Because of the Thanksgiving holiday in the U.S., TI E2E™ design support forum responses may be delayed from November 25 through December 2. Thank you for your patience.

  • State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 47 subscribers
  • Views 708 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMH6629: Unexpected output clipping in circuit with BUF602

Robert Godin
Prodigy 50 points
Part Number: LMH6629
Other Parts Discussed in Thread: BUF602, , OPA300, OPA301

Hi all,

I designed an active filter circuit using LMH6629 and BUF602. I am using a dual power supply at +/- 2.5 V. I'd expect to be able to get an output voltage of around +/- 1.7 V (assuming 0.8 V of voltage swing headroom) but my signal is getting clipped at +/- 0.8 V. Any ideas why this might be the case? Here are the relevant bit of the schematics (I can send the full schematic if necessary).

                                     

  • 0
    Guru 48395 points

    Hey Robert, your images did not come through - use the little paperclip in the ribbon to insert files - hopefully a TINA file - but, 

    The LMH6629 and BUF602 are not RR out, watch their headroom requirements. 

  • 0 in reply to Michael Steffes
    Prodigy 50 points

    Hi Michael,

    I've been using LTSpice and not TINA for simulations, sorry. Here's the schematic in a few different formats (some small differences between the gray simulation and the white schematic for building the PCB like the switch to change the gain with different resistors). Hopefully one of these has all the info you're looking for. I can send the .asc LTSpice file as well if its useful (its where the gray background image comes from).

    For the headroom, from the specs sheet, the output range for LMH6629 on a 5V supply is 0.8 - 4.2 V so I should be able to get - 1.7 V to + 1.7 V with the +/- 2.5 V right? For the BUF602 I'm not seeing any clear numbers. On page 15 of the specs sheet it states 'Under no-load conditions at +25°C, the output voltage typically swings closer than 1.2V to either supply rail; the +25°C swing limit is within 1.2V.' So I'd expect to be able to get at least - 1.3 V to + 1.3 V. Does anyone have more info on BUF602 or real world experience with it? Or maybe an alternative (unity gain buffer, at least 40 MHz bandwidth)?  

    Cheers,
    Robert

  • 0 in reply to Robert Godin
    Guru 48395 points

    Spending time over on the dark side I see (actually, I could open a LTSPice file, but this is enough for now) 

    First, let's talk about the circuit - High gain AC coupled, 

    1. Your first stage shows a 50ohm source, but then shunt 50ohm on the inverting nodes to ground - that would work better as a series input 50ohm (noise gain of 16V/V) and give much lower output noise

    2. You are using the BUF602 to imbed an RC filter - don't really need that 1st 500ohm to ground on teh LMH6629, unnecessarily loaded and not part of the filter -maybe go to 200ohm . Two sets of single pole stages? ok, kind of noisy - could do a lot better with an RLC and your DC source R is pretty high giving offset through the Ib * Rsource at the BUF602 input, 

    3. again, at the BUF602 output, that 50ohm to ground is not needed as the series input gain 50ohm will look like a 50ohm load to the BUF output, right now you are showing a 25ohm AC load, that will kill your AC and HD. 

    Anyway, the headroom is output current dependent (physically) but you do not have a DC load on the BUF (you do have a lot of gain in the chain), 

    You should plan on a 1.3V I/O headroom - so on 2.5V supply, a max output of +/-1.2V is what you can get - if you have a lot of DC offset, will clip on one side more. 

  • 0 in reply to Michael Steffes
    Prodigy 50 points

    Hi Michael,

    Thanks again for your help. Sorry if I'm misunderstanding something but this is what I got from your comments. An updated schematic is below.

    1. Noted. I've removed the shunt resistor to ground. To clarify, the 50 ohm resistance in series (R12 in the figure below) is to simulate the 50 ohm output impedence of a transconductance amplifier this is connected to. I had the shunt resistance in there as I thought that it can be important for reducing reflections in high-speed applications. I'm aiming to go as fast as 40 MHz with this.

    2. You point to a 500 ohm resistor but I don't see which one this is. The only one of that value Is this the gain R22 from the picture with white background but this doesn't seem to line up with the rest of your thought. Or do you mean the 50 ohm R9?

    The idea for the input to BUF602 is that the RC circuit from C2 and R9 is to remove any output offset from the first LMH6629 and then the second order RC circuit with R4, C4, R5, C1 is a low pass smoothing filter. In the full design I have switches to go to different sets of resistors to change the cut off frequency of the filter. The target is to have a wide range of cut off frequencies (roughly 1 kHz to 40 MHz using resistors of about 150k to 25 ohm) and a steep roll off. Were you thinking of a first order RLC filter? This would have a less steep roll-off than the 2nd order RC currently in there (confirmed by simulations). 

    3. I removed the 50 ohm to ground on the BUF602 output. 

    Maybe I'll go up to +/- 3.3 V on the BUF602 then. The LMH6629 is limited to 5.5 V so I'm already almost there. In the first board the DC offset is low and the clipping is symmetric.

    Cheers,
    Robert

  • 0 in reply to Robert Godin
    Prodigy 50 points

    A small addition:

    I'm looking at alternatives to the BUF602 and the OPA300 seems promising. It's 150 MHz GBW, unity-gain stable and has output to 100 mV of the rails.

  • 0 in reply to Robert Godin
    Guru 48395 points

    Well Robert, your interstage filter is then more than what you show - 

    1. Since you are switching source R around in the two pole RC, a CMOS input might be better and have more I/0 range. Since you follow this with a gain of about 31, you can't have much swing out of the buffer AC wise so slew rate is not too important there. 

    2. Not clear how you are getting into clipping issues on that buffer stage as the AC swing has to be pretty low and you are AC coupled. into it with a ground reference through a variable DC R apparently. 

    3. Yes, a >150MHz CMOS unity gain stable device would remove clipping from this stage. The OPA300 looks good. 

  • 0 in reply to Michael Steffes
    Prodigy 50 points

    Hi Michael,

    Thanks again for your comments. I went with the OPA301 in the end as I don't need the enable pin from the OPA300. The (simplified) revised schematic now looks like this. We'll see where the signals clip when the board is made.

    Cheers,
    Robert