Power amp measurement system

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I decided to upgrade my poweramp measurement system.

The devices:

1/ ultra low distortion sine wave oscillator - I bought Victor's 1k and 10k
2/ dummy load - designed and made by me
3/ passive voltage divider - under development
4/ passive notch filter- schematic Victor, made by me
5/ post notch buffer - under development - powered by two 9V battery
6/ PC sound cards I have
- ESI Juli@ PCI
- Asus Xonar STX II
- X-Fi USB HD modded Improving A/D perfromance of Sound Balster X-fi Music (SB1240) - and a Puzzle!!! - diyAudio

attachment.php


the dummy load
 

Attachments

  • DUMMY LOAD LKA.JPG
    DUMMY LOAD LKA.JPG
    51.2 KB · Views: 353
  • dummy_load_1.jpg
    dummy_load_1.jpg
    324.3 KB · Views: 407
  • dummy_load_2.jpg
    dummy_load_2.jpg
    253.4 KB · Views: 326
  • dummy_load_3.jpg
    dummy_load_3.jpg
    134.8 KB · Views: 220
  • dummy_load_4.jpg
    dummy_load_4.jpg
    264.1 KB · Views: 220
Ground loops aren't a big problem in my system. The voltage divider has a GND switch. For floating oscillators grounds are connected. If you use the sound card's output, the ground has to be disconnected at the poweramp side. (or coupled via 1k resistor )
 
Zeners are optional because the divider has fixed value. In the past I burned two sound cards, so I put zeners everywhere. The roll-off is much higher, above 1MHz, the divider output impedance is below 100R.
The heatsink is not grounded, here is a meauserement of my non-gnfb amp, mains related frequencies are below -120dB.

Those HS200 are not non-inductive. It may not matter if it's low enough, I'm just picky.

I'm planning to use switchable (ON/OFF) RC compensation network. First I need to measure the HS200 inductance.

very nice work!;)

TYVM
 

Attachments

  • KLA200 FFT1K 20Vrms 4R LEFT.png
    KLA200 FFT1K 20Vrms 4R LEFT.png
    79.3 KB · Views: 227
you are right

test1 : generator = victor's 10k (output imp 600R)
picture 1 : no zeners
picture 2 : two bzx83C3V3 zeners at the sound card input (input imp 5k6)

test2 : loopback (output imp 100R)
picture 3 : no zeners
picture 4 : two bzx83C3V3 zeners at the sound card input (input imp 5k6)
 

Attachments

  • Julia FFT10k victor.png
    Julia FFT10k victor.png
    76.2 KB · Views: 339
  • Julia FFT10k victor zeners.png
    Julia FFT10k victor zeners.png
    77.3 KB · Views: 326
  • Julia FFT10k loopback zeners.png
    Julia FFT10k loopback zeners.png
    77.4 KB · Views: 322
  • Julia FFT10k loopback.png
    Julia FFT10k loopback.png
    76.7 KB · Views: 333
Member
Joined 2004
Paid Member
For the clamping you would be better off with biased diodes. Use some 1n4148 reverse biased (batteries might work even) to limit the peak voltage. Very little current and not much cap modulation since the bias voltage will be larger than the signal.

In any case a significant overload with fry the diodes since you don't have current limiting in either case. You could use some small fuses. . .
 
I tried to find a diode with the lowest reverse current, ended up with dual BAV199 with max. reverse voltage 85V for 10 cents (for LKA - available in Czechia by GME). Two 12V A23 batteries last over a year (ideal for balanced inputs of Juli@ with max. voltage each lane at 10V), a check-up voltmeter costs 1 dollar from ebay. Of course a serial resistor limiting the current is important.
 
AX tech editor
Joined 2002
Paid Member
I have had excellent results with the attached. Set up a threshold voltage with biased zeners, in this example 4.7V. Then use ultra-low leakage diodes like the FDH300 as diode clamps. No influence down to -130dB. (The LEDs are just for power indication and can be deleted of course.

Jan
 

Attachments

  • clamp-1.PNG
    clamp-1.PNG
    12.3 KB · Views: 292
  • clamp-2.PNG
    clamp-2.PNG
    7.3 KB · Views: 122
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.