I picked up some Bose 901 original's yesterday, but they came without the equalizer that is so important for making them sound right.
I am thinking about trying to build the equalizer for them myself.
Before I start, let me get a few things out of the way.
1) I am aware that no other company faces the amount of derision from audiophiles and audio-snobs on the net. I, too, am one of those snobs. I made the purchase with the intent of restoring and selling them.
2) On the other hand, these are supposed to be one of the best speakers that Bose ever made. I am curious how they were meant to sound and, if nothing else, they are an interesting piece of audio history.
3) I know that the EQ curve can be approximated by a regular EQ, but I don't have one of those. I am thinking an EQ box, even a homemade one, will make them easier to sell.
My goal is to make an EQ box for less than $100, as that is about the going rate on the 'bay. I have a little experience with a soldering iron, but the only circuit I've ever built from scratch is a c-moy headphone amp.
The original EQ had a couple of switches: a below 40Hz bass-reduction switch (read bass-boost off), a treble reduction switch (ditto), and a couple of treble "contour" settings on a rotary switch. The only feature I am interested in keeping is the below 40Hz switch since this allows the speakers to be used with lower powered amplifiers (<200W).
Also, I would like to use op-amps to simplify things if possible. That way I could power it off of a 12V wallwart instead of building a dedicated PS.
My first step will be to try to simplify the circuit as much as possible--focus on one channel and remove the parts of the circuit that will be switched off. I am pretty sure I know enough about electronics to do that.
Trouble is, I don't know enough about electronics to break down the circuit and figure out exactly what its doing to the waveform. Any suggestions on how best to proceed?
The SM is available at hifiengine.com. Here is the schematic: http://i1051.photobucket.com/albums.../bose_901-I_eq_schem_zpsa28902fc.png~original
I am thinking about trying to build the equalizer for them myself.
Before I start, let me get a few things out of the way.
1) I am aware that no other company faces the amount of derision from audiophiles and audio-snobs on the net. I, too, am one of those snobs. I made the purchase with the intent of restoring and selling them.
2) On the other hand, these are supposed to be one of the best speakers that Bose ever made. I am curious how they were meant to sound and, if nothing else, they are an interesting piece of audio history.
3) I know that the EQ curve can be approximated by a regular EQ, but I don't have one of those. I am thinking an EQ box, even a homemade one, will make them easier to sell.
My goal is to make an EQ box for less than $100, as that is about the going rate on the 'bay. I have a little experience with a soldering iron, but the only circuit I've ever built from scratch is a c-moy headphone amp.
The original EQ had a couple of switches: a below 40Hz bass-reduction switch (read bass-boost off), a treble reduction switch (ditto), and a couple of treble "contour" settings on a rotary switch. The only feature I am interested in keeping is the below 40Hz switch since this allows the speakers to be used with lower powered amplifiers (<200W).
Also, I would like to use op-amps to simplify things if possible. That way I could power it off of a 12V wallwart instead of building a dedicated PS.
My first step will be to try to simplify the circuit as much as possible--focus on one channel and remove the parts of the circuit that will be switched off. I am pretty sure I know enough about electronics to do that.
Trouble is, I don't know enough about electronics to break down the circuit and figure out exactly what its doing to the waveform. Any suggestions on how best to proceed?
The SM is available at hifiengine.com. Here is the schematic: http://i1051.photobucket.com/albums.../bose_901-I_eq_schem_zpsa28902fc.png~original
Well, sorry you invested in these. I had a friend that bought a pair of 901's in 1969 and we listened to them extensively. I was totally unimpressed. The problem could have been the five sided bedroom his father had built for him, but I much preferred my pair of LWE III's in my square bedroom. The LWE III's were cheaper, too.
I would suggest buying quality microphone and doing frequency sweeps in the open air to determine what frequency compensation is required. You could build an anechoic chamber to do this, but a quiet spot in the open air is a lot cheaper in my location.
