You can take a 9v wall wart and as long as it’s isolated from ground, flip the voltage to get negative bias. Sorry I forgot that the gate bias needs to be negative to pinch off an N JFET. It’s opposite for P JFET.
There are handy DCDC charge pumps that make small negative voltages too.
You can test it out quickly with a 9v battery.
There are handy DCDC charge pumps that make small negative voltages too.
You can test it out quickly with a 9v battery.
Attachments
Last edited:
You can use the -ve output from one of these, fed by your 19v to 24v supply to generate -15v.
https://www.mouser.com/ProductDetail/Murata-Power-Solutions/MEA1D2415SC?qs=RiwdeSZ004R7Dze/uEhrfQ==
https://www.mouser.com/ProductDetail/Murata-Power-Solutions/MEA1D2415SC?qs=RiwdeSZ004R7Dze/uEhrfQ==
Thanks X. I knew there must have been a reason I wasn’t seeing a much more dramatic decrease in the current. I may circle back to try the negative bias scheme later.
For now, making progress on the second channel. I’ll have to wait for the next pair of ERSE inductors to arrive for the final install. For now I’ll test the new channel out with the Hammond 159ZC.
For now, making progress on the second channel. I’ll have to wait for the next pair of ERSE inductors to arrive for the final install. For now I’ll test the new channel out with the Hammond 159ZC.
Attachments
I ran some simulations with the negative bias.
For the LuFo Lite, the -ve bias seems to work at about -900mV to -960mV at the JFET gate and is what you need to get about 1.4A bias current assuming 1ohm total DCR from the chokes, no resistors needed.
Removing the resistor seems to have decreased the distortion and increased the headroom.
Since only tens of micro amps are needed, a 1.5v alkaline AA battery would be an excellent source of -950mV when used with a 15k resistor and 10k pot set at 5k to get —950mV. The battery is quiet and would last for weeks as current draw of JFET gate is very low.
For the LuFo Lite, the -ve bias seems to work at about -900mV to -960mV at the JFET gate and is what you need to get about 1.4A bias current assuming 1ohm total DCR from the chokes, no resistors needed.
Removing the resistor seems to have decreased the distortion and increased the headroom.
Since only tens of micro amps are needed, a 1.5v alkaline AA battery would be an excellent source of -950mV when used with a 15k resistor and 10k pot set at 5k to get —950mV. The battery is quiet and would last for weeks as current draw of JFET gate is very low.
Thought I would hack this together quickly this afternoon and hack it was indeed.
Not sure how I managed but when I fired it up, I had - 5 V on the gate. I shut it down quickly but when I fired it up with the appropriate voltages, I am not getting any current flow through the source. I suspect I killed the jfet.
I will remove it all and set up a new jfet or try the other side first to make sure it works.
More haste, less speed as they say.. dB
Not sure how I managed but when I fired it up, I had - 5 V on the gate. I shut it down quickly but when I fired it up with the appropriate voltages, I am not getting any current flow through the source. I suspect I killed the jfet.
I will remove it all and set up a new jfet or try the other side first to make sure it works.
More haste, less speed as they say.. dB
This bias scheme doesn't seem to work. I am experiencing pretty much what jwjarch did with the initial schema.
I started off at -.5 and slowly increased to - 1.5 to no avail. It was drawing 5.8A regardless if I shut off the negative bias or cranked it up to - 2V.
I had 0.1 resistors to measure current. I shut it down as the voltage started to sag and the smps went I to self protect mode.
I started off at -.5 and slowly increased to - 1.5 to no avail. It was drawing 5.8A regardless if I shut off the negative bias or cranked it up to - 2V.
I had 0.1 resistors to measure current. I shut it down as the voltage started to sag and the smps went I to self protect mode.
That’s very strange. A negative voltage should be able to pinch it off and turn it off basically.
I just double checked the simulation and the setpoint at the gate is actually -290mV not -900 mV. But if you had it at -5v and it still did not shut off? What are you using for the negative bias voltage?
Did you remove the 100k to ground?
I just double checked the simulation and the setpoint at the gate is actually -290mV not -900 mV. But if you had it at -5v and it still did not shut off? What are you using for the negative bias voltage?
Did you remove the 100k to ground?
