Hi everyone
I have these Neumann KH120 studio monitors, I quite like the sound but there are some reliability issues.
A while ago one of the monitors died. The problem lies within the smps.
Before that I had a customer bringing another one with the exact same problem, and on the internet there are several reports of smps failures.
I am not inclined in paying the official repair centre to fix the monitor and then having the same problem again at some point in the future.
I won't get much money for the broken speaker so I am starting to think I could have a play with it removing the whole psu + amp board and converting the speaker to passive.
I am totally new to crossover design so I am wondering wether this project would be worthwhile at all.
Very keen in knowing any opinions from who is more experienced in this subject
Here are some specs for the KH120 active monitor
NEUMANN
I have these Neumann KH120 studio monitors, I quite like the sound but there are some reliability issues.
A while ago one of the monitors died. The problem lies within the smps.
Before that I had a customer bringing another one with the exact same problem, and on the internet there are several reports of smps failures.
I am not inclined in paying the official repair centre to fix the monitor and then having the same problem again at some point in the future.
I won't get much money for the broken speaker so I am starting to think I could have a play with it removing the whole psu + amp board and converting the speaker to passive.
I am totally new to crossover design so I am wondering wether this project would be worthwhile at all.
Very keen in knowing any opinions from who is more experienced in this subject
Here are some specs for the KH120 active monitor
NEUMANN
No, easy answer 
.
Difficult one: Maybe
Without the propper equipment (measuring microfon, tsp measuring possibility) and time and skill this task will not be easy.
Thing is that both drivers have their own crossover and amplification, this to compensate in passive is an interessting task, at least. Not impossible but you have to have knowledgeof what you are doing.
Besides this, an passive X-over is most likely not as accurate as the propper one which is implemented from Neumann.
But if you can live with some limitations in overall flatness and maybe soundquality this task could be done.
I'll give to consider, that most likely also the bass response is adjusted with the active crossover, for the bass driver in this paticiular enclosure and an passive one will most likely not compensate this. Which means you would have to measure the tsp, either way and maybe build a bigger box for the drivers.
But if all this does not bother you, give it a try
The other thing you could do, is ask around the board if someone with knowledge is around your place and maybe able to fix the broken smps for a reasonable price instead of wasting your money at the "original vendor".
My guess is, that like at our town, there are some indepent A/V repair shops where you also could go and aks if they would change the smps, if this is realy the one which makes all the trouble.
Greets Swany
.Difficult one: Maybe
Without the propper equipment (measuring microfon, tsp measuring possibility) and time and skill this task will not be easy.
Thing is that both drivers have their own crossover and amplification, this to compensate in passive is an interessting task, at least. Not impossible but you have to have knowledgeof what you are doing.
Besides this, an passive X-over is most likely not as accurate as the propper one which is implemented from Neumann.
But if you can live with some limitations in overall flatness and maybe soundquality this task could be done.
I'll give to consider, that most likely also the bass response is adjusted with the active crossover, for the bass driver in this paticiular enclosure and an passive one will most likely not compensate this. Which means you would have to measure the tsp, either way and maybe build a bigger box for the drivers.
But if all this does not bother you, give it a try
The other thing you could do, is ask around the board if someone with knowledge is around your place and maybe able to fix the broken smps for a reasonable price instead of wasting your money at the "original vendor".
My guess is, that like at our town, there are some indepent A/V repair shops where you also could go and aks if they would change the smps, if this is realy the one which makes all the trouble.
Greets Swany
Yes, I agree with the above. One could easily bypass the SMPS and tap connections to the power supply of the amplifier directly, so that another functional power supply maybe used in its place. Now, if the amplifier and smps are married to each other in some way (not separable), you could also isolate the amplifier input/outputs, enabling the use of another amplifier (and power supply).
Within any SMPS, the most unreliable parts are the electrolytic capacitors, followed by transistors, diodes, other passives, HF transformer and finally the mechanical parts. PCB reliability depends on the conditions and they can, at times, exhibit secondary damages like breakage, charring etc.
Besides, if this kind of damage is commonly occurring, then a method that fixes the same could also bring in a lot of income as similarly damaged pieces would be plentiful and inexpensive. I hope you're aware of the SMPS sub-forum on Diyaudio.