WIthout the bose equalizer unit, you are back to using a bunch of drivers to derive a flat frequency curve by electronics. And that involves engineering from scratch, or mathematically analyzing the schematic diagrams. As you have surmised, op amp equalization would be a lot cheaper than buying a lot of non-polar electrolytic caps and large current inductors.
Picking the design might be easier if you had a 16 band or 20 band equalizer to set up a curve with knobs, before commiting to the soldering iron.
I would suggest buying quality microphone and doing frequency sweeps in the open air to determine what frequency compensation is required. You could build an anechoic chamber to do this, but a quiet spot in the open air is a lot cheaper in my location.
WIthout the bose equalizer unit, you are back to using a bunch of drivers to derive a flat frequency curve by electronics. And that involves engineering from scratch, or mathematically analyzing the schematic diagrams. As you have surmised, op amp equalization would be a lot cheaper than buying a lot of non-polar electrolytic caps and large current inductors.
Picking the design might be easier if you had a 16 band or 20 band equalizer to set up a curve with knobs, before commiting to the soldering iron.
Last edited:
Oh gee.
This is pretty simple.
The Bose 901 requires about 12dB of bass boost at about 35Hz, iirc... the highs had two settings, iirc, more and even more.
Today I'd just take *any* inexpensive EQ, Behringer is a best bet, and dial in the EQ settings. The manual did indeed have curves, so you could match them.
You DO NOT want a "shelving" EQ, which would be a straight LP (low pass), since that will put in way too much subsonic gain, and cause the drivers to exceed their excursion capabilities. The equivalent would be a highpass (aka "subsonic filter") and a LP ("bass boost"). That is a bandpass filter.
As far as measurement, there is a vast difference between in room measurement and outdoor measurement. A "flat" response outside is likely to be very bass boosted in the room. Also the distance between the 901 and the wall has a definite effect on the bass.
So, there are a variety of free software that will do acoustical measurements, and you can use almost any microphone you like, assuming it is reasonably decent. The Panasonic WM-61a is like a few $$ and is very flat. Much info on using it, especially on Sigfried Linkwitz' site. A straw, battery, and a connector and you have a measurement system.
Since you don't care much about absolute and true readings, all you need to do is to A) get the bass and highs to be approximately "in line" with the midrange level and B) have it sound ok in your room, to your ears.
I would suggest measuring these drivers at a very close distance to the drivers, and as far as possible from any surfaces. Even within 6" is fine... keep in mind that for highs, small movement of the mic position changes what you see. Ur trying to get things "in the ballpark".
So, you would dial in ur boost on your adjustable equalizer and watch the effect on the curve. Changing the slope matters, as does the center frequency, and "Q" or width of the EQ.
With a modern EQ system in theory you can do a better job than Bose originally did. Especially since you can dial it in precisely for your room and your ears.
Keep in mind that you NEED POWER to run 901s properly.
I suggest 200watts RMS minimum, IF you like to run at SPLs that anyone would consider not background levels.
As you increase the boost, the requirement for amp power goes up considerably.
Keep in mind that for each 3dB of boost (EQ) you need to DOUBLE your amplifier power. Doing the math you can see that *if* you had to run at say 4 watts average to make 90dbSPL, 9dB of boost requires 32 watts. That doesn't sound too bad. But that's at 90dB. If you want peaks at 100dB you have to go up 9dB from there. So 9dB up from 32 watts is 64,128, 256watts!
Hope this helps...
_-_-
This is pretty simple.
The Bose 901 requires about 12dB of bass boost at about 35Hz, iirc... the highs had two settings, iirc, more and even more.
Today I'd just take *any* inexpensive EQ, Behringer is a best bet, and dial in the EQ settings. The manual did indeed have curves, so you could match them.
You DO NOT want a "shelving" EQ, which would be a straight LP (low pass), since that will put in way too much subsonic gain, and cause the drivers to exceed their excursion capabilities. The equivalent would be a highpass (aka "subsonic filter") and a LP ("bass boost"). That is a bandpass filter.
As far as measurement, there is a vast difference between in room measurement and outdoor measurement. A "flat" response outside is likely to be very bass boosted in the room. Also the distance between the 901 and the wall has a definite effect on the bass.
