EMRFD Message Archive 13205

Message Date From Subject
13205 2016-10-12 16:02:19 sa5pmg Specan - questions
Hi all,

I think this is the correct forum to ask questions about the Specan- W7ZOI analyzer reboot construction by VU2ESE. If not, please let know where.

All these questions are part of my learning process to understand the construction and the various tradeoffs made.

Lets start with "1st OSC", the Si570 based VFO.

1a. What is the expected output level into the 1st mixer?
My calculations get it to around 4dBm.

1b. What about using both clock outputs on the Si570 and join the signal in a transformer?
Good or bad idea?

2. Is the purpose of the 33Ohm resistor between Si570 getting close to 50Ohm input impedance to the MAV-11?
I found somewhere that the output impedance of Si570 is 12 Ohm, so I guess a resistor value closer to 38Ohm would be better, right?

3. Why is the MAV-11s DC bias different between U4 and U2?

4. The output attenuator is mentioned as 6dB so would it be better to use a 37.4 Ohm resistor?
Same question also applies the five? other places throughout the schematic where the 150, 33, 150 Ohm PI attenuator is used.

Schematics errata
5. There are two L5 inductors, one on each MAV-11.

6. Label for LO input is missing on mixer 2. And LO input capacitor lacks value. Is the capacitor really needed? Should the value be 0.01uF?

7. L10 and L11 miss values.

2nd Osc
8. What is the expected output level?
My plan is to use a 100MHz TTL/HCMOS oscillator instead and construct accordingly. Or is it a bad idea?

2nd IF
9. Expected gain?

Best regards
Marcus, SA5PMG
13219 2016-10-26 16:56:52 Marcus Gustafsson Re: Specan - questions
Hi again,

almost two weeks and not a single answer, a bit disappointing. Well, I
guess I learn by doing instead.

I have started my build and along the way I plan to learn to cad PCBs and
handle SMT components so I started with the last IF amplifier in 1206
resistors, capacitors and SOT-23 BJTs. I also tried a 1206 5% inductor for
L17 but the filter peaked at 11.2MHz so I changed it to a toroid (T37-6 18
turns 29AWG) and now the amplifier has peak gain at 12MHz. And the gain is
high, around 40dB.

Attached is a picture of the build.
Ashar: Please contact me regarding making PCB designs available.

Stay tuned for more module builds.

Marcus, SA5PMG

13223 2016-10-26 20:18:16 Ashhar Farhan Re: Specan - questions


apologies for not answering earlier. I opened your mail while travelling and then promptly forgot about it after I came back home. Here are my inputs.

I would preface them to say that some more simplifications can be done. These are only suggestions, I haven't tried them out. 

1. The U2 could be replaced by just a 6db pad. Low noise figure is not of much use. Indeed, the dynamic range would actually go up if you replaced U2 with a pad. Similarly, a CMOS version of Si570 may be such a bad idea. Try it if you have it handly. The Si570 can drive the 6db pad through the 33 ohms resistor directly. 

2. You could replace the second oscillator with an Si5351 and use another channel of the Si5351 as the tracking generator!

13226 2016-10-27 10:41:31 sa5pmg Re: Specan - questions

thanks for getting back. Apologies accepted, don't worry we all done the same thing.

1a. No post 1st mixer amplifier and only keep the 6dB pad.
I certainly try that.

1.b Si570 CMOS
I only have the LVDS version at hand.

2. Feature creep, I like it!
How would one get harmonic free output from the Si5351? Using two outputs and a mixer and down convert to 0-70MHz?

3. MAV-11 bias and Arduino noise.
I make sure to shield the Arduino properly and maybe use pass-thru capacitors, or actually I'm planning to use a STM32F103 but I'll do proper shielding. In the PCBs I've designed I tried to mimic Mini-circuits eval boards schema.

