19V should be fine for test. As it looks it's not loops issue. So power supply is your next suspect.
Hm, it’s strange … 27.5V AC at the transformer input (between each side and center tap) which seems fine, but around 36.5V DC at the output, not 24V. Damaged LM317 perhaps?
Last edited:
I just replaced the LM317. It was defective indeed, it had a short. Now I am getting 24.6V at the output and the hum is gone completely
Thanks so much for your help!
I notice that it is sold in package of 4gram.
Any idea of how much i need? Have never worked with this before - so I am blanc!
Kind regards,
Thomas
And thanks for this! I wired it exactly like this now. I disconnected the signal ground from the NAC42 chassis and instead I connected the shield of the 4-pin cable to its metal housings. So now both chassis are connected only by this cable. Without the cable there is no connection between signal ground and chassis/mains earth at all.
The only thing that’s still a bit annoying now is a short loud and deep noise through the speakers when I turn the NAP-200 on.
This is the normal "turn-on" thump from a power amplifier when there is no automatic delay to speaker connection, as is now normal because a speaker relay and protection system is typically fitted to commercial amplifiers. Many older amplifiers have this same minor fault - sometimes the thump occurs when the amplifier is turned off instead but it is seldom a serious problem. It can sometimes be reduced in severity and sometimes it is due to unwise mods but I've seldom needed to do anything about it.
It's plenty for 4 transistors and mica washers. Only a very thin film of it is needed; any more will be squeezed out when the transistors are screwed down.
Whether it is enough for this and also between a metal bar and the chassis I'm not sure; depends on the dimensions. JimSalabim might be able to tell you how much he used.
The amount also depends on how smooth your metalwork is. The smoother and flatter the surfaces the less grease is required.
Last edited:
As Ian says you may have to put up with that. The original Naim's do this. How much of it is from the NAC-42 powering up? If you disconnect the NAC is the thump as loud?
Ah, I see, thanks! Not really a big problem here as well, because I use a switch to connect the speakers either to the NAP-200 or to my 5.1 system. I can just switch it to the other side while turning the NAP on.
4 gram is enough to do the job several times over if you don't squish the stuff everywhere. Be careful with the type though, some high performance grades are conductive because of a high silver powder content.
Conductive paste is fine with plastic bodied CPUs but you get nasty crackles and shorts where the paste bridges between the collector and heatsink if you are careless with application.
Again, only use the paste with MICA washers. It is useless with silicone rubber washers
Last edited:
I have bought 8g of grease and more than half of it is still left. At last I removed it again from the transistors, because I had no mica washers, only rubber/silicon washers. But I used it between the metal bar and the chassis. 4g should be enough for both the metal bar and the transistors.
I would avoid the Arctic stuff. It is not suitable for high power use e.g. it will pump out quite soon when used with IGBT modules etc. So why risk it with medium power?
The thermal conductivity of the paste is far less important the its ability to fill the gaps and stay there. Metal to metal contact has the lowest thermal resistance and ideally the paste is just there to fill the micro-gaps.
Stick with something proven like DC340 IMO.
The thermal conductivity of the paste is far less important the its ability to fill the gaps and stay there. Metal to metal contact has the lowest thermal resistance and ideally the paste is just there to fill the micro-gaps.
Stick with something proven like DC340 IMO.
The specialized thermal compounds used by PC gamers are very much proven and are used for high performance CPUs and GPUs which have to dissipate considerably higher powers than a Naim clone is ever likely to.
Here's my old tube of Dow Corning 340. This is what I used to use for power transistors before Arctic MX-4. I still use the 340 for large areas like between heat sink fin blocks and chassis because these areas are less important and the Arctic is 10x more expensive. I also considered the Noctua NT-H1 which I use in my PC but it was 50% more expensive again and I didn't think it was worth it.
Attachments
Last edited: