Topics

Simple UBITX Test Set


Scott McDonald
 

Jerry,

You're right, I think, a simple cheap standardized UBitx test set would really be a great tool, not just for a Ubitx but for anyone starting out without any RF test equipment.

I started down that road for a club talk last year and built a single 2N3904 oscillator with a color burst crystal that gave signals in or very close to all the HF bands except 60 meters.  

By juggling a few values I was getting reliable output on all those freqs at -20 dBm plus or minus probably 5 dBm, and this was reproduceable with 5 different color burst crystals (tho from the same batch).  Not enough to drive a diode ring mixer but otherwise pretty useful and consistent, I think close enough that you could count on the output being what you expected without measuring it.  

As an option adding a buffer with another 2N3904 would make it more useful, or maybe just use one of Diz's inexpensive new SMT HF amp kits instead of reinventing the wheel.

Added an IF crystal from Kits and Parts (it was for a V3) with a cap and coil to fudge it a bit and got a useful IF signal at the same level. 

Added a couple 50 dB Pi attenuators to give a choice of 50 over 9, S9 and a few tenths of a microvolt for signal level.  Standard resistors used that way yielded decent attenuators plus or minus 3 or 4 dB.

Did it on a piece of perf board using 0.1 inch headers and jumpers to switch crystals and attenuators so there were no switches or other hardware needed.

It was dirt cheap and pretty handy, adding the diode detector would be simple, and the attenuators possible useful with that, same for adding an audio generator.  That's got to be a 10 dollar or less project with bulk parts purchases.

It would make a great first build project too.  

If I had any board skills I would make a board and put in Osh Park... :)  But if anyone with some skills wants to collaborate on it, I'd be happy to get together, as I expect people troubleshooting with a known test kit would make the great advice you, Evan, Allison and others give on here even quicker and more powerful.  

73 Scott ka9p


-----Original Message-----
From: Jerry Gaffke via groups.io <jgaffke@...>
To: BITX20@groups.io
Sent: Sun, Oct 25, 2020 10:50 am
Subject: Re: [BITX20] My son's V6

The primary selling point of the uBitx is its simplicity,
would be good to keep it that way.
The circuit improvements moving from V4 to V5 are a major win,
badly needed and no significant added complexity.

Though I would have kept the +12V for the IRF510's separate, much safer to 
debug all but the final amp of the transmitter with those IRF510's disabled.
And allows flexibility in setting the output power by adjusting the IRF510 supply voltage.
A simple way to adjust RF gain would be good.

Moving from the V5 16x2 LCD to the V6 TFT display adds lots of overhead
to the Nano's firmware.  I'd prefer the uBitx had kept the 16x2, at least as an option. 

An SWR meter and antenna tuner with a suitable dummy load for the uBitx 
might be a good accessory.  Audio and RF signal generators, a diode RF probe
(or AD8307), complete instructions to debug the rig, those would be most welcome.   
Perhaps make them all part of a single accessory?  An audio CW filter too?

Jerry, KE7ER


On Sat, Oct 24, 2020 at 05:49 PM, Bob Lunsford wrote:
Using a tuner on an antenna that is designed/dedicated to a specific band or frequency would theoretically funnel the signal to that specific frequency but it would also, theoretically, suppress received signals up and down frequency from the tuner's settings. Since it is also a good approach to getting as much power as possible from a QRP radio to the antenna by allowing the transmitter's output stage to see a much better load and lessen the complications of a mismatch, it would/should be an automatic part of the system unless weight or system complexity are concerns. It is undoubtedly a prime candidate for experimentation to see if a tuner does indeed reduce the influence of a high powered transmitter either nearby in frequency or location.
 
Thanks, Curt, for your feedback on this. Sometimes a simple solution may be best even if not a complete solution. I agree that AGC may be nice but it is not always the best or necessary. For example, on my G90 if you turn off the AGC it reveals many weak stations and even if some others come booming in, it's a small price to pay to guarantee that all stations on the [net] frequency are heard. Unless it is truly excessive in volume, of course, and this is "in the ear of the hearer."
 
Bob — KK5R


Jerry Gaffke
 

A standardized way to debug a uBitx would be a big plus for this forum.

Some kind of RF source is required that hits the operating frequencies of interest
plus the two IF frequencies would be good.
Perhaps using harmonics from a crystal oscillator is sufficient, though not always ideal.
Colorburst crystals are now archaic, but easy enough to get.
Plan B might be a $1 Si5351, plus of course an extra $10 for Nano and display and knobs.

VK4PP laid out a Raduino replacement with pads for an AD8307, though far as I know
nobody ever tried using the AD8307.     https://groups.io/g/BITX20/message/42251
That looks like a really noisy spot for the AD8307.
With proper layout (perhaps a smaller processor) this could be made to work.

Jerry, KE7ER



On Sun, Oct 25, 2020 at 09:50 AM, Scott McDonald wrote:
Jerry,
 
You're right, I think, a simple cheap standardized UBitx test set would really be a great tool, not just for a Ubitx but for anyone starting out without any RF test equipment.
 
I started down that road for a club talk last year and built a single 2N3904 oscillator with a color burst crystal that gave signals in or very close to all the HF bands except 60 meters.  
 
By juggling a few values I was getting reliable output on all those freqs at -20 dBm plus or minus probably 5 dBm, and this was reproduceable with 5 different color burst crystals (tho from the same batch).  Not enough to drive a diode ring mixer but otherwise pretty useful and consistent, I think close enough that you could count on the output being what you expected without measuring it.  
 
