Re: Ebay green and black ADF4351 module performance
Completed the tuning of the green modules and they are a massive improvement over the black modules.
The output signal level is at least 5dB higher.
The 25MHz, 50MHz, 75MHz,etc? spurs are no longer there
Noise is still a bit worse but is secondary to the reduction in spurs.
Now I need to get some ultra low noise 3.3V regulators
|
The benefit of switchable attenuation for spectrum analyzer measurements
During measurements there may be certain spurs that do not have an obvious cause. Are they caused by limitations of the SA? Or are they present in the input signal? An example is this two tone measurement of the input IIP3 of a mixer  The many spurs below -70dB are cause by bad shielding of the two signal generators. Without these connected the noise at about -100dB is without spurs The SA automatically finds the peaks and calculates the input IP3 in two independent ways, the results should be equal but there is some difference. Left IIP3 is calculatec at +9dB where right IIP3 is calculated at +7.7dB But can we be sure the IIP3 of the mixer is indeed around +8dB? The simplest way to know is the add attenuation before and after the mixer. Attenuation after the mixer did not change anything (as it should) but -10dB attenuation before the mixer resulted in a very different picture. The measured levels are increase by the level of attenuation to keep the displayed levels equal so the noise floor moves up about 10dB  The results (15dB improvement of IIP3)? is not entirely what was expected as every dB reduction of the input signal level should? increase the IIP3 with one dB. There is still more to investigate and learn.
|
Hello
Earlier post regarding DC analysis of step attenuators was in relationship to design
and building jumper-based attenuators.?
In the past I have built step attenuators using DPDT slide switches and DPDT rocker switches, but over time and with frequent use these usually develop Instabilities, especially if the switches
are of low quality.? Present approach is to use simple jumpers in place of the switches.? Prototype
build and testing seems to indicate that these may be better for long-term use on the test bench.??
|
Re: Very basic tests of step attenuator
Alan
Fantastic!? That information is what I was looking for when I posted the original question. Now to work up some ADC code for an Arduino so the testing process can be automated.?
This is all related to making a few step attenuators using SMT parts.? Tolerances (and
probability of errors) seems to warrant a testing procedure to verify each attenuation step. Higher attenuation steps may be beyond the Arduino ADC resolution range but I can
work around that by adding DC amplification.
Thanks,
Arv? K7HKL _._
On Fri, Mar 29, 2019 at 6:08 AM Alan < g8lco1@...> wrote: The attenuation at DC is exactly the same as at LF as the stray capacitance? and inductance will? be insignificant up to around 10MHz .? I would suggest that a 50 ohm resistor is inserted in series with the power supply to the? input?? and the output has a 50 ohm resistor to ground.? I would use two 1% 100 ohm? ? W ??wire ended metal film resistors in series for each termination. ? If the direct voltage is 2V to the? input side termination and you use a DMM , set the voltage across the 50 ohm series R to 1.000V then the voltage across the input to the atten gives? the ratio of the input resistance to the 50 ohm resistor ?to around 1% accuracy. The calibration of the DMM is unimportant, the linearity of the DMM should be very good, much better than it¡¯s accuracy. ? Then you can check the voltage ratio between the input and output to get the attenuation.? Turn the attenuator around for? checking the output side input resistance. ? The limitation of this method is thermal voltages adding or subtracting from voltages. As these are around a few micro Volts and you are using around a Volt of drive? the errors are very small for home use. Thermal emf¡¯s ?can be cancelled ?by reversing the supply Heating in the resistors is easy enough, with 1V across 50 ohms it¡¯s 1V * 20mA=20mW. At ?2.236V you reach 100mW, 3.535V is 0.25W,? 5.0V is 1/2W ( 5*0.1 W)? 7.707V( root 50)= 1W? so 2V should be safe without too much thinking. ? It¡¯s the linearity of the DMM that matters not the attenuator ( if you don¡¯t change ranges ). That should be better than 1 in a thousand? reading around half scale on a 1999 count meter. ( cheap meters are dual ?slope) so reading a 20dB pad to about? 1%.of ratio. A 19999 count would read the attenuation to 0.1% of ratio ?. I would be inexpensive to invest in some 0.1% 100 ohm resistors for many reasons! ? Alan G8LCO ? ? ? ? ? Sent from for Windows 10 ? ? Can I test the DC attenuation of a step attenuator by using just a DC source voltage and DC voltmeter across the 50 ohm termination? This would not validate LF to HF performance but just the DC characteristics. The math seems to indicate that this is possible and that it is maybe a rather simple
