I forgot to mention... I have no hum, at least not with that frequency.
What I measured is about 1,5V with a frequency of 454,545 kHz, what I'm also wondering about.
What I measured is about 1,5V with a frequency of 454,545 kHz, what I'm also wondering about.
Hi guys,
I have a problem also with a hum in my 4 LPUHP. I've buy 2 transformers rated for 150VA 4*15V. When I connect only 1 amp to the transformer the amp is dead silent and is working really nice. As soon as I connect the 2 others secondaries to the second LPUHP I have a hum (50hz I guess)
Anyone has an idea how can I fix this issue?
I have a problem also with a hum in my 4 LPUHP. I've buy 2 transformers rated for 150VA 4*15V. When I connect only 1 amp to the transformer the amp is dead silent and is working really nice. As soon as I connect the 2 others secondaries to the second LPUHP I have a hum (50hz I guess)
Anyone has an idea how can I fix this issue?
Hi multisync,
I'll be building an overclocked/overamped LPUHP soon and was updating the BOM to meet the requirements but :
- I couldn't find 35V caps with 7.5mm space between both connectors, only 10mm
- Could you be more specific regarding the new resistor values ?
All in all, it would be very nice of you if you could upload the modified BOM
Cheers!
As I measured my LPUHP's, the HF is the same on the Input & Output on the Buffers, and also on the Output of the LME49990! They all are exactly the same.
Owen, is it possible, that this issue of still because of the shorting under C104? Or is the opamp just oscillating?
Do others have the same problem?
The power section is OK, I oscilloscope'd them, too.
Owen, is it possible, that this issue of still because of the shorting under C104? Or is the opamp just oscillating?
Do others have the same problem?
The power section is OK, I oscilloscope'd them, too.
according to the data sheet for the regulator. If the Ref voltage is 2.37V and the ref current resistor is set to 2.37k, which it is, then the desired r value is R=V(desired-2.37) or R=22-2.37 or 19.63K. I picked the nearest lower value at 19.6kohm. This gives me 21.9V +/- a tiny bit depending on chip tolerance, temp, tolerance of whatever or the moon cycle. For the capacitors I have a couple of projects so I bought qty 50 of 4700uf at 35v, a bit bigger in size so that all 4 don't fit between the regulators so I bent the leads a bit. Not very pretty but works. I put a bit of silicone to hold the capacitors in place. I am not concerned about the extra inductance a cm or 2 will give.
on another note I bought some expensive Antek transformers and all the 120 hz hum is gone. Any residue hum is now 60 hz and is lost in the noise of the scope. The analogue scope I have can do 1 milli volt/division. And the regulator that I thought wasn't working well is now working. WTF. The transformers weren't too expensive and the shipping wasn't that bad, but what killed me was the $44 for duty and customs paper work provided by UPS. A giant rip off. Duty if any should have been about $3 at most.
on another note I bought some expensive Antek transformers and all the 120 hz hum is gone. Any residue hum is now 60 hz and is lost in the noise of the scope. The analogue scope I have can do 1 milli volt/division. And the regulator that I thought wasn't working well is now working. WTF. The transformers weren't too expensive and the shipping wasn't that bad, but what killed me was the $44 for duty and customs paper work provided by UPS. A giant rip off. Duty if any should have been about $3 at most.
Hey guys,
unfortunatelly nobody is giving constructive answers anymore. What's up? Owen?
As I see, others seem to have some issues with transformers, I'm not getting help for some weeks now. Did the project die?
I have another question... what is the 49990 in the SE-version for? Does it make sense to put the signal directly in, without the 49990?
Regards,
Stammheim
Hi Gentlemen,
Project definitely isn't dead... but my modules are on loan right now, so I have nothing constructive to offer at this point beyond what I have already said.
There are a few issues at play here:
1. I specifically said at the onset that each supply phase required a dedicated winding. If you are currently trying to make some arrangement other than that work, then you need to be prepared to sort it out on your own. I cannot test every conceivable permutation people will come up with, so stray from the BOM and schematics at your own risk.
2. I believe some people have got this working (one transformer, two amps) which means if you don't have it working, you're doing something wrong. This cannot be sorted out with words in a post. You need diagrams, and you need to be concise and methodical in your testing and verification. Just saying "I tested everything and a regulator gets hot" isn't helpful. There are dozens of different connection possibilities, and only one will work correctly. Are you 100% sure you have it right?
3. I posted simulations of this setup a while back which should help you sort it out. You need to follow them very closely. If a regulator is getting hot, it's because it's seeing a short on the output and going into current limiting mode. This short is likely being caused by incorrectly wired AC inputs to the amp which is connecting one of the regulator outputs to GND. This may or may not have damaged the regulator. The only way to tell is to replace the reg and see if the problem goes away.
4. There's a chance that this setup is just not possible without severely degraded performance. The simulation doesn't account for inter-winding capacitance or a plethora of other possible issues that might arise in the real world. What the simulation does tell us is that there doesn't appear to be any short circuit, and that the voltage rails should all be working. There may be severe hum or ground loops depending on the input and output configuration, caused by things outside the scope of the simulation.