As an option adding a buffer with another 2N3904 would make it more useful, or maybe just use one of Diz's inexpensive new SMT HF amp kits instead of reinventing the wheel.
 
Added an IF crystal from Kits and Parts (it was for a V3) with a cap and coil to fudge it a bit and got a useful IF signal at the same level. 
 
Added a couple 50 dB Pi attenuators to give a choice of 50 over 9, S9 and a few tenths of a microvolt for signal level.  Standard resistors used that way yielded decent attenuators plus or minus 3 or 4 dB.
 
Did it on a piece of perf board using 0.1 inch headers and jumpers to switch crystals and attenuators so there were no switches or other hardware needed.
 
It was dirt cheap and pretty handy, adding the diode detector would be simple, and the attenuators possible useful with that, same for adding an audio generator.  That's got to be a 10 dollar or less project with bulk parts purchases.
 
It would make a great first build project too.  
 
If I had any board skills I would make a board and put in Osh Park... :)  But if anyone with some skills wants to collaborate on it, I'd be happy to get together, as I expect people troubleshooting with a known test kit would make the great advice you, Evan, Allison and others give on here even quicker and more powerful.  
 
73 Scott ka9p


Ashhar Farhan
 

I can document what we use at HF Signals. But , in effect it is like this ...
1. We have a dummy load (20, 1K 2/watt resistors) and an Rf sniffer with a 10K and a 50 ohms in series. the 50 ohms goes to the oscilloscope.
2. A 10 MHz crystal oscillator with a trimmer to net it exactly to 10 MHz. we calibrate it every month.
3. A zener diode biased with 1 ma of current as a noise source. it amplified by two stage of feedback amplifier and feb back to the transceiver through the dummy load through a 10pf. there are two 1N4148 back-to-back diodes to prevent high RF from entering the noise amplifiers. 
in rx mode, we can net the frequencies against the 10 Mhz oscillator.
the noise source allows us to set the BFO using the online BFO alignment tool.
on tx, we can monitor the RF to check for proper modulation.
We check it on 3.5 Mhz, 7 Mhz, 14 Mhz and 28 Mhz to exercise all the relays and LPFs.
- f

On Mon, Oct 26, 2020 at 12:22 AM Jerry Gaffke via groups.io <jgaffke=yahoo.com@groups.io> wrote:
A standardized way to debug a uBitx would be a big plus for this forum.

Some kind of RF source is required that hits the operating frequencies of interest
plus the two IF frequencies would be good.
Perhaps using harmonics from a crystal oscillator is sufficient, though not always ideal.
Colorburst crystals are now archaic, but easy enough to get.
Plan B might be a $1 Si5351, plus of course an extra $10 for Nano and display and knobs.

VK4PP laid out a Raduino replacement with pads for an AD8307, though far as I know
nobody ever tried using the AD8307.     https://groups.io/g/BITX20/message/42251
That looks like a really noisy spot for the AD8307.
With proper layout (perhaps a smaller processor) this could be made to work.

Jerry, KE7ER



On Sun, Oct 25, 2020 at 09:50 AM, Scott McDonald wrote:
Jerry,
 
You're right, I think, a simple cheap standardized UBitx test set would really be a great tool, not just for a Ubitx but for anyone starting out without any RF test equipment.
 
I started down that road for a club talk last year and built a single 2N3904 oscillator with a color burst crystal that gave signals in or very close to all the HF bands except 60 meters.  
 
By juggling a few values I was getting reliable output on all those freqs at -20 dBm plus or minus probably 5 dBm, and this was reproduceable with 5 different color burst crystals (tho from the same batch).  Not enough to drive a diode ring mixer but otherwise pretty useful and consistent, I think close enough that you could count on the output being what you expected without measuring it.  
 
As an option adding a buffer with another 2N3904 would make it more useful, or maybe just use one of Diz's inexpensive new SMT HF amp kits instead of reinventing the wheel.
 
Added an IF crystal from Kits and Parts (it was for a V3) with a cap and coil to fudge it a bit and got a useful IF signal at the same level. 
 
Added a couple 50 dB Pi attenuators to give a choice of 50 over 9, S9 and a few tenths of a microvolt for signal level.  Standard resistors used that way yielded decent attenuators plus or minus 3 or 4 dB.
 
Did it on a piece of perf board using 0.1 inch headers and jumpers to switch crystals and attenuators so there were no switches or other hardware needed.
 
It was dirt cheap and pretty handy, adding the diode detector would be simple, and the attenuators possible useful with that, same for adding an audio generator.  That's got to be a 10 dollar or less project with bulk parts purchases.
 
It would make a great first build project too.  
 
If I had any board skills I would make a board and put in Osh Park... :)  But if anyone with some skills wants to collaborate on it, I'd be happy to get together, as I expect people troubleshooting with a known test kit would make the great advice you, Evan, Allison and others give on here even quicker and more powerful.  
 
73 Scott ka9p


Gordon Gibby
 

Ashhar, do you think it would be worthwhile to make a small circuit board and sell that circuitry as an optional test unit?