That was my error. I started out with the voltage at zero, thinking I had to dial it in more negative to find the correct spot. And obviously the moment it powered up, it was conducting like crazy. It is odd that it didn't shut it off as I pushed up the negative voltage but I am wondering if my devices need a higher negative voltage. The original channel, I thought I had blown, wasn't conducting with a high negative voltage at the gate and didn't seem to start conducting as I decreased the negative voltage (became more positive). I thought I reduced it all the way down to zero. I can try this aagain.
The 5V is a seperate supply on the smps it measures appropriately at the gate when set up as a voltage divider with a 10K resistor and a 20K potentiometer (all I have on hand). The circuit as drawn did not give the desired negative voltage.
The 5V is a seperate supply on the smps it measures appropriately at the gate when set up as a voltage divider with a 10K resistor and a 20K potentiometer (all I have on hand). The circuit as drawn did not give the desired negative voltage.
Thanks for your work on this dBel84; just to clarify, your error was leaving the 100K to ground in place?
To clarify all my confusing posts - my error was not setting up the circuit correctly but I was still able to get -5 V on the gate.
So here is a summary of the various things I have tried.
Using the SMPS 5V supply for the negative gate voltage with the circuit as illustrated by X but with a 20K bourns pot instead of the 10K pot.
The negative voltage does indeed control the current flow through the jfet.
the pinch voltage on my jfets is closer to -1.9V
The behavior is rather rapid - what I see is no current flow and as i slowly drop the negative bias, it suddenly starts conducting and it pulls the negative bias positive and the unit is conducting 5+ amps, I am not able to pull it back by dialing up the negative gate bias - which remains positive
This behaviour is replicated when using a 9V battery
To prep the circuit for this test, I disconnect the 100K resistor and the 22pF as well as the 7R5 source resistor and put a 0.1R resistor following the choke to measure the current .
I have tried keeping the circuit completely separated ( no common ground ) as well as using a common ground with no change.
I may still be doing something dodo-esque but it is eluding me at this time. It works perfectly when rewired as original.
..dB
So here is a summary of the various things I have tried.
Using the SMPS 5V supply for the negative gate voltage with the circuit as illustrated by X but with a 20K bourns pot instead of the 10K pot.
The negative voltage does indeed control the current flow through the jfet.
the pinch voltage on my jfets is closer to -1.9V
The behavior is rather rapid - what I see is no current flow and as i slowly drop the negative bias, it suddenly starts conducting and it pulls the negative bias positive and the unit is conducting 5+ amps, I am not able to pull it back by dialing up the negative gate bias - which remains positive
This behaviour is replicated when using a 9V battery
To prep the circuit for this test, I disconnect the 100K resistor and the 22pF as well as the 7R5 source resistor and put a 0.1R resistor following the choke to measure the current .
I have tried keeping the circuit completely separated ( no common ground ) as well as using a common ground with no change.
I may still be doing something dodo-esque but it is eluding me at this time. It works perfectly when rewired as original.
..dB
Thanks for trying the negative bias mod out, dBel. I’m wondering if the jfet is either on or off with no in between. Is that possible?
I got the second channel finished today. Listening to all my usual tracks. In short, the sound is very nice. Easy to listen too. With the WHAMMY preamp there is just enough pot for my normal listening levels which are typically a bit more than your average listener IMO. Here’s a quick clip. At first listen the amp seems stout enough for use with high sensitivity speakers. One of the best parts, it is 100% dead silent when not playing. I can’t even tell it’s on. I’ll post some more extensive impressions soon!
LuFo Lite Amplifier - Design by XRK971 - YouTube
I got the second channel finished today. Listening to all my usual tracks. In short, the sound is very nice. Easy to listen too. With the WHAMMY preamp there is just enough pot for my normal listening levels which are typically a bit more than your average listener IMO. Here’s a quick clip. At first listen the amp seems stout enough for use with high sensitivity speakers. One of the best parts, it is 100% dead silent when not playing. I can’t even tell it’s on. I’ll post some more extensive impressions soon!
LuFo Lite Amplifier - Design by XRK971 - YouTube
Attachments
Last edited:
Thanks dBel! I agree. I’m going put some hours on it to let everything settle in. First impressions: The highs are crystal clear and detailed. Mids and lows seem a bit thin. That’s compared to the Babelfish J2 I’ve been listening to for the past month or so, though. So not really a fair comparison. But I would saw the LuFo Lite can hold its own.