• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Steve Bench's tube crossover

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2004
Paid Member
I know someone who has, and that person is quite pleased with it.

Note that Steve's site is gone as of the end of this month - if you need anything I would grab it now. Select run from the windows start menu, type cmd and navigate (cd or dir commands as needed) to the right directory then you can use wget to grab it all at one time, just remember use this syntax:

wget -mr h tee tee pee colon slash slash the_name_of_a_site.whatever (com, net, org, etc.)

Sorry the automatic url parser forced me to create the above travesty, but I think it is kind of humorous.. ;) Hope this is obvious.. :devilr: This will completely recursively mirror the site in question to your hard drive. (The location where wget lives on your drive actually.)

For windows go here: http://gnuwin32.sourceforge.net/packages/wget.htm

For linux go here: http://www.gnu.org/software/wget/
 
It looks like the terms Steve has put on his site will let other people mirror it if they attribute him. ("Feel free to put these schematics or other information on your web sites, but give the author credit.")

I have the webspace to support it, I might look at mirroring all of the content in the next couple of days.
 
I need to get schematics off of another computer on to this one to post them but I have been playing with the design in LTspice. My topology has a 50Hzish LP for the subwoofer and a 50-200 bandpass for the helper woofer. The full range will be run um... full range as it starts rolling off at about 200 hz anyway but I will include a simple low order passive high pass that can be switched in.

I am investigating a couple of things unique to my situation.

One is that I wanted to tune the high pass part of the band pass to a higher Q that normal because the normal configuration gives a big round peak in the middle of the passband which is not good in this case. I did not have a formula for the Q of a 3rd order S-K so I inferred some things from the Q factor in a 2nd order and started diddling with the resistor values. By increasing the value of the first resistor to ground in the high pass portion and adjusting the feedback resistor as necessary to keep the turnover frequency where I wanted I was able to get the desired boost right before cutoff. When I followed this with a more standard low pass section at the 200Hz point I got a very nice slope of about 3dB/8va downward as frequency increased in the pass band. This along with the use of highish Q driver should provide good integration in the open baffle speakers I am using.

I found that I could diddle the caps similarly in the lowpass section to get a lower than normal Q resulting in a steeper slope but I liked the looks of the 3dB slope better and I think that it will be superior in the long run. Of course experimentation of the actual circuit will tell the story in the end with the proof being in the pudding and all that.

The second thing I looked at was the tube choice. I had been designing with the idea of using 6N1Ps as the final CF in each band but really wanting to use those in a phono amp project I decided to try the sim with 12AX7s of which I have four on hand. This is probably one of the worst tubes to choose but there they sit on my bench so what the heck. If it doesn't work it won't exactly blow up so why not give it a try.

Given the lowish frequencies involved I thought that the AX7s might work OK if obviously not optimally. Sims seem to support this. Even into a tough load of 10k I was able to maintain good slopes (though with about 1.5dB less gain). It required a somewhat larger but still manageable (2.2uf) output cap for the subwoofer section and the woofer filter seemed to be just fine with as low as 0.68uf which would allow a poly cap to be used there. Seems like a reasonable compromise.

By wiring for the 12A family I can start with the AX7s I have on hand and switch to the more capable 12AT7 at a later date by simply switching tubes and reducing Rk to draw more current.

At first both the woofer and subwoofer will be driven by sand but eventually the woofer will be a tube amp. With this in mind I tried the bandpass with a 47k load and it tightened up the curves so that there was no discernible difference before and after the output cap.

Just wanted to let you know where I was heading with this. Will get schemos up as soon as I can.
 
I have been slowly transferring stuff over to this computer via thumb drive. I don't have the latest diagram but the attached will suffice for discussion.

In the current simulation the 6N1Ps have been replaced with 12AX7. I changed R3 to 380K. I don't remember for sure but R11 may have been adjusted slightly. Last night I simulated various capacitances across R100 and R16 to simulate input capacitance on amplifiers and interconnects. C8 has also been changed to 0.68uf.

With input capacitance of less than 100pf or 200pf the output is nearly unchanged. A loss of a few tenths of a dB starts to show up at around 1000p but the shape of the curve remains the same. When we get up above 10,000pf or so the shape of the curve starts to show quite noticeable deformity.

I suppose that square wave response would not be very pretty but given that the upper corner frequency is about 200Hz this may be a non-issue.

So what kind of input capacitance should I expect with average 3' interconnects feeding a solid state integrated amplifier?

I suppose the next step is to do similar runs using 12AT7s at about 8ma.
 

Attachments

  • tubecrosoverax7input.pdf
    12 KB · Views: 215
Thanks for the mirroring service!

The attached shows the LTspice results. The upper curves are before the output caps and the lower ones after. The rolloff in the subwoofer output below 20Hz is intentional. By adjusting the first two resistors in the highpass section of the woofer output I could get a nice "mesa" shaped curve with flat pass band but I wanted the tilt for my application. The actual sub to woofer crossover point is a bit higher than it looks at first since the gain in the subwoofer output is a bit lower. When the sub amp is adjusted to give similar output the cross will move up a few Hz.
 

Attachments

  • tubecrosoverexp_12ax_grid_output.pdf
    75 KB · Views: 169
Steve Bench Website

Hi All:

Just wanted to give my THANKS to both Steve Bench and TheVoice at greygum.net for making Steve's website available. There's so much kwell stuff there for a nuby like me it would have been a shame to lose it.:clown:

Mike
If there's no sound in a vacuum, where'd the music come from...?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.