On Oct 25, 2020, at 15:00, Ashhar Farhan <farhanbox@...> wrote:


I can document what we use at HF Signals. But , in effect it is like this ...
1. We have a dummy load (20, 1K 2/watt resistors) and an Rf sniffer with a 10K and a 50 ohms in series. the 50 ohms goes to the oscilloscope.
2. A 10 MHz crystal oscillator with a trimmer to net it exactly to 10 MHz. we calibrate it every month.
3. A zener diode biased with 1 ma of current as a noise source. it amplified by two stage of feedback amplifier and feb back to the transceiver through the dummy load through a 10pf. there are two 1N4148 back-to-back diodes to prevent high RF from entering the noise amplifiers. 
in rx mode, we can net the frequencies against the 10 Mhz oscillator.
the noise source allows us to set the BFO using the online BFO alignment tool.
on tx, we can monitor the RF to check for proper modulation.
We check it on 3.5 Mhz, 7 Mhz, 14 Mhz and 28 Mhz to exercise all the relays and LPFs.
- f

On Mon, Oct 26, 2020 at 12:22 AM Jerry Gaffke via groups.io <jgaffke=yahoo.com@groups.io> wrote:
A standardized way to debug a uBitx would be a big plus for this forum.

Some kind of RF source is required that hits the operating frequencies of interest
plus the two IF frequencies would be good.
Perhaps using harmonics from a crystal oscillator is sufficient, though not always ideal.
Colorburst crystals are now archaic, but easy enough to get.
Plan B might be a $1 Si5351, plus of course an extra $10 for Nano and display and knobs.

VK4PP laid out a Raduino replacement with pads for an AD8307, though far as I know
nobody ever tried using the AD8307.     https://groups.io/g/BITX20/message/42251
That looks like a really noisy spot for the AD8307.
With proper layout (perhaps a smaller processor) this could be made to work.

Jerry, KE7ER



On Sun, Oct 25, 2020 at 09:50 AM, Scott McDonald wrote:
Jerry,
 
You're right, I think, a simple cheap standardized UBitx test set would really be a great tool, not just for a Ubitx but for anyone starting out without any RF test equipment.
 
I started down that road for a club talk last year and built a single 2N3904 oscillator with a color burst crystal that gave signals in or very close to all the HF bands except 60 meters.  
 
By juggling a few values I was getting reliable output on all those freqs at -20 dBm plus or minus probably 5 dBm, and this was reproduceable with 5 different color burst crystals (tho from the same batch).  Not enough to drive a diode ring mixer but otherwise pretty useful and consistent, I think close enough that you could count on the output being what you expected without measuring it.  
 
As an option adding a buffer with another 2N3904 would make it more useful, or maybe just use one of Diz's inexpensive new SMT HF amp kits instead of reinventing the wheel.
 
Added an IF crystal from Kits and Parts (it was for a V3) with a cap and coil to fudge it a bit and got a useful IF signal at the same level. 
 
Added a couple 50 dB Pi attenuators to give a choice of 50 over 9, S9 and a few tenths of a microvolt for signal level.  Standard resistors used that way yielded decent attenuators plus or minus 3 or 4 dB.
 
Did it on a piece of perf board using 0.1 inch headers and jumpers to switch crystals and attenuators so there were no switches or other hardware needed.
 
It was dirt cheap and pretty handy, adding the diode detector would be simple, and the attenuators possible useful with that, same for adding an audio generator.  That's got to be a 10 dollar or less project with bulk parts purchases.
 
It would make a great first build project too.  
 
If I had any board skills I would make a board and put in Osh Park... :)  But if anyone with some skills wants to collaborate on it, I'd be happy to get together, as I expect people troubleshooting with a known test kit would make the great advice you, Evan, Allison and others give on here even quicker and more powerful.  
 
73 Scott ka9p


Scott McDonald
 

Another simple approach (for people like me that don't program well) is to just put a 2d Raduino in a breakout box.

This generates all the signals, digital and RF, you need to check full functioning of the radio's Raduino; and more importantly

when you go into the cw transmit mode, becomes a relatively high level general purpose HF signal generator reading out the frequency directly.

Considering you can get a W0EB v6.1 Raduino clone shipped for $35 US ( plus display, I think) that does all this with the 5351 already soldered on the board, it's a reasonably cost effective piece of test equipment most anyone can get running.

I use this one to check out any new software or displays before I try them on my radio.  The box pictured is just stupid overkill because of too much lockdown time on my hands, but it's been really useful as a back up signal source on other projects too.  It evolved from a blown BITX 40 Raduino over time, but a bare Raduino, basic display and some headers, pins and jumpers in a big Altoids tin will do all the same things.



Scott ka9p

-----Original Message-----
From: Jerry Gaffke via groups.io <jgaffke@...>
To: BITX20@groups.io
Sent: Sun, Oct 25, 2020 1:52 pm
Subject: Re: [BITX20] Simple UBITX Test Set

A standardized way to debug a uBitx would be a big plus for this forum.

Some kind of RF source is required that hits the operating frequencies of interest
plus the two IF frequencies would be good.
Perhaps using harmonics from a crystal oscillator is sufficient, though not always ideal.
Colorburst crystals are now archaic, but easy enough to get.
Plan B might be a $1 Si5351, plus of course an extra $10 for Nano and display and knobs.

VK4PP laid out a Raduino replacement with pads for an AD8307, though far as I know
nobody ever tried using the AD8307.     https://groups.io/g/BITX20/message/42251
That looks like a really noisy spot for the AD8307.
With proper layout (perhaps a smaller processor) this could be made to work.

Jerry, KE7ER



On Sun, Oct 25, 2020 at 09:50 AM, Scott McDonald wrote:
Jerry,
 
You're right, I think, a simple cheap standardized UBitx test set would really be a great tool, not just for a Ubitx but for anyone starting out without any RF test equipment.
 
