That's great. Makes me want to take mine apart and modify them. Thanks for sharing.
Hi,
I'm happy I found time to try this out. I don't understand why there isn't more info about this reg here. It's simple and it works so well. The 4V dropout for the mosfet isn't so great, but then it has low output impedance so that's a fair trade off.
I let it burn in overnight and had a listen today - no significant difference as far as I can tell - can't do A-B testing to be sure though. However, it's driving 4 op amps with high psrr ( LME49990 ) so that's not too surprising I guess.
Some tech details : my psu is giving +/-15.55 V output unloaded and it drops to 15.50V with a 200mA load. The ccd into the error amp is E-152 so that's 1.5mA +/-10% so I chose the feedback resistors / potential divider as 1K/10K so the current running through that roughly matches @~1.4mA (15.5/11). This also gives just under 10% gain for the two LM329, which are the Vref. For the RC filter after the Vref, I used a 1K res and a 100uF cap. There's a soft start but it's still quick enough. For the caps across Vref, I replaced the electro and used 10uF tantalum and 0.1uF X7R. I haven't tried changing any of the other electrolytics yet - Panasonic FC - but I may if the signature changes.
Anyway, technically, it's lower noise but in my use, that actually hasn't made a heap of difference. It might in more sensitive applications like an I/V for a DAC.
Cheers
I'm happy I found time to try this out. I don't understand why there isn't more info about this reg here. It's simple and it works so well. The 4V dropout for the mosfet isn't so great, but then it has low output impedance so that's a fair trade off.
I let it burn in overnight and had a listen today - no significant difference as far as I can tell - can't do A-B testing to be sure though. However, it's driving 4 op amps with high psrr ( LME49990 ) so that's not too surprising I guess.
Some tech details : my psu is giving +/-15.55 V output unloaded and it drops to 15.50V with a 200mA load. The ccd into the error amp is E-152 so that's 1.5mA +/-10% so I chose the feedback resistors / potential divider as 1K/10K so the current running through that roughly matches @~1.4mA (15.5/11). This also gives just under 10% gain for the two LM329, which are the Vref. For the RC filter after the Vref, I used a 1K res and a 100uF cap. There's a soft start but it's still quick enough. For the caps across Vref, I replaced the electro and used 10uF tantalum and 0.1uF X7R. I haven't tried changing any of the other electrolytics yet - Panasonic FC - but I may if the signature changes.
Anyway, technically, it's lower noise but in my use, that actually hasn't made a heap of difference. It might in more sensitive applications like an I/V for a DAC.
Cheers
Last edited:
Just did a noise test on the Kubota from Jims. Seems very quiet at -84dBu (50mA load) My equipment isnt highly calibrated but give or take its a good result. It needs at least 10 v drop to work correctly, 15v dropped the reading by a few extra db.
My test one did have an odd low frequency noise drift. it didnt effect the above result but was visible on the spec analyser. Didnt have much time to investigate and it certianly would not be audible.
My test one did have an odd low frequency noise drift. it didnt effect the above result but was visible on the spec analyser. Didnt have much time to investigate and it certianly would not be audible.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.