if your Source resistors are identical value, then you will find that bias voltage will be identical when output offset is zero.
If your Source resistors are slightly different, then when output offset is zero you will see a different voltage across the source resistors. To avoid this measurement confusion/error, you need to accurately match your source resistors.
That measurement difference does not affect performance, you still strive to arrive at zero output offset.
When you get there, you can measure across the jFET source resistors to see what level of current is flowing through the jFETs.
That is essentially full bias. That is very good.
Ignore the fact that N- and P- channel measurements are a little different, that is completely normal, they will never be exact match. The important thing is DC offset.
Set bias and offset with input shorted. Then Power off. Attach a source, power on, and measure bias and offset. If good, attach test speakers and listen.
Well, that was whatI did before...still 2.6V.
What I did is a little sensitivity test: I used a variac to lower the Supply voltage. When I use 28V instead of 33.5V every thing is like in the schoolbook. Does not matter if I am using input grounded or not. From 29-29.5 V I got immediately 0.5V offset...
Soooooo, I guess there is no choice, I need to include the cascode. I somehow managed to loose the 2SC4793 and 2SA1837 from the diyaudio kit...I assume I can use any of those I can buy from mouser or have the got a matching as well ? Any specific watt specs on the cascode resistors or are 0.5 w types ok ?
Thx for all your support...
What I did is a little sensitivity test: I used a variac to lower the Supply voltage. When I use 28V instead of 33.5V every thing is like in the schoolbook. Does not matter if I am using input grounded or not. From 29-29.5 V I got immediately 0.5V offset...
Soooooo, I guess there is no choice, I need to include the cascode. I somehow managed to loose the 2SC4793 and 2SA1837 from the diyaudio kit...I assume I can use any of those I can buy from mouser or have the got a matching as well ? Any specific watt specs on the cascode resistors or are 0.5 w types ok ?
Thx for all your support...
Mistake in ordering Jfets for F5 Turbo
or kind of... Instead of ordering the matched pairs Jfets sets, I ordered the packages containing the "Grade A" LSJ74 and LSK170 : a package containing 8 of each...I could have saved some $$$. Do I have to match these? Or no because they are grade A? Would I have to make the exchange? Not for the saving but for convenience in assembly. Thanks
or kind of... Instead of ordering the matched pairs Jfets sets, I ordered the packages containing the "Grade A" LSJ74 and LSK170 : a package containing 8 of each...I could have saved some $$$. Do I have to match these? Or no because they are grade A? Would I have to make the exchange? Not for the saving but for convenience in assembly. Thanks
So....finally it works...I have implemented the Cascode now and now the DC is the same, no matter if input is grounded or now....it is just burning in...comparisonof sound against my tube amps will be done tomorrow...
...small grain of salt: With R5/6=1k, I can only bias up to 340mV....still pretty warm, so for the moment I leave it like it is, but wanted to run it at 380mV...rail voltage is 33V.
...small grain of salt: With R5/6=1k, I can only bias up to 340mV....still pretty warm, so for the moment I leave it like it is, but wanted to run it at 380mV...rail voltage is 33V.
By the way: The amplification is for my System a bit too much. Is there a smart way to get it down a bit ?
"Smart" meaning: not jus destroy input voltage with a divider, but changing for example components to trade in gain for lower distortion by having for example higher local feedback like you would do it with an un-bypassed cathode resistor in a tube circuit ?
"Smart" meaning: not jus destroy input voltage with a divider, but changing for example components to trade in gain for lower distortion by having for example higher local feedback like you would do it with an un-bypassed cathode resistor in a tube circuit ?
Last edited:
