Member
Joined 2003
If there is no electrical problem with your amp (ie. oscillations), I would say that there is way too much thermal paste between the chip and the heatsink. Too much goo will actually decrease the thermal conductivity between the chip and the heatsink (because the goo is not as thermally conductive as metal), and could possibly be the reason behind the heating. Is it the chip you are measuring as being hotter, or is the heatsink itself hotter? It looks like the chips are nice a flush against the heatsink, but if there is even the smallest gap between the chip and sink there will be very little thermal conductivity between the two.motzu' said:
Are the resistors in the feedback path high tolerance 0.1% resistors as stated in the schematic? If not, there is potential for oscillations causing heat as well.
Hi All,
Carlos has some very interesting experiences with snubbers no doubt and others as well I am sure. But it is not an accident they are not included in the circuit.
This is not conjecture, but design. This is not promises of "tighter, clearer, better" it is execution.
But for now please try the snubber if your curiosity has you, just don't expect the promised improvements, or you will be sure to hear them.
Adding the snubber will not hurt (well not much anyway), so no harm will come to try it, but if you wish to make comparisons make sure you review/measure the modified amp directly against an unmodified amp before you subscribe to subjective impressions and grandiose promises.
The reason is, it is easy to fool yourself with euphonics simply because you expect a change to occur. It is pure psychology.
Just remember this is very much an engineered amp, and it has been executed in a precise and deliberate way. Remeber that adding or changing components without a thorough understanding of all of the principles involved as well as a thorough undstanding of the circuit is something that is fun, but not necessarily better.
Anyway I will not denegrate those who try, God knows I did. With every value Carlos has suggested, with side by side comparisons. In every case the amp has always performed well, but the snubbly version certainly did not perform better in any way that I could hear, nor anyone else who listened.
If the snubbers really had any positive effect in this circuit, they would be on the PCB, at least mine. The absence is not because they were not tried, quite the opposite.
Cheers!
Russ
Carlos has some very interesting experiences with snubbers no doubt and others as well I am sure. But it is not an accident they are not included in the circuit.
This is not conjecture, but design. This is not promises of "tighter, clearer, better" it is execution.But for now please try the snubber if your curiosity has you, just don't expect the promised improvements, or you will be sure to hear them.
Adding the snubber will not hurt (well not much anyway), so no harm will come to try it, but if you wish to make comparisons make sure you review/measure the modified amp directly against an unmodified amp before you subscribe to subjective impressions and grandiose promises.
The reason is, it is easy to fool yourself with euphonics simply because you expect a change to occur. It is pure psychology.
Just remember this is very much an engineered amp, and it has been executed in a precise and deliberate way. Remeber that adding or changing components without a thorough understanding of all of the principles involved as well as a thorough undstanding of the circuit is something that is fun, but not necessarily better.
Anyway I will not denegrate those who try, God knows I did. With every value Carlos has suggested, with side by side comparisons. In every case the amp has always performed well, but the snubbly version certainly did not perform better in any way that I could hear, nor anyone else who listened.
If the snubbers really had any positive effect in this circuit, they would be on the PCB, at least mine. The absence is not because they were not tried, quite the opposite.
Cheers!
Russ
Carlos - I'll make a try with snubbers. About the psicho effect we shall see. Finnaly we are DIY'ers so there is room for experience, any time
All the components were selected as in Mauro specs. Yes the resistors in feedback are 0.1% . In fact they were the only resistors brought very late because i need to place a special order for them.
About the thermal paste , the pictures were token in a earlier stage of development, when the contact between the chips and radiator was imperfect.
In that moment the chips were getting hot but the radiator , not.
Now I've rectified with sand paper the PCB border and in this way I've moved closer to the radiators. The interface between the chips and the radiator is one piece ox 2x3cmx1mm cooper with a tiny layer of Titan thermal grease/paste on both sides.
Still the left chip is getting warmer than the right one. Unfortunattely i don't have an oscilloscope, but in the audible range there is no problem or hum
Dcibel - the radiator is warm as well as the chip.
What about the bass ? Should I move to REv_C ? Still I don't get it why the amp is - pretty - warm when there is no signal.
Thank you everyone for the answers
All the components were selected as in Mauro specs. Yes the resistors in feedback are 0.1% . In fact they were the only resistors brought very late because i need to place a special order for them.
About the thermal paste , the pictures were token in a earlier stage of development, when the contact between the chips and radiator was imperfect.