EDIT: I'm of the opinion that all active settings need not necessarily have an exactly equivalent passive network. However, opinions vary.
Last edited:
These monitors use some steep filter slopes and dsp to make them work the way they do. I dont think you'll get similar results trying to run these passive. Your best bet will be fixing the smps and finding out why they are failing that way. Sometimes just basic component upgrades in the smps (better electrolytic
capacitors) will get them working reliably again.
capacitors) will get them working reliably again.
Thanks everyone for the replies.
The smps is on the same pcb as the amplifier section. It is a multilayer board, with poor / non existing labelling on the components.
It would be much easier if the psu were separate and only 2-layer so could cut tracks if needed - I also thought of building an external linear psu and forget about the smps but didn't seem that easy back then, perhaps I will have another look.
I could give another try at fixing the smps, but the previous one I have worked on it was a bit of a nightmare and I just ended up wasting time and money and sent it to the official repair. There is a flat rate of £240 for this repair and I feel any other repair store will quote around the same price...
The smps is on the same pcb as the amplifier section. It is a multilayer board, with poor / non existing labelling on the components.
It would be much easier if the psu were separate and only 2-layer so could cut tracks if needed - I also thought of building an external linear psu and forget about the smps but didn't seem that easy back then, perhaps I will have another look.
I could give another try at fixing the smps, but the previous one I have worked on it was a bit of a nightmare and I just ended up wasting time and money and sent it to the official repair. There is a flat rate of £240 for this repair and I feel any other repair store will quote around the same price...
Dear reamp,
Even if it's a multilayer PCB, its should still be possible to disable the on-board SMPS and get back to operation (as long as the PCB is not burnt) by a) Removing the rectifier diodes and/or output choke and b) removing the ferrite transformer while choosing not to apply power to it.
Also, even with multilayer PCBs, designers seldom use the inner layers for the SMPS/power sections as the clearance between layers is often not sufficient for the voltages involved. Multilayer PCBs are primarily adopted by designers to make the design more difficult (if not impossible) to copy. I have seen multilayer designs wherein the inner layers were being used entirely as ground/power planes!
So on the whole, it's definitely worth trying.
I think your PCB looks like this. You can try checking the central area with electrolytic capacitors including the big input side filter cap. If there's no choke to be seen, then its a flyback converter and you've to remove the transformer (ETD core) in the middle.
All the best.
Even if it's a multilayer PCB, its should still be possible to disable the on-board SMPS and get back to operation (as long as the PCB is not burnt) by a) Removing the rectifier diodes and/or output choke and b) removing the ferrite transformer while choosing not to apply power to it.
Also, even with multilayer PCBs, designers seldom use the inner layers for the SMPS/power sections as the clearance between layers is often not sufficient for the voltages involved. Multilayer PCBs are primarily adopted by designers to make the design more difficult (if not impossible) to copy. I have seen multilayer designs wherein the inner layers were being used entirely as ground/power planes!
So on the whole, it's definitely worth trying.
I think your PCB looks like this. You can try checking the central area with electrolytic capacitors including the big input side filter cap. If there's no choke to be seen, then its a flyback converter and you've to remove the transformer (ETD core) in the middle.
All the best.
Last edited:
Just had a chat with UK service department (sennheiser UK)
They have a flat rate for repairs and they change the whole pcb.
I will give another shot at repairing the smps and if that doesn't work I can try make an external linear supply. Will open a thread in the smps section to see if I can get some help.
In the worst case if nothing is successful and the pcb ends up being beyond repair I can still send it to the repair centre and they will change the board for the usual flat rate fee.
I would much prefer find a way to use a linear power supply, much more reliable and easy to service in my experience.
They have a flat rate for repairs and they change the whole pcb.
I will give another shot at repairing the smps and if that doesn't work I can try make an external linear supply. Will open a thread in the smps section to see if I can get some help.
In the worst case if nothing is successful and the pcb ends up being beyond repair I can still send it to the repair centre and they will change the board for the usual flat rate fee.
I would much prefer find a way to use a linear power supply, much more reliable and easy to service in my experience.