I started down that road for a club talk last year and built a single 2N3904 oscillator with a color burst crystal that gave signals in or very close to all the HF bands except 60 meters.  
 
By juggling a few values I was getting reliable output on all those freqs at -20 dBm plus or minus probably 5 dBm, and this was reproduceable with 5 different color burst crystals (tho from the same batch).  Not enough to drive a diode ring mixer but otherwise pretty useful and consistent, I think close enough that you could count on the output being what you expected without measuring it.  
 
As an option adding a buffer with another 2N3904 would make it more useful, or maybe just use one of Diz's inexpensive new SMT HF amp kits instead of reinventing the wheel.
 
Added an IF crystal from Kits and Parts (it was for a V3) with a cap and coil to fudge it a bit and got a useful IF signal at the same level. 
 
Added a couple 50 dB Pi attenuators to give a choice of 50 over 9, S9 and a few tenths of a microvolt for signal level.  Standard resistors used that way yielded decent attenuators plus or minus 3 or 4 dB.
 
Did it on a piece of perf board using 0.1 inch headers and jumpers to switch crystals and attenuators so there were no switches or other hardware needed.
 
It was dirt cheap and pretty handy, adding the diode detector would be simple, and the attenuators possible useful with that, same for adding an audio generator.  That's got to be a 10 dollar or less project with bulk parts purchases.
 
It would make a great first build project too.  
 
If I had any board skills I would make a board and put in Osh Park... :)  But if anyone with some skills wants to collaborate on it, I'd be happy to get together, as I expect people troubleshooting with a known test kit would make the great advice you, Evan, Allison and others give on here even quicker and more powerful.  
 
73 Scott ka9p


Jerry Gaffke
 

Would be good if the standard test set did not require a $400 scope.
I'd be shooting for under $10 in parts if we expect everybody having trouble
with a uBitx to obtain it.

Curious that a broadband noise source is good enough for this.
Ideal if that is truly good enough.
We could put down pads for a $1 Si5351 and somehow drive that i2c bus
from the uBitx's Nano, perhaps bit-blasting over a couple pins usually used
by the display so we don't disturb the Raduino's Si5351.

The Si5351 in a QFN-20 is now $2 from Mouser, it has an A0 pin to 
flip the i2c address so we could use the same two wires as the Raduino's Si5351.
Perhaps some new Raduino should use the QFN-20 so we always have an extra output
for use as an RF signal source?
The QFN-20 has separate pins for each output, reducing crosstalk.

An AD8307 could be monitored using a cheap DVM, far more sensitive than a diode RF probe.
Ebay has AD8307 chips for $0.50 each if you get 10, reports are that they work fine.
Harbor Freight here in the US has suitable DVM's on sale for a dollar at times.

We also need a Nano pin to use as a standard audio source into the mike amp.
Perhaps use CW-TONE for this, though there are plenty of
other possible pins that could be temporarily borrowed.

A bunch of resistors for use as calibrated audio and RF attenuators,
and a 50 ohm dummy load.
 
Jerry, KE7ER



On Sun, Oct 25, 2020 at 12:21 PM, Gordon Gibby wrote:
Ashhar, do you think it would be worthwhile to make a small circuit board and sell that circuitry as an optional test unit?
 


Jerry Gaffke
 

Yet more stuff one might weigh this test set down with:

A capacitance meter:
 https://groups.io/g/BITX20/message/52471

A 74LV8154 prescaler, so a Nano IO pin could serve as a frequency counter:
  https://hackaday.com/2016/09/09/very-simple-pc-frequency-counter-works-up-to-100mhz/
I'd consider adding an input buffer, but even without a buffer it should do fine watching an Si5351 output.

We're still under $10 in parts.

Jerry, KE7ER


On Sun, Oct 25, 2020 at 01:16 PM, Jerry Gaffke wrote:
Would be good if the standard test set did not require a $400 scope.
I'd be shooting for under $10 in parts if we expect everybody having trouble
with a uBitx to obtain it.

Curious that a broadband noise source is good enough for this.
Ideal if that is truly good enough.
We could put down pads for a $1 Si5351 and somehow drive that i2c bus
from the uBitx's Nano, perhaps bit-blasting over a couple pins usually used
by the display so we don't disturb the Raduino's Si5351.

The Si5351 in a QFN-20 is now $2 from Mouser, it has an A0 pin to 
flip the i2c address so we could use the same two wires as the Raduino's Si5351.
Perhaps some new Raduino should use the QFN-20 so we always have an extra output
for use as an RF signal source?
The QFN-20 has separate pins for each output, reducing crosstalk.

An AD8307 could be monitored using a cheap DVM, far more sensitive than a diode RF probe.
Ebay has AD8307 chips for $0.50 each if you get 10, reports are that they work fine.
Harbor Freight here in the US has suitable DVM's on sale for a dollar at times.

We also need a Nano pin to use as a standard audio source into the mike amp.
Perhaps use CW-TONE for this, though there are plenty of
other possible pins that could be temporarily borrowed.

A bunch of resistors for use as calibrated audio and RF attenuators,
and a 50 ohm dummy load.


Allen Hill
 

Check out the TX/RX test circuit in January 2015 issue of QRP Quarterly on page 6 (shared by K1MBO).
Very simple and low cost.
This particular issue is the free download from their website. 

QRPARCI.ORG

73
KI4QCK


Jerry Gaffke
 

Page 6 has the test set Allen mentioned:  https://www.qrparci.org/qqsample/qqsample.pdf

A 7.050mhz crystal oscillator with an attenuator suitable for checking out a 40m receiver.
And a 50 ohm dummy load for the transmitter.

