Andrew something you may know - In a post recently I read, EUVL eluded to the fact that not all ceramics are the same, if you're critical about measurements then choice matters even within the same class.
Can you advise when to use X7R or C0G, do you worry or study the specs of ceramics?
I generally choose the cheapest/best price break usually AVX or Kemet, EUVL mentioned a particular series of Murata that is now extinct, I think CJ8 that had good noise/frequency specs.
Can you advise when to use X7R or C0G, do you worry or study the specs of ceramics?
I generally choose the cheapest/best price break usually AVX or Kemet, EUVL mentioned a particular series of Murata that is now extinct, I think CJ8 that had good noise/frequency specs.
For DC rail decoupling, X7R. I usually use Kerry, AVX or Murata, but have used Samsung, Panasonic, TDK and Wurth without problems.
For compensation, EQ networks or filtering, only COG or NP0
Never use X7R for anything but decoupling DC rails. They have a wide initial tolerance spread (-20% to +50%) but the real problem is they have a high voltage coefficient that in some cases can be as much as 50% of the stated value from 20% to 100% of the rated working voltage. So, if you pass an AC signal through them, the value is changing dynamically with the voltage.
I made the mistake of using one in one of the early EF3 amps and when driving it into a load, the caps would sing at the test frequency from the piezo electric effect.
COG/NP0 have none of these problems, but get expensive in high values.
For compensation, EQ networks or filtering, only COG or NP0
Never use X7R for anything but decoupling DC rails. They have a wide initial tolerance spread (-20% to +50%) but the real problem is they have a high voltage coefficient that in some cases can be as much as 50% of the stated value from 20% to 100% of the rated working voltage. So, if you pass an AC signal through them, the value is changing dynamically with the voltage.
I made the mistake of using one in one of the early EF3 amps and when driving it into a load, the caps would sing at the test frequency from the piezo electric effect.
COG/NP0 have none of these problems, but get expensive in high values.
Last edited:
Thanks for that XYour work it notably good. It's fun to see the chips snap into place under tension. I'm finishing that exact bal board today.
Andrew those are 0603 parts, I did mine by hand and lost a resistor at the end. It arrived today.
Thanks for that X
Andrew those are 0603 parts, I did mine by hand and lost a resistor at the end. It arrived today.
I never thought I’d fall prey to solder envy ....
As Passive said, only 0603 - easy peasy.
0402 is another matter...
Just this week, a friend asked me to replace one resistor on his board. It’s an 0201! Now that is small. I have to use my microscope for that one. I also did not have any 0201’s in my bench. They could be mistaken for dust.
Lead free can actually be quite lower temp. It is made of mostly tin and added silver and bismuth etc. it doesn’t flow as nicely and grows dendritic whiskers over time. These can actually short.
Billions of $ have been lost on satellites that once placed in orbit have circuits that shorted out due to tin whiskers years later and caused failures. Pb solder does not have this issue. All this was because we were pushing for Pb-free PCBs so that landfills would not be polluted. Guess what, satellites won’t ever go to a landfill! At least not on earth.
Billions of $ have been lost on satellites that once placed in orbit have circuits that shorted out due to tin whiskers years later and caused failures. Pb solder does not have this issue. All this was because we were pushing for Pb-free PCBs so that landfills would not be polluted. Guess what, satellites won’t ever go to a landfill! At least not on earth.
Hello Chris,
No DC protection. The output devices used in the X-Altra HPA-1 are rated at 8 Amps, the supply voltages are low (+-18 V Max) vs the rated Vce (200V IIRC) and the short circuit output current is 300 mA. The devices are on a generous heatsink so there is little chance of a failure, assuming it’s well constructed and thoroughly tested.
Most opamps are incredibly reliable so IMV you should be safe.
No DC protection. The output devices used in the X-Altra HPA-1 are rated at 8 Amps, the supply voltages are low (+-18 V Max) vs the rated Vce (200V IIRC) and the short circuit output current is 300 mA. The devices are on a generous heatsink so there is little chance of a failure, assuming it’s well constructed and thoroughly tested.
