Bose 901 Modifications

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Serie IV EQ

Well, I'm glad for you.

EQ S-IV
Today I did this simple change of OpAmps.
No doubt, the signature of the sound is different with 5532.
As expected, more middles. But good clean mid range. Especially, the strings emerge, and sound pleasant. The voices too. Much more details. Somebody feel lack of low end with this OpAmp. Maybe it's true, but I'll remind you, the S4 EQ is with the biggest low and hi gain, of all other EQs. So, things are OK for me.
My opinion is, the mid-range is one of the strong points of 901s. And by this way, with NE5532 we underline it. For that, who still fell a lack of bass, just push the ton-control 1 degree up.
Comparing.
I own 2 EQ S4, one is not moded. And make quick changes with connections.
Even, to be sure, few times I replaced 5532 in the moded with sockets EQ, with the original TL072. But then, both sound equal.
So, everyone may test this simple, few cents mod, and, if want, to share his own impressions.

The other idea, to null the offset and avoid decoupling caps, for now is hard to do, because of lack of space around ICs. I want to make 8-pins adapters, for 2 single OpAmps.
 
S-I EQ

The last I did for now, is changing the 8 ceramic capacitors in the supply lines of ICs, with WIMA MKS. Just a theoretical improvement, nothing noticeable.

But let me leave a side S4 EQ, for a moment.
The problem is, the system S1 EQ/S2 speakers, become to sound worst, comparing to S4EQ/ S3 speakers. The voices are not so clear, miss some details. I'm 90% confident, the reason is not in drivers.

Well, please help me to analyze the scheme of S1 EQ. And try to find together, the reason of that sound.
My few points of supposed reasons are:
1. The topology of scheme, and feedbacks?
2. The type of resistors - carbon ?
3. The unchanged film capacitors in RCR and CRC groups?
4. The type, brand, of transistors used. May they exert such influence on sound quality?
So, what are your suppositions and advices? How to begin?
 

Attachments

  • 901-I-Second.png
    901-I-Second.png
    366.4 KB · Views: 371
Last edited:
S-1 EQ

First I did was to change all transistors with BC550, known as good audio.
This did clear the sound one idea.
Other, I replaced one of the existing ceramic capacitors, with WIMA MKS. For the other, 10pF, I have no substitution.
Also discovered, the first input capacitor is critical for the sound quality. I was replaced it with WIMA MKS, but the sound was dark. After few changes, I found something what I liked.
For all other film capacitors, I may tell, if we have no proved good audio capacitors, better keep the originals.
The sound now is again very close to the moded S-4. But is a little softer, smoother. It is not bad at all.
Why is it so? May be because of carbon resistors. I know some opinions, carbons sound soft, while metal-films are more sharp.

This is for now.
Enjoy the music!:violin:
 

Attachments

  • P1010487.JPG
    P1010487.JPG
    363 KB · Views: 334
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Has anyone tried skipping the Bose EQ units and gone straight to a miniDSP and then use a mic with REW to get their own custom EQ curve? It seems that would be the ideal way to go and is probably a lot less expensive (and easier and more flexible) than upgrading caps and transistors in the analog EQ box.

I heard that SL finally built an Orion that doesn't use his analog EQ boxes but now uses miniDSP and he likes them.
 
xkr971: Yes, I've been doing exactly that for years (Behringer DEQ2496 and/or MiniDSP 2x4) to EQ room + Bose 901 at same time. I think I even tried to dissuade our soldering-iron wielding friend earlier in this thread :) I guess some people are old-school (Herr Linkwitz) and would rather build than hack. God bless them, if it weren't for them there would be no hardware for us software people to wreak our unseen havoc upon :darkside:
 
Hi Soldermizer,
Yesterday, before I read Your post, I did the same.
My configuration was: PC->USB/SPDIF-> DEQ2496/SPDIF-> DAC TDA1541(ECC83 tube out)->AMP
My classic configuration is: PC->USB/SPDIF->DAC TDA1541(ECC83 tube out)->BOSE EQ->AMP
So, what to say you? Sound is very clear, but...
There is some sharpness, some lack of dynamic. The voices are some like synthetic.
There can't be comparison with EQ S4, at all. The word for it is just - presence.
Well, I am aware, that this can be, because of coloration of the sound with BOSE, some harmonics and etc. But there is more life in it.
Enough digits in the sound with the digital sources. May be, future digital processes add
errors. Jitter is also big enemy. Don't know.
The advantage of this test, for me, is, that make me to think about future cleaning the sound of S-1 EQ.
By the way, can you tell me? Some late series of 802 call their EQs sound processors. Are they real DSP, or simply analog EQ and X-over?
 
ctef: You may use any EQ you like with your Bose 901s. If you have the time, perhaps you can try a blind A-B test between a DSP and the Bose Active EQ. To be fair, you would need to have the DSP match the Bose curve as closely as possible. I never tried that, because I like the flexibility the DSP allows. With modern test equipment, it should be possible to conduct very precise tests of a Bose EQ and see if there is any thing special going on: perhaps phase changes that do not show up in a frequency test?
And no, I don't know anything about the series 802 EQ.
 