In that moment the chips were getting hot but the radiator , not.
Now I've rectified with sand paper the PCB border and in this way I've moved closer to the radiators. The interface between the chips and the radiator is one piece ox 2x3cmx1mm cooper with a tiny layer of Titan thermal grease/paste on both sides.
Still the left chip is getting warmer than the right one. Unfortunattely i don't have an oscilloscope, but in the audible range there is no problem or hum
Dcibel - the radiator is warm as well as the chip.
What about the bass ? Should I move to REv_C ? Still I don't get it why the amp is - pretty - warm when there is no signal.
Thank you everyone for the answers
"My_Ref" is NOT a Gainclone !!!
For the sake of correctness, once again: Mauro's "My_Ref" design is NOT a Gainclone !!!
It may be "categorized" as a chip amplifier (in that it make use of a chip amp), but definitely it has NOTHING to do with a Gainclone other than using (also) the LM3886!
So, please, please, STOP referring to it as a gainclone!
motzu' said:
For the sake of correctness, once again: Mauro's "My_Ref" design is NOT a Gainclone !!!
It may be "categorized" as a chip amplifier (in that it make use of a chip amp), but definitely it has NOTHING to do with a Gainclone other than using (also) the LM3886!
So, please, please, STOP referring to it as a gainclone!
motzu-
I would try a few things to investigate. Try disconnecting your input and output wires from the board, then power the amp up. Does it still get hot the same way? (This is to rule out possible wiring issues.)
Also, it's worth the effort to reexamine the bottom of your circuit board very closely, looking for small solder bridges between joints. They are very easy to miss (I do).
Assuming neither of these bears fruit, you could try swapping the LM318s between the channels to see if the issue follows them. If so, replace the "hot" 318 (I know it's not the 318 getting hot, but one could be "bad" and cause the problems).
Keep us posted, and good luck!
I would try a few things to investigate. Try disconnecting your input and output wires from the board, then power the amp up. Does it still get hot the same way? (This is to rule out possible wiring issues.)
Also, it's worth the effort to reexamine the bottom of your circuit board very closely, looking for small solder bridges between joints. They are very easy to miss (I do).
Assuming neither of these bears fruit, you could try swapping the LM318s between the channels to see if the issue follows them. If so, replace the "hot" 318 (I know it's not the 318 getting hot, but one could be "bad" and cause the problems).
Keep us posted, and good luck!
hi motzu',
1. all the connections, input RCA, and Power output have to be isolated by the chassis. Am seems me that you have the power GND connected to the chassis...
2. Check the RCA connections, I have seen a red connected to the GND...
3. Check all the resistors, ( R9 and R32 =47Kohm 01 % ) R5 and R28 seem me from 4k7 but they have to be 22Kohm, checks and measures all the values...
4. The heatsink has to be ALWAYS connect OR to chassis (main gnd) OR to trafo GND.
Your problem is a caused oscillation by LM3386, because you have not put the correct resistors. Uses the "original" My_rev revA circuit to check all...
FOR ALL:
My_ref it's a mid-difficulty electronic project, and requires an adeguate ability
BEFORE turn on the circuit, verify all the components and the connections are correct
ciao
Mauro
1. all the connections, input RCA, and Power output have to be isolated by the chassis. Am seems me that you have the power GND connected to the chassis...
2. Check the RCA connections, I have seen a red connected to the GND...
3. Check all the resistors, ( R9 and R32 =47Kohm 01 % ) R5 and R28 seem me from 4k7 but they have to be 22Kohm, checks and measures all the values...
4. The heatsink has to be ALWAYS connect OR to chassis (main gnd) OR to trafo GND.
Your problem is a caused oscillation by LM3386, because you have not put the correct resistors. Uses the "original" My_rev revA circuit to check all...
FOR ALL:
My_ref it's a mid-difficulty electronic project, and requires an adeguate ability
BEFORE turn on the circuit, verify all the components and the connections are correct
ciao
Mauro
Agreed, no GC I have heard sounds like this.
I will try the zener voltage mod and the snubber mod ONLY because both are relatively minor additions / changes that are reversible.
Last night I played the Audio Quest recording of "The Mighty Sam McClain(?)".
My wife listened with me. She quit doing that 10 yrs ago. I asked why she was "here", and she said she was listening. I discreetly said "WHAT!". And she replied "this sounds nice".
