[tangent]

 
 
Quote:

LED1 is not turned on when i turn on the amp

 
Are you certain you didn't put it in backwards?
 
 
Quote:

isn't the rotating knob should rotate in steps?

 
Probably not. I've never heard of a "stepped" trim pot. Without the part number, it's impossible to say more. (Bourns makes a lot of trim pot types.)
 
 
Quote:

the knob keeps rotating when i rotate them

 
If you listen very carefully, you should be able to hear a faint click when you are trying to push the pot adjustment past its limit point. I've broken cheaper trim pots by doing that too often. I don't think I've ever broken a Bourns that way, but best not to do it.

 

[vodkex]

Beautiful amps guys! I've made a CMOY following your schemes tangent (etched my own PCB) and I have to say that it sounds wonderful. I'm using some modded HD202's (opened them up among other things), plus quite a modded xtrememusic. Nice combo! Anyway, I'd like to build something more powerful which will give me some room to upgrade, since I'll probably upgrade to some HD650s in the future

I live in Argentina so it's difficult for me to find quality components, the only way to get decent opamps, rail splitters, buffers is through TI's free samples program. Right now some TLE2426s and BUF634s are on their way home, takes them three days to get from the USA to down here. I also ordered some OPA827s to try the "FET magic" I keep on reading around here... I already have some OPA1612s, which I've used in my CMOY and my x-fi. Beautiful opamp!

I've been reading your PIMETA schematics and I think it's just complicated enough that I, following the schematics, can turn them into a single layer PCB on my own. (The PPA is too complex for me to design and debug, I won't go that far right now...) This would be my second amp. I think I'll build the v2; but I see that you're using other buffers (the LMH6321, not the BUF634). Can I adapt BUF634s on the opamp outputs instead of the LMH6321, without ill effects on the circuit? If it's not possible, I see the v1 already has them on the schematics, and that'd be the way to go.

I've also read your notes on the V2, and it seems you've tailored it to use FET opamps like the OPA827 instead of bipolar opamps like the OPA1612. Would it be a safe bet to just use the 827s and not even bother with the 1612/pair of 1611s? What do you think? Thanks in advance.

 

[i_djoel2000]

Quote:

 
 
 
Are you certain you didn't put it in backwards?
 
you're right! i thought the longer leg was the cathode. silly me..lol
 
Probably not. I've never heard of a "stepped" trim pot. Without the part number, it's impossible to say more. (Bourns makes a lot of trim pot types.)
 

 
this one: http://my.element14.com/jsp/search/productdetail.jsp?SKU=9353186

 
i just checked it with my multimeter, the pot is still working. but there is no mark indicator to inform us on which impedance we are now. i have to change it manually using multimeter everytime i change them to know the exact value i have at the moment.
 
it's pretty annoying when you can't just play with them on the go, or without using multimeter. because you want the opamp to get to the maximum performance, yet you don't want to go to deep biasing them.
 
 
Quote:

i just checked it with my multimeter, the pot is still working. but there is no mark indicator to inform us on which impedance we are now. i have to change it manually using multimeter everytime i change them to know the exact value i have at the moment.

it's pretty annoying when you can't just play with them on the go, or without using multimeter. because you want the opamp to get to the maximum performance, yet you don't want to go to deep biasing them.
 

 

[Avro_Arrow]

Maybe you put the LED in backwards..?
The Trim Pot is 25 turns, not 25 steps.
That means you have to turn the adjustment
25 turns to go from the highest to lowest setting
and visa versa. The is a clutch in the pot so if
you turn it past the end it will not be damaged.
 
Follow Tangent guide on how to properly
set the trim pot.

 

[tangent]

Quote:

I live in Argentina so it's difficult for me to find quality components

You'll have to do the same thing we do here in the US: mail-order them.
 
 
Quote:

the only way to get decent opamps, rail splitters, buffers is through TI's free samples program.

 
That is simply not true:
 
DigiKey Argentina
RS International
Farnell Argentina
 
 
Quote:

Can I adapt BUF634s on the opamp outputs instead of the LMH6321, without ill effects on the circuit?

 
As you've noted, we used to use that in PIMETA v1. There are no changes in PIMETA v2 that prevent you from using this other buffer. The main reason to use the LMH6321 instead is that it sounds better.
 
 
Quote:

Would it be a safe bet to just use the 827s and not even bother with the 1612/pair of 1611s?

 
Yes.
 

 

[vodkex]

Thank you tangent! I'll keep those in mind next time I'm in the need of parts. On the other hand, I've got some more questions, if you don't mind. I've already designed the entire amp, one PCB for the PSU and another for the amp itself and its 3 channels.

What's up with RBIAS in the v2 schematic? I don't understand how should I wire it up in the PCB. I mean, it's a trimpot, one pin goes to ground, the other's the input and the middle one is the wiper. Am I right?