Most opamps are incredibly reliable so IMV you should be safe.
Last edited:
Lead free can actually be quite lower temp. It is made of mostly tin and added silver and bismuth etc. it doesn’t flow as nicely and grows dendritic whiskers over time. These can actually short.
Billions of $ have been lost on satellites that once placed in orbit have circuits that shorted out due to tin whiskers years later and caused failures. Pb solder does not have this issue. All this was because we were pushing for Pb-free PCBs so that landfills would not be polluted. Guess what, satellites won’t ever go to a landfill! At least not on earth.
Yes, pure lead or lead-tin alloys can be as high as 250C but most are around 183C. They add silver and sometimes copper to lower the melting point (weird since both have melting points well above 183 - it’s a job for a metallurgist to explain that!). I am probably a bit jaundiced on this because I bought a roll of lead free about 5 yrs ago and it was virtually unusable. I vowed never to use lead free again.
YMMV
Happy days Sonidos. Love it when a Mouser or Digikey box arrives
Forgot to recommend a iron tip cleaner, use regularly to keep your tip clean. You'll notice the tip not picking up solder properly when using for a while, a quick dip in the cleaner and you're good.
https://www.amazon.com/Thermaltroni...jbGlja1JlZGlyZWN0JmRvTm90TG9nQ2xpY2s9dHJ1ZQ==
Forgot to recommend a iron tip cleaner, use regularly to keep your tip clean. You'll notice the tip not picking up solder properly when using for a while, a quick dip in the cleaner and you're good.
https://www.amazon.com/Thermaltroni...jbGlja1JlZGlyZWN0JmRvTm90TG9nQ2xpY2s9dHJ1ZQ==
Hi, thanks for this headphone amp, this is going to be my first diy project
Would it be possible to use the capacitor Dersonic CC1H470JC74DCH4B10MN instead of MSR 80-C0805C470J5G? Because JLCPCB is offering the assembly service for the other one.
Hello,
I looked at the data sheet, unfortunately the part is not SMD, but through hole. You will need an 0805 47 pF NP0/COG capacitor rated at 50V or higher.
Nothing exciting to report. Chassis should arrive from Italy in a matter of days. Halfway done with the amp board. The only excitement was that I never noticed it at the time, but my tube of ChipQuik is Pb free. I had some nasty solder bridges after coming out of the oven. Mix in the fact that I have a new T12 soldering station, and what happens? One chip removed with a lifted pad. Sigh. Fortunately I can tack on a wire to pad 2 of the IC. Oh, and a new tube of ChipQuik is on order, this time SnPb mix. Maybe next week I have everything I need to power on. I'm learning a ton now, including getting nice fillets on my smd resistors and caps.
What's a good rule of thumb for a fuse? I was thinking a 1 amp/250 V slow-blow type.
What's a good rule of thumb for a fuse? I was thinking a 1 amp/250 V slow-blow type.
Javier, I think on the mains side, a 200mA slo-blo is what you need. This is at least 2x a 25 VA current draw. It’s there for gross failure/fire prevention. If you are using a 15 VA transformer, a 100 mA slo-blo will suffice.
I haven’t yet done my SMD video - I will try to find some time this week.
I haven’t yet done my SMD video - I will try to find some time this week.
Andrew, always appreciate your explanations. This fills another knowledge gap. Yeah, it's the 25VA Talema transformer.
BTW, both chips had these solder bridges. On the 2nd chip, I flooded the bridged pins with leaded solder, added flux, then applied the wick. It sucked that excess solder right off in a heartbeat. I lost one chip but saved the other.
BTW, both chips had these solder bridges. On the 2nd chip, I flooded the bridged pins with leaded solder, added flux, then applied the wick. It sucked that excess solder right off in a heartbeat. I lost one chip but saved the other.