Women have WAY better hearing than men, and for her to say it sounds nice, over rides ANYTHING I may have said.
I will try the zener voltage mod and the snubber mod ONLY because both are relatively minor additions / changes that are reversible.
Last night I played the Audio Quest recording of "The Mighty Sam McClain(?)".
My wife listened with me. She quit doing that 10 yrs ago. I asked why she was "here", and she said she was listening. I discreetly said "WHAT!". And she replied "this sounds nice".
Women have WAY better hearing than men, and for her to say it sounds nice, over rides ANYTHING I may have said.
Pictures of MyRef Rev. A
http://tommytube.com.moranplanet.com/MyRef/
I've finally got my amp up and running and placed in my big rig. It's about 95% completed. You can see some pix at the above URL. It's working well and doesn't have any hum at all. Rev. A components.
My issue from a week again was a dodgy solder joint which was causing the left channel to heat up and oscillate. Finally found that and it's working fine now.
My last bit of problem was my power switch had a broken pin so I'm awaiting a new one. That's why the power is wire nutted. I also haven't punched a hole for a fuse yet...which I probably will add in so it's externally accessible.
I pretty much used the values stated on the BOM, but choose different parts since that's something I really enjoy doing. I used 15,000uF electrolytics on the power side. These are bigger than the normal 10,000 cans. To make them fit all I had to do was rotate the outside cap 180 degrees and run a jumper from the inner node to the outside pin on the cap. The advantage of this is it leaves room to run a screwdriver in to tighten/loosen the screws on the LM3886/heatsink.
The heatsink was the hardest part of the kit. Getting it tapped and placed was a challenge. I originally started using a smaller heatsink, but switched it out when they started getting really hot. I like to crank it up and wanted to have plenty of sinking for the heat.
The external cap board is for the input capacitors. I used a Rel Multicap MKP 1.0uF in parallel with a Russian teflon .1uF cap. I have tons of nice weird caps on hand, so I look for special opportunities to use them whenever I can.
I used 160VA toroids from Plitron. All wiring is cryo'ed from Cryo-parts.com...which is my favorite place for tweaky stuff. The case is a Parmetal case off Ebay painted Hammertone light green. I've installed a cryo'ed IEC connector and am using a well burned in VHAudio Flavor 4 power cord (which usually gives great low end and imagining).
If you look at the main picture of the amp, you will see three other tube amps that I frequently use....a VAC, a Mac MC240 and a ST35 dynaclone with EL84 tubes. I also have a TEAC tripath amp and a 41Hz Amp3 tripath. So my sound impressions are based on comparisons to mainly tripath SS and tube sound.
The MyRef is very clean and detailed sounding. Very punchy sound with plenty of volume. The midrange is well fleshed out with a bit of a rising top end on certain material. The bass is decent and has a bit of slam, but I'm hoping it'll develop a bit more as it settles in. It's a more up front sound than the Tripath amps I've used. The Tripath stuff is a bit more laid back and softer sounding. The MyRef is almost front hall in it's perspective. While it certainly doesn't have the warmth of an EL84, it does have a slightly warm sound and isn't sterile by any standard. Just very accurate and laid bare. Very clear window into the performance....it's accuracy is startling on certain material.
These are initial impressions and since I am a firm believer in burn in, I am expecting to get improvements over the next 100 to 200 hours. Typically I find bass to strengthen and tighten up and any rising top end/harshness to eventually even out and subside. I am very happy with this amp and am looking forward to playing the sh*t out of it all this weekend. Quite a fun project and Mauro should be commended for turning us onto his design. Also Russ thanks for doing the PCB.
I'm still undecided whether I should add in a Kookaburra pre since my VAC preamp is so fun to use.
http://tommytube.com.moranplanet.com/MyRef/
I've finally got my amp up and running and placed in my big rig. It's about 95% completed. You can see some pix at the above URL. It's working well and doesn't have any hum at all. Rev. A components.
My issue from a week again was a dodgy solder joint which was causing the left channel to heat up and oscillate. Finally found that and it's working fine now.
My last bit of problem was my power switch had a broken pin so I'm awaiting a new one. That's why the power is wire nutted. I also haven't punched a hole for a fuse yet...which I probably will add in so it's externally accessible.
