Few days ago I got AD8302's evaluation board and firstly I'd like asking if there are examples for getting started with this board...
My first question is..
Any signal that I fed at INPA or INPB must be in the range of 223 uV to 223mV??
Yes, the input range on INPA and INPB is from 223 uVrms to 223 mVrms, which is equal to 316 uVp to 316 mVp, or 632 uVpp ot 632 mVpp...you have to be careful which voltage you specify.
The eval board should be set up such that all you have to do is power the board up, and apply to input signals, and it should give give a magnitude slope of 30 mV/dB and a phase slope of 10 mV/degree. So you if change the amplitude of one signal by one dB, and leave the other signal's magnitude constant, the output VMAG should change by 30 mV. Similarly, if you change the respective phase by 1 degree, VPHS should change by 10 mV.
the result Hi,
When I have no input signals... should I get something at Vmag and Vphs pins?? Or ideally the result is 0V??
If the signals are the same – in this case noise – then the magnitude ratio should be 0 dB and Vmag should be about 0.9V. This can be seen on TPC 1 in the datasheet.
Again when both input signals are noise, the Vphs should be about 1.7V as seen in TPC 25 in the datasheet.
If you are measuring 0V on either of those outputs with no signal on the input, something is wrong with your setup or you have a bad part.
Hope this helps,
From: Mresau19 firstname.lastname@example.org
Sent: Monday, May 02, 2011 7:24 AM
To: Dobler, Joel
Subject: New message: "AD8302 -EVAL"
Analog Devices EngineerZone<http://ez.analog.com/index.jspa>
reply from Mresau19<http://ez.analog.com/people/Mresau19> in RF Components - View the full discussion<http://ez.analog.com/message/24159#24159
Thankyou, I asked you that question because of I'm using a RC test circuit for measuring phase and magnitude, but when I put the DC Voltimeter I see a variable voltage. I'm applying an input signal with 100 mVRMS (according to the range 223uV - 223 mV ) and 1 kHz .... I was thinking the DC voltimeter measures noise, I don`t know.... Is supposed that Vmag and Vphs is direct voltaje, isn't it?? Thus, I can´t see the slope 30 mV/dB
You will measure a voltage, the change in the output voltage corresponding to a change in input amplitude should be 30 mV/dB assuming the other input is not changing.
Thankyou, can you explain me an easy way to verify those 30 mV/dB. Rigth now I have a function generator, AD8302 -EVAL and osciloscope. Please explain me step by step how to probe this board, what arragements I need to do... Please it is important maybe I have no idea how to get good results so I'd like to follow your instructions.
I saw in datasheet the equation 8a:
Vmag= RFISLP log (Vina / Vinb) + VCP
Then if Vina = 150 mVRMS, Vinb = 600 uVRMS. Is correct this way for solving that equation:
RFISLP = 30 mV/dB
VCP = 900 mV
Vmag = (30 mV /dB) * log (150 mVRMS/ 600 uVRMS) + 900 mV = 971 mV
Is this procedure OK?? I'd like to compare the theorical values vs measured values...
Please, explain me an easy way for getting this measured values in AD8302-EVAL.
I don’t think that equation is actually how the part is implemented…or there is an error in the datasheet. I’ll have to get back to you on that…
I think my last response is correct, because if you look at TPC 1, if you have a +6 dB magnitude ratio (twice the voltage), you’ll get a little less than 1100 mV. It’s a little hard to extract the exact values from the graph.
And if you have a -6 dB (half the voltage) magnitude ratio, you get about 700 mV.
Which is a lot different than:
30 mV/dB * log10(2) + 900mV = 909 mV, or
30 mV/dB * log10(0.5) + 900 mV = 891 mV
Sent: Monday, May 02, 2011 5:41 PM
reply from Mresau19<http://ez.analog.com/people/Mresau19> in RF Components - View the full discussion<http://ez.analog.com/message/24223#24223
If your function generator has amplitude control in dB, then it should be easy: change the amplitude on one signal by 1 dB and the output should change by 30 mV.
If your function generator only has amplitude control in volts, then it’ll be a little harder:
1) Make both signals the same amplitude; you should see 900 mV at the output.
2) Now make one of the signals twice as large in voltage; the output should change by 6 dB * 30mV/dB = 180 mV. Depending on which input signal you change, will determine whether the output is 1080 mV or 720 mV.
Sent: Monday, May 02, 2011 2:33 PM
reply from Mresau19<http://ez.analog.com/people/Mresau19> in RF Components - View the full discussion<http://ez.analog.com/message/24203#24203
Joel, considering your experiment what is the frequency of your input signals for getting the 1080 mV or 720 mV??
You should be able to do this experiment at any frequency of operation that the AD8302 is specified at. The linearity doesn’t start to have appreciable error until the bigger magnitude ratios, of 15 (-15) and above. Please see TPC 1 in the datasheet.
Sent: Wednesday, May 04, 2011 11:06 AM
reply from Mresau19<http://ez.analog.com/people/Mresau19> in RF Components - View the full discussion<http://ez.analog.com/message/24413#24413
Few days ago I saw the eval-board's performance and the results that I obtained was diferent respect to datasheet and I don't know why????. I send you an excel file with this measurements. The coaxial cables used are RG-58 and two function generators wich have attenuation's control in decibels.
Please let me know exactly what is performing different from the datasheet verses on the eval board you received. Also, a schematic of your setup is useful.
Sent: Monday, June 06, 2011 7:31 AM
reply from Mresau19<http://ez.analog.com/people/Mresau19> in RF Components - View the full discussion<http://ez.analog.com/message/26280#26280
The schematic is the circuit of measurement mode... I'm interested in the range 1 kHz - 10 MHz with this board, although the excel file that I sent you only referrs to frequencies from 10 kHz to 10 MHz...... However this board has Cc=1nF with fhp = 2MHz so my question is can I consider the results listed on excel file despite they don't correspond to 10mV/º and 30mV/dB?????
Is supposed that eval board is designed for working at any frequency isn't it???
Another question that I though:
Is it OK if I use RG-58 coaxial cables??? I got this cables due to they have an impedance of 53 ohms so I believed the eval board needed to be coupled at 50 ohms so I decided to use this cables......
Or do I need ANOTHER coaxial cable with diferent impedance or simply any screened cable????
I asked this situation because I saw the input signals (generator's signals) get attenuated when I put the RG-58 coaxial cables referred to INPA and INPB??
Then probably they may cause the eval board doesn't work very well and the measurements that I sent you in the excel file are the result of this situation...
Can I use the ad8302-eval in teh range 1 kHz - 10 MHz???
You should be able to use the AD8302 down to 1kHz by placing big enough caps on C1 and C5 for the input signal coupling and C4 and C6 for the input offset correction loops of each log amp. 1uF should be big enough for all caps, but I might try 10uF on C4 and C6 just to get the high pass corner down low enough that it won't effect your signals much.
We have set up the AD8304 EVAL as a frequency discriminator. The inputs are at 1920Mhz plus or minus a differential frequency on one input. We get the expected 1.8 Vpp "triangle" waveform at the phase VPHS output up to about 2-3Mhz with no capacitor at the PFLT pin. With a 10pF cap at PFLT the amplitude starts to drop at output frequencies well below 1MHz. Looking at the datasheet "Small Signal Bandwidth" is 30Mhz, however the slew rate is 25V/uS which seems to be a limitng factor. What conditions need to be met for a constant 1.8Vpp "triangle" output waveform up to 5MHz or better?
Retrieving data ...