2m Rx 1st IF Frequency

693 views
Skip to first unread message

NZ0I

unread,
Dec 25, 2016, 9:39:16 PM12/25/16
to Receiver Development Platform
Using a first IF of 70 MHz would require a VFO of 74-78 MHz. Those two frequencies (IF and VFO) seem a little too close together, and I have some concern about leakage of the VFO into the first IF Amp possibly causing de-sense or other issues. What do you think about using a 45 MHz first IF? That would be the input frequency going into the SA605, and there are quite a few examples of the SA605 being used successfully for sensitive receivers at that frequency, some of them using an IF (2nd IF in our case) of 10.7 MHz.

DigiKey has some (pricey) xtal filters for 45 MHz, one example: http://www.digikey.com/product-detail/en/ecs-inc/ECS-96SMF45A30/X714CT-ND/274092. But building our own filter at that frequency shouldn't be very difficult if we don't find a reasonable xtal alternative.

A 45 MHz first IF would put the VFO at 99-103 MHz, which I don't foresee being problematic.


NZ0I

unread,
Dec 25, 2016, 10:29:00 PM12/25/16
to receiver-devel...@googlegroups.com
And since we are now talking about having a double-conversion receiver, we might as well examine going to a 2nd IF of 455 kHz. Most SA605 example designs tend to use that as the (in our case 2nd) IF frequency, and the literature indicates that the SA605 should be more stable and sensitive with that IF. I would think that a 455 kHz IF would also be a better choice for use on the 80m band as well, for the same reasons.

Mouser has some choices of 455 kHz ceramic filters, one example: http://www.mouser.com/ProductDetail/Murata-Electronics/CFUKG455KE4XR0/ (This post was edited to include a non-end-of-life product example from Mouser.)


NZ0I

unread,
Dec 26, 2016, 9:16:52 AM12/26/16
to receiver-devel...@googlegroups.com
[Note: this post has been edited to include a corrected block diagram.]

It seems that NF would benefit by moving the LNA to be before the BPF in the signal chain. While there seem to be many conflicting blog posts on the subject, this link seems to be pretty authoritative on the matter: http://www.ti.com/lit/an/slaa652/slaa652.pdf

Matching between the antenna and LNA is essential and must remain since the LNA doesn't have 50-ohm input impedance, but we should try to reduce loss in the matching network. Below is a revised block diagram that assumes a first IF of 45 MHz and and a second IF of 455 kHz. The biggest remaining question mark is the 1st IF amplifier: do we even need it? With the additional gain we should get from the SA605 by going to 45 MHz / 455 kHz IF frequencies, perhaps the first IF amp is not required, and all that is needed is a matching network for the 4.7k input impedance of the SA605 at 45 MHz.


Gerald Boyd

unread,
Dec 26, 2016, 12:38:19 PM12/26/16
to Receiver Development Platform
I think going to the 45 MHz if is a good idea. Concur with the analysis that the Lo may be to close to the IF if it was 70 MHz.

We may still need the 45 MHz IF as going direct to 455 kHz ( having only a 455 IF) could result in image issues on 2 meters.
It may be possible to only have a 45 MHz IF and do away with the 455 IF as that would put the image out of the RF front end passband response. We could use the other Gilbert cell device as a product detector is sa602?


Sent from my iPad
--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/e7f6874a-d977-4bdc-a4bb-0e6e13e2254e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Charles Scharlau

unread,
Dec 26, 2016, 3:02:43 PM12/26/16
to Gerald Boyd, Receiver Development Platform
Having a single-conversion receiver with an IF of 45 MHz, using Gilbert Cell mixers, would be something like what we'd get if we converted our original design (SA605 only) from 10.7 MHz to 45 MHz IF. If we can't get the desired sensitivity using 10.7 MHz with the SA605, I think we'd have similar difficulty (maybe worse) using 45 MHz and SA602.

If we still want to go with a single-conversion design, but improve the sensitivity over the SA605-only approach, then we might want to try adding just the Match, LNA, and BPF, shown in the block diagram. That should give a low-noise 10dB boost to the signal before it arrives at the SA605. We could increase the gain by another 10 to 20dB without too much difficulty by adding one more amplifier stage if necessary. But I would suggest this pre-amped single-conversion 10.7 MHz IF approach only if we are "close" to having a viable receiver with the SA605 and 10.7MHz IF all by itself. 

