View Full Version : 250ti charge coupled networks - bi amped ?

07-18-2010, 12:42 PM
I'm still gathering the parts list and it seems more fitting to continue here. This is continuing from my original post:


Regarding the 1.0 mH 7.5ohm, I have decided to use a 1mh .7 ohm dcr and add a 6.8ohm resistor in series. The parts availability is better and seems acceptable to do this.

I'm contemplating on ordering a speakon 8 conductor cable and associated hardware. This will go from this external network to the speakers. The input side of the network will be 5 way posts / banana jack hardware.

The padding resistors will most likely be interchangeable via terminal blocks.

I have spent some time researching the possibility of an active crossover. This would eliminate quite a few passive network parts. The good ones seem to be very expensive and I don't have the ability/equipment to measure the setup. I would still end up with the upper frequencies handled by a passive network. To add to the complexity, there is already an active crossover for the subwoofer. Just seems like more room for errors at this point.:dont-know:

The main reason for the bi-amp quest is to use lower power tube amps and sweet sounding SS on the upper frequencies. The LF will be powered by a professional SS amp.

Would any of the network component values need to change if I separated the LF section from the rest of the network?

The components are expensive, so I want to do it right the first time.

Again, I'm using the schematics posted here:
http://www.audioheritage.org/vbullet...ull=1#post2306 (http://www.audioheritage.org/vbulletin/showthread.php?289-L250-gt-250Ti-Upgrade-and-Differences&p=2306&viewfull=1#post2306)

Thanks for these btw. Comments and advice is always welcome:yes:

07-18-2010, 01:56 PM
Here's what happens when you remove the series 80.0 uF capacitor pair and the parallel 2.0 mH inductor. You can also play around with the 5.1 ohm resistor on the conjugate shunt to affect the knee of the curve. Removing it completely results in the last plot.

07-18-2010, 02:08 PM
Doing it right by removing all the loses would require a network rework. It can be done but I don't have time to do it nor do I have the driver set anymore to make it as accurate as possible.

For example, the 108H network would simply be a 1.3 mH series inductor and a conjugate consisting of a 24.0 uF biased capacitor pair and a 10 ohm resistor. That yields a real nice voltage drive (based on an 8 ohm dummy load) that removes the 10 dBV loss built into the stock network. The black plot is 0.32 V input compared to the green plot which is a 1 V input to show the curve tracking more easily. The second graph shows the "low loss" network with a 1 V input.

Merely pulling the fixed resistor pads on the 104H and 044Ti unfortunately will not result in a viable voltage drive, hence the rework required of those filters as well.

07-18-2010, 02:47 PM
The difficulties are understood. Thanks very much for looking into it.

May I take another approach to the problem?

Just to clarify my earlier post, I'm not planning on using active crossover at this time. Maybe much later.

Would a charge coupled 250ti bq schematic get us closer to being accurate? The 035tia driver seems to be very close to the 044ti.

07-30-2010, 06:17 PM
I was trying to ask what the effects on the network would be if I separated the LF section. Similar to what was built for jblnut. The idea of trying different amps for upper has been a goal since the thought of the 250. The L250 is great as it is, but I can't resist tinkering with everything that I own. I bought a blemished pair to support that nature.

Anyhow, I have ordered most of the parts for the posted schematic. I will give the network a try in both configurations. I have gathered as much as I have time for on component selection. Hopefully everything will turn out great one way or the other.

Thanks for the posts of above. That satisfied my curiosity of going the active route.