So, there are a variety of free software that will do acoustical measurements, and you can use almost any microphone you like, assuming it is reasonably decent. The Panasonic WM-61a is like a few $$ and is very flat. Much info on using it, especially on Sigfried Linkwitz' site. A straw, battery, and a connector and you have a measurement system.
Since you don't care much about absolute and true readings, all you need to do is to A) get the bass and highs to be approximately "in line" with the midrange level and B) have it sound ok in your room, to your ears.
I would suggest measuring these drivers at a very close distance to the drivers, and as far as possible from any surfaces. Even within 6" is fine... keep in mind that for highs, small movement of the mic position changes what you see. Ur trying to get things "in the ballpark".
So, you would dial in ur boost on your adjustable equalizer and watch the effect on the curve. Changing the slope matters, as does the center frequency, and "Q" or width of the EQ.
With a modern EQ system in theory you can do a better job than Bose originally did. Especially since you can dial it in precisely for your room and your ears.
Keep in mind that you NEED POWER to run 901s properly.
I suggest 200watts RMS minimum, IF you like to run at SPLs that anyone would consider not background levels.
As you increase the boost, the requirement for amp power goes up considerably.
Keep in mind that for each 3dB of boost (EQ) you need to DOUBLE your amplifier power. Doing the math you can see that *if* you had to run at say 4 watts average to make 90dbSPL, 9dB of boost requires 32 watts. That doesn't sound too bad. But that's at 90dB. If you want peaks at 100dB you have to go up 9dB from there. So 9dB up from 32 watts is 64,128, 256watts!
Hope this helps...
_-_-
Also, while I am still thinking about this...
...there are multiple threads on the 901s here, check them.
...one idea is to eliminate the need for HF EQ by installing an array of tweeters. One would make a box that sits on top of the 901s and has some combination of reflecting and direct tweeter... one would probably put a pad on the front firing tweeter in the case where one used say 2 rear facing and one front facing. But as the number of rear facing goes up the need for padding the front one decreases - but since tweeters are typically more sensitive (higher output for given input) one might need to pad the entire tweeter array or use a separate amp for the tweeters (not a bad idea). Depending on what one measured, and where the xover point is it may be sufficient to use a small inductor of proper value to roll off the 901.
_-_-
...there are multiple threads on the 901s here, check them.
...one idea is to eliminate the need for HF EQ by installing an array of tweeters. One would make a box that sits on top of the 901s and has some combination of reflecting and direct tweeter... one would probably put a pad on the front firing tweeter in the case where one used say 2 rear facing and one front facing. But as the number of rear facing goes up the need for padding the front one decreases - but since tweeters are typically more sensitive (higher output for given input) one might need to pad the entire tweeter array or use a separate amp for the tweeters (not a bad idea). Depending on what one measured, and where the xover point is it may be sufficient to use a small inductor of proper value to roll off the 901.
_-_-
Thanks for all the input! These are some interesting ideas. I found the service manual and it has a chart with some response levels at various test frequencies. I'm not 100% how to interpret this, but I used these numbers to approximate a curve on a software equalizer. They sounded okay, but none too impressive. Of course I had to tone down the bass since I only have about 80W to play with right now. I need to put my 200 watter under the knife before I can use it again.
@indianajo, I'm not too invested in these so far. Picked them up for $20 so don't feel too bad about my investment. I could easily sell the stands alone for twice that.
@bear, I'm not sure exactly what you mean by a "shelving" EQ. I thought that basically all EQ's (graphical EQ's, not tone control boards) were a series of band-pass filters. Are you saying that in some EQ's the lowest freq tone control is just a simple low-pass filter? If this is the case, maybe one solution would be to run it through a preamp that has a rumble filter.
Also, I think the curves in the manual are only meant to illustrate how the various "contours" on the EQ affect the sound. As I understand it, these "contours" are simply additional equalization on top of the standard equalization that give the speakers a somewhat flat response.
