly

[AB2EZ]

Buddly

Will do!

Stu

[K9ACT]

I built a SoftRock version for my S40B and have become addicted to it.  I can not listen to another station without looking at it also.

I notice this one has one more torroid to wind and I find these a far bigger pain than the surface mount devices.  It also costs "a lot more".  I think the Lite is still only $13.

It's hard to believe what this does for an S40 type receiver for $13.  The filtering flexibility puts it in the same class as a Flex and if you use Power SDR  software and don't try to talk, you will think you have one., sort of.

js

[AB2EZ]

I received my soft66db2 kit in the mail today. It took just about 1 week... which is excellent considering that it was sent by ordinary mail to keep the cost down.

It took me about an hour to put it together. I'm using a pre-programmed crystal oscillator that I purchased from Digikey ($7.09 programmed to the frequency I specified):

Digikey part # CPPT1-HT06P
5.0 volt TTL 100 ppm
programmed for 1.776 MHz (444 kHz x 4)

[Remember to cut off or bend over pin 1. You want it floating, and the board will ground this pin if you plug it into the 14 pin dip socket. The other three pins are "plug and play" with this board.]

I'm not using any band-limiting components (i.e. small toroidal transformers and chokes). I'm just using the board with a isolation transformer constructed with one the the 1" diameter type 43 toroidal cores I have on hand for other purposes. 5 turns on the input and 5 turns on the output of insulated hookup wire. The large area of each turn eliminates the requirement for a large number of turns to get a sufficient magnetizing inductance at 455kHz.

It works like a charm. The programmable oscillator is right on frequency.

Attached is a .jpg picture. The silver colored object plugged into the 14 pin DIP socket is the programmable oscillator.

Stu

[WA1GFZ]

Stu,
The softrock guys are playing with the Si570 source. I read it is quite clean.
and doesn't have the spurs of a DDS.
I also read the Flex 5000 has more spurs than the the SDR1000

[AB2EZ]

This unit works just fine with the 50 kHz output of my Drake 1A (keeping in mind that the receiver's i.f. filter is, by design, fairly narrow):

1. I made up a little board to take the output from the i.f. amplifier of the Drake 1A (.jpg attached). I mounted this board (in a completely reversible way, of course), under the chassis, next to the i.f. amplifier tube in the Drake 1A, and used the 1A's 12.6 vac filament supply bus to power the board. I inserted the input of the board in series with the existing 150kohm grid leak resistor of the Drake 1A. The 50 kHz output exits an existing (stock) hole on the back of the Drake 1A via a piece of RG-174 coaxial cable.

2. I did away with the input coupling transformer completely, and I coupled into the mixer board through a .1 uF capacitor (.jpg attached). The 1000 ohm resistor is part of the 26dB input attenuator... to reduce the 1 volt level of the Drake's add-on i.f. output to the 50mV level appropriate for the mixer board.

3. I'm using the circuit shown the third .jpg attachment to produce the 61kHz x 4 = 244kHz signal that serves as the local oscillator for the mixer board. Since I elected to set the local oscillator to 244kHz (61 x 4 kHz), i.e., 11kHz higher in frequency vs. the 50 kHz i.f., the I and Q 11 kHz outputs of the mixer board are flipped over in the frequency domain. For example, if you want to use the SDR software to demodulate a lower sideband signal, you have to select the upper sideband mode. I could have used 156kHz (39 x 4 kHz)... but I was playing it cautious by keeping the local oscillator frequency further away from the mixer output frequency.

Stu

[AB2EZ]

I am updating one of the three .jpg images I posted above. See the attached .jpg image.

I changed the design of interface from the Drake 1A's 50kHz output (~ 1 volt peak) to the input of the mixer board... as shown on the attached .jpg image.

1. I removed the 47 ohm resistor on the mixer board that was shown in the previously posted design.

The nature of the design of the mixer board is such that it is better to have no dc path from the 2.5 volt reference point on the board to the input of the "analog switching" device that performs the multiplication operation. This forces the op-amps at the output of the mixer board to have unity gain at DC whether or not the "analog switching" devices are gated "on" or "off" by the local oscillator.

2. The new design employs a 3.3k ohm series resistor and a 0.1uF series coupling capacitor at the input. These are shown in the attached diagram as located on a separate board, but I actually placed them on the mixer board using existing holes and traces.

This causes the op-amps at the output of the mixer board to each produce a net voltage gain (Drake 1A i.f. output voltage => op amp output voltage) of 5100/3300 ~ 1.5. By adjusting the value of the 3300 ohm resistor, one can make the net gain higher or lower.

Of note:

When I connected both outputs of the mixer board to the stereo inputs of the sound card of my computer... an extraneous oscillation (~ 1MHz) would appear on both outputs. When I connected only one output (either one) of the mixer board to the sound card, there was no problem. This is probably an artifact of the circuitry at the input of the sound card. I solved this problem by putting a 10kohm resistor in series with each of the outputs of the mixer board, and by placing a 100pf capacitor to ground at the output of each resistor.