What should I do here? As I understand it, both the wiper (middle) and either left or right pins are connected (and shorted, maybe?) to v+, while the other remaining pin goes to RBLIM and then it connects to Q1's emitter pin. I think I don't have a use for the SBIAS switch in the schematics, so I left it out. What's the correct way to wire this trimpot into the PCB? Is it as I have described? I have a feeling it's not like that, so that's why I'm asking. If you can shed some light on this one, I'll be really grateful.

 

[tangent]

 
 
Quote:

both the wiper (middle) and either left or right pins are connected (and shorted, maybe?) to v+, while the other remaining pin goes to RBLIM and then it connects to Q1's emitter pin

 
That is indeed what the schematic says.

 

[fylde]

Hello,
 
 
First post here.
 
I have just received the board and parts for the PIMETA V2 and have completed the build. To my total amazement it worked first time with no problems and the DC offset seems well with the acceptable limits per the guidance notes. The amp sounds great and I have never attempted anything like this before.
 
One question, if anyone would be so kind, is regarding the class A biasing. I have followed the notes to activate this and have installed the relevant parts. If I adjust the trim pot I can't detect any difference when adjusting either way. I did have trouble with my soldering iron when soldering the Q1, Q2 and Q3 parts and I wondered, if I have killed any of these parts would the amp still work, allbeit without the class A biasing, or would I not be getting any sound at all?
 
 
Many thanks in advance.

 

[MisterX]

Did you install a jumper between the two through hole pads located below the RBLIM resistor?
 
http://tangentsoft.net/audio/pimeta2/bitmaps/brd-2.01.png

 

[fylde]

Thank you for the reply.
 
 
Yes I installed that jumper and also the resistor RBLIM. I realised the jumper could have been left out an I could have used RBLIM to span the gap but I have left the jumper and RBLIM as they are.

 

[tangent]

When you say you cannot "detect" a difference, do you mean you cannot hear it or you cannot measure it with a DMM?
 
If you just can't hear the difference, beware that class A biasing isn't a night-and-day kind of thing.  Some op-amps are so designed that you might get no benefit from additional biasing at all.

 

[fylde]

Hello Tangent,
 
 
Thank you for your reply.
 
 
My multimeter has now gone on the blink so I haven't measured the bias. I couldn't detect the difference when listening and from reading the notes I thought that I should detect a slight difference in volume and possibly quality / character of sound.
 
My main question is whether the amp would still operate at all even if I had killed any of the Q1, Q2 and Q3 parts? Or would it work as if the bias was disabled? Apologies if this is a question that cannot be answered given the limited information I have given.
 
The OP amp I am using for OPALR is AD823ANZ per the parts list and OPA134PA for OPAG. Further listening has shown me that this chip does sound a tad edgy with my Grados so I may try the other recommended DIP 8 OPALR OPA2132PA or even AD8620ARZ with an adapter socket.
 
I would like to add the unity bass boost also, but am unsure what type of capacitor should be used and what sort of range of capacitance I should be looking for?
 
 
 
Many thanks again.
 
 

 

[i_djoel2000]

Quote:

When you say you cannot "detect" a difference, do you mean you cannot hear it or you cannot measure it with a DMM?
 
If you just can't hear the difference, beware that class A biasing isn't a night-and-day kind of thing.  Some op-amps are so designed that you might get no benefit from additional biasing at all.

hi tangent, can you mention some opamps that benefit greatly from class A biasing? i'm now using between OPA827, LT1364, or LM4562 (depending on mood), and i couldn't really notice any difference when i play with the trimpot
 

 

[tangent]

Quote:

My main question is whether the amp would still operate at all even if I had killed any of the Q1, Q2 and Q3 parts?

You can kill a transistor in probably half a dozen different major ways.  If I remember my combinatorics correctly, that makes for 6 * 5 * 4 = 120 unique failure mode combinations.
 
So no, you haven't given enough information.
 
 
Quote:

I would like to add the unity bass boost also, but am unsure what type of capacitor should be used and what sort of range of capacitance I should be looking for?

 
See the bass boost docs for the PPA.  The same principles apply.
 
Quote:

can you mention some opamps that benefit greatly from class A biasing?

I've never bothered to sit down and compare op-amps purely in terms of "biasability".
 
Generally speaking, the lower the device's idle current, the more likely it is to benefit from a little extra bias on the output stage.  This is why the 13 mA Iq AD843 is famously among those chips not really helped by extra biasing.  It already has plenty designed into it.
 

 

[fylde]

Thank you for the reply.
 
 
 
I borrowed a multimeter and have measured the voltage going through RBLIM and this does change as RBIAS is adjusted. At the maximum I get 0.964 V. I do not have the knowledge to do any further tests on my PIMETA V2 build and the class A biasing so any further pointers would be gratefully received.
 
 
Many thanks.