Using High Pass Filter with BHK 250


I’m currently using my BHK 250 with Vandersteen 5 speakers. Recently I tried a Real Time Analyzer (RTA) app for my iPad. The RTA gives a frequency response measurement which can help you identify frequency anomalies with your room. While I am a beginner on using this app I noticed a higher than expected level of low bass (below 60 Hz). This surprised me so I dug a little deeper and discovered a potential problem, read on …

For those unfamiliar the Model 5 speakers they have a built in subwoofer. For the subwoofer to properly function the input of the amplifier must be filtered to roll off the bass (the subwoofer amplifier then fills in for roll off). To accomplish this Vandersteen provides a filter box to fit between the preamp and amp. The filter box is adjustable and the adjustment is based on the input impedance of the amplifier (I’m using balanced cables so the input impedance of the BHK 250 is 200K ohms). The ultimate goal is to have a 3 dB down point at 100 Hz (relative to 1000 Hz). To measure this 3 dB down point you do it electrically rather than acoustically (to avoid the room effect). The short version of the procedure is to input a 1000 Hz signal, measure output voltage at speaker taps, then input a 100 Hz signal and measure output voltage at speaker taps again. The output level at 100 Hz is supposed to be 0.707 of the output level at 1000 Hz.

When I set the filter box for 200K I get a reading about 0.95, in other words very little roll off at 100 Hz. While I couldn’t easily figure out where the 0.707 (i.e., 3dB down) point was, I would expect it is somewhere in the 60-70 Hz range. The result of this improper roll off is accentuated bass. If this is in fact what is happening you would think I would have heard it. Well I might have, but, the frequencies are quite low so there isn’t a lot of musical content plus my room is small so very low bass is not the best. Anyway, I played with the filter box and found the best I could do was 0.841 with it set for 400K impedance which is the highest it will go. Something is still wrong …

So now I have to figure out why this is happening. My first reaction is the stated output impedance of the BHK 250 is wrong, but I would suspect this logic leads to an output impedance of 800k to 1Meg! Doesn’t seem likely. I don’t recall anyone else mentioning this so I’m putting the question out to the “experts” - Can someone explain to me what is happening?

New Forum - Tips and Tricks, Problems?

Well, perhaps they need you to put in the individual input impedance which is 100kΩ. When we say it is 200kΩ we are measuring across the balanced input which places its 100kΩ in series, thus 200kΩ. I’ll bet they want the individual leg impedance.

Try entering 100kΩ And see if that doesn’t fixc it.


Actually it says on the filter box “Amplifier Imput Impedance is the Sum of the Positive and Negative Leg” so the 200K figure is the correct one. Additionally, I tried the higher impedance and it got me closer to the correct roll of. Going lower in impedance should lessen the roll off, though I did not try it. My main concern, which may be unfounded, is there is something about the output impedance of the BHK that is different from other amplifiers. Perhaps Bascom will see this and comment. I could also call Vandersteen to see if they can make sense of it, though I’m not sure how I should ask the question.


If you want to lower the crossover point with Vandersteen subwoofers either internal or external you lower the impedance setting. To raise the crossover point go higher. It is worth going at least two places in both directions though you may never get the .707exactly with any of them. You can just call Richard on the phone or e-mail him. Another option is to ask one of the long time dealers like John Rutan at Audio Connection. I used a pair of 2WQ’s for over 15 years with a least a dozen speaker and amp combos including three Vandersteen models and it was never as cut and dry at it was made to seem.


That’s good information to know, I did figure most of it out through analysis. The problem is I have gone up two settings from the recommended and it still is not where it should be. I understand that I won’t necessarily get the 0.707 figure, but I should be able to get within, say, 10%. Right now I can only get within 20% and I can’t go any further unless there is an unmarked setting. Before calling Vandersteen I wanted to make sure there wasn’t something about the BHK 250 that was unusual and was the root of the problem. So far that hasn’t happened so I may call Vandersteen tomorrow to see if there could be a problem with the filter boxes or I’m making a fundamental measurement error.


What measurements did you get when you went down the two steps?


None that I know of. The input impedance of the BHK is set with fixed shunt resistors so it’s pretty straight forward.


Thanks Paul! That’s one less thing to worry about.

Dawkinsj, I did not try any lower impedance settings, didn’t seem appropriate. Regardless I will try them later today just to verify.


I’m pretty sure I have solved this problem! When I was adjusting the filter boxes for the BHK 250 I was using the “Vandertones” files which are put out by Vandersteen for tuning the speakers. Since I seemed to be at an impasse with the BHK 250 I played around with my old amp, an Audio Research Classic 60, which I thought worked correctly. Without getting into details, I discovered that using a signal generator (rather than Vandertones) I was able to get the expected result. So the next logical step is to use the signal generator on the BHK, and surprise, it worked! I got a reading very close to 0.707 using the 200K setting on the filter so it was correct after all. Discovering this lead me to check voltage levels of the signals being used. The signal generator was less than 1% difference, but when I checked the Vandertones I discovered the 100 Hz tone was 33% higher than the 1KHz tone, aha! This must be where the problem is. I first verified I had the latest version of Vandertones, then I contacted Vandersteen. Richard said there has to be a problem with my files, they should be the same level. How this happened is a mystery, he thought a bad download, but I have one nagging thing…

The main use of Vandertones is for setting the 11 bands of the subwoofers, but there are a few other tracks. The Vandertones are to be used with a Radio Shack analog sound level meter, which I have used previously for this task. The problem with the Radio Shack meter is it is non-linear and needs compensation to read correctly. So the Vandertones are “adjusted” to compensate for the non-linear operation of the Radio Shack meter. The only thing I could think of is the level of the 100 Hz tone might have this “adjustment” but Richard said “No”. The files have been posted on the Vandersteen website for about six years and have not been altered in that time so it seems unlikely no one found the problem before so I have to agree, but…

Anyway, since I have a “fancy” RTA app with calibrated mic I probably will just let this go. Thanks to dawkinsj and Paul for their assistance!


I have added the ability to accept a post as the answer/solution to the original question posed by a thread.

As the OP (topic creator) you will have a button that allows you to accept a subsequent post as an answer:



Fascinating. Thank you for letting us know.