That would definitely be better than nothing at all for debugging a uBitx,
but hardly sufficient.

Could tell if the receiver works on 40m.
And could drive the uBitx power amp to check it out on 40m as well
if you do away with the cute trick of using one connector for
both signal generator and dummy load.

Jerry, KE7ER


On Sun, Oct 25, 2020 at 02:25 PM, Allen Hill wrote:
Check out the TX/RX test circuit in January 2015 issue of QRP Quarterly on page 6 (shared by K1MBO).
Very simple and low cost.
This particular issue is the free download from their website. 

QRPARCI.ORG

73
KI4QCK


Gerald Bramwell <gerald.bramwell@...>
 

Hi all,
This might be of interest from Banggood:-

CJMCU-5351 Si5351A Si5351 I2C 25MHZ Controller Clock Generator Board 8KHz To 160MHz

Gerald RS49875


Jerry Gaffke
 

I guess Gerald is pointing at this for $11.99:
   https://usa.banggood.com/CJMCU-5351-Si5351A-Si5351-I2C-25MHZ-Controller-Clock-Generator-Board-8KHz-To-160MHz-p-1256227.html?cur_warehouse=CN

They also have this for $3.99, if you can wait for a boat from China:
   https://usa.banggood.com/CJMCU-5351-Si5351A-Clock-Generator-Signal-Generator-Clock-8KHz-160MHz-p-1613912.html

There is cheaper still, Banggood probably buys from somebody here:
   https://www.aliexpress.com/af/si5351a%25252dbreakout%25252dboard.html?d=y&origin=n&SearchText=si5351a-breakout-board


Also check out offerings from EtherKit, Adafruit, and QRPLabs.
https://www.adafruit.com/product/2045
https://www.etherkit.com/rf-modules/si5351a-breakout-board.html
https://www.qrp-labs.com/synth.html

Note that most of those won't have the series caps we want when driving a mixer.

Most of the search hits for "si5351 board" seem to be going to Adafruit.

Si5351's are cheap, but if it's used we need to a microprocessor, display, knobs,
to configure it over the i2c bus.


If all you want is a crystal oscillator at a fixed frequency to check a receiver,
something like Mouser 520-5032MV-80-CNT would be easy to dead bug to a bare PCB.
Simpler than the K1MBO design to get going.

Jerry, KE7ER




On Sun, Oct 25, 2020 at 03:42 PM, Gerald Bramwell wrote:

Hi all,
This might be of interest from Banggood:-
 

CJMCU-5351 Si5351A Si5351 I2C 25MHZ Controller Clock Generator Board 8KHz To 160MHz

Gerald RS49875


Shirley Dulcey KE1L
 

That's an Asian clone of an Adafruit board that costs $7.95. It does include the SMA connectors, which are not included with the Adafruit board, but it's still overpriced. The Adafruit design is open source hardware, so the clones are legal so long as they don't pretend to be actual Adafruit boards. Note the absence of a logo on the Banggood board.

As KE7ER pointed out, similar boards from other Asian sellers (even another Banggood listing) are available for MUCH less than $12. Some of them even include an 8 pin header strip as well as the SMA connectors.

All the boards give their upper frequency limit as 160 MHz because Adafruit first sold them with the original Si5351A, which had that as its limit. (It was not enforced by the chip and they could go somewhat higher.) That was replaced by the Si5351A-B, which is now used in all current boards from all sources; that updated chip has a published limit of 200 MHz and can be pushed considerably higher to nearly 300 MHz. The boards are no more limited to 160 MHz than the bare chip is, which is to say not at all.

It's possible that Asian-made clones of the Si5351A might appear at some point, and those might not reach the same frequencies that the authentic Silicon Labs chips do. Clones of some popular chips have appeared, including microcontrollers, but I have heard no reports to date of clones of SiLabs clock generators. Authentic Si5351A chips in the 10MSOP package cost 92 cents in quantity one from major distributors, so there isn't much incentive to counterfeit them.

If you ever need to buy the chip separately to build or repair something, what you want is specifically an Si5351A-B-GT or Si5351A-B-GTR. No additional numbers; those signify chips with pre-programmed start frequencies and/or alternate I2C addresses. The difference between the two is that the version with the R at the end comes on tape and reel rather than in a tube. Same chips either way, so buy whichever is more convenient for you, available when you order, or cheaper.

On Sun, Oct 25, 2020 at 8:34 PM Jerry Gaffke via groups.io <jgaffke=yahoo.com@groups.io> wrote:

I guess Gerald is pointing at this for $11.99:
   https://usa.banggood.com/CJMCU-5351-Si5351A-Si5351-I2C-25MHZ-Controller-Clock-Generator-Board-8KHz-To-160MHz-p-1256227.html?cur_warehouse=CN

They also have this for $3.99, if you can wait for a boat from China:
   https://usa.banggood.com/CJMCU-5351-Si5351A-Clock-Generator-Signal-Generator-Clock-8KHz-160MHz-p-1613912.html

There is cheaper still, Banggood probably buys from somebody here:
   https://www.aliexpress.com/af/si5351a%25252dbreakout%25252dboard.html?d=y&origin=n&SearchText=si5351a-breakout-board


Also check out offerings from EtherKit, Adafruit, and QRPLabs.
https://www.adafruit.com/product/2045
https://www.etherkit.com/rf-modules/si5351a-breakout-board.html
https://www.qrp-labs.com/synth.html

Note that most of those won't have the series caps we want when driving a mixer.