|
Re: Very basic tests of step attenuator
The attenuation at DC is exactly the same as at LF as the stray capacitance? and inductance will? be insignificant up to around 10MHz . ?I would suggest that a 50 ohm resistor is inserted in series with the power supply to the? input?? and the output has a 50 ohm resistor to ground.? I would use two 1% 100 ohm? ? W ??wire ended metal film resistors in series for each termination. ? If the direct voltage is 2V to the? input side termination and you use a DMM , set the voltage across the 50 ohm series R to 1.000V then the voltage across the input to the atten gives? the ratio of the input resistance to the 50 ohm resistor ?to around 1% accuracy. The calibration of the DMM is unimportant, the linearity of the DMM should be very good, much better than it¡¯s accuracy. ? Then you can check the voltage ratio between the input and output to get the attenuation.? Turn the attenuator around for? checking the output side input resistance. ? The limitation of this method is thermal voltages adding or subtracting from voltages. As these are around a few micro Volts and you are using around a Volt of drive? the errors are very small for home use. Thermal emf¡¯s ?can be cancelled ?by reversing the supply Heating in the resistors is easy enough, with 1V across 50 ohms it¡¯s 1V * 20mA=20mW. At ?2.236V you reach 100mW, 3.535V is 0.25W,? 5.0V is 1/2W ( 5*0.1 W)? 7.707V( root 50)= 1W? so 2V should be safe without too much thinking. ? It¡¯s the linearity of the DMM that matters not the attenuator ( if you don¡¯t change ranges ). That should be better than 1 in a thousand? reading around half scale on a 1999 count meter. ( cheap meters are dual ?slope) so reading a 20dB pad to about? 1%.of ratio. A 19999 count would read the attenuation to 0.1% of ratio ?. I would be inexpensive to invest in some 0.1% 100 ohm resistors for many reasons! ? Alan G8LCO ? ? ? ? ? Sent from for Windows 10 ?
toggle quoted message
Show quoted text
From: Arv EvansSent: 27 March 2019 20:01 To: [email protected]Subject: [HBTE] Very basic tests of step attenuator ? Can I test the DC attenuation of a step attenuator by using just a DC source voltage and DC voltmeter across the 50 ohm termination? This would not validate LF to HF performance but just the DC characteristics. The math seems to indicate that this is possible and that it is maybe a rather simple
|
Re: Useful as RF signal generator?
I did not measure the black boards, only got one of the green boards so far.
Some hints to reduce spurs:
1. Use as high Fcomp as possible.This is done by using as low R
divider as possible. A ADF4351 can use Fcomp on 32MHz, this isn't that
much higher than 10MHz, but any move helps.
2. Use a hard limiter on the Ref input. This is specially important if
you are using a DDS or Si5351 to generate adjustable frequency. A hard
limiter can with some adjustments supress unwanted spurs as much as
20dB. This means also driving the PLL input as hard as the datasheet
allows to reduce spurs.
3. Use a bandpass filter on the Fref input of the module. This way you
reduce the amplitude of the spurs outside the bandwidth of the filter,
should be combined with 2 above.
4. Reduce the loop filter bandwidth as much as you can. The ideal
thing is to optimize this for best PN, but that requires a repeatable
setup to measure PN over a large frequency range, and most people
don't have that.
5. Follow the power supply requirements, and use low noise regulator.
Large capacitors and Capacitance multiplier is a easy way to reduce
noise on the supply lines. This is more to improve phase noise than
spurs, but crosstalk can generate spurs. Having the VCO on a separate
supply rail than the digital stuff helps.
ADF5355 looks like a nice approach to reduce the amount of work in
making a signal generator. A bit expencive, but should be interesting.
