I have the Gotham Audio library. Basically every Neumann lathe document ever produced. I don't want everything floating around the web. Gotta keep the punters thinking it's magic. I give the info to those who need it. I'm happy to scan whatever you want to look at. I just don't want it posted.mediatechnology wrote: Thanks for all the info Paul. The SE66 schematics are very hard to find online.
Vintage Windings doesn't seem to have it.
Neumann SE66
Neumann SE66
Neumann SE66
You mean it isn't???Gold wrote:Basically every Neumann lathe document ever produced. I don't want everything floating around the web. Gotta keep the punters thinking it's magic.
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Neumann SE66
You may still be able to find some LF356 in TO-99 cans. I may have some.
The LME49710 should be fine though.
The LME49710 should be fine though.
Neumann SE66
mediatechnology wrote:You may still be able to find some LF356 in TO-99 cans. I may have some.
The LME49710 should be fine though.
I think I have some TO-99 LF356. Mouser has TO-99 LF356 too. I'm not exactly sure I'm asking the right questions but here goes. Both current noise and voltage noise are determined by the source impedance, right? Is R5 33k seen as the source impedance to IC1 in parallel with the transformer Z and output impedance of the source signal?
You said this stage is difficult for any opamp. What would be the ideal set of parameters for an opamp be? Last question. Would the OPA627 be the best of the lot?
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Neumann SE66
The source impedance is complex and frequency-dependent due to the EQ.
Noise current and voltage are not the most critical aspect IMHO.
What makes this circuit difficult is the RIAA preemphasis and the bandwidth demands placed on the op amp.
A true inverse RIAA needs infinite bandwidth which is why we have a "Neumann pole."
The response can't rise forever.
I don't know that an OPA627 would be that good of a choice.
As I understand it it can be pretty difficult to compensate.
The last thing you want ahead of the cutterhead is HF instability.
The LME49720 does quite well in the PTS' inverse RIAA EQ.
The LF356 should be a huge improvement over the 709.
Noise current and voltage are not the most critical aspect IMHO.
What makes this circuit difficult is the RIAA preemphasis and the bandwidth demands placed on the op amp.
A true inverse RIAA needs infinite bandwidth which is why we have a "Neumann pole."
The response can't rise forever.
I don't know that an OPA627 would be that good of a choice.
As I understand it it can be pretty difficult to compensate.
The last thing you want ahead of the cutterhead is HF instability.
The LME49720 does quite well in the PTS' inverse RIAA EQ.
The LF356 should be a huge improvement over the 709.
Neumann SE66
Ahh that makes sense, and I actually understand it. Unlike voltage noise and current noise. Those are mysterious parameters to me. This is where my fundamental lack of understating electronics shines. I've been looking at the opamp data sheet graphs and I kind of get that but I don't understand the practical implications. The somewhat subjective stuff like when it's a problem and when it's not.mediatechnology wrote: What makes this circuit difficult is the RIAA preemphasis and the bandwidth demands placed on the op amp.
A true inverse RIAA needs infinite bandwidth which is why we have a "Neumann pole."
The response can't rise forever.
Neumann SE66
Yup, RIAA EQ is only defined up to 20 kHz (used to be only defined up to 15 kHz), but the common tweaker assumption is that the 75uSec pole continues up to light.mediatechnology wrote:The source impedance is complex and frequency-dependent due to the EQ.
Noise current and voltage are not the most critical aspect IMHO.
What makes this circuit difficult is the RIAA preemphasis and the bandwidth demands placed on the op amp.
A true inverse RIAA needs infinite bandwidth which is why we have a "Neumann pole."
The response can't rise forever.
To avoid easily observable errors below 20 kHz the response should extend higher. 200 kHz is a nice safe margin, but even a couple octaves above 20 kHz is quite good. (It doesn't hurt to have extra attenuation during playback for above band noise/signals, during encode a zero somewhere reasonable above 20 kHz is just fine IMO. )
In my judgement RIAA accuracy is over rated for importance in the context of audible factors surrounding vinyl playback (cartridges are a hot mess), but it is worth making things that we can control as good as practical.
JR
Cancel the "cancel culture", do not support mob hatred.
Neumann SE66
I looked through my stash. In TO-99 I have MC-1456CG, LM301H, LM302H, LM318H.
Of course I'd like to use what's on hand. If for no other reason then I doubt the parts will ever get used otherwise. The reason I have them is because Al Grundy never used them. I see the fate for them.
That said I want to use what's best. The LF356 looks better on paper to me than the MC1456. I haven't checked the others yet. It's a lot faster. They aren't too expensive and I only need four per rack.
Of course I'd like to use what's on hand. If for no other reason then I doubt the parts will ever get used otherwise. The reason I have them is because Al Grundy never used them. I see the fate for them.
That said I want to use what's best. The LF356 looks better on paper to me than the MC1456. I haven't checked the others yet. It's a lot faster. They aren't too expensive and I only need four per rack.
- mediatechnology
- Posts: 5466
- Joined: Sat Aug 11, 2007 2:34 pm
- Location: Oak Cliff, Texas
- Contact:
Neumann SE66
Well the LF356 is what Neumann ended up using and would be an improvement over the µA709.
I was surprised to see that TI still make it in TO-99.
I was surprised to see that TI still make it in TO-99.
Neumann SE66
Yeah, that makes the most sense. I guess the opamp stash will probably end up with whoever cleans out my shop...mediatechnology wrote:Well the LF356 is what Neumann ended up using and would be an improvement over the µA709.