[mendiola]

I have uploaded the new schematic and the board of the CMD v1.0
 
CMD: Cmoy Multiloop Desktop.
 
https://www.dropbox.com/s/tsuml8vn4e9aqdy/CMDV1.png
 
 
https://www.dropbox.com/s/z9cjbjtp9n59gw7/CMDV1.rar
 
Regards.
Alfredo Mendiola Loyola
Lima, Peru

 

[alel]

Hi is their anywhere I can get the pcb alone?

 

[3l3tric]

Question:
I'm about to start building a PIMETA (I've had the PCB for a while, just haven't gotten around to building the thing yet), and I had one question for people who've already built their's:
Has anyone used the OPA627 OpAmp for a battery powered PIMETA? I use a tube amp with my desktop setup, and I'm afraid I won't like the sound of the Analog Devices opamp due to its brighter nature (At least, according to tangent).
Right now, I'm considering building the Pimeta using 2 9V batteries in series for an 18V supply for the OPA627, but before I pulled the trigger, I wanted to know if anyone else has used a similar setup and how successful it was.
Thanks!

 

[tangent]

  Has anyone used the OPA627 OpAmp for a battery powered PIMETA?

 
The PIMETA is primarily designed as a battery-powered headphone amplifier, so when the docs tell you that the OPA627 is known to work, well, what more did you want?
 

I use a tube amp with my desktop setup, and I'm afraid I won't like the sound of the Analog Devices opamp due to its brighter nature (At least, according to tangent).

 
You need to listen to at least one of the chips with a lengthy review, then use that to calibrate what I wrote about that chip with what you hear. Only then can you hope to make good buying decisions based solely on my reviews.
 
It's really important here, because you're talking about ~$26 x 3 = $78 worth of chips. If you don't have any of the reference chips' sounds in your head, I'd think you'd want to start with something a little more reasonable.
 

I'm considering building the Pimeta using 2 9V batteries in series for an 18V supply for the OPA627

 
The docs already tell you that alkalines aren't powerful enough for this, and it isn't good to stack so many NiMH cells in series. (The more cells in series, the greater the risk of cell reversal if you let the battery voltage drop too far.)
 
If you're just itching to use the 2-battery input feature on the board, have you considered using two 4xAAA cell holders? That's not a lot bigger than a single "9V" NiMH, but will run a lot longer, and only strings 8 cells in series, which is what you get in the 9.6V flavor of NiMH "9V" battery.
 
If you're basing the decision to start with 18V on my clipping tests, you're misapplying the results. Those clipping tests were done with the op-amp directly driving headphones, which is not the case in the PIMETA, or in fact in any of the headphone amplifiers on my site other than the CMoy. In all the other cases, it is the buffers that carry the load, and there is no virtual ground shift to contend with. Hence, the op-amps will work to lower voltages in a PIMETA than the values I give for a CMoy Pocket Amplifer.

 

[3l3tric]

Is the PIMETA V2 site not working for anyone else?
Everything on Tangent's site seems to be giving a 403: Forbidden error.
I found the site on the Internet Archive, but I'm curious as to if the site is working for anyone else.

 

[tangent]

Sorry about that...file permission bug. It's fixed now.

 

[vixr]

  Sorry about that...file permission bug. It's fixed now.

yay!!!    thought I had been banned!

 

[omonoid]

tangent, can you help me troubleshoot? 
 
Notes:

1. Omitting buffers (for now)
2.     - omitted R6's
3.     - jumped R5's
4.     - jumped buffer input to R7 (see pic to confirm i did this right)
5.     - omitted C6g
6.     - installed C6 R/L (not sure if needed??)
7. Omitted input caps
8. Omitted R11G/R/L
9. Jumpered LED2 
10. LED1 hole damaged so soldered direct to RLED
11. Installed diodes (D1, D2)
12. op amp: OPA2132P
13. Configured for bias always-on
14. Used LM317 so i can trickle charge
15. Q2/3L/3R used 2N5088
16. Q1 used 2N5087
17. R1G is currently 4.1k and is two resistors in parallel since i forgot to order the right size

 
Testing:
Used a 9V wall wart connected to the WALL molex 2 pin connector (reads 9.1V before connected)
When connected but with the ALPs vol toggle OFF the voltage at the WALL molex pins reads 8.3V (why is there a V drop when it is off???)
When the ALPs is switched ON, the voltage at the WALL molex reads 8.06V and the LED1 turns ON
In attempting to "check the voltages at all of the DIP-8 chips’ supply pins relative to input ground (IG)...should be about half the PS voltage..." I find that all pads for both opamps (not yet installed) read 0V except pin 5 for one and pin 6 for the other, which read 8.57V
 
This is where i have stopped because it seems something isn't right. I went through and checked all my solder connections for shorts or discontinuities, but didnt see anything
 
I'm hoping since you have a much better understanding of the circuit, you could help me resolve this.
 
