I suppose if you had a variable resistor you could continuously change its resistance up and down until you got a meter reading of 8.00 mA.
Or if you have several non-variable resistors, I suppose you could try each one and write out a table of resistance versus measured value. Then interpolate or extrapolate the data to get 8.00 mA.
Or if you have several non-variable resistors, I suppose you could try each one and write out a table of resistance versus measured value. Then interpolate or extrapolate the data to get 8.00 mA.
Apologies in advance as this has been covered ad nauseam. I recently came into possession of 100 2sk170BL fets, and I’d like to replace the Q1’s in my build. I’ve seen the formula that Mr. Pass provided and built a breadboard to test them. Before I do anything regrettable, I want to double check that I’m measuring them correctly and provided a pic.
https://www.diyaudio.com/community/threads/2sk170-matching.141162/#post-1784617
9v battery
100ohm resister
1vdc=10ma
Therefore 0.942vdc=9.42ma correct?
https://www.diyaudio.com/community/threads/2sk170-matching.141162/#post-1784617
9v battery
100ohm resister
1vdc=10ma
Therefore 0.942vdc=9.42ma correct?
Attachments
Correct.
Time in the rig is very important... it's best to put on music with a consistent rhythm, or just sing yourself a song or say a short rhyme or something like that for cadence, and take the reading at the same time for each device. You'll find that they tend to change as they warm up and the number never really stabilizes, so taking a reading at the end of a set period is the right approach.
Time in the rig is very important... it's best to put on music with a consistent rhythm, or just sing yourself a song or say a short rhyme or something like that for cadence, and take the reading at the same time for each device. You'll find that they tend to change as they warm up and the number never really stabilizes, so taking a reading at the end of a set period is the right approach.
I would very much like to know your next step, to choose resistors for the JFETs, as I have not figured that part out yet.
You and me both
. I ordered a bunch of J113’s that should arrive tomorrow. My plan was to create another bread board mock-up, and play around with different resistor values until I find a compatible set for their respective Q1 idss.I have no shame so if I get stuck and/or figure it out, I’ll share with the thread.
That is the easy part!That's great.
And of course you are aware that the pinouts of the two parts are different.
I finally received some resistors for biasing Q2 and spent some of the holiday grouping my batch of 2sk170s in hopes of finding enough similar 8mA values to swap all the Q1s in the crossover. The attached pic is my attempt at biasing Q2 such that it’s value is a bit lower than the Q1 (it may be too close). I was a bit surprised that the bias resistor for Q2 (27.4ohm) is so much lower than Q1 (100ohm).
1. Does 27.4R Q2 bias resistor seem too low?
2. Is the voltage delta between Q1 and Q2 too close? Should I have a bit more headroom to ensure Q2 voltage does not exceed Q1?
1. Does 27.4R Q2 bias resistor seem too low?
2. Is the voltage delta between Q1 and Q2 too close? Should I have a bit more headroom to ensure Q2 voltage does not exceed Q1?
Attachments
Just for context to my previous post, I removed the Q2 transistor that was included in the kit and put it into the bread board to test it. Using a 47ohm resistor for bias instead of the 100ohm used in the kit, I got roughly 0.727 vdc. If this test looks sound, it would be much easier to just replace Q1 and Q2 bias vs Q1, Q2 and Q2 bias.
Attachments
Last edited:
The J113's in the kits are sorted by the pinch off voltage in .1 volt bins and matched to resistors to give about 10 ma bias current. You would want to run the 2sk170's as followers at roughly this figure and yours appears to be appropriate. No reason to spend extra money on 2sk170 as current sources. The bias current is not critical, but ideally you want q1 to have an idss similar to the bias value.
Here is my current (mis)understanding of the procedure to replace J113 with 2SK170BL in the 6-24 crossover:
"The J113's in the kits are sorted by the pinch off voltage in .1 volt bins and matched to resistors to give about 10 ma bias current. ....ideally you want q1 to have an idss similar to the bias value" - N.P.
Step 1: Sort Jfets. Pick JFETS with Idss close to 10mA, because N.P just wrote that Q1 should have a similar idss as the bias value (10mA?).