Hi,
I did the curve, by the graphic from their service manual. As close as can do. Not by ears.
Also with DEQ2496 we can test this analog EQs, electrically and see the curve, not only their sound with the micro.
But question is not just in equalization. A little more or less.
I speak about the signature of the sound we hear.
Probably there are phase changes, indeed.

P.s. By the way, I use my DEQ mainly for adjusting loudspeakers and their X-overs.
 
Last edited:
EQ S1

So, I did and the last step for now.
Spent a few $ and hours, and changed all carbon resistors with new metal-films. In 2 steps. First only one channel, for comparing, next and the other.
The expecting difference is fact.
Sound become clean, sharp, more dynamic.
Yes, the carbon capacitors sounded soft, but for me this softness = dullness.
Now this cleanness is comparable to cleanness of S4 EQ.
And something else. The bass changed. Become more fundamental and deeper. Before under 40Hz was empty, now frequencies about 30Hz are noticeable. With analyzer, too.

In conclusion, what to say?
If some of you is going to renew his old EQ S1-2, the first things to change are capacitors. And next the carbon resistors. And you'll find your 901s in new light.
May be, someone'll feel the new sound a little sharp. But for me it isn't problem. I simply turned 2 points backward the ton-control knob.
If I can use figurative language, it is same, as to look out of well-washed widow.:eek:
 

Attachments

  • P1070411.JPG
    P1070411.JPG
    501 KB · Views: 386
But, do I like what I see? ;)
Well, now a little measurements.
Here you can see the real-life diagrams.
What is my comment? Obviously the S1 is more noisy.
My big regret is, I didn't measure the S1 before the changes, and after every step.
For EQ S4, you can see 2 lines. The yellow is EQ in original shape. The green is with NE5532 and refreshed capacitors.
 

Attachments

  • EQ S1.JPG
    EQ S1.JPG
    86.9 KB · Views: 399
  • EQ S4.JPG
    EQ S4.JPG
    87.9 KB · Views: 291
Is the noise audible with the EQ in the circuit? I never had a problem with noise using Bose EQ, but yes in some installations with a GEQ or PEQ. Generally I think this is a gain staging problem (noise added by preamp or a cheap $20 GEQ bought off Ebay.) The problem is if you need 30-40 dB of gain that low noise floor gets elevated too :(
 
Could one just use a miniDSP to do the EQ and skip the Bose analog EQ completely? Are the EQ curves well-known?.
I just now show the real frequency curves of the EQs. They are very close to the curves in the service manuals. You can work with them.

Is the noise audible with the EQ in the circuit?

No, there is no hearing noise. The main noise level is under 100dB. Eh, well, if I fully open the volume of my 150W amp, without no records, I hear some noises.
A little problem I see with S1 50 Hz. Must work on it. You see, moded S4 is with better result at 50Hz.
And again, the most important. The first 4 series are very sensitive to level of the input signal. Increasing it over 0.5V, and noise and distortions increase dramatically.
 
Phase response

And now, someone was asking about the Phase response of both EQs.
I measured it twice, one time with Spectralab, and other with Steps(Arta) software. Results was more or less the same.
Here you are.
The left is S1, the right is S4.
 

Attachments

  • Bose Phases.JPG
    Bose Phases.JPG
    91.3 KB · Views: 371
Founder of XSA-Labs
Joined 2012
Paid Member
Zobsky,

I think if we implement the EQ via miniDSP, once has to be careful to use PEQ's very judiciously to give a linear phase response between 500 Hz and 5 kHz where all the phase important spatial imaging info is important. It looks like the Bose boxes maintain semi linear phase in the 600 Hz to 10kHz range:

424183d1403024747-bose-901-modifications-bose-phases.jpg


So basically try to have very low Q (broad) PEQ's, cut peaks where possible, do not fill valleys, and touch 500 Hz to 5kHz as little as possible. I think a +gain low-shelf PEQ at 100 Hz and +gain high-shelf at 6 kHz might work. Try to do it with no more than 3 PEQ's to get this shape:

424087d1402941816-bose-901-modifications-eq-s1.jpg
 
Hi,
I'll give it another shot after I'm done with my K-tube experiments. Broadly speaking, my initial tinkerings with the minidsp / 901 were along the lines you describe, but the EQing never quite seemed enough in the bass region (compared to the original EQ box which had a rather drastic mid-cut).
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.