Most of the search hits for "si5351 board" seem to be going to Adafruit.

Si5351's are cheap, but if it's used we need to a microprocessor, display, knobs,
to configure it over the i2c bus.


If all you want is a crystal oscillator at a fixed frequency to check a receiver,
something like Mouser 520-5032MV-80-CNT would be easy to dead bug to a bare PCB.
Simpler than the K1MBO design to get going.

Jerry, KE7ER




On Sun, Oct 25, 2020 at 03:42 PM, Gerald Bramwell wrote:

Hi all,
This might be of interest from Banggood:-
 

CJMCU-5351 Si5351A Si5351 I2C 25MHZ Controller Clock Generator Board 8KHz To 160MHz

Gerald RS49875


Bob Lunsford
 

Here's a (pricey) idea of what might work bu tfor our purpose, we need a simple oscillator with a controlled output.



A simple 1MHz crystal oscillator with a system of attenuating the output should be rich with harmonics that could be used for signal detection purposes but it would be helpful if there were some kind of modulation options so it is more easily detected aside from merely turning it off and on again. What say others in the group?

Bob — KK5R

On Sunday, October 25, 2020, 8:34:34 PM EDT, Jerry Gaffke via groups.io <jgaffke@...> wrote:


I guess Gerald is pointing at this for $11.99:
   https://usa.banggood.com/CJMCU-5351-Si5351A-Si5351-I2C-25MHZ-Controller-Clock-Generator-Board-8KHz-To-160MHz-p-1256227.html?cur_warehouse=CN

They also have this for $3.99, if you can wait for a boat from China:
   https://usa.banggood.com/CJMCU-5351-Si5351A-Clock-Generator-Signal-Generator-Clock-8KHz-160MHz-p-1613912.html

There is cheaper still, Banggood probably buys from somebody here:
   https://www.aliexpress.com/af/si5351a%25252dbreakout%25252dboard.html?d=y&origin=n&SearchText=si5351a-breakout-board


Also check out offerings from EtherKit, Adafruit, and QRPLabs.
https://www.adafruit.com/product/2045
https://www.etherkit.com/rf-modules/si5351a-breakout-board.html
https://www.qrp-labs.com/synth.html

Note that most of those won't have the series caps we want when driving a mixer.

Most of the search hits for "si5351 board" seem to be going to Adafruit.

Si5351's are cheap, but if it's used we need to a microprocessor, display, knobs,
to configure it over the i2c bus.


If all you want is a crystal oscillator at a fixed frequency to check a receiver,
something like Mouser 520-5032MV-80-CNT would be easy to dead bug to a bare PCB.
Simpler than the K1MBO design to get going.

Jerry, KE7ER




On Sun, Oct 25, 2020 at 03:42 PM, Gerald Bramwell wrote:

Hi all,
This might be of interest from Banggood:-
 

CJMCU-5351 Si5351A Si5351 I2C 25MHZ Controller Clock Generator Board 8KHz To 160MHz

Gerald RS49875


Jerry Gaffke
 

I think new board designs should consider the Si5351 in a QFN-20 package.
Last time I looked, the MSOP-10 package that everybody uses now (including the Raduino)
was around $1, all the QFN-20 variants were $10.  But the QFN-20's are now $2 on Mouser.

The QFN-20 has the following advantages over the usual MSOP-10:
  8 outputs instead of 3.   So could have signal generator outputs for testing on the uBitx.
  4 VDDO pins to supply power to the outputs instead of 1, reducing crosstalk
 
In the QFN-20 packages, you have the following choices:
  The Si5351A-B-GM has an A0 pin to select between a couple i2c chip addresses
  The Si5351B-B-GM has a VC control pin into a digital VCXO at the PLLB position.
  The Si5351C-B-GM has a CLKIN pin that takes an external reference between 10mhz and 100mhz,
    perhaps from a GPS module.  The A and B parts are spec'd only for use with a 25 to 27mhz crystals.

The SSEN pin (spread spectrum enable on Si5351A and Si5351B) is likely not of much interest to a ham.
The INTR pin (interrupt the host on loss of lock,  Si5351C only) might be of interest, 

Unfortunately, the QFN-20 has the same 0.5mm pin pitch as the MSOP-10, which is pretty tight.
The QFN-20 only has one ground pad, which is the large center square pad, could be accommodated
by drilling a large plated through hole, and soldering that big pad from the back of the board.
Or you could use solder paste and a toaster oven.

AN619 for the Si5351 says max VCO is 900mhz, output msynth of 4, so that suggests a max output of 225mhz.
But the datasheet does show 200mhz max as KE1L states, which is pretty fast for a CMOS output buffer.
Hans Summers, G0UPL of QRP-Labs found he could press the internal VCO up toward 1200mhz
to give an output of around 290mhz before it pooped out. 
Some versions of the Nanovna firmware try to run it at 300mhz max, but not all Si5351 parts can do this.

Hans found he could accurately lock two of the outputs in quadrature down to about 3.5mhz.

Some good discussions of the nitty gritty details around the Si5351:
https://groups.io/g/BITX20/message/51935
https://groups.io/g/qrptech/topic/53562857#324

Jerry, KE7ER


Ashhar Farhan
 

Jerry,
We do sell a test set, it is called Antunio!
Except for the dummy load, the Antuino is really all you need to align any HF radio set. It can be used to test any radio well into UHF range as well as individual filters, including duplexers, amplifiers, mixers, etc. It can also measure input or output impedance match, antenna vswr and even act as a vtvm.
End of the plug.
- f



On Mon 26 Oct, 2020, 4:02 AM Jerry Gaffke via groups.io, <jgaffke=yahoo.com@groups.io> wrote:
Page 6 has the test set Allen mentioned:  https://www.qrparci.org/qqsample/qqsample.pdf

A 7.050mhz crystal oscillator with an attenuator suitable for checking out a 40m receiver.
And a 50 ohm dummy load for the transmitter.