Edit: Although, this can't be true since the J113 were not close to 10mA in Idss when I measured them.
Step 2a): Populate the PCB with all components - except Q1 and Q1 Bias resistor.
Step 2b) Solder ONE sacrificial 2sk170BL in place of Q1. Different pinout than J113!
Step 3: Find Q1 bias resistor resistance:
Step 3a): Solder a variable resistor in place of ONE Q1 Bias resistor on the 6-24 crossover PCB.
Step 3b): Turn variable resistor until the bias current is close to 10mA. Where to measure?
Step 3c): Read value of variable resistor with multimeter.
Step 3d): Replace with fixed resistor.
Questions:
Where do I measure the current across the Q1 JFET and Q1 Bias resistor?
Answer: I could just populate with original Q1 J113 and its Q1 Bias resistor and find out where I read 10mA. Does anyone have better idea?
Which Q1 Bias resistor restistance would be a good start for the variable resistor, not to blow up too many JFETs?
Feel free to correct and guide me (out).
"The J113's in the kits are sorted by the pinch off voltage in .1 volt bins and matched to resistors to give about 10 ma bias current. ....ideally you want q1 to have an idss similar to the bias value" - N.P.
Step 1: Sort Jfets. Pick JFETS with Idss close to 10mA, because N.P just wrote that Q1 should have a similar idss as the bias value (10mA?).
Edit: Although, this can't be true since the J113 were not close to 10mA in Idss when I measured them.
Step 2a): Populate the PCB with all components - except Q1 and Q1 Bias resistor.
Step 2b) Solder ONE sacrificial 2sk170BL in place of Q1. Different pinout than J113!
Step 3: Find Q1 bias resistor resistance:
Step 3a): Solder a variable resistor in place of ONE Q1 Bias resistor on the 6-24 crossover PCB.
Step 3b): Turn variable resistor until the bias current is close to 10mA. Where to measure?
Step 3c): Read value of variable resistor with multimeter.
Step 3d): Replace with fixed resistor.
Questions:
Where do I measure the current across the Q1 JFET and Q1 Bias resistor?
Answer: I could just populate with original Q1 J113 and its Q1 Bias resistor and find out where I read 10mA. Does anyone have better idea?
Which Q1 Bias resistor restistance would be a good start for the variable resistor, not to blow up too many JFETs?
Feel free to correct and guide me (out).
Last edited:
@ one and only, appreciate your help. Since I don’t have enough 2sk170’s over the 10mA mark, I’ll use the batch of 12 near 8mA to replace all the Q1’s and lower the Q2 bias resistor from 100R such that it’s idss is a bit lower…I’m getting .761vdc on my “test bench”. I did a quick google search of pinch off and most of the results were rather nefarious, so I’ll leave it alone.
Attachments
@edwinjones4 : Something isn't right. If the original Q2 bias resistor is 100R for Q2 to conduct a current of 10mA, then you need to increase the resistor value if you want to lower that current.
thanks Dennis. is the method of testing the Q2 current the same as the 2sk170? I realize the pinout is different. Do you need to connect the Q2’s gate and source together for testing as one does for testing the 2sk170’s? See pics for how I’ve tested the original Q2 with 100ohm resistor. It results in roughly 1.9 vdc. Using a 47ohm resistor cuts the current in half.
Attachments
Please see the attached. For the Q2 bias resistor you want to connect it between the gate and the drain of the jfet.
The current flowing through the jfet is the voltage across R1 divided by R1. So you adjust R1 to get the current you desire (say, 8mA).
Notice that Idss corresponds to the case when R1 = 0 with a resistor on the side of the Drain just for current measurement.
The current flowing through the jfet is the voltage across R1 divided by R1. So you adjust R1 to get the current you desire (say, 8mA).
Notice that Idss corresponds to the case when R1 = 0 with a resistor on the side of the Drain just for current measurement.
Attachments
Let's not get too excited. The bias resistor for q1 is not very important to the performance. It's the resistor for q2 that's important. I have other examples where I omit the q1 bias resistor altogether. For the 2sk170 just set it at 0 or low value and don't worry about it