DIY biamp 6-24 crossover

Jump to Latest
#1,561
If you bought the kit from DIYAudio, the the JFET's have bias resistors which match the JFET'S CCS at around 8mA. You would want to replace them with the same type you have in circuit. Do you know what the value of the bias resistors are for that filters CCS?

I have dozens of the JFET's from Nelson with the bias resistors as supplied by the DIYAudio store. I'd be happy to mail you replacements if needed.
 
#1,563
Actually I've just realised that this is an old problem which was never resolved! A couple of years ago when the problem surfaced I replaced all the transistors in the faulty section; this didn't fix it, so, in my time honoured head-in-the-sand approach I put the board aside and built up the other board I happened to have lying around. Job done. However, I now want to run two crossovers for a three way system.

I'm not sure that it is the transistors, as they were all replaced. Unless of course I have two independently faulty transistors?? The fault sounds a bit like what you get if the DC offset is wrong on your amp - a buildup of charge, a discharge, rinse and repeat.

Anyway, the bias resistors are 68ohms. I guess I'll try replacing them again - if someone has some available that match that value that'd be grand.

Many thanks!
 
#1,565
elwood625 said:
It would change the crossover center point frequency.
Click to expand...
Care to evolve?
24000.png
2400.png
 
#1,566
stajo said:
Hi. About example component values. For 1000 Hz cutover the examples uses 4,7 nF and 24k base values for xover comps. Would it affect the circuit performance to use for example 47 nF and 2,4k instead?
Click to expand...
I too had this question, but decided not to ask. I'll stick with the convenience (for now) of the 50k trim pots.

Are you asking because you are concerned about noise? For what it's worth, the resistors are low noise (max. 1.5 μV/V for R > 1 MΩ) as per data sheet. I suspect if keeping the pots, most noise would be generated there?
 
#1,567
Yes, maybe to give some benefit to thermal noice but mainly because I have built other xovers with those values so I have alot of comps to match from. Those were opamp based so my question was more about if this circuit would be affected in terms of current/drive issues related to those values. My plan is to breadboard with fixed values and compare.
 
#1,570
I measured 66Ohm Rout of the buffer stage (without the output protection resistor). I tried to incorporate the Rin of my soundcard (10k) but I am not sure if I did it correctly.

Distortion vs .load (load without the 10k Rin of my soundcard). The absolute values are maybe not correct but you can see the trend.
THD % Load
0.0067 10k
0.0125 3.6k
0.088 1k
 
Last edited:
  • Like
Reactions: stajo
#1,572
So what do you guys think of basing a fixed xover on:

Dale CMF5511K300FHEB 1/2W 11.3Kohms 1% and Wima FKP2C021001G00HSSD 63V .01uF 2.5% using one resistor for R1 and R2 and two serial resistors for the rest and one cap on every position except C6 and C8 that get dual in series?

Exact center point for crossing over is less important then phase alignment etc, but I aim for ~1000 Hz.

Easy to source and decent quality for the buck and only two models.

xpass.png
 
#1,573
Novice here, so take my thoughts with a grain of salt. I think limiting to a couple of resistor values would be fine for a "good enough" approach if you aren't super worried about perfect phase alignment. These simulations work obviously as technically correct if your drivers and cabinet had perfectly flat responses.

Of course in the real world, if you are concerned about good phase alignment you will need a full suite of near and far measurements of all drivers from 0-180 degrees around the speaker, crossover component values optimised by simulation software and using ears from there. After all that, those resistor values will not be looking so neat, surely.
 
You must log in or register to reply here.
Top Bottom