switch Negative voltage from positive logic?

[stormbringer]

Hello.

I'm currently experimenting with a channel switching circuit. I got a working part using jfets to 1. with 0 voltage on gate let signals pass, 2. with -12 block the signal.

i have also made a circuit from 3 momentary buttons that sets a 74hc74 Flip flop to on/off for each channelwith the help of a 74hc08 quad AND Gate, while disabling the other "channels" latches. This one running on +5, output goes to a bc337 which lights a led for the currently selected channel.

But, now i need to merge the 2 circuits somehow, getting 0v when output is high on the bc337, and -12 when low, so that i can actually switch the jfet the way i want. anyone that can help me achieve this? Been googling forever...

Edit:

Just tried setting up an lm358 as a comparator, which atleast gave me +10.5 and -12 out depending on the on/off state of the bc337. But i really need to get that +10.5 down to 0, to keep the circuit from oscillating like mad.

[Enzo]

I don't know if this answers your question, but...

One trick is to use one JFET in series with a channel, and another from signal to ground in the other.  COnnect both their gates to the same control voltage.  Now when voltage is on the gate, they shut off.  SO the series one won;lt let signal pass, and the shunt one will stop shorting across the signal.  Remove the gate voltage, and the series one now lets signal pass in its channel, and the shunt one grounds off the signal in its channel.

Another idea is to have the control line go from -12 to +12.  Connect it to the gates of two JFETs in series with their channel signals.  But make one an N channel JFET, and the other a P channel JFET.  Now when the control is on +12 one of the JFETs turns on but not the other, thenb switch the voltage to -12, and the opposite occurs.

[J M Fahey]

You can merge your actual circuit almost unchanged if you mate CMOS (which can supply anything between 0 and +15V) and P channel FETS which will love them.
No need for negative supplies.

*OR*

Refer your NCH FETs to the +15V rail , then any gate voltage available goes between 0V and -15V *for them*.
You can use a lower "+Voltage" reference, but avoid using less than 5V.

[stormbringer]

Thanx guys!

Will try this out when i get home from work!

[stormbringer]

Thanks for the advice there JM. I got it working in my simulator by offsetting the signal to half supply voltage!

Also found Teemus amazing book which helped alot! Thanks again!

[J M Fahey]

Have a look at Boss pedals. They often use NCh FETS, obviously for cost and availability reasons, and refer them either to +9V or 4.5V .

[stormbringer]

Did go through a bunch of boss pedals. thanx for the tip! Seems i got the jfet switching up to speed now in my sim. So i'm moving on to refine the logic now.

What i was planning for my preamp is 3 channel buttons, 2 boost buttons, 1 for each of the high gain channel, and a delay switch (will include a delay instead of reverb, as i play with a mic on my amp 99 % of the time anyway, where reverb will just weaken the signal/introduce phase errors. a small delay however is nice to have during clean parts and solos.

now, this might be more complex than it needs to.. But here's how it works:

Channel 1, 2 and 3 switches are momentary, because i want separate buttons for all 3, not having to switch to channel 2, and then a toggle between 2 and 3.
Boost 1, 2 and delay could be latched, but have been made momentary in the schematic. I think this simplifies having the front panel switches in paralell with the footswitch.

Edit:
The bypass line is when no channel is selected, this behaviour is intended, as i usually stomp the high gain channel off in some riffs where there's alot of rhytmic muting to avoid monitor feedback.

The problem with the momentary switches is obviously LED indicators. I'm not really feeling up to pulling 6 extra wires out just for the LEDs. Anyone have a practical example circuit with 3 channels? I know some mesa footswitches have individual buttons for 3 channels, but havent been able to find a schematic. this circuit is just how i solved it on my own. without LEDS. Now there's probably more logic than needed. (i was actually almost ready to do this with a pic16 when i saw the IC count).