One idea is just to have some fun and see where it gets you. Try running the bass-mid full range and no tweeter. Sometimes this works better than it should. Next calculate 4 kHz 8 ohms for the tweeter blocking capacitor from Xc=1/2Pifc. This is a safe value. If it sounds too bright reverse phase the tweeter to see if it's doing anything. For me the next bit is easy. Find a choke of lets say 100 uH with thick wire. That goes in series with the bass-mid. Easy for me is I have them. Now is make or brake time. Is it sounding awful or interesting. Radio voice sometimes sounds more awful than music. After that it's wherever it goes and needs imagination. Try slightly different caps after that.
Rega showed me a speaker like this in the late 1970s. Roy twisted the wires for speed of listening and tried one thing a day. These became Linn Kans is my guess based on BBC LS3/5A. Chalk and cheese speakers. Linn chums heard and liked them. Most important is to protect the tweeter with a small capacitor value. Quality of parts at first is not as important as right value. Resistors in series with the tweeter are cheap and fun if too bright an easier than bass chokes to try on the pocket. Don't be surprise if the tweeter is a puzzle. Strictly speaking this is the wrong way to do it. I suspect AR18s were designed this way. I dare say if this gets you somewhere others can help. It's like Chess. Book or instinct. Or both.
BTW. Cheaper Non Polar capacitors can be very OK. Sometimes they even sound best. I would say if the theory i getting ahead of you it's a good time to study. It's then time to pull your hair out. Remember at that point it's not too important you win. As others have said the drivers will likely suit active best. Maybe not?
Forgive typos as battery is very low.
Rega showed me a speaker like this in the late 1970s. Roy twisted the wires for speed of listening and tried one thing a day. These became Linn Kans is my guess based on BBC LS3/5A. Chalk and cheese speakers. Linn chums heard and liked them. Most important is to protect the tweeter with a small capacitor value. Quality of parts at first is not as important as right value. Resistors in series with the tweeter are cheap and fun if too bright an easier than bass chokes to try on the pocket. Don't be surprise if the tweeter is a puzzle. Strictly speaking this is the wrong way to do it. I suspect AR18s were designed this way. I dare say if this gets you somewhere others can help. It's like Chess. Book or instinct. Or both.
BTW. Cheaper Non Polar capacitors can be very OK. Sometimes they even sound best. I would say if the theory i getting ahead of you it's a good time to study. It's then time to pull your hair out. Remember at that point it's not too important you win. As others have said the drivers will likely suit active best. Maybe not?
Forgive typos as battery is very low.
What kind of amp, and expected optimum voltage?
That will give you a pretty good idea of the power supply specifications.
Go for a reputed brand like Meanwell to replace, if you are sure the amp is okay.
If not, data on active crossover circuits is available, even the old ST TDA 2030 data sheets had those. Not a very big job.
And if you really want to hack it, find LED drivers to use as SMPS...
That will give you a pretty good idea of the power supply specifications.
Go for a reputed brand like Meanwell to replace, if you are sure the amp is okay.
If not, data on active crossover circuits is available, even the old ST TDA 2030 data sheets had those. Not a very big job.
And if you really want to hack it, find LED drivers to use as SMPS...
Very true. Unless you are very experienced in cross-over design, running them passive is doomed to fail.
Jan
That section is actually called "Power Supplies" and SMPS is just a convenient term, hope you already knew that.
However, a small problem with LED drivers is that many (but not all) are constant current supplies (vs. constant voltage.), which in turn is because it's the LED current (not voltage) that controls their luminous flux output.
My point exactly, constant voltage units may not respond as fast to load variations as constant current units.
LED drivers from old units do not work on newer lamps as they are bulky, or the new designs are driver on board which means the lamp PCB has a direct mains input. No driver is needed.
So they are being scrapped. Which means available for a song.
I feel they would work, no harm in trying.
LED drivers from old units do not work on newer lamps as they are bulky, or the new designs are driver on board which means the lamp PCB has a direct mains input. No driver is needed.
So they are being scrapped. Which means available for a song.
I feel they would work, no harm in trying.
