B1 Buffer Preamp

[WithTarragon]

Originally Posted by AndrewT

there's something wrong.
The transformer balanced output signal is not referenced to ground. It should be floating.
You should not be able to measure half voltage from one pole of the transformer to ground.

If instead of using a pair of B1s, you used another transformer at the input end, then again you would be floating the balanced inputs.

It's when you replace the transformers, at either end, with electronics that the ground references come in. That's the big disadvantage of electronically balanced signal transmission. Transformer coupled signals float and give best attenuation of interference.

I am afraid that I worded my question poorly. I will be tapping directly from the output of the DAC chip and bypassing the existing analogue output (opamps, RC filters etc). I referred to the output of the DAC as being "balanced". I should have referred to it as a differential output.

I will be getting the galvanic isolation when I take the outs of the DAC (+ and -) and I will not be using the center tap on either the primary or secondary side. When I measured the voltage from a single leg, I of course had the voltmeter connected to one leg and a chassis gnd (in that case only).

I am greedy about getting as much voltage out since the B1 provide no gain.

I looked at the Whitlock / Jensen papers and they are helpful.

[AndrewT]

Have a look at Jung. He explains the >=1M for input impedance to <=50ohm for source impedance ratio for best rejection. Same concept to Whitlock but with different words.

[tessier]

Hi

I will build the B1 line preamp

I've buy ten 2SK170-BR, but they all have an IDSS of arround 2.7 ma. Is it a too low IDSS to get low noise and low thd ?

I've use a transistor analyser with a precision of 5% to measure the IDSS.

Where can I ad a low pass filter in the B1 circuit ?

Thanx

Paul

[tessier]

Hi

No reply to my questions ?

Thanx

Paul

[Salas]

2SK170 are either GR, BL, or V. Real Toshiba GR span 2.6mA to 6.5mA and tend to bunch half that way as random samples. You probably got fakes. Even if they are selected near IDSS samples from bottom of the barrel real 2SK170GR such low transconductance is no good for B1. You would need 7 to 9mA BL samples for good dissipation and trasconductance balance.

P.S. Which transistor analyzer that measures IDSS?

[tessier]

Originally Posted by Salas

2SK170 are either GR, BL, or V. Real Toshiba GR span 2.6mA to 6.5mA and tend to bunch half that way as random samples. You probably got fakes. Even if they are selected near IDSS samples from bottom of the barrel real 2SK170GR such low transconductance is no good for B1. You would need 7 to 9mA BL samples for good dissipation and trasconductance balance.

P.S. Which transistor analyzer that measures IDSS?

Hi

How about using two of my 2SK170 in parallel ?

I used a M Cube semiconductor analyser.

Thanx

Paul

[Salas]

No, don't do that, they will never be 2 exactly same ones. Their capacitance will double also. Short g,s pins to a 9V battery's - and put a mA meter between + to drain, verify the analyzer for now.

[tessier]

Hi

I've done the test you suggested and it have arround 3 ma for all my 2SK170

How about changing the values of some resistors in the circuits so I can use my low IDSS 2SK170 ?

Thanx

Paul

[Russellc]

Originally Posted by tessier

Hi

I've done the test you suggested and it have arround 3 ma for all my 2SK170

How about changing the values of some resistors in the circuits so I can use my low IDSS 2SK170 ?

Thanx

Paul

Just get the proper fets. If they are fake, this is just the beginning of a nightmare that would otherwise be a super simple build. 2SK170 is not the least bit rare and are available.

Russellc

[Salas]

Originally Posted by tessier

Hi

I've done the test you suggested and it have arround 3 ma for all my 2SK170

How about changing the values of some resistors in the circuits so I can use my low IDSS 2SK170 ?

Thanx

Paul

Shoot them some clean macro pictures front & back, we will be able to advise if real Toshiba, and if they are valid save them for something else. For B1 get BL as Russellc wrote.

P.S. That's 11% error, don't trust the analyzer for any real IDSS matching in the future.