the Pimeta-V2 thread - Page 20

Hello,

I have just built my Linkwitz Crossfeed circuit as supplied by Tangentsoft and I am a little unsure if I have installed the jumpers correctly. I wanted to use the board in a set configuration of low crossfeed so have not installed switches. I have soldered one jumper from S1L1 to S1L2 and a second jumper from S2E1 to S2E2. I have tested the circuit and the bass just disappears and the presentation is a little harsh.

I have re-read the article on the Tangentsoft website and in particular the bit that says "If you only want one crossfeed setting, you can leave out the R1As and jumper from each S1C to the corresponding S1L" which I think suggests that my jumpers are incorrect.

Any help is appreciated.

Many thanks

Yes, you've wired it incorrectly all right. To get the effect you want, the correct jumper configuration is:

S1C1 -> S1L1

S2C1 -> S2E1

S1C2 -> S1L2

S2C2 -> S2E2

Thank you for clearing that up for me!

The crossfeed circuit has really transformed the sound, it sounds so much more spacious and even, for some reason, tamed some of the hard edge to the upper frequencies I was experiencing.

I would still like to try the unity gain bass boost and have read the information on the PPA page, but when I use the bass boost calculator and enter my resistor values it seems as though I would have to use a different R7 with much higher resistance than the one I have used per the schematic (100 Ω). All my resistor values are as per the schematic exept R3 and R4 (2K and 10K). Is there anywhere I can find out what R7 values I can go up to safely?

Many thanks.

The PPA bass boost calculator doesn't work for the PIMETA circuit. I pointed people to the PPA docs only for advice on cap types.

If you want to use the calculator, you'll have to build up the PPA circuit atop the PIMETA circuit, as described in the PIMETA docs.

Thank you for your reply.

Do the notes on the capacitor values recommended (0.01UF to 0.1UF) still apply to the Pimeta?

I am just trying to find out what sort of range of capacitance I need so I can try a few different values and see what works for me.

Many thanks

I've built a CMOY, and am planning building a PIMETA next.   I also have an E9 for comparison.   Can anyone speculate on what sonic differences these circuit level changes make:

What does the 3ch topology do for the sound?

What does biasing into class A do sonically?

Does the buffer matter as much if I am driving high impedance (600ohm) head phones? .

I know what all of these changes do electrically, but have no idea what they might do for the sound.

Quote:

Originally Posted by fylde 

Do the notes on the capacitor values recommended (0.01UF to 0.1UF) still apply to the Pimeta?

What's important the ratios of the cap and the resistor paralleling it.  If you use the same resistor value as recommended for the PPA, then yes, the cap value range is the same.  By default, we recommend a lower R4 value about half that in the PPA, however, so you might need to use a cap value twice what you'd use in a PPA. Quarter the resistor value, quadruple the cap value, etc.

Many thanks again for your help, I understand now.

I just changed my R3 and R4 to increase the gain to compensate for the crossfeed board I have installed, R3 was as per the schematic 1K and R3 was not as per the schematic at 2K for a gain of 3.

I used the calculator to go for a gain of 6 and it gave me an R3 vlaue of 2K and R4 10K so I have installed these. I may be imagining it but it possibly doesn't sound as good as before but I can't put my finger on it, I looked back through the notes in the part selection guide for these resistors and it does mention keeping R3, R5 and R6 the same and just adjusting R4 for the gain.

Have I made an error in keeping R5 and R6 as per the schematic and changing R3 & R4 instead of just R4 as suggested?

Many thanks

I think it's more likely that you just don't like the sound of the crossfeed circuit.

Thak you for the reply.

I did think that the crossfeed sounded great, then I installed different R3 & R4 to compensate for the loss in gain and further listening was wondering if I had lost something. If changing R3 as well as R4 doesn't throw anything out in relation to R5 & R6 then thats all I was concerned about.

I have just taken delivery of a couple of switches that I shall wire up to the crossfeed circuit so I can get a better idea when comparing.

Many thanks.

Quote:

Originally Posted by ShadowKntSDS 

I've built a CMOY, and am planning building a PIMETA next.   I also have an E9 for comparison.   Can anyone speculate on what sonic differences these circuit level changes make:

 

What does the 3ch topology do for the sound?

What does biasing into class A do sonically?

Does the buffer matter as much if I am driving high impedance (600ohm) head phones? .

 

I know what all of these changes do electrically, but have no idea what they might do for the sound.

 

 

I have a basic CMOY I got off ebay which then prompted me to build the PIMETA V2. The PIMETA sounds much cleaner and drives my Grado SR80s much better than the basic CMOY. Sorry i am not much good at going into detail here but I am glad I bulit the PIMETA.