If we aren't close to having a working receiver with the SA605 and 10.7MHz IF, then I think we might be better off going to a double-conversion design. We could use the SA602 in either (or both) of the last two mixers. But I think we will want to use a more conventional DBM in the first mixer, since that should help improve dynamic range at higher signal levels... though I have seen a simple ARDF receiver design that "deafens" an SA602 used as the first mixer, so maybe that would be satisfactory. But the double-conversion with a 45 MHz first IF, and 455 kHz second IF, allows us to use a proven design in the later stages, so there should be fewer concerns about being able to achieve adequate sensitivity with that design, I think.

Things that might get us close (< 30dB) to the desired sensitivity with just the SA605 and single 10.7 MHz IF:

1. Use proper matching and signal levels for LO and BFO
2. Match the RF input into the SA605
3. Reduce the gain-killing resistor short of creating instability*
4. Increase the supply voltage of the SA605 from 5V to 7.5V (it can handle up to 8V)

*Note: reducing the gain-killing resistor too much can result in decreased sensitivity - if the SA605 becomes unstable the self oscillations overpower the mixers/amplifiers causing de-sense.

We are getting to point where we will need some answers about what the SA605 can do, before we can decide how to proceed.







On Mon, Dec 26, 2016 at 12:38 PM, Gerald Boyd <wb8...@icloud.com> wrote:
I think going to the 45 MHz if is a good idea. Concur with the analysis that the Lo may be to close to the IF if it was 70 MHz.

We may still need the 45 MHz IF as going direct to 455 kHz ( having only a 455 IF) could result in image issues on 2 meters.
It may be possible to only have a 45 MHz IF and do away with the 455 IF as that would put the image out of the RF front end passband response. We could use the other Gilbert cell device as a product detector is sa602?


Sent from my iPad

On Dec 26, 2016, at 7:16 AM, NZ0I <charles....@gmail.com> wrote:

It seems that NF would benefit by moving the LNA to be before the BPF in the signal chain. While there seem to be many conflicting blog posts on the subject, this link seems to be pretty authoritative on the matter: http://www.ti.com/lit/an/slaa652/slaa652.pdf

Matching between the antenna and LNA is essential and must remain since the LNA doesn't have 50-ohm input impedance, but we should try to reduce loss in the matching network. Below is a revised block diagram that assumes a first IF of 45 MHz and and a second IF of 455 kHz. The biggest remaining question mark is the 1st IF amplifier: do we even need it? With the additional gain we should get from the SA605 by going to 45 MHz / 455 kHz IF frequencies, perhaps the first IF amp is not required, and all that is needed is a matching network for the 4.7k input impedance of the SA605 at 45 MHz.


--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/AD79C5B7-4352-46C9-BC40-149711A79E21%40icloud.com.

Gerald Boyd

unread,
Dec 26, 2016, 6:39:15 PM12/26/16
to Receiver Development Platform

Charles,

 

Yes we need to get the answer on the SA 605.

 

I have unpacked the parts from digikey for the trying the matching network. Can wire it up tonight.

 

Attached is the schematic showing the jumpers as if the circuit was in the IF board. I plan to wire it on the test adaptor board between the SMA and the SA chip socket.

Attached is a photo of a schematic without the configuration jumpers. Is this correct?

Jerry

--

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

--

You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

 

--

You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/CAK-GiwZP3EVfN0FwGx5JBxVGxBXSOcUZm7bFMxJ_sKracg03HQ%40mail.gmail.com.

sch.jpg
hand wired on test board.jpg

Charles Scharlau

unread,
Dec 26, 2016, 7:54:45 PM12/26/16
to Gerald Boyd, Receiver Development Platform
Yes, you've got the correct circuit minus the jumpers. It is simply the "tapped-C network" as described in the SA605 Application Note 1994. I followed NXP's procedure for determining the component values.


--

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

--

You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/006d01d25fd1%242ecfce60%248c6f6b20%24%40covad.net.

NZ0I

unread,
Dec 28, 2016, 11:51:36 AM12/28/16
to Receiver Development Platform
Here's an updated block diagram of the 2m receiver design. The main change is that the front end now includes a .5dB - 31dB attenuator, and a single gain stage (that can be bypassed) following the mixer. This should give us a great deal of control over the gain we get. This front end design can provide from -45dB of loss to +35dB of gain, with a low noise figure, and high linearity.

