I guess I'm less impressed than Demian is -- I'm trying to figure out how you got 50dB of attenuation at 2kHz with a peak that broad, unless there's some limiting function at work in the sim. I also don't want my oscillator loaded with 600 ohms, but then that's a personal choice. In any case, if that attenuation can be real, then you've got a winner.
This is a quick sim of a filter synth I tried. Its a 3 pole Butterworth. It uses 1 Henry inductors. I first tried an elliptic but that needed 500H inductors. The passband is 100 Hz. The 1 Henry inductors are available from Triad. They almost certainly will have much higher distortion than any other element in this design. I don't think there is a practical 1 H air core. It may be the size of a Tesla coil.
Attachments
It's not practical because the inductors are huge. The nonlinearity of iron core inductors really mucks up the AC response.
Last edited:
cute. The dips are set at the second and third harmonic. Those inductors are still huge, and they would need iron cores to be something you can put into a box. I can't even find a place to buy them, since the biggest I can get from Parts Express are about 10mH. I don't know if the iron cores were included in the simulation.
@dimitri -- not sure why we suddenly started talking about notch filters -- but the active Twin-T (RC only, positive feedback) is just better all the way around. But to clean up residual oscillator distortion, a low-pass or bandpass is needed. A simple LC peaking filter, one part each, has sufficient Q and selectivity to do the job. With a pretty good oscillator to start with -- an HP 239 or 339, or a KH 4402 -- an extra 10-20dB of suppression will take you mostly into noise.....
Neat Dimitri, I never saw that when I did my stuff. A 4" by 4" coil with 1300 turns of #16 wire is about 100mH and 4 or 5 Ohms (or am I way off). In the second one the coil does not see fundamental? and maybe could be cored??
Last edited:
You are probably correct, but sometimes I would like to use something I trust - like 50 Ohm fixed coaxial attenuator 🙂
I don't think an active opamp bandpass filter would add that much noise. I've been simulating a two stage filter, and the NE5534 is again a very good choice. I like this because I don't have to make a huge investment in copper wire in order to make a passive one.
Definitely easier to make. Although the 5534 is one of my faves, I'd suggest an LME49710 or its dual (4562?) or LT1468 for minimum distortion, which is the issue at stake here.
Thanks. Yeah, the LME49720 is a good opamp.
I just hate waiting around for parts to show up in the mail. That's one of the downsides of this hobby: virtually everything has to be mail ordered.
Low distortion oscillator
Hello,
Have you looked at the oscillator circuit in the Amber 3501?
Manual is available from Index of /manuals
I also have a printed copy of the manual, as well.
I have one of these and I measure its internal oscillator at ~0.0008% and it has not been adjusted since is it was done at the factory back in the 80's.
Only problems, that I have had is a couple of tantalum caps blow, easy to find them, just by looking around and smelling 🙂 Last week one went inside the shielded DC/DC converter pcb.
The majority of the parts are NE5534 parts!! & old Motorola transistors, 2N4401/3
Regards
Rick
Hello,
Have you looked at the oscillator circuit in the Amber 3501?
Manual is available from Index of /manuals
I also have a printed copy of the manual, as well.
I have one of these and I measure its internal oscillator at ~0.0008% and it has not been adjusted since is it was done at the factory back in the 80's.
Only problems, that I have had is a couple of tantalum caps blow, easy to find them, just by looking around and smelling 🙂 Last week one went inside the shielded DC/DC converter pcb.
The majority of the parts are NE5534 parts!! & old Motorola transistors, 2N4401/3
Regards
Rick
I just picked up one of those Ambers as well. I have not figured out the 2nd harmonic cancelling circuit yet. There are two on eBay right now for pretty good prices.
Searched eBay, did not find an Amber 3501, what am I doing wrong?
Yes the Amber designer/s knew their analog circuit design very well. I wish I was as smart😱
I like the pre-amp design and the output amplifier.
Oscillator circuit , obviously works well!!Yet another one to figure out.
I start looking at the parts, the only two that I saw that are out of date. CA3280 OTA went obsolete in 2005. The MPQ6542 is obsolete as well.
I think that the pre-amp can be adapted for a good phono pre-amp
The output amplifier would make a good headphone amp.
I plan on using PSPICE to sim these circuits just to learn from them. Maybe make them on day!!
I have used HP 333,339,8903, I still like the Amber!!
Rick
Yes the Amber designer/s knew their analog circuit design very well. I wish I was as smart😱
I like the pre-amp design and the output amplifier.
Oscillator circuit , obviously works well!!Yet another one to figure out.
I start looking at the parts, the only two that I saw that are out of date. CA3280 OTA went obsolete in 2005. The MPQ6542 is obsolete as well.
I think that the pre-amp can be adapted for a good phono pre-amp
The output amplifier would make a good headphone amp.
I plan on using PSPICE to sim these circuits just to learn from them. Maybe make them on day!!
I have used HP 333,339,8903, I still like the Amber!!
Rick
Rick, please think twice. Four-gang potentiometer is a real PITA. They put consumer type and there is no way to replace it 🙁 I still have the unit 🙁
Yes I missed that, V2015(20K), 2 gangs on the osc & 2 on the notch filter, oh, well, still using a two gang pot for an oscillator only design is do able, don't you think? After all that is what this discussion is about, not making a distortion analyzer set.
Rick
Rick