I haven't fully explored this thread (much less this forum on the following topics) so please forgive any waste-of-bandwidth. Anyway ...

WRT tweaking v2 ...

(1) How much extra benefit can be obtained further bypassing (beyond Tangent's topology/schema) various opamps (either buffers or main)? E.g., for the main L/R opamp, I am using 10uF BlackGate electros + AVX 0.1uF PPS smd's. Not a whole lot of room, but the electros have improved SQ.

(2) C6 (NP0 or C0G) seem to be located – PCB footprintwise -- quite far from opamp pins (esp. C6 R/L). Has anyone experimented with moving R6/C6 closer -- make a diff.? Also, given the signal-path route of this cap, I wonder how much benefit can be obtained using Styroflex (shown below) – anyone experiment with this type?

BTW ... for you fellow op-amp rollers ... I've just installed LT1364, which is (shudder!) a bi-polar device. Haven't 'scoped the output yet but, even simply 'dropped in' into Tangent's default topology, they seem to work and work well! Prev. running AD825 (best IMO!), but also tried AD8620 and OPA2132.

@alphaman: have you tried ad827 on both signal and ground opamp? i like them better than the ad8620/8610 combo

Quote:

Originally Posted by i_djoel2000 

@alphaman: have you tried ad827 on both signal and ground opamp? i like them better than the ad8620/8610 combo

Thx for reminding me of this OPA ... no I haven't, tho' I have it in my parts bin. (Feedback on this soon ...)

About the AD8610/20 ... 

GOOD: Pace, rhythm, acceleration and timing (aka PRaT) is among the best (like AD825).

OK: Detail, image focus

BAD: Graininess, unrefined, "loud" (as opposed to fwd or lively) sounding.

I'm not sure there is any way to tame these BAD qualities. Certainly, my added bypassing, noted above, didn't help. There is also op-amp snubbing, which I've tried in other projects, but not in Pimeta due to limited space.

I have some voltage limits, too, so I've gotta be somewhat selective with OPAs. FWIW, I use two parallel 9.6V Imedion NiMH batts. (Parallel for longer-duration pwr).

BTW: (Mostly a query for Tangent, I guess, tho' anyone's reply is welcome) ... how important is it to have "matching" OPAs for BOTH gnd and main L/R? E.g. AD8610/8620; OPA132, 2132; etc.?

Notes on tweaking op-amp performance via non-std. bypassing and snubbing:

47-100 uF high-quality electro cap from each rail to gnd (hard to do in Pimeta). Add a very small cap across the rails (1nF MKT). Snubbers: from each rail to ground (0.22R+3.3nF). 

Quote:

Originally Posted by alphaman 

Thx for reminding me of this OPA ... no I haven't, tho' I have it in my parts bin. (Feedback on this soon ...)

About the AD8610/20 ... 

GOOD: Pace, rhythm, acceleration and timing (aka PRaT) is among the best (like AD825).

OK: Detail, image focus

BAD: Graininess, unrefined, "loud" (as opposes to fwd or lively) sounding.

I'm not sure there is any way to tame these BAD qualities. Certainly, my added bypassing, noted above, didn't help. There is also op-amp snubbing, which I've tried in other projects, but not in Pimeta due to limited space.

 

I have some voltage limits, too, so I've gotta be somewhat selective with OPAs. FWIW, I use two parallel 9.6V Imedion NiMH batts. (Parallel for longer-duration pwr).

 

BTW: (Mostly a query for Tangent, I guess, tho' anyone's reply is welcome) ... how important is it to have "matching" OPAs for BOTH gnd and main L/R? E.g. AD8610/8620; OPA132, 2132; etc.?

 

Notes on tweaking op-amp performance via non-std. bypassing and snubbing:

47-100 uF high-quality electro cap from each rail to gnd (hard to do in Pimeta). Add a very small cap across the rails (1nF MKT). Snubbers: from each rail to ground (0.22R+3.3nF). 

have you tried to upgrade your resistor? i use full takman metal film resistors 1/4W on the signal and ground path and i did not find those problems with the ad8620/8610 combo, not grainy at all and very transparent sounding. i know tangent will deny this statement, but, why don't you just give it a try? there must be good reason why gaincard uses premium quality resistors in their signal path i know it's like 8 times more expensive than RN55 (which is only around 0.1 cent a piece), but i did not regret buying those resistors at all

as for the ad8620, i just found it too bright for my current setup, that's why i switched to opa827