I’m struggling to get two different power amps working well together in a horizontal bi-amp setup. Could I ask for some guidance, please?
The system prior to a recent change was:
• HF channel = Fi 2A3 monos (now using 46) on Altec 1505B / 288H
• LF channel = Don Allen SE 807 stereo on 65 Hz hypex midbass horn / JBL 2220J
Using miniDSP 2x4 HD, I needed to lower HF channel outputs by 10dB or so. With that, plus some HF delay and EQ top and bottom, I thought I was doing a decent Jean Hiraga impression.
Recent change, resulting fail, and reason for this post was swapping the HF amps for Don Allen SE 801A monos. miniDSP gain settings have flipped, with the LF channel now the one requiring attenuation (~3dB). But, even with that level match, it’s not working optimally. HF channel has lost some vitality—lacks a bit of the life and energy it used to have. Low testosterone levels or something. (Likewise, I found myself increasing high-shelf gain by a few dB.)
I’m assuming these 2-stage 6BC8/801A amps are simply a poor gain match for the 3-stage 1/2-6SN7/6J5/807 (pentode-wired) amp. Does that sound right? If so, wondering if I could add a mild step-up transformer before the 6BC8 without too much downside—or are there better ways to address this?
Thanks for any suggestions—they’ll be hugely appreciated.
The system prior to a recent change was:
• HF channel = Fi 2A3 monos (now using 46) on Altec 1505B / 288H
• LF channel = Don Allen SE 807 stereo on 65 Hz hypex midbass horn / JBL 2220J
Using miniDSP 2x4 HD, I needed to lower HF channel outputs by 10dB or so. With that, plus some HF delay and EQ top and bottom, I thought I was doing a decent Jean Hiraga impression.
Recent change, resulting fail, and reason for this post was swapping the HF amps for Don Allen SE 801A monos. miniDSP gain settings have flipped, with the LF channel now the one requiring attenuation (~3dB). But, even with that level match, it’s not working optimally. HF channel has lost some vitality—lacks a bit of the life and energy it used to have. Low testosterone levels or something. (Likewise, I found myself increasing high-shelf gain by a few dB.)
I’m assuming these 2-stage 6BC8/801A amps are simply a poor gain match for the 3-stage 1/2-6SN7/6J5/807 (pentode-wired) amp. Does that sound right? If so, wondering if I could add a mild step-up transformer before the 6BC8 without too much downside—or are there better ways to address this?
Thanks for any suggestions—they’ll be hugely appreciated.
Generally, when bi-amping, I like to use the same amplifier for everything. This alleviates a lot of the problems that you're having and will give you a much more consistent result. I think you just have drastically different gain situations and given the many other factors that determines an amplifiers performance, you can have a lot of other problems arise.
Just make sure that the amplifier has enough power to deliver to the lower frequency speakers. You will use a lot less power for the tweeters, but that's fine since you know it will be driven similarly. This takes the amplifier out of the question (mostly) and now you are just dealing with everything in your DSP.
Using high frequency delay is normal if the tweeter is flush mounted with the woofer. Also, I recommend only using the EQ in the Input section instead of the output section. If you apply EQ, it should effect both drivers in the box equally. You will have phase issues if you only EQ the tweeter and those frequencies are still acting in the woofer.
If you do decide to use different amplifiers for the highs and lows in the speaker then make sure you put the more powerful amplifier to the woofers.
Also, if you're using a Planar or ribbon tweeter then you should put a 30uf capacitor in line with the tweeter to avoid any low frequency damage when the amplifier is turned on. This capacitor is large enough to not effect the crossover.
One question-where is the volume set on your amplifiers? Normally, when bi-amping, amplifiers are turned all of the way up and the DSP is used to shade the gain appropriately. But be careful, because this can apply a lot of power to your speakers and risk blowing them up.
Just make sure that the amplifier has enough power to deliver to the lower frequency speakers. You will use a lot less power for the tweeters, but that's fine since you know it will be driven similarly. This takes the amplifier out of the question (mostly) and now you are just dealing with everything in your DSP.
Using high frequency delay is normal if the tweeter is flush mounted with the woofer. Also, I recommend only using the EQ in the Input section instead of the output section. If you apply EQ, it should effect both drivers in the box equally. You will have phase issues if you only EQ the tweeter and those frequencies are still acting in the woofer.
If you do decide to use different amplifiers for the highs and lows in the speaker then make sure you put the more powerful amplifier to the woofers.
Also, if you're using a Planar or ribbon tweeter then you should put a 30uf capacitor in line with the tweeter to avoid any low frequency damage when the amplifier is turned on. This capacitor is large enough to not effect the crossover.
One question-where is the volume set on your amplifiers? Normally, when bi-amping, amplifiers are turned all of the way up and the DSP is used to shade the gain appropriately. But be careful, because this can apply a lot of power to your speakers and risk blowing them up.