I thought it might be interesting to try to build the dedicated hardware EQ since the circuit itself didn't seem too complicated. Especially with most of the switches removed. I did make a simplified circuit diagram: http://i1051.photobucket.com/albums/s440/amerloc85/audio-misc/bose_901-I_eq_schem-simplified_zps49ae545b.png~original I was hoping to be able to reverse engineer it and use some op-amps, but I suppose it wouldn't be too hard to replicate as-is. On the other hand, maybe it just wouldn't be worth the effort.
@indianajo, I'm not too invested in these so far. Picked them up for $20 so don't feel too bad about my investment. I could easily sell the stands alone for twice that.
@bear, I'm not sure exactly what you mean by a "shelving" EQ. I thought that basically all EQ's (graphical EQ's, not tone control boards) were a series of band-pass filters. Are you saying that in some EQ's the lowest freq tone control is just a simple low-pass filter? If this is the case, maybe one solution would be to run it through a preamp that has a rumble filter.
Also, I think the curves in the manual are only meant to illustrate how the various "contours" on the EQ affect the sound. As I understand it, these "contours" are simply additional equalization on top of the standard equalization that give the speakers a somewhat flat response.
I thought it might be interesting to try to build the dedicated hardware EQ since the circuit itself didn't seem too complicated. Especially with most of the switches removed. I did make a simplified circuit diagram: http://i1051.photobucket.com/albums/s440/amerloc85/audio-misc/bose_901-I_eq_schem-simplified_zps49ae545b.png~original I was hoping to be able to reverse engineer it and use some op-amps, but I suppose it wouldn't be too hard to replicate as-is. On the other hand, maybe it just wouldn't be worth the effort.
Sure but you still double the power to go up 3dB.
So it works out that 100x the power you go up 20dB.
Here's a chart:
Watt to dBm conversion table
Power (mW) Power (dBm)
0.00001 W -20.0000 dBm
0.0001 W -10.0000 dBm
0.001 W 0.0000 dBm
0.01 W 10.0000 dBm
0.1 W 20.0000 dBm
1 W 30.0000 dBm
10 W 40.0000 dBm
100 W 50.0000 dBm
1000 W 60.0000 dBm
10000 W 70.0000 dBm
100000 W 80.0000 dBm
1000000 W 90.0000 dBm
Power (mW) Power (dBm)
0.00001 W -20.0000 dBm
0.0001 W -10.0000 dBm
0.001 W 0.0000 dBm
0.01 W 10.0000 dBm
0.1 W 20.0000 dBm
1 W 30.0000 dBm
10 W 40.0000 dBm
100 W 50.0000 dBm
1000 W 60.0000 dBm
10000 W 70.0000 dBm
100000 W 80.0000 dBm
1000000 W 90.0000 dBm
The contours show, irrc, the different settings of the switches.
That means EQ is either "normal" or "more".
The curves are the actual EQ applied, not a depiction of how it sounds.
Assuming you followed the "normal" curve (lowest boost) you ought to get reasonably good sound, assuming you have positioned the speakers properly in the room - that means about 10-20" from a flat relatively hard wall, not much between the speakers.
You can duplicate the original box, but that seems foolhardy today since you can do a better job with a DSP based EQ (like a Behringer) than the original.
Even a good parametric equalizer would likely be better than the original.
You could get improved distortion figures using modern opamps in the original EQ circuit.
The 80watts ought to be sufficient as long as you don't try to crank up the volume. They will sound a bit thin unless you have sufficient bass boost.
That means EQ is either "normal" or "more".
The curves are the actual EQ applied, not a depiction of how it sounds.
Assuming you followed the "normal" curve (lowest boost) you ought to get reasonably good sound, assuming you have positioned the speakers properly in the room - that means about 10-20" from a flat relatively hard wall, not much between the speakers.
You can duplicate the original box, but that seems foolhardy today since you can do a better job with a DSP based EQ (like a Behringer) than the original.
Even a good parametric equalizer would likely be better than the original.
You could get improved distortion figures using modern opamps in the original EQ circuit.
The 80watts ought to be sufficient as long as you don't try to crank up the volume. They will sound a bit thin unless you have sufficient bass boost.