4. 2nd oscillator, 100MHz crystal
It's a SMD DIL format EPS
13292 2016-11-23 12:34:47 sa5pmg Re: Specan - questions
Hi all again,

I've started my Specan build. I have a  STM32duino, LCD display and Si570 module and its amplifier up and running but I suspect something is wrong since the output from the Si570 and its amplifier is low? Or are the square wave fooling me but the level still varies a lot over the 110-180MHz range and might be a bit on the low side for a +7dBm mixer. Also the AD8307 handles up to 500MHz which the square wave odd harmonics passes out side around 167MHz, it that the explanation
 or parts of it?

Measuring output from my Si570 module which have a 10nF capacitor (C4) and a 38.3Ohm resistor (R4) in series:
With a AD8307 power meter:
MHz    dBm
110    -3.4
150    -6.6
180    -7.1

With oscilloscope BNC cable to 50Ohm input
MHz    Vpp mV
 10    350
110    333
150    305
180    241

The MAV-11 module I built has the following gain with -10dBm input (measured on a HP VNA at work):
MHz gain dB
---    ----
100    6.1
110    6,1
144    6
180    5.9
200    5.9

These values looks alright to me.
And with SI570 input and MAV-11 together measured with AD8307
MHz    dBm
110    2.8
150    0.5
180    -1.3

which to me indicates that the MAV-11 does its work with the Si570 input too.

Anyone with a clue what could be wrong? I can only think about various work-arounds such as use a higher gain amplifier like ERA-3 for instance or maybe try a LVDS to CMOS converter such as FIN1002.

A bunch of pictures are attached below.
/Marcus, SA5PMG
13293 2016-11-23 13:13:26 Dave Re: Specan - questions
I'm not familiar with that Si570 module. The Si570 in the picture is
a LVDS. I only see one output connected. Looks like the board is
designed for CMOS. I think the 38.2 resistor should be 0 for LVDS if
the load is 50 ohms.

Dave - WB6DHW

13294 2016-11-23 15:22:43 sa5pmg Re: Specan - questions

I've followed the schema at http://hfsignals.blogspot.se/p/specan-reboot-of-w7zoi.html, also see previous messages in this conversation about Si570 output and differential mode. The module is my own design.

Anyway, trying a 0 Ohm resistor is an easy experiment. I'll try that, thanks.

Marcus, SA5PMG
13295 2016-11-23 16:01:36 Dave Re: Specan - questions
  The 38.2 is probably OK for CMOS since it's output impedance is low.  LVDS actually like 100 ohms between the 2 outputs.  The 2nd output should probably have a .1 uF and a 50 ohm resistor to ground.  The interfacing is explained in the attachment I included on the previous message.

Dave - WB6DHW

13296 2016-11-23 23:30:23 Ashhar Farhan Re: Specan - questions

which version of si570 are you using?  the values suggest that it is the LVDS version. The tapering output suggests that there is a low pass filtering of the output. The Si570 is a digital part where the output IS flat across its operating spectrum. I hope you are using chip capacitors and very small leads. It could be that the cap and the resistor are forming some kind of a trap. I used chip caps and chip resistors

- f

13297 2016-11-24 03:22:16 sa5pmg Re: Specan - questions

I'm using the LVDS version. Not sure what happened to the pictures I attached and uploaded yesterday. I'm using SMD 1206 parts, the capacitor is a 10nF 2% NP0 one.

Anyway, a good clue. I investigate further by measuring directly on the Si570 outputs.

Dave: I cannot find any attachment you are referring to but I don't get along with Yahoo UI.

13298 2016-11-24 09:31:21 sa5pmg Re: Specan - questions
Hi again, I have now mounted a FIN1002 by some serious SMD patching (at least to me :-), now the output is fairly stable. Vpp varies from 2.37V at 10MHz to 2.45V at 110MHz to drop down to 2.41 at 180MHz.

Next step is to build a small DC blocking 3-4dB attenuator to adjust the input level and impedance to the mixer input.

Attaching a couple of pictures.
13299 2016-11-24 12:31:25 Dave Re: Specan - questions
  Yahoo must have stripped off the attachment.  Here is the URL to it:

Dave - WB6DHW