There are some chinese boards around, but its probably worth it,
altough more expencive, to make your own. That way you would get the
benefit of knowing that things are done correctly with regards to the
power supplies.
73 de Thomas LA3PNA.
ons. 27. mar. 2019 kl. 12:45 skrev <erik@...>:
toggle quoted message
Show quoted text
@Thomas
On Sat, Mar 23, 2019 at 03:47 PM, Thomas S. Knutsen wrote:
Most of the black boards have better RF design than the green one. The
green ones have some part of the PLL loop filter running under the
capacitors that AC couples the output.
Did you measure black boards? If so, what spur performance did you get, in particular the ref clock spur is degrading performance to unacceptable level. Can you suggest improvement options?
--
With Best regards, Thomas S. Knutsen.
Please avoid sending me Word or PowerPoint attachments.
|
Re: Ebay green and black ADF4351 module performance
On Wed, Mar 27, 2019 at 08:42 AM, <erik@...> wrote:
The instability on +5dB output level was caused by reflections from an unused output.
When properly terminated the instability disappeared and spur level reduced.
Was just reading about pass through terminators (high impedance DUT to lower impedance test equipment input) and also the typical terminators that are specific to the impedance of the DUT and test equipment.? There was an EEVblog#652 video recommendation on the Tekscopes groups.io group and I found even more information on the EEVblog#652 Forum discussion and examples with other good videos.? Not quite stripline or microstrip theory (if that's really applicable to PCB design for UHF on down)... though gets the idea.?
|
Re: Very basic tests of step attenuator
That is correct, in a regular, resistive attenuator with resistor, you can determine the attenuation by taking 3 measurements and solving for 3 unknown resistors. 73 de Thomas LA3PNA. ons. 27. mar. 2019 kl. 21:01 skrev Arv Evans <arvid.evans@...>:
Hello
Can I test the DC attenuation of a step attenuator by using just a DC source voltage
and DC voltmeter across the 50 ohm termination?
This would not validate LF to HF performance but just the DC characteristics.
The math seems to indicate that this is possible and that it is maybe a rather simple
spreadsheet application.
Arv K7HKL
_._
--
With Best regards, Thomas S. Knutsen.
Please avoid sending me Word or PowerPoint attachments.
|
Very basic tests of step attenuator
Hello
Can I test the DC attenuation of a step attenuator by using just a DC source voltage
and DC voltmeter across the 50 ohm termination? This would not validate LF to HF performance but just the DC characteristics. The math seems to indicate that this is possible and that it is maybe a rather simple
spreadsheet application.
Arv? K7HKL _._
|
Re: Ebay green and black ADF4351 module performance
First problem solved.
The instability on +5dB output level was caused by reflections from an unused output.
When properly terminated the instability disappeared and spur level reduced.
Now into loop filter design
|
Re: Useful as RF signal generator?
@Thomas
toggle quoted message
Show quoted text
On Sat, Mar 23, 2019 at 03:47 PM, Thomas S. Knutsen wrote:
Most of the black boards have better RF design than the green one. The
green ones have some part of the PLL loop filter running under the
capacitors that AC couples the output.
Did you measure black boards? If so, what spur performance did you get, in particular the ref clock spur is degrading performance to unacceptable level. Can you suggest improvement options?
|
Re: Ebay green and black ADF4351 module performance
These are the steps I tried till now:
- Decoupling VCO and output supply: No improvement yet??(not yet improving phase noise)
- Replacing the long charge pump trace with shielded connection: No impact for now (not yet improving phase noise)
Decoupling the internal 10MHz XCO and using an external reference showed weird locking behavior. The PLL will only lock on certain reference frequencies around 10MHz but on these frequencies the spur performance is much (-40dB) better.
Next step is to calculate and replace the loop filter components and to enable the fast locking functionality to have extra loop filtering when locked
|
Re: Interest in ADF4351 evaluation SW from AD for Arduino interface?
Yeah I¡¯m interested¡ ? ? Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ ? Owner - Operator Big Signal Ranch ¨C K9ZC Staunton, Illinois ? Owner ¨C Operator Villa Grand Piton ¨C J68HZ Soufriere, St. Lucia W.I. Rent it: Like us on Facebook! ? Moderator ¨C North American QRO Group at Groups.IO. ? email:? bill@... ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of erik@...