I built a hard wired equalizer circuit for my 301s. They are hanging in my shop. The biggest limitation they have in my setup is dynamic range and dynamic compression. They sound decent in my system.
Everybody that hears them can't believe it. People that have 301s always ask why theirs don't sound like mine. I just tell them that it's because I hung them from the ceiling, just like the display in the store.
Bose 301s and 901s do need a decent amplifier. Without that, they will sound thin and anemic. I'm driving mine with an old Nakamichi Stasis amp that I got for free and repaired myself. The Nak can deliver big current. Without good current reserve, the Bose speakers will fall flat on their face if you try to play them loud. They are very inefficient, they have a very ragged frequency response, their impedance varies from 3 ohms to 9 ohms not counting woofer resonance (yes I measured mine), and the 301s really don't handle a lot of power. The reps that do the store setups know all this and they know how to make them sound good in the store. (Not allowing A/B comparison to other more capable speakers helps too.) You can do everything they do in the store at home if you have the knowhow, money, and space. (Not having a female around helps too because females hate big speakers and intrusive installations.)
I thought that my equalizer was simple, but the Bose one linked to is way simpler than mine. That would be easy as pie to build.
Everybody that hears them can't believe it. People that have 301s always ask why theirs don't sound like mine. I just tell them that it's because I hung them from the ceiling, just like the display in the store.
Bose 301s and 901s do need a decent amplifier. Without that, they will sound thin and anemic. I'm driving mine with an old Nakamichi Stasis amp that I got for free and repaired myself. The Nak can deliver big current. Without good current reserve, the Bose speakers will fall flat on their face if you try to play them loud. They are very inefficient, they have a very ragged frequency response, their impedance varies from 3 ohms to 9 ohms not counting woofer resonance (yes I measured mine), and the 301s really don't handle a lot of power. The reps that do the store setups know all this and they know how to make them sound good in the store. (Not allowing A/B comparison to other more capable speakers helps too.) You can do everything they do in the store at home if you have the knowhow, money, and space. (Not having a female around helps too because females hate big speakers and intrusive installations.)
I thought that my equalizer was simple, but the Bose one linked to is way simpler than mine. That would be easy as pie to build.
Okay, ...
I believe I can add something to this discussion.
I've tried a couple of methods of equalizing the 901 (series II):
1. Implementing EQ (and optionally room correction) via a INUKE DSP amp.
2. Using the original factory EQ.
I could get method (1) to sound decent but never as good as (2).
Also, as already mentioned in this thread, .. the 901 (especially the series I and II) need power. I've tried various amps including a modest NAD C720BEE (50WPC), Mcintosh MC2100 (105 WPC) and a INuke 3000DSP (about 400 WPC, I'm guessing) and only the INuke managed to wake them up. I didn't believe that the difference would be that large
I believe I can add something to this discussion.
I've tried a couple of methods of equalizing the 901 (series II):
1. Implementing EQ (and optionally room correction) via a INUKE DSP amp.
2. Using the original factory EQ.
I could get method (1) to sound decent but never as good as (2).
Also, as already mentioned in this thread, .. the 901 (especially the series I and II) need power. I've tried various amps including a modest NAD C720BEE (50WPC), Mcintosh MC2100 (105 WPC) and a INuke 3000DSP (about 400 WPC, I'm guessing) and only the INuke managed to wake them up. I didn't believe that the difference would be that large
The original Bose EQ sounds better than a multi-band equalizer because it has only 2 peaks - near the treble and bass extremes. From the peak at the bass, the slope is smooth down to the midrange and then goes smoothly back up to the treble peak.
On a graphic EQ, each slider introduces a peak or valley in the frequency response which gets nastier as the gain or cut is increased.
On a graphic EQ, each slider introduces a peak or valley in the frequency response which gets nastier as the gain or cut is increased.
901s can sound great if set up properly. If they are placed haphazardly in an acoustic environment, they will not sound so great.