NZ0I

unread,
Dec 28, 2016, 10:41:52 PM12/28/16
to Receiver Development Platform

Here is a front end schematic that implements that portion of the latest block diagram. Many of the component values have yet to be determined, but I think the ICs and topology are a reasonable first cut. The attenuator is capable of over 30dB of attenuation in 0.5 dB steps. This should let us step the attenuation smoothly across the entire dynamic range in small increments. The idea is for the user to not experience any sudden large swings in signal level when the attenuation level is changed. Instead of the user having to keep track of attenuation levels, the user will instead hear a tone in the earphone: the pitch of that tone will increase (or decrease) with the attenuation setting to give an indication of distance to the transmitter (the display can show the actual estimated distance from the transmitter based on a calibration performed by the user). The received signal audio will also play through the headphones, and its strength will be used when taking bearings. 


We can experiment with a variety of concepts (including the traditional "whoopee mode"), but the concept above (though complex to describe) will actually be very intuitive if it is implemented successfully. The slowly-changing pitch will provide a constant indication of distance from the transmitter, while bearings can be taken without concern about what attenuation level is currently being used. It would be desirable for the attenuation level to quickly ramp down when a weaker transmitter comes on the air. A long press on the "trigger" button can be used to manually force the receiver to reset the attenuation level, and the processor can then automatically reset the gain at 1-minute intervals after a manual reset.


If we decide to stay with a single conversion design, then the front end will get simpler: the VFO matching and mixer will go away, but otherwise I think the same basic front end design can be used - the LNA, the BPF, the gain block, and the attenuator.


Adding the LT5537 RF log detector to the front end might ensure we can get high performance at very strong signal levels. I think I will try to add that to the design as an option... but we'll need an ADC channel to read it.




On Sunday, December 25, 2016 at 9:39:16 PM UTC-5, NZ0I wrote:

Gerald Boyd

unread,
Dec 29, 2016, 5:47:15 PM12/29/16
to Receiver Development Platform
On needing another adc for the log detector I see possible 2 options

Opting 1 use an analog switch controlled by one of the unused outputs from the i2c digital expander to switch between the voltage divider that is used to ID the module and the log detector.

Option 2 if we could find a cheep i2c EE rom this could be used to ID the module and also save calibrates for when used with the receiver. This could free up the analog channel used by the ID voltage divider.

Will look over rest of the circuit.

Jerry

Sent from my iPad
--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/5c370c06-2bdc-442a-9c9e-199df55f5808%40googlegroups.com.

Charles Scharlau

unread,
Dec 29, 2016, 8:13:15 PM12/29/16
to Gerald Boyd, Receiver Development Platform
Both are excellent ideas. The analog switch approach is easy enough, but the I2C EEPROM seems like an excellent fit, potentially useful for storing calibration data, has a small footprint, uses no more current than the voltage divider, and remarkably cheap: http://www.digikey.com/products/en?keywords=24AA00T-I%2FOTCT-ND

I think that's the route to take. Thanks for the suggestion!


To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/A4AD69E4-D3C5-4CB5-82AF-4FD530E5E24B%40icloud.com.

Charles Scharlau

unread,
Dec 29, 2016, 8:32:00 PM12/29/16
to Gerald Boyd, Receiver Development Platform
For the same price we can get 1k of EEPROM with its own unique serial number pre-programmed. Why not!

Gerald Boyd

unread,
Dec 30, 2016, 10:21:09 PM12/30/16
to Receiver Development Platform
Now that's a cool feature a built in serial number.
I like it.
Jerry

Sent from my iPad

NZ0I

unread,
Dec 30, 2016, 11:02:37 PM12/30/16
to Receiver Development Platform
I've updated the 2m front end schematic. It now has a first cut at all the component values. The BPF has been re-done, small re-design of mixer input matching, added the RF log detector circuit, the EEPROM, and some test points. The full schematic on the share drive has also been updated.



On Sunday, December 25, 2016 at 9:39:16 PM UTC-5, NZ0I wrote:
Reply all
Reply to author
Forward
0 new messages