| ??? 01/12/06 20:21 Read: times |
#107198 - Updated Schematic Responding to: ???'s previous message |
I know you have spend lots of time and efforts, but unfortunately the circuit needs some changes: Thanks Dear Kai, But definitely I'll have to sped a lot more and I could not get here without your help. But the buffer following R5 and R6 needs to have a much higher input impedance than what an AD844 can offer. So, please use the FET-buffer at this place, shown in the schematic of bitscope. There's hardly an alternative to such a FET-buffer at this place. Non-Inverting Input resistance of AD844 is specified at 7M ohm min. and 10M ohm typical ,Problem here that you are talking about should be that we are supplying this 10M ohm input by a High Impedance signal ? In the Bitscope website a JFET buffer has been used. I remember that the author said A very High impedance input(~10^12 M ohm) OP-Amp have to be used after this JFET buffer since they are not good at providing enough output current, So i omitted the idea and fed the input signal directly to the Op-Amp input like what DSOA mk3 and Bitscope BT300 did. My hesitate about using the FET input buffers is because they may need a lot of knowledge to design and their performance may degrade drastically over the time and temperature range. What do say about these problem ? f you want to handle a signal range of 0.01V to 5V in 1:2:5 steps, as it common practice with standard scopes, then you need three separate dampening paths in parallel at input of FET-buffer, which provides a dampening of 1, 10 and 100. This also, because the AD844 being used later in the signal chain can only handle signals smaller than +-1V at input So yo mean , we divide the input signal So we would try to have not more than 0.5 Vp-p at the first Op-Amp's input ? By the X2 gain of the Op-Amp then We will have Max. 1Vp-p at It's output Then this would be divided by 1 or 2 or 5 and fed to the second Op-Amp. This should give us these Voltage swings at input : +/- 50mv (at max x20 gain) ,+/- 100mv, +/- 250mv, +/- 250mv, +/-500mv, +/-1v, +/-2.5v, +/-5v, +/-10v, +/-25v (at Gain X 2/50) Thanks for your great suggestion. This will give me up to .3mV precision (in X20 Gain). About the Offset Shifting, My 20 MHz standalone Scope have a ground adjust dial. Sometimes i fed a 12VDC to my scope with V/Div placed at the 10mV section .Then I lower the signal till it gets visible on the screen. This gives me a nice way to see how much noise exists in this signal. I like to have this feature in This DSO , I know i could do this by using the AC Input but I think maybe this feature could be useful somewhere, but I can't think of anything right now ! :) This should be hard to implement I think ? Here is the new Design : http://www.8052.com/users/far...pdated.GIF - I have connected Both the input and output unused pins of the 4052 to the Analog ground , is it alright ? - By removing the 'additional cap between inverting input and signal ground' then we have lower bandwidth .I would like to have the possibility to increase the bandwidth , So could i do this later by soldering a simple Cap ? (After this much effort i got interested into 40MSPS TDA8703) - Offset section is incomplete also , please verify if the resistor/capacitor values are correct. - FET buffer is incomplete ,cause I don't know how to build a FET amplifier. What N-channel FET i have to look for ? I'll take a look at some providers website to see what I could find available. - Are the Resistors' value In feedback of the Op-Amps correct ? Again I appreciate all your help, Farshid |
