It may be beneficial to reverse the phase of the tweeter at the output of the power amp when using a 12 dB/Oct crossover.
https://www.diyaudio.com/community/threads/diy-biamp-6-24-crossover.357657/page-70#post-7437609
Elwood625 however is writing about this filter being unity gain in the pass band:
https://www.diyaudio.com/community/threads/diy-biamp-6-24-crossover.357657/page-70#post-7437843
Anyway, I guess all is sorted out.
What I was thinking about by “drive” is the ability to drive a lower impedance than the normal 10kOhms or more, like the old (inverting) “Otala-Lohstroh” amplifier.
Elwood625 however is writing about this filter being unity gain in the pass band:
https://www.diyaudio.com/community/threads/diy-biamp-6-24-crossover.357657/page-70#post-7437843
Anyway, I guess all is sorted out.
Trying to suss out the current requirements for the 6-24 as I intend to power with a shunt regulator.
Do I have this right.....given the J113 are biased to 10mA.....x 24 devices =240mA.
Seems a lot so I may have this well wrong!?
Do I have this right.....given the J113 are biased to 10mA.....x 24 devices =240mA.
Seems a lot so I may have this well wrong!?
So 120mA total?! That sounds a bit more like it. A bit of a noob with this sort of thing and I try not to just ask all the time but to search or learn how to calculate an answer, but had failed this time.!
I've been using the green bipolar electrolytic 10u/25V MUSE BP DC decoupling capacitor at the input buffers (where it connects to the 1Meg bias resistor to Vref. I've built three boards with them, and found that sometimes (perhaps 20% of the time) there seems to enough leakage current through the cap to 0V ground through the input POT to drag down the bias point for the top Jfet to around a couple of volts instead of Vref. I've found that it is fixed by replacing with smaller value/higher voltage same brand e.g. Muse BP 1u/50V. I searched this thread and not been able to find a similar occurrence so just wondering whether anyone has noticed this issue or it's just me. I've tried measuring leakage current through the "faulty" caps out of circuit with ohm meter but they seem to show up as open circuit which is what they should be. Weird.
Update: I decided to measure cap leakage current by using a 9V battery through a 1 meg series resistor also in series with a 100 ohm shunt resister to measure the voltage in order to ascertain leakage current values flowing. Testing my batch of Muse 10u/25 caps, I found that slightly more than half of them had remnant leakage of around 5uA after about 1 minute. The others showed no leakage in this circuit after about 30 seconds. I'll therefore select the low leakage ones for this position (or better perhaps smaller uF values), while the others 10u/25V are ok for output dc blockers. Not sure if my batch is representative for the brand or if they are dodgy production (they were quite cheap from china).
From Nichicon datasheet https://eu.mouser.com/datasheet/2/293/e_ues-3082323.pdf
"Leakage Current※: After 1 minute's application of rated voltage at 20°C, leakage current is not more than 0.03CV or 3 (μA), whichever is greater."
the max is then 0.03CV for 10u/25V or 7.5uA which would drop 7.5V through the 1meg resistor and mess up biassing of the top Jfet on the input buffers.
Perhaps 1u/25V could be a better choice for the input buffer to get more more reliable biassing.
"Leakage Current※: After 1 minute's application of rated voltage at 20°C, leakage current is not more than 0.03CV or 3 (μA), whichever is greater."
the max is then 0.03CV for 10u/25V or 7.5uA which would drop 7.5V through the 1meg resistor and mess up biassing of the top Jfet on the input buffers.
Perhaps 1u/25V could be a better choice for the input buffer to get more more reliable biassing.
"Perhaps 1u/25V could be a better choice"
Perhaps not using possible fake capacitors is more better choice. 😏
Perhaps not using possible fake capacitors is more better choice. 😏
The NP schematic shows 0.1uf on the input. Not sure why 10uf is in the BOM.
Using an RC filter calculator, I'm not sure whether we use the 1k or the 1M ohm for input impedance. Assuming 1meg and so with 0.1uf gives cut off at 1.5Hz. So we can probably go comfortably lower than 10uf!?
Using an RC filter calculator, I'm not sure whether we use the 1k or the 1M ohm for input impedance. Assuming 1meg and so with 0.1uf gives cut off at 1.5Hz. So we can probably go comfortably lower than 10uf!?
I believe papa once stated 10uF a typo.
Anyway, use 1 Meg for the calculation.
Check also out https://en.m.wikipedia.org/wiki/JFET#:~:text=Because a JFET in a,as input to the gate.
Anyway, use 1 Meg for the calculation.
Check also out https://en.m.wikipedia.org/wiki/JFET#:~:text=Because a JFET in a,as input to the gate.
I have not seen this issue addressed here. Lots of folks using this for a subwoofer xover. If using a single sub, as I am, the L and R must be summed and then only one channel of the xover will suffice. Is there a nifty way to do this in the circuit of the xover, or should I just hang the summing parts off the input RCA's before the xover? I'm wondering if they can be summed behind the input buffers by simply connecting one channel to the other. Will the buffers serve to isolate the summing from upstream stereo signals?
I have another question on a topic that I have not seen discussed here, or anywhere else that I can find.
If I want to create a bandpass, can the HP be second order and the LP fourth order? Does this screw up the phase of the cascaded filters?
If I want to create a bandpass, can the HP be second order and the LP fourth order? Does this screw up the phase of the cascaded filters?
I recently decided to upgrade my 6-24 crossover by replacing the J113 JFET's with LSK170's purchased from the diyAudio Store. (as recommended by Nelson in several posts)
The quad sets of LSK170's from the store are specified as having Idss of 8-11mA. The parts I received had measured Idss values of: 4.7, 6.5, 5.3, 4.9 mA All parts well BELOW the 8-11ma Idss store spec. The J113's I am replacing all measure 7 - 8 mA Idss.
QUESTIONS: Should I return / exchange these parts ? Or, use them at say a low bias of maybe 4 mA ? I was hoping not to have to change the existing 100 ohm bias resistors.
NOTE: I measured all the LSK170's using the procedure outlined here:
https://www.diyaudio.com/community/threads/diy-biamp-6-24-crossover.357657/post-7099039
I tested with VDS set to 10V per the LSK170 datasheet.
The quad sets of LSK170's from the store are specified as having Idss of 8-11mA. The parts I received had measured Idss values of: 4.7, 6.5, 5.3, 4.9 mA All parts well BELOW the 8-11ma Idss store spec. The J113's I am replacing all measure 7 - 8 mA Idss.
QUESTIONS: Should I return / exchange these parts ? Or, use them at say a low bias of maybe 4 mA ? I was hoping not to have to change the existing 100 ohm bias resistors.
NOTE: I measured all the LSK170's using the procedure outlined here:
https://www.diyaudio.com/community/threads/diy-biamp-6-24-crossover.357657/post-7099039
I tested with VDS set to 10V per the LSK170 datasheet.
Robin. I found your post when researching how I might be able to implement a Linkwitz Transform for my sealed box woofer, by using one of Nelson Pass’s ASP boards.
I’m stuck in the same place you were. That is “I think I need some gain”, but only unity gain on this board.
Did you find a way to do it?
Curious as well, I have my finger hovering over the add to cart button on some of these super shiny matched jfets.
This article by Nelson Pass may help with your decision.