I'm not familiar with the brands and models of amps you're using but as a general rule when bi-amping particularly with a DSP crossover you'll get your best gain staging by running the DSP outputs at unity gain and taking your attenuation at the inputs to the amplifiers. If your amps don't have level controls it's probably worth adding some or at least building some fixed input pads rather than taking any significant attenuation in the DSP.
The reason I lean towards attenuating within the DSP is that you know it will be consistent. With amplifiers, especially as they age, the analog parts can be inconsistent and if you turn two of them (or two channels) up half way then it might not be the same output. With DSP, attenuating 10db on each channel will be consistent. If you turn the amplifiers all of the way up then you will more likely be putting them in an equal position. All pro-audio gear works this way.
For home audio gear it could be different. Additionally, it is still best to run the DSP at as close to nominal as possible so that gain structure is best. Plus, there is often a pre-amp before the DSP so that would be the best place to attenuate what is entering the DSP. So maybe it's best to turn the amps up, run the DSP at nominal then control the volume of what is entering the preamp. It all depends on your system. But ultimately, it is important to have closely or perfectly matched amplifiers for bi-amping. I think the problem with the original post is drastically mismatched power amplifiers.
For home audio gear it could be different. Additionally, it is still best to run the DSP at as close to nominal as possible so that gain structure is best. Plus, there is often a pre-amp before the DSP so that would be the best place to attenuate what is entering the DSP. So maybe it's best to turn the amps up, run the DSP at nominal then control the volume of what is entering the preamp. It all depends on your system. But ultimately, it is important to have closely or perfectly matched amplifiers for bi-amping. I think the problem with the original post is drastically mismatched power amplifiers.
I appreciate that. And I actually have bi-amped these same speakers with two of the same direct-coupled 2A3 amps. The current project simply has different parameters and goals.
Even if flush-mounted doesn’t really apply here, woofer does sit about 3 feet behind tweeter. Channel delay is the whole reason I’ve embraced DSP.
Did not know this and figure it would have taken me years to learn it. I get it. Now, here’s a follow-up: if my 12 dB/oct crossover freq is 450 Hz, and my low and high shelf filters start at 65Hz and 5KHz respectively, does the same caution hold? (For practical purposes, I’m asking whether I should add the “other shelf” to both of the existing?)
Indeed—now we’re back to what I said earlier about different parameters and goals.
Neither has adjustment capability. So yeah, let’s call that “all of the way up.”
Thank you, @ThatSoundsGood—very helpful.
Last edited:
So, this system has existed in probably two dozen slightly different configurations over time—always the same idea, but a different amp here, a different driver there. The current configuration is actually the one in which I’ve gotten closest to unity at DSP output (LF = -3.0dB; HF = 0.0dB). That isn’t to be argumentative at all—I’m sure you’re correct. It’s just that I’m not sure how much closer I could get.
Likewise, I’m not sure if that really gets to the heart of it. It’s more the fact that my gain settings went massively upside down with the recent HF amp change. HF speaker is somewhere between 5dB and 10dB more sensitive than LF speaker; and the immediately prior amp pairing had HF attenuated by 10dB—totally expected.
The fact that the new HF amp would gobble up 13dB (causing the LF speaker now to need attenuation) was totally unexpected. As was the reduction in sound quality, which sounds like a gain problem. Now, I’d like to figure out what can be done, if anything, to recapture SQ … even if that means ruining the close-to-unity gain structure I have now. That looks more like a sad coincidence to me. Given the differing sensitivities between channels, isn’t it the wrong result?
No level controls on either amp. This actually sounds like a vote for my step-up transformer idea—is it?
Last edited:
I would put both amps into the system at different times and measure the response at the amp output terminals. The level isn't important when doing this, you just want to see whether the output impedance of the amp has altered the response by interacting with the speaker impedance.
I'm learning that this is a different kind of project. Very cool.
I wouldn't worry about putting the low and high shelf filters on both drivers. That is fine for each driver.
Also, there is a time and place to only use EQ on one driver, but it is generally better to start by doing it to both of them, especially in the regions where they overlap.
Last edited:
I second this idea. It makes sense that you would have to attenuate the tweeter, but whatever is causing the attenuation of the woofer sounds like a problem. Is there a drastic impedance mismatch between the speaker and amp? Is there just something wrong with the amplifier?
Those are really close. They look to be within 1db to me. They are different but that almost wouldn't be audible if the graph is correct. Different amplifiers can create a different responses in speakers but it won't normally be drastic. Your original problem sounds drastic.
I don't understand, that was my original problem.
I think this kind of difference will be audible on a good system. In any case amplifier Zo might be higher than 1 ohm and the tweeter in question will have a different impedance curve.
In particular, compression tweeters tend to have difficult resonance peaks.
I think this kind of difference will be audible on a good system. In any case amplifier Zo might be higher than 1 ohm and the tweeter in question will have a different impedance curve.
In particular, compression tweeters tend to have difficult resonance peaks.
Sorry, I misunderstood the situation. I was referring to the original amplifier problem brought up by Jim Hodgson where the amplifiers were drastically different.