I finished a Le Monstre a while back. Love the sonics, but it has a low level grundge. Playing around with current levels shows the problem is power supply related, higher currents increase the noise.
The power supply was changed to reduce this, but the effect was minimal. Tried shielding the power transformer and chokes with mu-metal and found it did not help either.
The current power supply is way over built. The filter is a C-L-C-L-C. The values are 22k ufd, 390 uH, 47k ufd, 2.7 mH, 94k ufd per rail. This should give extremely low noise.
The noise is worse on power up, and is much lower after a couple of hours of playing. The speakers are 100 dB efficient, and go to 40 Hz. It is not audible at listening distance, but easily heard at 3 feet.
I have read of people bypassing some of the resistors in the driver stage. But I am not really sure which ones, (2K to ground) and do not know if this is my problem.
I am sure it is not a ground issue, the level always drops after playing for a while. And the bias current would not make a big difference like it does. The grounds are all floating ala passlabs circuitry. This does really help.
George
The power supply was changed to reduce this, but the effect was minimal. Tried shielding the power transformer and chokes with mu-metal and found it did not help either.
The current power supply is way over built. The filter is a C-L-C-L-C. The values are 22k ufd, 390 uH, 47k ufd, 2.7 mH, 94k ufd per rail. This should give extremely low noise.
The noise is worse on power up, and is much lower after a couple of hours of playing. The speakers are 100 dB efficient, and go to 40 Hz. It is not audible at listening distance, but easily heard at 3 feet.
I have read of people bypassing some of the resistors in the driver stage. But I am not really sure which ones, (2K to ground) and do not know if this is my problem.
I am sure it is not a ground issue, the level always drops after playing for a while. And the bias current would not make a big difference like it does. The grounds are all floating ala passlabs circuitry. This does really help.
George
analog_sa said:
It may very well be a ground issue as a higher current will also increase the current through a ground loop producing a larger 'error' voltage drop.
Mysterious
Have you considered oscillation as well?
The ground was where I started. Increased the value of the resitors used to tie circuit board groung to electrical. Going from 0 ohms to 10 ohms to 30 ohms made no difference. And I could not measure any voltage across the reistor. Looks to have almost no current flow.
One item not explained. The circuit boards are from Kristijan. Used the "newer" transistors suggested on the website.
The dc offset was around 4 volts when first powered up. The balancing pot has it nulled out, but this required a lot of shift from center. The rails have shifted so there is about 300 mv difference between the +/- rails with respect to ground. The total offset drifts around a lot when powered up cold. The hum is highest when the offset is largest.
There may be some oscillation going on, but nothing gets warm. And the sound is excellent. Just more low level hum than I like.
I really thought beefing the power supply would quieten it down. Then installing the mu-metal. But neither seemed to help much.
George
Hi,
I'm building a monster too. I have to put it in a nice case.
I've some problems when building it. First, it oscillated at high frequency. I solve the problem by adding a 56pF cap between base and collector of the npn driver stage.
the schematic I used is there:
http://belinfabien.free.fr/amplificateur/le monstre/schema/schema_ampli.png
second, the ground is very important to avoid hum and noise.
I used 4mm² to ground all. The caps are as near as possible of the amplification cards. Each channel has his own ground, between the two filtering caps. They are tied together by a 4mm² wire. I have only one ground connection with the preamp, and it is 4mm² too (no ground in signal wires).
i hope this will help you.
I'm building a monster too. I have to put it in a nice case.
I've some problems when building it. First, it oscillated at high frequency. I solve the problem by adding a 56pF cap between base and collector of the npn driver stage.
the schematic I used is there:
http://belinfabien.free.fr/amplificateur/le monstre/schema/schema_ampli.png
second, the ground is very important to avoid hum and noise.
I used 4mm² to ground all. The caps are as near as possible of the amplification cards. Each channel has his own ground, between the two filtering caps. They are tied together by a 4mm² wire. I have only one ground connection with the preamp, and it is 4mm² too (no ground in signal wires).
i hope this will help you.
Panelhead said:
Found it! The caps used are all four pole Jensen. Rereading the tech help on the website was no help. But went over to the DNM website and there is much more information listed. They call thier version T network.
They show using a resistor between the diodes and the first cap to damp ringing between the caps and transformer. The website it does not degrade the sonics.
Used a pair of 0.2 ohm Dale LVR-5 current sensing resistors. These are trick, non-magnetic wire wounds there are designed for current sensing.
The hum dropped a bunch. Now my ear has to be six inches to hear any noise from the speaker. It sounds like the the dirty grit before, but many dB's lower level.
I can easliy live with this. On to my next project.
George
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.