I uploaded some pictures as well (see link to google drive since i dont have permission to post them here). Hopefully the pics plus the notes above can give you an idea of where i am and what may be wrong.
 
 
Pictures:
https://drive.google.com/folderview?id=0B_Z_i1f1IePDeVlxTG1VM3F6N28&usp=sharing
 
 
Thanks,
 
Jake

 

[tangent]

 
jumped buffer input to R7 (see pic to confirm i did this right)

 
I think it would have been simpler to just jump from pin 3 to 6 on each BUF SO-8 footprint, but buzzing it out on a bare board here says the pins you chose instead do also connect to the same locations.
 
Your use of bare wires for jumpers is a worry. You have to be very careful that they're not touching anything. Where I have to jump over exposed metal, I prefer to use insulated wire for jumpers, stripping it at the tips only.
 
A much bigger problem is how messy that soldering job is. Carbonized flux can be conductive, flux can trap conductive debris, and flux can conceal solder bridges. Even if it ends up not helping your immediate symptom, a good cleaning will at least add a touch of pride to the build.
 

1. LED1 hole damaged so soldered direct to RLED

 
Have you watched my soldering videos?
 

Used a 9V wall wart  

 
Unless your battery supply is going to be something marginal like 4 x AAA NiMH, that's not enough voltage to charge the battery. The charge controller section of the docs should have made this clear.
 
That doesn't explain your present problems, but it's a risk of problems down the road. There is a whole section of the docs on choosing a power supply for the amp, which your design choices seem to ignore.
 

connected to the WALL molex 2 pin connector

 
That's not what your pictures show. The orange/white pair is connected to BATT.
 
Why are you using Molex connectors but soldering directly to them, anyway? You should be doing it one way or the other: hard-wire it or use connectors. If the orange/white pair is just for testing, I'd still make up a female Molex KK adapter. If you continue doing electronics, you may reuse that adapter later.
 

why is there a V drop when it is off???

 
D1 and IC1.
 

When the ALPs is switched ON, the voltage at the WALL molex reads 8.06V

 
That means you're either using an unregulated power supply or it's too weak for the job. I highly doubt the latter, unless you've shorted something. The existing troubleshooting guide tells you how to test the board's current draw, which will tell you if your board is drawing too much power for your wall wart.
 

In attempting to "check the voltages at all of the DIP-8 chips’ supply pins relative to input ground (IG)...should be about half the PS voltage..." I find that all pads for both opamps (not yet installed) read 0V except pin 5 for one and pin 6 for the other, which read 8.57V

 
I'm guessing this is due to your R1G hack. It appears you've misread the board labeling for the scratchpad area, thinking the two long strips are IG. The docs make this clearer: there are only two IG pads in the scratchpad area. Using them for R1G is a bad idea. Just piggyback the two resistors on the existing R1G pads: solder one in place, then solder the other on top.

 

[omonoid]

Thanks for taking the time to give feedback.
 

  I think it would have been simpler to just jump from pin 3 to 6 on each BUF SO-8 footprint,

Ill do that since i want to redo my jumps with insulated wire
 
Quote:

 I prefer to use insulated wire for jumpers, stripping it at the tips only

will do
 

 Have you watched my soldering videos?

No, 'll check them out. I kinda just whipped this up with tools and parts I had, so I definatley could have done better with more ideal tools and planning
 
Quote:

  a good cleaning will at least add a touch of pride to the build

good point. isopropyl alcohol should clean it up right?
 
Quote:

 Why are you using Molex connectors but soldering directly to them, anyway? 

 
The orange/white soldering on the BATT was just a quick and dirty way to test with a battery, but it turns out I dont have any fully charged 9V batteries, so I used a molex adapter with the two pin WALL to connect to my wall wart. I should get a 4 pin molex to use with the BATT connector too, but i dont currently have one.
 

 I'm guessing this is due to your R1G hack. It appears you've misread the board labeling for the scratchpad area

Hmmm ill check it out
 

 The existing troubleshooting guide tells you how to test the board's current draw

Wasn't aware of this, i'll go through the steps after I fix everything else addressed above
 
I haven't put much thought into the trickle charging. I will definitely reread the appropriate sections before i ask more on that topic. My priority is to get it working first. 
 
About my op amp selection (OPA2132P) I had these already from a previous project that i didnt end up pursuing. After reading your op amp selection guide i determined ( a few months ago) that these would work (can't remember my exact logic). Do you agree that these are an ok selection for my build? If I decide to add the buffers later will they still be ok?
 
I get that the buffers help take some of the work from the op amps, but i am a bit shaky on what specs of the op amp make it ideal for not needing buffers. I would assume it should be an op amp capable of not saturating at the requested headphone vol, and also be somewhat low power consumption. I'll re-brief myself on op amps in the meantime. 
 
I bought components and planned this out like 6 months ago (and took minimal notes) so i am still trying to remember why i made certain part selections. 

 

Ill report back with what I get after I address everything above (probably not til next weekend)
 
Thanks again for helping me out. I'm learning alot from this project
 
--Jake

 

[tangent]

  isopropyl alcohol should clean it up right?

 
Tangent Tutorial #05.
 

About my op amp selection (OPA2132P) I had these already from a previous project 

 
Why the plural? ("These")
 
You should be using one dual-channel op-amp (e.g. OPA2132) and one single-channel op-amp (e.g. OPA132). If you were hoping to use two OPA2132s, it's not going to work.
 

Do you agree that these are an ok selection for my build? If I decide to add the buffers later will they still be ok? 

 
They're barely usable without the buffers, at 9V. See my clipping test results.
 
With the buffers, you might be able to drain a 9V battery dry before the op-amps start clipping.
 
Also keep in mind that a NiMH "9V" will only be at or above 9.0 V briefly, if ever.
 

i am a bit shaky on what specs of the op amp make it ideal for not needing buffers. 

 
You're not going to find such a spec in a datasheet. The best you can hope for is a front-page marketing claim that it's made for driving headphones directly, and you should take that with the same size grain of salt you use prophylactically for dealing with all marketing claims.
 
This is why I did the clipping tests in my op-amp review article. It provides information you could not get before, short of doing the test yourself.

 

[omonoid]

 You should be using one dual-channel op-amp (e.g. OPA2132) and one single-channel op-amp (e.g. OPA132)

 
You are right. They I mistook them as the same when quickly reading off the tiny label on the op amp. Like i said i gathered these components a while ago, so i am re-famliarizing my self the the design.
 
It sounds like leaving the buffers out is not as good of a long term idea as i originally gathered. I was hoping to optimize it for portable use, but after reviewing the "power supply matters" I'm thinking about ditching batteries for now and just trying to get it working on off a 9 or 12V supply first.
 
 
Could powering off the usb bus be done effectively if the voltage were boosted?  
 
The more I think about how i will use this amp, the more i realize having it portable isnt really a priority. Having it powered of my pc could be useful because that is my main music player.
 
Do you think i should try to get it working as is, then add buffers, or purchase buffers then try to finish assembling the amp? I'd rather stick with the op amps i already have, and it sounds like buffers are optimal for my config.
 
Thanks,
Jake

 

[tangent]

  Could powering off the usb bus be done effectively if the voltage were boosted?  

 
http://www.head-fi.org/t/386517/usb-powered-pimeta
 

 Do you think i should try to get it working as is, then add buffers, or purchase buffers then try to finish assembling the amp?

 
A PIMETA without buffers is a glorified CMoy pocket amp. Is that what you were after?

 

[omonoid]

Thanks for all the help. Now that i am on winter break (im an undergrad ee student) I have had time to actually learn about the circuit and have decided to keep the OPA2132/132 and add buffers and modify some values to best suite my 36 ohm impedance headphones. I also have access to campus bench testing gear and multisim, so i might as well make use of it and optimize components. I plan on eventually adding trickle charge and other tweaks, but want basic functionality first. I ended up desoldering the buffer jumpers and some other things i want to change. I am debating whether I should start with a fresh pcb, or use the possibly damaged one that I have. I am still learning about the theory of the amplifier, and for fun have started drafting up a spice (multisim) model to try to visualize how changing certain values affects the signal. I was wondering if you had an already made model that I could compare to. In the meantime i am reading all the content on your site, which is proving to be very useful, so thanks for organizing it.

 

[tangent]

  I am debating whether I should start with a fresh pcb, or use the possibly damaged one that I have.

 
Are you considering doing your own layout, or were you hoping to get another board from me?
 
The first option is of course open to you, and may make an interesting project, particularly if you do something a bit different with the circuit design or layout. A simple clone is boring. The main argument against it is the cost, particularly if you make yours a 4-layer board, as the official boards were.
 
As for the second option, original unpopulated PIMETA v2 boards are no longer available for sale, and haven't been for nearly a year.
 

I was wondering if you had an already made model that I could compare to. 

 
No.
 
I only use circuit modeling programs as a glorified calculator. I can also see a use for such programs in IC design, where it is simply too expensive to just go try an idea.
 
Neither situation really applies to the PIMETA v2. Calculations are either basic applications of Ohm's Law or elementary op-amp theory, and it is neither terribly expensive or time consuming to just try things with it.
 
If you think modeling is fast, it's probably because you aren't doing it right. I'll bet you haven't modeled all the parasitics, or swept the model over input offset voltage, or applied ripple to the power rails, etc.