Sent: Tuesday, March 26, 2019 4:36 AM
To: [email protected]
Subject: [HBTE] Interest in ADF4351 evaluation SW from AD for Arduino interface? ? AD has create a nice open source SW package for evaluation of the ADF4351.
It is intended for use with the AD ADF4351 evaluation board but these boards are very expensive and use a complex FPGA to interface between USB and the ADF4351.
I modified the ADF4351 evaluation SW to make use of an Arduino based interface between USB and the AD4351 so you can test the chinese ADF4351 boards
As Arduino you can use a 5 volt version such as a nano or uno with level shift resistors or a 3.3 volt Arduino such as the zero without level shifters
If there is interest I can make the modified evaluation SW and the Arduino SW available to this group.
|
Re: Interest in ADF4351 evaluation SW from AD for Arduino interface?
|
Re: Interest in ADF4351 evaluation SW from AD for Arduino interface?
I'd like to have a copy and details relating to the mod since seems interesting from a programming exercise and for use also once I have the ADF4351 and other modules as I assume will function with other ADF modules.?
Great to read and see you worked on as I was wondering about the ADF4371 capabilities to create a simpler evaluation like cost effective board.? ?Awesome man!
|
Interest in ADF4351 evaluation SW from AD for Arduino interface?
AD has create a nice open source SW package for evaluation of the ADF4351.  It is intended for use with the AD ADF4351 evaluation board but these boards are very expensive and use a complex FPGA to interface between USB and the ADF4351. I modified the ADF4351 evaluation SW to make use of an Arduino based interface between USB and the AD4351 so you can test the chinese ADF4351 boards As Arduino you can use a 5 volt version such as a nano or uno with level shift resistors or a 3.3 volt Arduino such as the zero without level shifters If there is interest I can make the modified evaluation SW and the Arduino SW available to this group.
|
Ebay green and black ADF4351 module performance
I was hoping the many via's on the green module where a sign of craftsmanship but I was disappointed.
The black module generates many substantial spurs when used above 2GHz.
The green module generates even 20dB louder spurs and when the output is driven at full power its no longer stable.
The only advantage of the green module seems to be its a two layer design so it will be easier to modify the board to decouple the VCO supply from the rest of the supplies.
More work to do.....
|
Re: eBay SA complete, suggestions for improvement?
I am asking Neils. It was his message I quoted/included in mine
and he stated (in that message) that he is using a PI to control
his specan.
toggle quoted message
Show quoted text
Not sure who you are asking.
My sw is also on windows using .net.
|
Re: Useful as RF signal generator?
Most of the black boards have better RF design than the green one. The
green ones have some part of the PLL loop filter running under the
capacitors that AC couples the output.
A 2 layer design should be sufficient, if done correctly, although a 4
layer design is probably easier to get working. Do pay attention to
the layout and neither should be a problem.
73 de Thomas.
l?r. 23. mar. 2019 kl. 23:08 skrev jafinch78 . <jafinch78@...>:
toggle quoted message
Show quoted text
I was thinking the cheap Chinese chips might be counterfeits also and not meeting specifications of the actual AD chips.
Based on eBay policy... I think these can be refunded and not have to be returned also.
I was wondering about the board design too and wondering if needed to can in more a stripline or microstrip type enclosed design or technically... if I understand correctly use multilayer PCB designs and similar theory to spec the board more critically... versus just dual layer.
Is there something going on with the black board versus green board designs in general or more a board design performance for each module... I'm probably trying to over generalize?
--
With Best regards, Thomas S. Knutsen.
Please avoid sending me Word or PowerPoint attachments.
|
Re: Useful as RF signal generator?
The below article got me thinking also about not just multilayer PCB design for shielding better with the stripline or microstrip methods thinking... also the layered designs for the traditional components to embed... which isn't that complicated and like double sided boards if using an oven with different melting point solders so highest melting point components soldered first.?
|
jafinch78 .
K9HZ