Euvl, if that is the case i have to re dimension the filter. Most likely the cap is too big.
Shaman, the cap has to be a film-foil like the FKP2 or KP1830. The coil is our own design.
The LM49713 sounds better then a simple B1 style follower. The follower can be improved though much by cascoding. I will post a circuit i tried.
Shaman, the cap has to be a film-foil like the FKP2 or KP1830. The coil is our own design.
The LM49713 sounds better then a simple B1 style follower. The follower can be improved though much by cascoding. I will post a circuit i tried.
Sorry Euvl, my mistake. The coil has to be 10x bigger. I did a mis calculation because i have to scale the frequency plot 10x down. i will be back in a minute.
This should work. This has a bit more group delay ( 28usec to 5usec @ 20kHz ) and a little bit less overshot in the impulse. The impulse is a bit wider though. Part of the post ringing is from the 48kHz sampling rate i used.
The buffer can use any high Gm N-channel Fet and the floating cascodes could be made with even higher pinch off voltage Fets. It has very low distortion and is very high speed.
The buffer can use any high Gm N-channel Fet and the floating cascodes could be made with even higher pinch off voltage Fets. It has very low distortion and is very high speed.
This one I like, much better. 😉
For those who do not want to use SMD, I suggest you use 2SK372BL or 2SK369BL (or-V) to replace the BF862.
Also personally I use J111 for cascoding, but J310 is also fine. Both available in either SOT23 or TO92.
Thank you Joachim. And I presume R6 is Zout of the ES9022, i.e. inside the chip itself.
Modules built by me have MKP2 4.7n as C2 already. So you only need to add from R5 onwards.
Patrick
For those who do not want to use SMD, I suggest you use 2SK372BL or 2SK369BL (or-V) to replace the BF862.
Also personally I use J111 for cascoding, but J310 is also fine. Both available in either SOT23 or TO92.
Thank you Joachim. And I presume R6 is Zout of the ES9022, i.e. inside the chip itself.
Modules built by me have MKP2 4.7n as C2 already. So you only need to add from R5 onwards.
Patrick
Last edited:
Yes, thats right. I like the buffer too but trust me, the simple B1 sounds tired against it.
I build it with 2SK369BL. That Fet is very good.
I build it with 2SK369BL. That Fet is very good.
😉
I no longer know who published it first -- Curl, Borbely, Dennis Feucht, maybe Horowitz & Hill, or some German EE bible ?
Patrick
I no longer know who published it first -- Curl, Borbely, Dennis Feucht, maybe Horowitz & Hill, or some German EE bible ?
Patrick
I would say Tietze-Schenk. This circuit is as old as N-Channel Fets. I assume somebody did it with tubes ages ago.
I am searching for inductors that could work. For example two of this in series.
They are available with 12mH too but then the self resonance is lower.09P 5,6M - Stehende Induktivität, 09P, 5,6mH - Fest-Induktivitäten, radial bei reichelt elektronik
The resistance is higher then i simulated but that is not particular critical.
I am searching for inductors that could work. For example two of this in series.
They are available with 12mH too but then the self resonance is lower.09P 5,6M - Stehende Induktivität, 09P, 5,6mH - Fest-Induktivitäten, radial bei reichelt elektronik
The resistance is higher then i simulated but that is not particular critical.
No, they are not wide band enough. We need a core material that is good to at least 20kHz. There are materials that are good at high frequency for switch mode supplies for example but they are no good at audio frequencies. This is difficult.
We could ask Haufe, Pikatron, Sowter, Lundahl or Jensen but then it gets expensive.
I pay 100,-$ for a 24mH audio inductor from Sowter.
We could ask Haufe, Pikatron, Sowter, Lundahl or Jensen but then it gets expensive.
I pay 100,-$ for a 24mH audio inductor from Sowter.
Last edited:
I think it is the frequency where the inductance is tested.
See here :
http://www.fastrongroup.com/image-s...pdf?type=dataSheet&productType=seriesCategory
Patrick
See here :
http://www.fastrongroup.com/image-s...pdf?type=dataSheet&productType=seriesCategory
Patrick
So rated L at 20kHz, rated Q at 59kHz, and self resonance at 525kHz.
Seems quite OK (not perfect), and not expensive.
Patrick
Seems quite OK (not perfect), and not expensive.
Patrick
Yes, that could work. Worth a try. The circuit has to be measured then if there is any problem. The coil has to be put in a small Mumetall container.
- Status
- Not open for further replies.