Anyone work this out?????

georgehifi · 2017-06-01 4:37 am

What does the NE5532's do???

Cheers George

georgehifi · 2017-06-01 4:46 am

Just got this from someone else, but he doesn't sound too sure either.
Hi George,

Looks like soneone's idea of improving cmrr for an unbalanced output.

First time I've seen it though.

Cheers George

abraxalito · 2017-06-01 5:10 am

That kind of configuration looks rather like a GIC (generalized impedance converter). I haven't plugged in the specific Zs to work out what kind of impedance is being created there, but if you take a look at the schematic here you may notice similarities - https://electronics.stackexchange.com/questions/130632/general-impedance-converter

The last page (p10) of the PCM63 datasheet has a similar opamp configuration for the output filtering.

georgehifi · 2017-06-01 5:26 am

Hi abraxalito, did you see the 5v - + analog dac connections on the first NE5532 for VCC+ and VCC- tags?

It's a very nice sounding player, the Shanling CD-T100 with PMD200 as well.

.Cheers George

abraxalito · 2017-06-01 5:36 am

No, I only see +15,-15V being fed to the NE5532s. Via inductors.

CharlieLaub · 2017-06-01 6:14 am

At first looked much like a frequency dependent negative resistor to me... but I'm not sure because it has a capacitor where that circuit branch connects to ground (Z5 in the image below) whereas a FDNR has a resistor there. Anyway, it is some kind of generalized impedance converter of this form:

Oops I see Abraxalito beat me to this conclusion.

Edit: On second thought maybe I was right. Here is an example of FDNRs used for filtering:
https://www.silvaco.com/examples/gateway/section1/example7/index.html
https://www.silvaco.com/examples/gateway/section1/example7/GIC.png

Also, see:
http://www.diyaudio.com/forums/part...rs-frequency-dependent-negative-resistor.html

georgehifi · 2017-06-01 6:29 am

abraxalito said:
No, I only see +15,-15V being fed to the NE5532s. Via inductors.

georgehifi · 2017-06-01 6:37 am

CharlieLaub said:
At first looked much like a frequency dependent negative resistor to me... but I'm not sure because it has a capacitor where that circuit branch connects to ground (Z5 in the image below) whereas a FDNR has a resistor there. Anyway, it is some kind of generalized impedance converter of this form:

Oops I see Abraxalito beat me to this conclusion.

Edit: On second thought maybe I was right. Here is an example of FDNRs used for filtering:
https://www.silvaco.com/examples/gateway/section1/example7/index.html
https://www.silvaco.com/examples/gateway/section1/example7/GIC.png

Also, see:
http://www.diyaudio.com/forums/part...rs-frequency-dependent-negative-resistor.html

I see the output connected to the +input of the second 5523, but where's the input?

Cheers George

abraxalito · 2017-06-01 6:38 am

The schematic looks like it was created in Protel. In Protel pin names within the outline of ICs are not the same as net names (attached to wires). So those VCCs you've circled aren't the same.

The input's going to pin5, U19B on the top one, U18B on the bottom.

georgehifi · 2017-06-01 6:43 am

abraxalito said:
The schematic looks like it was created in Protel. In Protel pin names within the outline of ICs are not the same as net names (attached to wires). So those VCCs you've circled aren't the same.

The input's going to pin5, U19B on the top one, U18B on the bottom.

So would they be test V points, or maybe inputs from somewhere, that I asked above?

Can that they wouldn't be inputs as they are at v rail potential.

Cheers George

CharlieLaub · 2017-06-01 6:44 am

georgehifi said:
I see the output connected to the +input of the second 5523, but where's the input?

Cheers George

It's a grounded FDNR. If you look at the figure I posted, the connection to the rest of the circuit is done at left, next to Z1, and is grounded at right. Just think of it like a component created by a subcircuit that has properties (resistance and reactance) that are not possibly in a real world component.

One connection to the FDNR in the schematic that you posted is, for example, done between R148, R144, and R126. That FDNR consists of U19A, U19B, C150, R137, R139, R142, and C154 and connects to "AG" (analog ground).

abraxalito · 2017-06-01 6:46 am

Could well be test points for production/service yes.

Incidentally I reckon you might get a SQ improvement by reducing the 220pF feedback caps around the first OPA2604s. Set them much lower, say 22pF.

georgehifi · 2017-06-01 6:50 am

CharlieLaub said:
It's a grounded FDNR. If you look at the figure I posted, the connection to the rest of the circuit is done at left, next to Z1, and is grounded at right. Just think of it like a component created by a subcircuit that has properties (resistance and reactance) that are not possibly in a real world component.

One connection to the FDNR in the schematic that you posted is, for example, done between R148, R144, and R126. That FDNR consists of U19A, U19B, C150, R137, R139, R142, and C154 and connects to "AG" (analog ground).

So what do you think it does in laymans terms Charlie?

Cheers George

georgehifi · 2017-06-01 7:31 am

This is what I found a grounded FDR does, a bloody sharp LP filter. If indeed that's what it is.
https://www.researchgate.net/public...-impedance-and-phase-of-circuit-in-Fig-3a.png

Also just found this, if it's the same thing. Must admit this player does sound very good.

http://www.diyaudio.com/forums/digi...-using-fdnrs-nos-dac-anti-imaging-filter.html

Cheers George

abraxalito · 2017-06-01 8:01 am

Looks like a high Q (say about 10) low pass filter. You can get that with a 2nd order Sallen-Key provided your active element has enough GBW. No intrinsic need for FDNR.

The FDNR in your original schematic looks to be integrated into a MFB low pass in a config I've not come across before.

georgehifi · 2017-06-01 8:14 am

abraxalito said:
Looks like a high Q (say about 10) low pass filter. You can get that with a 2nd order Sallen-Key provided your active element has enough GBW. No intrinsic need for FDNR.

The FDNR in your original schematic looks to be integrated into a MFB low pass in a config I've not come across before.

You know how I like the OPA627 Abrax, can you see any problems subbing the OPA2604's for them, when I get it??

Cheers George

abraxalito · 2017-06-01 8:17 am

No, other than you'd need single-dual adaptors as dual OPA627s don't exist.

You could try out ISL28210s as they're duals. Still would need adaptors though as they're only available in SMT.

georgehifi · 2017-06-01 8:22 am

abraxalito said:
No, other than you'd need single-dual adaptors as dual OPA627s don't exist.

You could try out ISL28210s as they're duals. Still would need adaptors though as they're only available in SMT.

single-dual adaptors

got them.

georgehifi · 2017-06-01 8:22 am

Are the ISL28210s unity gain stable?

Cheers George

Calvin · 2017-06-02 10:28 am

Hi,

its a GIC lowpass filter synthesizing a frequency dependent negative resistance.
see an play with the asc

jauu
Calvin

The FDNR in your original schematic looks to be integrated into a MFB low pass in a config I've not come across before.

You probabely missed out the Agnd node between R144/R118/C130

So that R144 is in parallel with the GIC-filter and R126/C130 make up a passive 6dB filter.
They tame the GIC´s peaking response with the post filtering.
C117 in U17A´s feedback loop is so small it just limits bandwidth and doesn´t belong to the post filter.

Anyone work this out?????

Member

Attachments

Member

Member

Member

Member

Member

Member

Attachments

Member

Member

Member

Member

Member

Member

Member

Member

Member

Attachments

Member

Member

Member

Member

Attachments