I pretty much used the values stated on the BOM, but choose different parts since that's something I really enjoy doing. I used 15,000uF electrolytics on the power side. These are bigger than the normal 10,000 cans. To make them fit all I had to do was rotate the outside cap 180 degrees and run a jumper from the inner node to the outside pin on the cap. The advantage of this is it leaves room to run a screwdriver in to tighten/loosen the screws on the LM3886/heatsink.
The heatsink was the hardest part of the kit. Getting it tapped and placed was a challenge. I originally started using a smaller heatsink, but switched it out when they started getting really hot. I like to crank it up and wanted to have plenty of sinking for the heat.
The external cap board is for the input capacitors. I used a Rel Multicap MKP 1.0uF in parallel with a Russian teflon .1uF cap. I have tons of nice weird caps on hand, so I look for special opportunities to use them whenever I can.
I used 160VA toroids from Plitron. All wiring is cryo'ed from Cryo-parts.com...which is my favorite place for tweaky stuff. The case is a Parmetal case off Ebay painted Hammertone light green. I've installed a cryo'ed IEC connector and am using a well burned in VHAudio Flavor 4 power cord (which usually gives great low end and imagining).
If you look at the main picture of the amp, you will see three other tube amps that I frequently use....a VAC, a Mac MC240 and a ST35 dynaclone with EL84 tubes. I also have a TEAC tripath amp and a 41Hz Amp3 tripath. So my sound impressions are based on comparisons to mainly tripath SS and tube sound.
The MyRef is very clean and detailed sounding. Very punchy sound with plenty of volume. The midrange is well fleshed out with a bit of a rising top end on certain material. The bass is decent and has a bit of slam, but I'm hoping it'll develop a bit more as it settles in. It's a more up front sound than the Tripath amps I've used. The Tripath stuff is a bit more laid back and softer sounding. The MyRef is almost front hall in it's perspective. While it certainly doesn't have the warmth of an EL84, it does have a slightly warm sound and isn't sterile by any standard. Just very accurate and laid bare. Very clear window into the performance....it's accuracy is startling on certain material.
These are initial impressions and since I am a firm believer in burn in, I am expecting to get improvements over the next 100 to 200 hours. Typically I find bass to strengthen and tighten up and any rising top end/harshness to eventually even out and subside. I am very happy with this amp and am looking forward to playing the sh*t out of it all this weekend. Quite a fun project and Mauro should be commended for turning us onto his design. Also Russ thanks for doing the PCB.
I'm still undecided whether I should add in a Kookaburra pre since my VAC preamp is so fun to use.
Russ White said:
Yeah, I'm a psycho.
Russ, you tend to treat everyone as kids, that's a very beginner mistake to make, to fool yourself because you expect a change.
Russ White said:
With my permission you could use them, you didn't ask.
Not that I own anything, it's just a question of principles.
My schematics are clear, and we are just talking about this because I shared it.
You didn't build any of my last PSUs, regulated and/or unregulated, so you can't comment on that.
Just let people try it, ok?
Maybe you like a certain kind of sound, or your amp just sounds better to you that way with your source and/or speakers.
Certainly for me a 'smoother' sound than a Yamaha amp is way out of question.
Hi Carlos,
It's a mistake people make all the time Carlos. Any honest person would admit that, especially when tests aren't done blind.
, and how would you assume to know what I have or have not built anyway? That is once again absurd.
There is one clear problem though... You have not even tried this amp, yet you claim to know what impact your PS mods will have on it... once again absurd.
And BTW, I think most people would not think this amp is overly warm at all. Just ask Troy, Brian, Ed.., George.. Etc... Etc... Etc..
Good day to you.
carlosfm said:
It's a mistake people make all the time Carlos.
Any honest person would admit that, especially when tests aren't done blind.Ok that is absurd... why on earth would someone ask your permission to implement an idea that is not yours?carlosfm said:
That is not even close to true, I very certainly have built them and even have a test rig just for that, just in case some miracle should actually happen.carlosfm said:
, and how would you assume to know what I have or have not built anyway? That is once again absurd.I stated right off that bat that people should try whatever they like, and still hold to that. People should try whatever they like.carlosfm said:
There is one clear problem though... You have not even tried this amp, yet you claim to know what impact your PS mods will have on it... once again absurd.
And BTW, I think most people would not think this amp is overly warm at all. Just ask Troy, Brian, Ed.., George.. Etc... Etc... Etc..
Good day to you.
Russ White said:Hi Carlos,
It's a mistake people make all the time Carlos.
Blind tests, sure. That would give another debate, and I'm not in the mood.
