Hey Tom, Just wondering what you think of this discussion:
https://duckduckgo.com/l/?uddg=http...e49860-do-they-sound-different-to-you.521972/
Seems that some swear by using the can version lme49720HA others think that the higher voltage part lme49860 is the best sounding. Probably doesn't matter in this use case but I thought you might find it interesting. I imagine it doesn't matter which one of these we use?
https://duckduckgo.com/l/?uddg=http...e49860-do-they-sound-different-to-you.521972/
Seems that some swear by using the can version lme49720HA others think that the higher voltage part lme49860 is the best sounding. Probably doesn't matter in this use case but I thought you might find it interesting. I imagine it doesn't matter which one of these we use?
The HP-22 is not sensitive to the opamp used. Of course, if you use an opamp with worse specs than the LME49720 you'll degrade performance. But as long as you use a unity gain stable opamp the HP-22 will be remain stable.
Some will always favour equipment that uses their pet parts. To them such equipment will always provide more PRaT, PLoP, or KäPLUÑK (or whatever the meaningless audiophile term d'jour is). I go by the end-to-end measured performance. Others are free to do as they prefer.
If you want to try different opamps, the HP-22 would be a good platform for that. I can come up with a theoretical argument for why the LME49860 would be better (higher voltage rails -> lower THD), but note that's it's SMD-only, so you'll have to use an adapter.
The TO-99 can package offers the lowest package stress. Package stress is a known contributor to DC offset in an opamp. However, there are better ways to deal with package stress than to put the part in a $10 package. There are anti-stress coatings, for example. One can also design the layout of the input stage of the opamp such that it is not as sensitive to package stress. Also note that package stress is just one type of stress that you're dealing with in an IC. I don't know if package stress would impact the distortion of an opamp. I doubt it, but, like I said, I don't know.
Tom
Some will always favour equipment that uses their pet parts. To them such equipment will always provide more PRaT, PLoP, or KäPLUÑK (or whatever the meaningless audiophile term d'jour is). I go by the end-to-end measured performance. Others are free to do as they prefer.
If you want to try different opamps, the HP-22 would be a good platform for that. I can come up with a theoretical argument for why the LME49860 would be better (higher voltage rails -> lower THD), but note that's it's SMD-only, so you'll have to use an adapter.
The TO-99 can package offers the lowest package stress. Package stress is a known contributor to DC offset in an opamp. However, there are better ways to deal with package stress than to put the part in a $10 package. There are anti-stress coatings, for example. One can also design the layout of the input stage of the opamp such that it is not as sensitive to package stress. Also note that package stress is just one type of stress that you're dealing with in an IC. I don't know if package stress would impact the distortion of an opamp. I doubt it, but, like I said, I don't know.
Tom
Thanks, Yeah I first thought the LME49860 was a 8Dip. I think I'll start out with the regular 49720, I might get a few of those adapters though if I feel like trying some exotic ICs later.
May I ask if anyone knows how can a amateur like myself ground a aluminum chassis properly? I am looking for a chassis while waiting for the eva board to be restocked on mouser. i saw mixed opinions on the net.
You can buy the EVM directly through TI. With shipping it comes to a few bucks more than Mouser, but you'll have it within days.
I usually drill a hole, sand off the anodization, and use a grounding lug. If you want to get fancy (which I recommend) use a lock nut. Keystone 908 or 7328 are good lugs for this.
Tom
I usually drill a hole, sand off the anodization, and use a grounding lug. If you want to get fancy (which I recommend) use a lock nut. Keystone 908 or 7328 are good lugs for this.
Tom
Thanks, mouser told me to send a email to change my order. I wonder if the freshly sanded Al oxidise lose its conductivity in long term. Also, say if I use a hammond enclosure, do I need to ground every single piece of the enclosure? I think the 220v worries me so much that I may consider 3d print plastic enclosure, which costs a lot🙁 . Am I overthinking a simple question?