Stammheim:
I don't mean to sound short with the above comments, but you have received a great deal of advice and feedback on this question, and short of having someone come over and physically fix it for you, I'm not sure there's much more we can offer. You have a schematic, and independent confirmation that it works. That should be all you need.
Perhaps if you ask nicely, Multisync can draw up exactly how he has his hooked up (between transformer and amp) and you can follow that. Or, alternatively, he can simply confirm if he has his connected as per the simulation I posted a while back.
As for the LME49990, it's what provides the voltage gain and the feedback loop, while the nested buffers simply boost the current output of the op-amp itself.
If you omitted the LME49990, you would get degraded performance (no feedback) and a gain of a little less than 1.
The oscillation you're seeing is also an issue with either a cold joint somewhere, or input and output wiring. I scoped both the SE and BAL input variations built as per the BOM, and there are definitely no oscillations on the output with or without a load.
Some photos of the test setup might help troubleshoot this.
Regards,
Owen
Project definitely isn't dead... but my modules are on loan right now, so I have nothing constructive to offer at this point beyond what I have already said.
There are a few issues at play here:
1. I specifically said at the onset that each supply phase required a dedicated winding. If you are currently trying to make some arrangement other than that work, then you need to be prepared to sort it out on your own. I cannot test every conceivable permutation people will come up with, so stray from the BOM and schematics at your own risk.
2. I believe some people have got this working (one transformer, two amps) which means if you don't have it working, you're doing something wrong. This cannot be sorted out with words in a post. You need diagrams, and you need to be concise and methodical in your testing and verification. Just saying "I tested everything and a regulator gets hot" isn't helpful. There are dozens of different connection possibilities, and only one will work correctly. Are you 100% sure you have it right?
3. I posted simulations of this setup a while back which should help you sort it out. You need to follow them very closely. If a regulator is getting hot, it's because it's seeing a short on the output and going into current limiting mode. This short is likely being caused by incorrectly wired AC inputs to the amp which is connecting one of the regulator outputs to GND. This may or may not have damaged the regulator. The only way to tell is to replace the reg and see if the problem goes away.
4. There's a chance that this setup is just not possible without severely degraded performance. The simulation doesn't account for inter-winding capacitance or a plethora of other possible issues that might arise in the real world. What the simulation does tell us is that there doesn't appear to be any short circuit, and that the voltage rails should all be working. There may be severe hum or ground loops depending on the input and output configuration, caused by things outside the scope of the simulation.
Stammheim:
I don't mean to sound short with the above comments, but you have received a great deal of advice and feedback on this question, and short of having someone come over and physically fix it for you, I'm not sure there's much more we can offer. You have a schematic, and independent confirmation that it works. That should be all you need.
Perhaps if you ask nicely, Multisync can draw up exactly how he has his hooked up (between transformer and amp) and you can follow that. Or, alternatively, he can simply confirm if he has his connected as per the simulation I posted a while back.
As for the LME49990, it's what provides the voltage gain and the feedback loop, while the nested buffers simply boost the current output of the op-amp itself.
If you omitted the LME49990, you would get degraded performance (no feedback) and a gain of a little less than 1.
The oscillation you're seeing is also an issue with either a cold joint somewhere, or input and output wiring. I scoped both the SE and BAL input variations built as per the BOM, and there are definitely no oscillations on the output with or without a load.
Some photos of the test setup might help troubleshoot this.
Regards,
Owen
Last edited:
Hi !
Thanks @multisync for the info 🙂
I've updated the BOM to try to match my requirements which are :
- get an overclocked LPUHP
- get the right gain regarding 4VRMS in from IVY III I/V
In red text are the parts that are different from the original BOM and I added a "Comment" column to give more details about why this component have been changed. Digikey BOM is available here as well : http://www.digikey.fr/short/h5fpn
As I am not very used to make this type of calcultation, I would be glad if someone here could help me to confirme everything is OK 🙂
Thanks!
Thanks @multisync for the info 🙂
I've updated the BOM to try to match my requirements which are :
- get an overclocked LPUHP
- get the right gain regarding 4VRMS in from IVY III I/V
In red text are the parts that are different from the original BOM and I added a "Comment" column to give more details about why this component have been changed. Digikey BOM is available here as well : http://www.digikey.fr/short/h5fpn
As I am not very used to make this type of calcultation, I would be glad if someone here could help me to confirme everything is OK 🙂
Thanks!
the bypass caps for the buffers don't have to be changed as the old ones were 25v I think. After the regulators the voltage should be 22. So 25 is fine. The .047 cap can be 25V as well, only 19 volt across it.
Hello back,
Owen, many thanks for your response!
Yes, of course I know that your layout is approved and I'm also sure that the amp is doing fine for most people. On the other hand I wonder why I have such problems 🙁
Of course I tried to help myself first before asking other people, but I can't go on now.
Today I did some tests comparing the board to the layout, seems ok. I then replaced the LME49990 trough a LME49710. I just fell of my chair - everything is alright, no HF. I cannot say if BOTH LME49990 are defective, but that seems impossible. I'll do some listening tests tomorrow, maybe the LME49710 will stay at his place.