Stu

[AB2EZ]

I have attached a photo of the mixer board (with the 3300 ohm resistor and the 0.1uF capacitor added at the input using existing holes and traces, and with no toroidal transformers or inductors); the auxiliary local oscillator board (which plugs into pin 8 of the stock 14 pin socket on the mixer board); and a board that contains a pair of 10kohm series resistors and a pair of 100 pf bypass capacitors, which provides a better interface from the mixer board to my computer's sound card.

Of note, to use this at 455kHz, all I need to do is to pull the green wire from pin 8 of the 14 pin dip package, and re-insert the 1776 kHz crystal oscillator into the 14 pin dip socket.

[WD5JKO]

Stu,

  I am excited to hear how your SDR project works out. Last Saturday I attended the Austin, Texas Hamcom swap meet, and Flex radio which is based here had two one hour presentations. One item presented stuck with me after the discussion:

Since the IF frequency is around 11 Khz using your sound card, you can get any bandwidth you want with good "brick wall" skirts so long as the RF source is wide band. With your Drake 1A 50 khz output you will already have a narrow band RF source, so the ability to open up and listen to a glorious AM signal with a 15 Khz bandpass will not exist.

I am considering a similar project using my Gonset G76 receiver and picking up at the 1st IF which is between 2-3 Mhz and it should be wide band.

At the following link is me on my Central Electronics 20A as received by a local ham with a Flex SDR 5000 set to 7 Khz wide LSB. There is also a recording clip of Bob Heil with his 20A (unknown receive situation). Bob is on AM, and I'm on LSB. The quality of a 20a is the same on AM as SSB so long as nothing is over driven. When I get my SDR working, I want to be able to listen to sideband signals if I choose to.

http://pages.prodigy.net/jcandela/20A_MP3/
 
So with a NB 50 Khz IF source for your SDR, how wide can you get?

Regards,
Jim
WD5JKO

[AB2EZ]

Jim

On the Drake 1A, the 50 kHz i.f. is fairly narrow (~ 3kHz)... so it is quite limited for AM operation. It does work ok (but not great) on AM if I move the Drake 1A's passband tuner to one side of the carrier (USB or LSB position). I was anxious to try this out with the 1A because the 1A doesn't have an AM detector. Now I can demodulate AM signals in AM or synchronous AM mode without having to zero beat each person in a round table. That said, it would be better to use the 1.1 MHz i.f. of the Drake 1A... and I will switch to that at a future date.

The configuration I ended up with (per my most recent posts) is good for any i.f. application... provided you use the right oscillator frequency. Assuming at least some amount of i.f. filtering in the receiver itself, you don't need any toroidal components.

Note that I also have a Drake R-4C, and its 50kHz i.f. bandwidth is controlled by the crystal filters that are installed. Mine will have an ~ 8kHz i.f. bandwidth (limited by the INRAD 8 pole 1st i.f. crystal filter) if I select the AM position (no AM filter) on the bandwidth selector.

For people who have a Drake 2A or a Drake 2B it would probably make sense to pick up the 455 kHz i.f. that is available at the Q-multiplier connector on the back of the receiver.

Right now, I'm really enjoying receiving AM on my Drake 1A, and looking at the spectral display on the computer... even with the constraints of the 50Hz i.f. filter in the 1A.

Best regards
Stu

[AB2EZ]

Perhaps my last update... just to close out one open item:

I am getting excellent results using the 50 kHz output from my Drake R-4C.

My R-4C has an 8-pole, 8kHz first i.f. filter (in place of the original 4-pole filter)... but in this application, one might do even better with the wider, stock filter. In any event, it works great in the AM position, where the only filter of significance is that first i.f. filter.

I took the i.f. output from the plate of V5 (6BA6 50 kHz  i.f. amplfier). That seemed the simplest way to get the i.f. output without disturbing the product detector or the a.m. detector... although you probably could pick it up in a lot of alternative places. As per the attached .jpg

Plate of V5 => 330 pf 500 volt mica capacitor => 470k ohm resistor  => 5600 ohm resistor => ground.

The point where the 470k ohm resistor and the 5600 ohm resistor come together feeds the input of an LM386N-1 8-pin DIP amplifier, which requires only +12 volts (any op-amp would do... set the gain to around 10-20).

The output of the amplifier is connected to ground via a 47 ohm resistor. The output of the amplifier is also capacitively coupled to a piece of RG-174 coaxial cable.

The coaxial cable feeds into the 50kHz version of the mixer (see attached .jpg) as described in my earlier posts for the case of my Drake 1A.

I hope that many of you who have not yet tried this will do so. Once you use the PowerSDR software to filter, demodulate, and display your classic receiver's i.f. output signal, you will find it hard to go back.

Best regards
Stu