Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

How to remove Pulse Ripples and Noise in the Circuit

Status
Not open for further replies.

sulemanzp

Newbie level 5
Joined
Apr 3, 2010
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
lahore
Activity points
1,378
I am working on HMCAD1520, for this I have to create a two stage input circuit, which translates 0v to 5v signal to -1v to +1v.
For this I created a circuit (PDF attached in the end) with AD8056 for -1 offset (adjustable with +/- 200mV), AD8009 for amplitude mapping (0v <-> 5v signal to 0v <-> 2v) and LMH6553 for single to differential driver.

The circuit is working fine, the only issue is ripples and noise in the circuit (input and output both sides) and these are not acceptable for me in the ADC.
I confirm that the Pulse Generator is fine, and outputting signal without ripples or noise.
Below i am attaching the screen shots of sin wave and pulses as well.

In all pictures Blue is Input and Yellow is Differential Output.
Sin Wave but noise is there.
sine_wave_1.png
Sin Wave 50ns/Div, clearly i can see the noise.
sine_wave_50nsdiv.png

Pulse Width 100ns, Look at the ripples.
pulse_ripple.png
couple of pulses
pulse_ripples.png

10 ns pulse width
pulse_50ns.png


I need help to remove the ripples, What i should do (add, remove or replace ) in the circuit to remove the ripple and what could be the cause for this.

Note: Schematic Updates:
R11 = R9 = 1000R,
R13 = R12 = 400R,
R1 = R2 = R3 = 0R soldered.

Thank you experts.
 

Attachments

  • SCHEMATIC1 _ PAGE1.pdf
    18.8 KB · Views: 156

Hi,

Please define your input signal (source): Frequency range, offset and amplitude.

The AD8009 has a signal gain of about 2. I wonder why you say 5Vpp --> 2Vpp

****
Pulses:
Maybe you expect that a square wave input impulse should result in a clean square wave output impulse.
There I see problems.
A square wave (let´s talk about 50% duty cycle, 10ns ON, 10ns OFF) has a lot of overtones.
The fundamental is 20ns = 50MHz (sine). And then there are 3rd, 5th, 7th, 9th, 11th... overtones, up to unlimited frequency.
The phrase "unlimited frequency" makes it obvious: It is impossible.
So in reality your HF signal runs through two amplifiers. Buth have limited frequency response. Therefore overtones are missing. The result is ringing.

****
Noise and distortion:
Your signal runs through two amplifiers. Both have limited frequency response, both introduce distortion, both introduce noise...
Therefore I´d try to use only one stage. I´d say you can do without the AD8009.

Noise:
Your OFFSET circuit has no low pass filter. Therefore every introduced noise will be amplified and added to the HF signal.
Noise sources are: +5V supply, -5V supply, pot, AD8056.
I strongly recommend NOT to use supply voltages to generate a reliable offset. Any drift in supply voltage (let´s say 2%) will generate a change in offset voltage (also 2%).
Use V_ref of the ADC instead, if available. Use low pass filters with fc of less than 10Hz.

****
Offset:
The LMH doesn´t tell about offset voltage and offset voltage drift.
I´d try to avoid the analog offset adjust circuit. Just add enough offset (generated from VRef), so that your signal is within the ADC input range. Then do the offset adjust on the digital side.
A simple "add". It is the most reliable way. No drift, no noise.

****
Add a capacitor at VCom, like shown in th LMH datasheet. In case you don´t need it you just don´t assmemble it.

****
Misc:
Your impedance matching of AD8009 circuit is not optimal.
Especially AD8009 is at the top end of it´s frequency, expect distortions and amplitude errors.
Maybe AD8009 and LMH6553 influence each other and decrease overall stability.
With those high frequencies you need a very good PCB layout. Good decoupling of supply voltages.

If possible: Add a small capacitor (10pF ...33pF) at each ADC input to GND. Especiall when there are two PCBs, one for amplifier, one for ADC. This should avoid high common mode frequency.

Klaus
 

We see oscillations around 14 MHz which won't be expected according to AD8009 and LMH6553 bandwidths. Most likely conclusion, it's caused by parasitic circuit components like wiring inductance, cable capacitance, generally bad circuit layout or unsuitable measurement setup. I see that some oscillations are already present in the "input" trace.

We would need the layout and measurement circuit details to better understand the problem.

Circuit noise can be estimated based on OP data and resistor values. The dimensioning doesn't look wrong at first sight, although an unnecessary noise contribution of the offset circuit can be probably eliminated.

In the present circuit, R4 is loading U2 output inappropriately, don't see what it's good for.
 

Hi,

In the present circuit, R4 is loading U2 output inappropriately, don't see what it's good for.
Maybe try: R2 = R8 = 50R, R4 = not assembled.


Klaus
 

@KlausST & @FvM ... Sorry Let me update now,, R10 is DNS here.
Sir R4 is the actually RM as per LMH6553 datasheet, at page 21, it says for 0db gain set RM to 26.7R (PowerPad IC).
Sir Tomorrow I will upload the updated schematic with layout files for Orcad.
 

If R2 is 0R, R4 should be omitted because the node is driven by the low impedance U2 output. R8 should be 0R too to achieve balanced output.

The datasheet circuit is different, driven by a 50 ohm source.
 

If R2 is 0R, R4 should be omitted because the node is driven by the low impedance U2 output. R8 should be 0R too to achieve balanced output.

The datasheet circuit is different, driven by a 50 ohm source.

Let me try by doing modification as you said and I will update here. Sir I am also not clear about offset circuit and ripples.
Sir am attaching the complete schematic, PCB and other files with. Software used, Orcad 16.
Note:
R12 = R13 = 400 Ohm
R9 = R11 = 1K Ohm
R10 Removed
R1 = R2 = R3 = OR
 

Attachments

  • Two Stage Input Circuit HMCAD1520.zip
    1.3 MB · Views: 111

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top