[sdiy] 2SK170 FET alternative

[Seb Francis]

Thank you Harry for this bit of information!  I did read 3 different 
LM311 datasheets and nowhere did it say that pins 5 & 6 needed shorting 
together if not used (or maybe it did say this, but I missed it .. it 
was late at night).

Also I wasn't 100% sure that the GND pin 1 should always be connected to 
0V (I couldn't work out why such a connection was needed as well as V+ 
and V-).

The schematic I have specifies the LM393, but doesn't show the power 
connections .. I'm assuming that the GND connection pin 4 should 
actually be connected to V- when used with split rail power (there isn't 
an actual V- connection for the LM393).

I would try a CA3080, but I have only a small supply of these with no 
possibility to get more without paying silly money..

BTW, why do you say that the TL072 sucks as a comparator?  Is it because 
of the relatively low slew rate (16V/us)?  Or is what the datasheet 
calls 'DIFFERENTIAL VOLTAGE AMPLIFICATION' (I guess this is open loop 
gain) too small?

Seb



harrybissell wrote:

>The LM311 should have pins 5 and 6 shorted together if not used.
>The LM311 is a nasty beast to tame IMHO as are most comparators...
>
>I would not overlook the CA3080 as a comparator. It has a slew rate
>of about 50V/us... plenty fast enough.  You need to limit the differential
>input voltage to 5V or less. IMHO the CA3094 OTA said you could just limit
>the current into the inputs to prevent damage if you go over 5V. I have used
>that technique on the CA3080, although it does not SAY you can do that in
>the data sheet.
>
>The TL072 sucks as a comparator, also IMHO.
>
>H^) harry
>
>Seb Francis wrote:
>
>  
>
>>jays at aracnet.com wrote:
>>
>>    
>>
>>>Seb,
>>>
>>>THAT parts are targeted for more of the high end.
>>>
>>>
>>>
>>>      
>>>
>>So I think they should be capable of a range of sounds from low end to
>>high end (as opposed to using low end stuff which can never sound high
>>end :)
>>
>>    
>>
>>>Remember that the 2SK170 takes less drive than the 2SK30. Maybe increasing the gate voltage on the 2SK30 would give you more range and harder compression. One thing to do is get a bunch of appropriately valued trim pots and put those in and tweak them to see what happens. This can lead to all sorts of fun. You might even stumble on to something you like. Or you might let some of the smoke out of the parts.
>>>
>>>
>>>
>>>      
>>>
>>To be honest, I'm not sure the adjustable slew-limiter part of the
>>circuit is working right at all .. the comparator doesn't seem to be
>>behaving itself (I'm using an LM311 and I'm not really sure how to
>>connect the extra pins balance, strobe, gnd, etc.) I think I might
>>rebuild it in the morning using a TL072 instead (and maybe use the other
>>half of the TL072 instead of the JEFT).
>>
>>Without the slew-limiter switched in, the compressor is working fine in
>>'auto' mode (using a trick with an opamp to make a 'non-linear
>>capacitor' for the RMS detector integration time).  But I don't really
>>like how this works for single drum hits, so I'll make it switchable to
>>a normal capacitor - actually the effect with a really small capacitor
>>is quite cool for low frequency things like bass kicks because you get
>>some of the rectified input signal modulating the VCA level - kind of a
>>metalic sound.  I guess I will have a rotary switch with a few different
>>capacitors (also including the auto mode non-linear capacitor circuit in
>>case I want to compress some full range music).
>>
>>    
>>
>>>Maybe in this case you are looking for more for soft distortion than compression. Like tubes or MOSFET clipping. One thing I looked at in the guitar world (I think it was called the Shaka) where there were a set of GE diodes and MOSFETs are put in series to form a soft distortion/clipping. Haven't tried it yet so don't know how well it works.
>>>
>>>
>>>
>>>      
>>>
>>Yeah, soft distortion is definitely something I'm looking for .. but
>>compression as well I think.  For example I like to put a bass kick
>>through a very fast compressor to get the body really full, then use
>>distortion to clip the initial clicky peak that the compressor lets
>>through.  This way you get a really full bodied drum, but without it
>>sounding too clipped.
>>
>>So, yes it's very likely there will be some distortion gizzmos in the
>>box as well :)  There seem to be plenty of circuits around for this kind
>>of thing on all the stompbox websites.
>>
>>Seb
>>    
>>
>
>
>
>  
>