A composite with LT1210 - another way of doing it

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I had an idea of improving results of my ultra-cheap amp with paralleled NE5532 and a nested feedback. Accidentally I found a bag with LT1210 samples in my drawer. There was not much time spent thinking after that. I pulled out output section from the prototype board (NE5532 x 5) and fitted an LT1210 instead. As a result, it does not look really pleasing, but it does the job!
Run a quick test. Harmonics were measured for 20kHz 5Vrms (+14dBVrms) output signal loaded on 33Ohm resistor (around 0.75W at load). For IMD I mixed 20kHz and 21kHz signals, 2.5Vrms (+8dBVrms) each at output. Results you can see on the screenshots. If we go from absolute to relative levels, we will get -149dB for second and -157dB for third harmonics of the 20kHz signal. For IMD test it will be -162dB for 19kHz and 22kHz products, -161dB for 1kHz product.
There is no conclusion at the end. Just an improved by ~9dB results on higher output levels comparing to the old prototype.
 

Attachments

  • pcb.jpg
    pcb.jpg
    150.4 KB · Views: 1,319
  • lt1210_20k21k_41k.PNG
    lt1210_20k21k_41k.PNG
    29.5 KB · Views: 449
  • lt1210_20k21k_1k.PNG
    lt1210_20k21k_1k.PNG
    27.8 KB · Views: 645
  • lt1210_20k21k.PNG
    lt1210_20k21k.PNG
    29.2 KB · Views: 656
  • lt1210_60k.PNG
    lt1210_60k.PNG
    30.7 KB · Views: 665
  • lt1210_40k.PNG
    lt1210_40k.PNG
    28.8 KB · Views: 1,024
Same thing driving 16.5Ohm load. Just a little bit worse for the second harmonic at 20kHz. The rest is pretty much the same.
 

Attachments

  • LT1210_16ohm_60k.PNG
    LT1210_16ohm_60k.PNG
    30.2 KB · Views: 106
  • LT1210_16ohm_20k21k.PNG
    LT1210_16ohm_20k21k.PNG
    29.1 KB · Views: 86
  • LT1210_16ohm_20k21k_1k.PNG
    LT1210_16ohm_20k21k_1k.PNG
    24.8 KB · Views: 72
  • LT1210_16ohm_40k.PNG
    LT1210_16ohm_40k.PNG
    29.8 KB · Views: 74
Last edited:
I haven't had it hooked up to a supply that delivers enough current to fully test its current capability yet, but I love it's very low THD and noise level.

As my plots show, it superseded my Gina24 sound card!!!

I just got some more to parallel a few of them. ;)

I used it to plot the impedance of a primary winding of one of my ESL step up transformers and it worked like a charm!!!

jer :)
 
Last edited:
No problems. Here it is.

There is nothing new. Just shifted accents a little bit the way I wanted. It is possible to increase the loop gain even higher, but not sure if it will bring real benefits.
Used it without output filter during my testing, but in the real world it is required. Could may sense readjust values to elevate input impedance to 3.3k or around it.
 

Attachments

  • sch_2.png
    sch_2.png
    28.7 KB · Views: 556
Thanks a lot, i'm far from your knowledge, but if i don't understand now all details, i'm learning easily when design interest me as i'm trying to understand harder. I didn't notice until now that there can be one "zobel//snubber" (however you call it) each side of this kind of output filter. For very high frequencies ?
What is the supply voltage in your test ?
 
This output filter was an attempt to optimize a decoupling of an amplifier output and damping of a cable and a driver resonance. This resonance frequency is usually not very high, around 1-2MHz.
Perhaps I can double check if those values are correct.
I was running it from +/-11.5V
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.