Funny problems 😉
Best regards,
Stammheim
Owen, many thanks for your response!
Yes, of course I know that your layout is approved and I'm also sure that the amp is doing fine for most people. On the other hand I wonder why I have such problems 🙁
Of course I tried to help myself first before asking other people, but I can't go on now.
Today I did some tests comparing the board to the layout, seems ok. I then replaced the LME49990 trough a LME49710. I just fell of my chair - everything is alright, no HF. I cannot say if BOTH LME49990 are defective, but that seems impossible. I'll do some listening tests tomorrow, maybe the LME49710 will stay at his place.
Funny problems 😉
Best regards,
Stammheim
Thank you so so much multisync for the advice it fixed my hum/buzz in my 4 LPUHP, you saved my life 🙂
I just soldered a 2.5mm² speaker cable between each GND speaker output of the 2 LPUHP that are on the same transformer and the noise is completely gone.
Replacing the LME49990 trough a LME49710 was successful, so I replaced the LME49990 trough a new LME49990. It seems, that both "old" LME49990 were defective!!! I cannot belive it, but it's true! These guys were responsible for the HF, which it not there anymore.
I connected both amps to one transformer and connecting both GND was successful, although the osci showed the following picture (2 Ch on output, one goes up, one down).
[img=http://abload.de/thumb/2014-03-2614.37.13bqkgy.jpg]
Further questions:
Is it a good idea to connect not only the GND, but also the rails of both amps to synchronize all voltages?
Id I'd like to use only for PS for testing, can I connect one PS-out of one amp to the other, without desoldering the regulators? Or will the regulators be destroyed?
I connected both amps to one transformer and connecting both GND was successful, although the osci showed the following picture (2 Ch on output, one goes up, one down).
[img=http://abload.de/thumb/2014-03-2614.37.13bqkgy.jpg]
Further questions:
Is it a good idea to connect not only the GND, but also the rails of both amps to synchronize all voltages?
Id I'd like to use only for PS for testing, can I connect one PS-out of one amp to the other, without desoldering the regulators? Or will the regulators be destroyed?
Last edited:
don't connect the regulators in parallel. There is a tiny difference in voltage between the regulators. Each regulator will try to correct the difference in the opposite direction. The regulator that is a bit higher will try to pull up the one that is low and the one that is low will try to pull down the one that is high. Normally when putting regulators in parallel there is a small series resistor, then the paralle connection.
PandazYa, glad to help. That is why we are all here to help each other. Before there was internet I used to have dreams about the problems I was trying to solve. The amazing thing was I would sometimes dream a solution but upon waking up I would forget what it was.
Many thanks for your answer!
Yes, that's what I assumed.
And what about a separate PS for both amps? Do I have to remove the regulators as they will get damaged, if I plug in the separate PS after the regulators?
You are absolutely right. I think many of us would never dare to that complex hobby. However, I hope that the thing called Internet will win through one day....
😉
Yes, that's what I assumed.
And what about a separate PS for both amps? Do I have to remove the regulators as they will get damaged, if I plug in the separate PS after the regulators?
You are absolutely right. I think many of us would never dare to that complex hobby. However, I hope that the thing called Internet will win through one day....
😉
LHUP combined wih BAL-BAL?
I just noticed that my LHUP board is a newly revised version with balanced input.
Since I have one BAL-BAL and two LHUP boards, technically I can build this setup:
- One box BAL-BAL that works as differential pre-amp ( balanced input & balanced output) and also work as phone amp
- Two LHUP monoblocks with differential input.
I assume there should be no issue in impendance. Maybe the gain on pre-amp side should be reduced for this setup. the two monoblocks don't need any attenuator anymore.
Do I miss something here? thanks.
I just noticed that my LHUP board is a newly revised version with balanced input.
Since I have one BAL-BAL and two LHUP boards, technically I can build this setup:
- One box BAL-BAL that works as differential pre-amp ( balanced input & balanced output) and also work as phone amp
- Two LHUP monoblocks with differential input.
I assume there should be no issue in impendance. Maybe the gain on pre-amp side should be reduced for this setup. the two monoblocks don't need any attenuator anymore.
Do I miss something here? thanks.
Last edited:
The pre-amp BAL-BAL really should be built with no gain in most cases.
Both versions of the LPUHP PCB can be built for BAL input or SE input - refer different tabs of the BOM here - https://docs.google.com/spreadsheet/ccc?key=0AnMLwB9mdXl2dDVuTjh6WnZmQ1VxYm1FVW9zRzRaMUE&usp=sharing - you can adjust gain and input type to suit your system.
That's cool.
Recently I bought Astell&Kern Ak100 Gen II. It has differential output.
I think it should be interesting to try.
Recently I bought Astell&Kern Ak100 Gen II. It has differential output.
I think it should be interesting to try.
Just wondering if there is any LPUHP pcb left. I would like to buy one. And who made the last run of PCB?
Thanks
SB
Thanks
SB