One thing you can't accuse me is not being honest.
Unless you have something to point me.
I'm doing blind tests for more years than you know what's hi-fi.
Russ White said:Ok that is absurd... why on earth would someone ask your permission to implement an idea that is not yours?
Well, that clearly makes the point, who's honest.
Instead of discussing, you should accept it.
I never asked money to anyone for using my public schematics, for personal or even commercial use.
But for commercial use I like to be contacted, and I like to be given credit for my schematics.
Brian (GT) contacted me to make the snubberized PSU boards. He used a schematic of mine, and I even helped him with the PCB layout.
Money? Me? No.
Of course this is out of the question for you, because you don't understand these principles.
Mauro also didn't invent anything, so I suppose it would be ok for anyone to copy his schematics and boards and sell them, without giving any credit for him, right?
PS: some time ago you posted the values of the snubbers you tried. Now you say you tried everything.
Hey, I'm a serious man. I don't lie and I have no patience for these discussions, when they are not serious.
I'd like to sidestep the whole snubber discussion, but would like to point out to everyone that Russ and I are not doing this to make any money. Both of use are, in fact, currently carrying quite a bit of debt for the efforts.
The goal of selling stuff through the web site is to avoid the whole Group Buy affair, which is hard to manage, and leaves a lot of people out, mostly because of simple timing. We want to make the stuff we offer available on a continuing basis, so everyone can get one. (Look around the forum and see how many times you see this: "Hey, am I too late to get in on this group buy?") Half the question we field through the web site ask just that.
The downside of doing it this way is that we have to pay for lots of stuff ahead of time, and hope people buy it. There is a tiny bit of profit, maybe a couple of dollars per kit, but that is more than erased by the credit card interest and many many hours that go in to getting everything ordered, packaged, discribed, shipped, supported, etc. I think the yearly fee for the web site alone probably erases any profit. Total up the parts in the kits and compare to how much they cost if you were to buy them. Not much difference, if any. I work a after hours contract job to get the money to front for all the parts.
Don't get me wrong, I am doing this because I enjoy it, but it's a mountain of work. The more popular it is, the more work.
The good part? Why are we doing it? Well, we get to design stuff, play with it, and share it. We have a great amp, a great preamp, and soon a great source selector. you can make your own boards, or get ours, buy your own parts, or get ours (well, the source selector pretty much has to be a double-sided board, unless it's really big, so that might be tougher, but we will share the schematic for it).
Sorry for the rant, just wanted to chime in about the "commercial profit" stuff.
The goal of selling stuff through the web site is to avoid the whole Group Buy affair, which is hard to manage, and leaves a lot of people out, mostly because of simple timing. We want to make the stuff we offer available on a continuing basis, so everyone can get one. (Look around the forum and see how many times you see this: "Hey, am I too late to get in on this group buy?") Half the question we field through the web site ask just that.
The downside of doing it this way is that we have to pay for lots of stuff ahead of time, and hope people buy it. There is a tiny bit of profit, maybe a couple of dollars per kit, but that is more than erased by the credit card interest and many many hours that go in to getting everything ordered, packaged, discribed, shipped, supported, etc. I think the yearly fee for the web site alone probably erases any profit. Total up the parts in the kits and compare to how much they cost if you were to buy them. Not much difference, if any. I work a after hours contract job to get the money to front for all the parts.
Don't get me wrong, I am doing this because I enjoy it, but it's a mountain of work. The more popular it is, the more work.
The good part? Why are we doing it? Well, we get to design stuff, play with it, and share it. We have a great amp, a great preamp, and soon a great source selector. you can make your own boards, or get ours, buy your own parts, or get ours (well, the source selector pretty much has to be a double-sided board, unless it's really big, so that might be tougher, but we will share the schematic for it).
Sorry for the rant, just wanted to chime in about the "commercial profit" stuff.
BrianDonegan said:
I agree, just forget about the snubber, move on.
I suggest the unfortunate member that used 15,000uf caps and reported his honest results (what proves that he's not biased with the new amp just because he's changing something on his system) to contact me personally and I'll try to help.
Also the LM318 op-amp PSU bypass, etc.
Well I think everyone should be thanked for their contributions, I have ordered my Rev C boards, but I also went to go and buy components to try the Carlos version for educational value, and so that I don't have to rely on theory and say so, but rather practical experience. Unfortunately I got some new games today so soldering is out of the question.....maybe tommorrow