Last edited:
Aluminum oxidizes. That's why you want a gas tight connection to it. Use a solder lug that's intended for the purpose.
In my general experience, the chassis screws have worn through the anodization when tightened, so I would expect all panels to be grounded. You can always check with a multimeter.
I recommend keeping at least 6.4 mm from any point connected to mains voltage to the chassis (or any metal the user can touch).
Tom
In my general experience, the chassis screws have worn through the anodization when tightened, so I would expect all panels to be grounded. You can always check with a multimeter.
I recommend keeping at least 6.4 mm from any point connected to mains voltage to the chassis (or any metal the user can touch).
Tom
Hello Tom and All,
This looks like a mature project. A couple of questions.
Does anyone have insights on a chassis, what chassis worked for you if you used one?
I liked the power toggle that lights up when it is turned on. What DPDT switch worked out for you?
Any impressions about the sound quality for you?
Thanks
DT
This looks like a mature project. A couple of questions.
Does anyone have insights on a chassis, what chassis worked for you if you used one?
I liked the power toggle that lights up when it is turned on. What DPDT switch worked out for you?
Any impressions about the sound quality for you?
Thanks
DT
I'm sure a ModuShop Galaxy would work. I also have a Hammond aluminum chassis that I plan to turn into an HP-22 amp and build video. I forget the exact model of the Hammond chassis but can dig that out if need be.
Personally I'd use a Preamp Power Supply. That'll allow you to use the nice anti-vandal switches, such as the E-Switch PV6-series.
Tom
Personally I'd use a Preamp Power Supply. That'll allow you to use the nice anti-vandal switches, such as the E-Switch PV6-series.
Tom
Now there's a build video. Thanks to @6L6 for the modelling. 🙂
The video shows how to take an off-the-shelf Hammond chassis and machine it to fit the HP-22 and Neurochrome Preamp Power Supply. All the drill templates and BOM are provided on the Resources tab of the HP-22 product page.
Tom
The video shows how to take an off-the-shelf Hammond chassis and machine it to fit the HP-22 and Neurochrome Preamp Power Supply. All the drill templates and BOM are provided on the Resources tab of the HP-22 product page.
Tom
Hi Tom,
Have you tried the LME49600TSBD?
It also looks like a good solution using LME49720 and LME49600 as composite amplifier. Also one side of LME49720 is for DC correction.
The figures in the user guide looks very promising.
Is there benefit to choose INA1620EVM to build HP-22?
Have you tried the LME49600TSBD?
It also looks like a good solution using LME49720 and LME49600 as composite amplifier. Also one side of LME49720 is for DC correction.
The figures in the user guide looks very promising.
Is there benefit to choose INA1620EVM to build HP-22?
I used the LME49600 in the HP-1: https://neurochrome.com/pages/hp-1 That's a great headphone driver. The circuit in the data sheet provides good performance, though it can be bettered. The LME49720 is not a good choice for a DC servo (even though that's what's done in the LME49600 data sheet). The OPA2277 or even the aging LF451 would be better due to the much lower input bias current.
Tom
I doubt that EVM will plug directly into the HP-22 board.
Tom
Just as you predicted...the lme49720 isn't a good option for DC servo.
The LME49600TSBD I got has 19.2mV and ~4mV DC offset. I'll perhaps try to replace it with OPA1656 to see what will happen.
The LME49600TSBD I got has 19.2mV and ~4mV DC offset. I'll perhaps try to replace it with OPA1656 to see what will happen.
I now wonder if the DC servo in the evaluation board make the offset even larger......
The DC servo test point I measured ~18mV and 5mV DC offset and it's close to the output offset.
The LME49720 itself only has 0.7mV Vos and the gain of the amp should be 2, so the original DC output would only be 1.4mV? Before consider the tolerance of resistance...
The DC servo test point I measured ~18mV and 5mV DC offset and it's close to the output offset.
The LME49720 itself only has 0.7mV Vos and the gain of the amp should be 2, so the original DC output would only be 1.4mV? Before consider the tolerance of resistance...