I was looking over the datasheet for the LME49710, and decided to replicate their test setup for THD%
Here's the test jig:
A 1kHz signal is swept from 10mV to 2V and fed to the non-inverting input of the DUT. Both V1 and V2 are 9V cells.
The THD% value on the y-axis is actually 1/100 of the value shown.
For some reason the results I show with the LME49710 are a bit better than those in the datasheet:
As you can see, the prosaic NE5534 does pretty well in comparison, but the LT1028 did much better than I had thought. I did not tweak the AD797 with its distortion nulling adjustment.
Here's the test jig:
An externally hosted image should be here but it was not working when we last tested it.
A 1kHz signal is swept from 10mV to 2V and fed to the non-inverting input of the DUT. Both V1 and V2 are 9V cells.
The THD% value on the y-axis is actually 1/100 of the value shown.
For some reason the results I show with the LME49710 are a bit better than those in the datasheet:
An externally hosted image should be here but it was not working when we last tested it.
As you can see, the prosaic NE5534 does pretty well in comparison, but the LT1028 did much better than I had thought. I did not tweak the AD797 with its distortion nulling adjustment.
OK, 100 pF are a lot for many opamps (stability), better to put 100 Ohm in series to the AP input.
Would be great if you can go up with level to 8Vrms to maximize THD+N and see what is possible.
Maybe i am wrong... I see -78dB THD at 1Vrms for the AD797. Samuel shows about 96 dB for the test. Can you please check this (pictures from Samuel)?
Would be great if you can go up with level to 8Vrms to maximize THD+N and see what is possible.
Maybe i am wrong... I see -78dB THD at 1Vrms for the AD797. Samuel shows about 96 dB for the test. Can you please check this (pictures from Samuel)?
Attachments
Hi Jackinnj, within this context, I am using 49710 cans as tape buffer in and out for my Telefunken studio recorder. It amplifies an approx. 10dB at the moment, but to what gain can I cranck that up without getting additional distortion or other misery, or is there really no limit other that hitting the supply rails at some point (+/- 15V) I in fact need an additional 6 to 10 dB. I can just increase the feedback R until I get there, or glue an other stage (similar) on top of it...but that gives me 2 x 49710 which then doubles the distortion, noise etc..
Thanks Leonard
Thanks Leonard
What do you mean by that? Generally speaking, 2 stages with 10 dB + 10 dB will perform better than one at 20 dB when it comes to distortion and bandwidth. It's just a matter of GBW.
On the results posted here, the AD797 must have been exceptionally noisy. As the curve seems to be right on top of the OPA134, I would make sure that this one hasn't been used again by mistake. The '797 ought to be somewhere near the LT1028.
BTW, you can clearly see why THD+N sucks for measurements like this - way too much N, way too little THD. The only two parts that do not appear to be noise-limited throughout are AD825 and LME49710. With the kind of supplies shown, it would have been feasible to to go up to about 4 Vrms, which would already have helped. Samuel Groner didn't do high-order FFT wih synchronous averaging for nothing though.
2 op amps can be used in a multiloop, sharing a common Global feedback - theory shows that all the gain in one loop is the way to get greatest distortion reduction for a given total gain
better system performance can also be had by choosing low noise input op amp and high current, high speed output op amp - such as a CFA DSL dirver
http://www.diyaudio.com/forums/solid-state/45794-high-loop-gain-composite-op-amp-circuits.html
better system performance can also be had by choosing low noise input op amp and high current, high speed output op amp - such as a CFA DSL dirver
http://www.diyaudio.com/forums/solid-state/45794-high-loop-gain-composite-op-amp-circuits.html
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.