That would definitely be better than nothing at all for debugging a uBitx,
but hardly sufficient.

Could tell if the receiver works on 40m.
And could drive the uBitx power amp to check it out on 40m as well
if you do away with the cute trick of using one connector for
both signal generator and dummy load.

Jerry, KE7ER


On Sun, Oct 25, 2020 at 02:25 PM, Allen Hill wrote:
Check out the TX/RX test circuit in January 2015 issue of QRP Quarterly on page 6 (shared by K1MBO).
Very simple and low cost.
This particular issue is the free download from their website. 

QRPARCI.ORG

73
KI4QCK


Bob Lunsford
 

This looks very interesting. One specification was not shown, however. In the Antuino write-up, you show it reading the SWR but also shown is the frequency readout. It then acts as a frequency counter, right? How accurate is it in that mode?

I'm considering getting one. Is it a kit or built and in a box? What kind of connectors does it have? Also, you mention in the write-up "by clicking on the Plot button" so does this mean that the unit works with a computer or is it possibly a touch-screen so you can tap the screen like with the V6...?

Hope you do not mind my questions.

Bob — KK5R

On Sunday, October 25, 2020, 11:30:46 PM EDT, Ashhar Farhan <farhanbox@...> wrote:


Jerry,
We do sell a test set, it is called Antunio!
Except for the dummy load, the Antuino is really all you need to align any HF radio set. It can be used to test any radio well into UHF range as well as individual filters, including duplexers, amplifiers, mixers, etc. It can also measure input or output impedance match, antenna vswr and even act as a vtvm.
End of the plug.
- f



On Mon 26 Oct, 2020, 4:02 AM Jerry Gaffke via groups.io, <jgaffke=yahoo.com@groups.io> wrote:
Page 6 has the test set Allen mentioned:  https://www.qrparci.org/qqsample/qqsample.pdf

A 7.050mhz crystal oscillator with an attenuator suitable for checking out a 40m receiver.
And a 50 ohm dummy load for the transmitter.

That would definitely be better than nothing at all for debugging a uBitx,
but hardly sufficient.

Could tell if the receiver works on 40m.
And could drive the uBitx power amp to check it out on 40m as well
if you do away with the cute trick of using one connector for
both signal generator and dummy load.

Jerry, KE7ER


On Sun, Oct 25, 2020 at 02:25 PM, Allen Hill wrote:
Check out the TX/RX test circuit in January 2015 issue of QRP Quarterly on page 6 (shared by K1MBO).
Very simple and low cost.
This particular issue is the free download from their website. 

QRPARCI.ORG

73
KI4QCK


 

Bob,

It has SMA connectors two of them. One output and one input.

When you press plot, it plots on its own screen. No PC required. No touch screen.

No freq. counter but you can measure indirectly in power mode.

It comes fully complete in a metal box..

Raj


At 26/10/2020, you wrote:
This looks very interesting. One specification was not shown, however. In the Antuino write-up, you show it reading the SWR but also shown is the frequency readout. It then acts as a frequency counter, right? How accurate is it in that mode?

I'm considering getting one. Is it a kit or built and in a box? What kind of connectors does it have? Also, you mention in the write-up "by clicking on the Plot button" so does this mean that the unit works with a computer or is it possibly a touch-screen so you can tap the screen like with the V6...?

Hope you do not mind my questions.

Bob — KK5R

On Sunday, October 25, 2020, 11:30:46 PM EDT, Ashhar Farhan <farhanbox@...> wrote:


Jerry,
We do sell a test set, it is called Antunio!
Except for the dummy load, the Antuino is really all you need to align any HF radio set. It can be used to test any radio well into UHF range as well as individual filters, including duplexers, amplifiers, mixers, etc. It can also measure input or output impedance match, antenna vswr and even act as a vtvm.
End of the plug.
- f



On Mon 26 Oct, 2020, 4:02 AM Jerry Gaffke via groups.io, <jgaffke=yahoo.com@groups.io > wrote:
Page 6 has the test set Allen mentioned:  https://www.qrparci.org/qqsample/qqsample.pdf

A 7.050mhz crystal oscillator with an attenuator suitable for checking out a 40m receiver.
And a 50 ohm dummy load for the transmitter.

That would definitely be better than nothing at all for debugging a uBitx,
but hardly sufficient.

Could tell if the receiver works on 40m.
And could drive the uBitx power amp to check it out on 40m as well
if you do away with the cute trick of using one connector for
both signal generator and dummy load.

Jerry, KE7ER


On Sun, Oct 25, 2020 at 02:25 PM, Allen Hill wrote:
Check out the TX/RX test circuit in January 2015 issue of QRP Quarterly on page 6 (shared by K1MBO).
Very simple and low cost.
This particular issue is the free download from their website.

QRPARCI.ORG

73
KI4QCK


Bob Lunsford
 

Thanks, Raj. I'm still wondering if you press the word "plot" on the screen or if there's a button on the case that is not visible. If you press "plot" on the screen, it must be sensitive to the finger's capacitance and goes to the plotting the signal. Regardless, I'm wondering if you "measure [frequency] indirectly in power mode," is it very accurate? Considering the many decimal places shown on the screen, I'm guessing it's pretty accurate. I see it's measuring in the HF range but does it also work at 2M for measuring, for example, the SWR of an antenna?