Are they accurate? Certainly not. If you measured their specs, they would look positively lackluster. Even with EQ they do not go over 15 kHz. They roll off sharply below 80 Hz. But set up properly, around two feet from a wall and corners, without furniture or curtains anywhere near them, they give a subjective soundstage that is very pleasing when playing rock (especially live recordings!) or "pop" music. Is it real? Not at all.
They require gobs of power to sound good. At the Bose store (I know guys, I know) the salesmen used to wire them up to a power chord and plug them into a wall socket (120 V). It's super cheesy I know, but the amazing fact is that they took it- the salesman would hook them right back up to the distribution center (amplifier used is apparently a well guarded secret) and they didn't skip a beat.
I made some "knock off" 901s in high school. I used three 8" full ranges; two rear facing and one forward facing. I used an Electro Voice horn tweeter up front (had to pad it way down of course). I had a long, folded vented port facing the rear. Since my tube amp had multiple output taps, I was able to wire the three speakers up to the multiple taps. These babies really rocked! All the drivers, crossovers, and L-pads cost under $100 in the 70s.
Just FYI, Parts Express sells replacement drivers for the 901s (I think they fit other models too). Replacement Speaker Driver for Bose 901 4-1/2" 1 Ohm 290-922 They are 1 ohm drivers, so DIY applications are limited. Us wise guys could build an amplifier to drive them though.
Last edited:
I did not suggest or recommend a graphic EQ.
I would not for this application.
I did say that you could make the equivalent or better EQ using a modern DSP based EQ or perhaps a parametric EQ.
It's important to discern between the different methods.
Assuming you can duplicate the curves there should be nil differences.
Good to see another Bose fan. I've been a fan of the Series I and II for years now. Having had a Behringer DEQ2496 "before" I got into 901's, I have always used it -- or later, a MiniDSP 2x4 -- to do the room EQ + EQ curve at the same time. This works perfectly. I just sold my last Bose (Series II) EQ on Ebay -- I hardly ever used it. Yes, you can use a regular old GEQ too but you'll get best results with 1/3 octave, PEQ and/or some test gear. Look for my other posts here if you want to see how I do it. Unlike 1968, amp power is stupid cheap nowadays, if you can bring yourself to say "Behringer" or similar 
Of course, many are here for the thrill of DIY. So go ahead and build a clone of the active EQ. But you will do a lot better -- and easier and more flexible -- with even a $20 used 10-band octave EQ. Yes, that works ok too.
Of course, you can buy the original thing on Ebay too...
Of course, many are here for the thrill of DIY. So go ahead and build a clone of the active EQ. But you will do a lot better -- and easier and more flexible -- with even a $20 used 10-band octave EQ. Yes, that works ok too.
Of course, you can buy the original thing on Ebay too...
Good to see another Bose fan. I've been a fan of the Series I and II for years now. Having had a Behringer DEQ2496 "before" I got into 901's, I have always used it -- or later, a MiniDSP 2x4 -- to do the room EQ + EQ curve at the same time. This works perfectly. I just sold my last Bose (Series II) EQ on Ebay -- I hardly ever used it. Yes, you can use a regular old GEQ too but you'll get best results with 1/3 octave, PEQ and/or some test gear. Look for my other posts here if you want to see how I do it. Unlike 1968, amp power is stupid cheap nowadays, if you can bring yourself to say "Behringer" or similar
Of course, many are here for the thrill of DIY. So go ahead and build a clone of the active EQ. But you will do a lot better -- and easier and more flexible -- with even a $20 used 10-band octave EQ. Yes, that works ok too.
Of course, you can buy the original thing on Ebay too...
Can you post your minidsp config or screenshots?
Thanks
my tutorial
http://bmmoser.myweb.usf.edu/shared/MiniDSP 2x4 as Bose 901 EQ/Custom EQ with the MiniDSP.doc
This has my unique curves. If you want the default Bose Eq curve I can send it or just search on the net.
http://bmmoser.myweb.usf.edu/shared/MiniDSP 2x4 as Bose 901 EQ/Custom EQ with the MiniDSP.doc
This has my unique curves. If you want the default Bose Eq curve I can send it or just search on the net.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.