Naresh,
Consider an amplifier at idle (or low power) powered from a constant current source, and think what would happen to the bus capacitor under such conditions (see picture)? And, I think you already understand that we can't omit the bus capacitor !
Now, let us be a bit optimistic and assume that the constant current LED driver has a voltage limit, in which case the power supply would be swinging in response to the fluctuating load current (or music power). However, the main question is if that's very becoming of a good power supply? Well, there's the PSRR, but does that mean that we should be presenting the amplifier with a fluctuating supply voltage?
Thanks and hope you understand my point.
One thing active allows is very low output impedance (< 0.1 ohms ). This might control the mid band better than placing a choke in series. I said 100 uH for that. Having a few 100 uH would be fun. 50 uH 100 uH and 200 uH using two devices should show a trend. The active impedance is like active energy recovery in cars. It has a feedback loop and can over compensate. Sometimes the choke helps due to it's DCR! Differing amplifiers behave very differently. I recently built an amplifer that should be hopeless for this. The old PW Texan driving a motor with massive extra feedback. It works great. It is point to point build that makes it more difficult in some ways. The motor is a very nasty speaker if you like. This motor has killed a few amplifiers before now. You never know until you think it out. Previously the old Hitachi MOS FET amp worked best. I even have built an active two pole filter into the power amp. That can be fun. Surprisingly that was a sucess! A very good option for sub woofers.
Some sounds in speakers is due to reflected mechanical enegy. An active amp might damp that a little ( better ). The combination of choke and box might be very different even if in theory similar.
This choke is unlikely to be the one you keep. It will show a trend. It is hyper low cost. The DCR is a little high. Often I find if that's all I have I still learn from it. They can be combined to arrive at a better value. DCR is simple ohms law. Inductors are similar. In theory we should make a vector diagram. Ears will say what you like and save you time. Try to keep DCR below 1 ohm even for initial tests.
Series and Parallel Inductors | Inductors | Electronics Textbook
https://cpc.farnell.com/panasonic/elc09d101f/choke-100uh-0-82a-10/dp/FT01265?st=choke
Used in blocks this looks interesting.
https://cpc.farnell.com/panasonic/elc10d561e/choke-560uh-0-560ohm-0-68a/dp/FT01303?st=choke
Expect a big upgrade when buying a bespoke air cored or audio grade choke from a specailist. However if it's nowhere near right with cheap chokes it never will be. Low level tests tell you most. That should be possible with cheaper chokes.
My old Dynaco speakers have no bass choke or bass mid crossover. It's said the bass-mid was engineered to work that way. The tweeter is a nice dome type typical of later speakers ( Audax, Seas, Peerless ). It has a sophisticated tweeter crossover. The speaker is more than usually descriptive of the amplifier used. This is most likely not tonal.
Some sounds in speakers is due to reflected mechanical enegy. An active amp might damp that a little ( better ). The combination of choke and box might be very different even if in theory similar.
This choke is unlikely to be the one you keep. It will show a trend. It is hyper low cost. The DCR is a little high. Often I find if that's all I have I still learn from it. They can be combined to arrive at a better value. DCR is simple ohms law. Inductors are similar. In theory we should make a vector diagram. Ears will say what you like and save you time. Try to keep DCR below 1 ohm even for initial tests.
Series and Parallel Inductors | Inductors | Electronics Textbook
https://cpc.farnell.com/panasonic/elc09d101f/choke-100uh-0-82a-10/dp/FT01265?st=choke
Used in blocks this looks interesting.
https://cpc.farnell.com/panasonic/elc10d561e/choke-560uh-0-560ohm-0-68a/dp/FT01303?st=choke
Expect a big upgrade when buying a bespoke air cored or audio grade choke from a specailist. However if it's nowhere near right with cheap chokes it never will be. Low level tests tell you most. That should be possible with cheaper chokes.
My old Dynaco speakers have no bass choke or bass mid crossover. It's said the bass-mid was engineered to work that way. The tweeter is a nice dome type typical of later speakers ( Audax, Seas, Peerless ). It has a sophisticated tweeter crossover. The speaker is more than usually descriptive of the amplifier used. This is most likely not tonal.