mikael__

2n3958 replacement
« on: November 22, 2020, 04:20:16 PM »
Hi,

I'm experimenting with a filter based on the Buchla 291 voltage controlled bandpass filter.
http://fluxmonkey.com/historicBuchla/buchlaFiles/Buchla_2910_150.jpg

The filter uses 2n3958 transistors for input and output buffer and I would like to replace them with something more available. I tried replacing them with a normal TL082 in LTSpice and it seemed to work just fine. However, I'm not knowledgable enough to pick up on the finer details of the characteristics. Is there another opamp that would be more suitable ? Is there a modern matched transistor pair I should try instead ?



squarewave

Re: 2n3958 replacement
« Reply #1 on: November 22, 2020, 04:56:10 PM »
That's a transistor not an op amp. But it looks like they're just being used as a simple buffers in which case you could use op amps instead (which would probably perform better especially with low frequency noise).

But this is where you have to decide: do you want to implement the circuit for the sake of implementing that circuit or do you want to implement something else? If the former, then you should just pony up the $ and get the 2n3958. If not, you could use some matched J201's or an op amp or follow that procedure to it's logical conclusion and find a modern voltage controlled bandpass filter based on LDRs or a proper VCA like THAT 2180.

Personally I have always been partial to replicating the part of the circuit that imparts it's character but then just to standard circuits for everything else. So in this case I would actually use the 2n3958 just because their weird and kind a "fancy" at $15 each and of course do the LDR, specific filter bits and control voltage stuff but wrap that in standard I/O circuits like diff input and impedance balanced out.

abbey road d enfer

Re: 2n3958 replacement
« Reply #2 on: November 22, 2020, 05:29:21 PM »
The filter uses 2n3958 transistors for input and output buffer and I would like to replace them with something more available.
You could use almost any current PNP transistor in this position. They are used to drive the LED's in the optocouplers. There is no audio signal into them.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

mikael__

Re: 2n3958 replacement
« Reply #3 on: November 22, 2020, 05:36:15 PM »
You could use almost any current PNP transistor in this position. They are used to drive the LED's in the optocouplers. There is no audio signal into them.

I think the LEDs are driven by the 2N3566s ? Those I have succesfully replaced :)

mikael__

Re: 2n3958 replacement
« Reply #4 on: November 22, 2020, 05:43:55 PM »
Personally I have always been partial to replicating the part of the circuit that imparts it's character but then just to standard circuits for everything else. So in this case I would actually use the 2n3958 just because their weird and kind a "fancy" at $15 each and of course do the LDR, specific filter bits and control voltage stuff but wrap that in standard I/O circuits like diff input and impedance balanced out.

Thanks for the inspiration. I have to admit that for some reason I thought that these were out of production since many years. I see now that Mouser has them in stock, at a steep price.

JohnRoberts

Re: 2n3958 replacement
« Reply #5 on: November 24, 2020, 08:51:41 AM »
Hi,

I'm experimenting with a filter based on the Buchla 291 voltage controlled bandpass filter.
http://fluxmonkey.com/historicBuchla/buchlaFiles/Buchla_2910_150.jpg

The filter uses 2n3958 transistors for input and output buffer and I would like to replace them with something more available. I tried replacing them with a normal TL082 in LTSpice and it seemed to work just fine. However, I'm not knowledgable enough to pick up on the finer details of the characteristics. Is there another opamp that would be more suitable ? Is there a modern matched transistor pair I should try instead ?
The 2n3958 dual N-channel JFETs could probably be replaced with two separate single N-ch JFETs with similar specifications. It looks like they are using the duals to generate a source follower buffer with low DC offset (the bottom device operates as a current source, to generate a similar voltage drop in the source resistor to create a first order cancellation of the Vgs voltage.) These current sources could be trimmed to minimize DC offsets with unmatched JFETS.

It seems that simple op amp buffers could be used instead. Now opamps are cheaper than exotic dual JFETS. I would add similar resistors in series with the op amp buffers outputs to exhibit similar source impedance for the following circuitry. 

JR

PS; For today's TMI I experiments with dual JFETs back in the 70s/80s trying to make inexpensive variable gain elements to use instead of early VCAs. I used one half of the dual JFET inside the feedback loop of an op amp to make a voltage controlled resistance. The other half of the JFET was used as a variable resistance to ground as a voltage controlled audio pad. It never graduated off my bench as a practical solution.   
It's nice to be nice....

mikael__

Re: 2n3958 replacement
« Reply #6 on: November 24, 2020, 11:58:57 AM »
Thanks John, very helpful!

In the meantime I found that Linear Systems does produce a replacement in a SOIC8 package, the LS3958. Think I will try these out before going the opamp route.

abbey road d enfer

Re: 2n3958 replacement
« Reply #7 on: November 24, 2020, 02:08:05 PM »
I think the LEDs are driven by the 2N3566s ? Those I have succesfully replaced :)
Aah, so the 2N3958 are the FET's. Did not se that.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.


 

Related Topics

  Subject / Started by Replies Last post
11 Replies
3782 Views
Last post September 14, 2004, 05:57:22 PM
by Gold
8 Replies
2838 Views
Last post July 22, 2005, 01:27:13 AM
by bradzatitagain
4 Replies
1510 Views
Last post February 28, 2006, 05:57:07 AM
by Analogique
7 Replies
2811 Views
Last post February 26, 2007, 10:20:06 AM
by edanderson