It's a box full of capabilities and must be of value when added to the tools in one's station. I'm very interested. I no longer buy a car on the basis of a sales brochure and also want to know as much as possible when buying test equipment.

Thanks again for your very helpful feedback.

Bob — KK5R

On Monday, October 26, 2020, 1:12:05 AM EDT, Raj vu2zap <rajendrakumargg@...> wrote:


Bob,

It has SMA connectors two of them. One output and one input.

When you press plot, it plots on its own screen. No PC required. No touch screen.

No freq. counter but you can measure indirectly in power mode.

It comes fully complete in a metal box..

Raj


At 26/10/2020, you wrote:
This looks very interesting. One specification was not shown, however. In the Antuino write-up, you show it reading the SWR but also shown is the frequency readout. It then acts as a frequency counter, right? How accurate is it in that mode?

I'm considering getting one. Is it a kit or built and in a box? What kind of connectors does it have? Also, you mention in the write-up "by clicking on the Plot button" so does this mean that the unit works with a computer or is it possibly a touch-screen so you can tap the screen like with the V6...?

Hope you do not mind my questions.

Bob — KK5R

On Sunday, October 25, 2020, 11:30:46 PM EDT, Ashhar Farhan <farhanbox@...> wrote:


Jerry,
We do sell a test set, it is called Antunio!
Except for the dummy load, the Antuino is really all you need to align any HF radio set. It can be used to test any radio well into UHF range as well as individual filters, including duplexers, amplifiers, mixers, etc. It can also measure input or output impedance match, antenna vswr and even act as a vtvm.
End of the plug.
- f



On Mon 26 Oct, 2020, 4:02 AM Jerry Gaffke via groups.io, <jgaffke=yahoo.com@groups.io > wrote:
Page 6 has the test set Allen mentioned:  https://www.qrparci.org/qqsample/qqsample.pdf

A 7.050mhz crystal oscillator with an attenuator suitable for checking out a 40m receiver.
And a 50 ohm dummy load for the transmitter.

That would definitely be better than nothing at all for debugging a uBitx,
but hardly sufficient.

Could tell if the receiver works on 40m.
And could drive the uBitx power amp to check it out on 40m as well
if you do away with the cute trick of using one connector for
both signal generator and dummy load.

Jerry, KE7ER


On Sun, Oct 25, 2020 at 02:25 PM, Allen Hill wrote:
Check out the TX/RX test circuit in January 2015 issue of QRP Quarterly on page 6 (shared by K1MBO).
Very simple and low cost.
This particular issue is the free download from their website.

QRPARCI.ORG

73
KI4QCK


Tom, wb6b
 

On Sun, Oct 25, 2020 at 05:34 PM, Jerry Gaffke wrote:
They also have this for $3.99, if you can wait for a boat from China:
Jerry,

Thanks for pointing those out. I just ordered some for my parts supply. Looks like the price is a 50% off promo price. 
For just a little more in shipping, less Than $0.50, it can ship Banggood Express, that is supposedly faster than Air. Still 15-20 days though.

Of course, I ended up buying some other things like a $1.99 stereo audio amp board and an 8 relay board (Might be good for switching between several pre-tuned matching networks at my antenna. A cheap remote "auto" match for multi band operation on one antenna.) And a tiny OLED display module to remind me that bigger displays are easier to read. 

On the subject of the test set. Would it be possible to add some jumper plugs to break the circuits at critical points or jumper across parts of the circuits and test points to the uBitx and produce a sketch that could be loaded into the Raduino for the express purpose of alignment and troubleshooting, assuming the uBitx is mostly functional. The idea would be to allow the built-in Raduino to be a signal generator for one section of the uBitx at a time, and possibility use a simple DVM and a RF detector probe to verify passband and centering of filters and such. You may still want an accurate signal source to verify the final frequency calibration. That could just be a simple xtal oscillator. 

Tom, wb6b


 

Bob,

you have to use the tuning knob to scroll through the menu, set the center freq and span.
then highlight plot and press the encoder knob to start the plot. In the older version it a
single scan but newer one does continuos scan just like an expensive spectrum analyser.

If I need to measure the freq, then I would feed the signal and do a sweep. Then center
the cursor on the peak and get a reading.. not very accurate but will do on many an
occasion. Error will be a few KHz.. the width of the IF filter in the antuino.

It works on VHF with no problems and is usable to 500MHz.

The main thing is hackable.. change what you want in hardware and software.

Raj

P.S we have an antuino group on groups.io

At 26/10/2020, you wrote:
Thanks, Raj. I'm still wondering if you press the word "plot" on the screen or if there's a button on the case that is not visible. If you press "plot" on the screen, it must be sensitive to the finger's capacitance and goes to the plotting the signal. Regardless, I'm wondering if you "measure [frequency] indirectly in power mode," is it very accurate? Considering the many decimal places shown on the screen, I'm guessing it's pretty accurate. I see it's measuring in the HF range but does it also work at 2M for measuring, for example, the SWR of an antenna?

It's a box full of capabilities and must be of value when added to the tools in one's station. I'm very interested. I no longer buy a car on the basis of a sales brochure and also want to know as much as possible when buying test equipment.

Thanks again for your very helpful feedback.

Bob — KK5R