Pimeta/Tread build questions and worklog
Sep 28, 2005 at 5:39 PM Thread Starter Post #1 of 61

darkisz

100+ Head-Fier
Joined
Aug 15, 2005
Posts
263
Likes
0
My audio DIY experience is limited to 2 Cmoys and some IC's. My soldering skills are comming along nicely. I've got all my parts, and I've read most everything on Tangent's site that relates to these 2 builds. However, my working knowledge of basic electric circuts is extremely limited, so I have a few general questions. I know I could find the answers by searching the forums some more, but I feel like I've been doing that for too long and now I just want to get my hands dirty and start this thing.

Details:
2xbuf634 on L/R, 1 on G
AD 843 for G
AD 823 or OPA2227 for L/R
Bias to class A - yes
LED circuts - no (not using batterys, so no point)
Alps RK27 pot
Caps from Tanget
Vishay Dale resistors
other parts per Tangents parts listing

On to the questions!

I plan to power the TREAD with an AC/DC 24v 100 mA wall transformer into THIS POWER JACK. I'm afraid that the jack might not be insulated, since I'd like to use a Hammond enclosure. Can anyone confirm my fears? If it is not isolated, would it be possible to isolate it with common household stuff? (electrical tape/plastic washers, ect) Does anyone see any other problems with this transformer/jack powering the Tread??

Is there a problem placing the TREAD right next to the Pimeta board in the case? I see some M3/PPA builds with STEPS that have metal shielding between the supply and the amp board. Since the transformer for my PS is in the wall plug, do I need to worry about interference with the tread being so close to the board?

I have already soldered all the parts onto the tread, it was a real pleasant experience working on PCB as opposed to the RS protoboard in the Cmoys. Here are a couple of shots of my handiwork. Let me know if you see any glaring errors, as I haven't done the testing yet
wink.gif
I have the Heatsink, but haven't installed it yet. I forgot to get some thermal paste, will Artic Silver work ok? I could probably dig up a few 'ketchup packs' of cheap white paste, if it matters.






I will probably use this thread as a worklog/help me thread, maybe it will prove useful for others in similar situations
 
Sep 28, 2005 at 6:25 PM Post #2 of 61
First, you probably don't even need the heatsink on the TREAD. But I think I see a problem with your chosen wallwart. Depending on the bandwidth setting that you chose for the buffers you might be asking for current supply problems. If set them to wide bandwidth mode, which you should be if you're going to bother stacking (IMHO) each one of the stacks might draw as much as 40mA! That means that just your buffers could be loading the power supply to it's max rated capacity, not good. If you haven't bought the buffers I'd suggest just going with singles and setting them to wide BW mode, that's what I run in my office.

As far as that power jack goes, it looks like there are provisions to isolate it but that it isn't an inherently isolated jack.

And no, there's no problem with having the TREAD in with the PIMETA, see the link in my sig for proof.

Nate
 
Sep 28, 2005 at 7:29 PM Post #3 of 61
Sep 28, 2005 at 8:26 PM Post #4 of 61
I knew I was missing something with the power (mA). Thanks Nate and Steve. I have since reread Tangent's 'Op-Amp Working Voltage Considerations'. My head is still spinning, so I need a bit more help.

Is bandwith determined by R11 value? If so, I was planning to run it with 100 ohm resistors on L and R, 220 on G since it isn't stacked. However, it seems I forgot to order any 220 ohm resistors (even for R11G)
frown.gif
So I have problems either way. Now, looking at Mouser, I can't find 220 ohm, but can find 215 and 221. Will either of these work?

I would like to stack the buffers, seeing as I already have them. It sounds like even with only 1 buffer per channel, I'd be pushing the limits of the 100mA wallwart. So, I need to pick up a new one. It looks like my Mouser choices are:

24v 400mA AC/DC
or
24v 500mA AC/AC

I'm leaning towards the AC/DC for the small advantage of less chance of ripple. I'll also pick the appropriatly sized snap in jack for isolation. Are these choices sufficient?
 
Sep 28, 2005 at 11:48 PM Post #5 of 61
Quote:

Originally Posted by darkisz
Thanks Nate


You're welcome.

Quote:

Originally Posted by darkisz
Is bandwith determined by R11 value?


Yup.

Quote:

Originally Posted by darkisz
Now, looking at Mouser, I can't find 220 ohm, but can find 215 and 221. Will either of these work?


The exact value isn't super critical, only that they're the same for both channels, so feel free to use either value. Your correct in assuming that you'd want to use different values for the L/R and G channels based on the stacking. Also, keep in mind that to keep all the buffers set up for a similar bandwidth you'll want to use as close to 2x the value for the ground channel.

Quote:

Originally Posted by darkisz
I would like to stack the buffers, seeing as I already have them. It sounds like even with only 1 buffer per channel, I'd be pushing the limits of the 100mA wallwart. So, I need to pick up a new one. It looks like my Mouser choices are:


If it were me I'd use the 400mA, the AC/DC will be much easier to work with and means fewer parts. And yes, 400mA should be more than enough. I think you'd be hard pressed to get a PIMETA pulling more than 250mA but don't take that as a challenge
smily_headphones1.gif
.

HTH,

Nate

P.s. If you ever get hung up needing a particular resistor value and don't want to order just one pack of resistors post a thread looking for the resistor and most likely you'll find someone (like me for example) who's got more than they'll ever need and will send you a few
wink.gif
.
 
Sep 29, 2005 at 3:36 PM Post #6 of 61
Quote:

Originally Posted by n_maher
Also, keep in mind that to keep all the buffers set up for a similar bandwidth you'll want to use as close to 2x the value for the ground channel.


It sounds like I should use 100ohms for R11 L/R and 200ohms for R11 Ground.

Thanks again, Nate.

I got some work done, most of the resistors are down, and the TLE. Here's some shots :







As always, and input is welcomed.
 
Sep 29, 2005 at 4:00 PM Post #7 of 61
I listened to my Pimeta last night (rolling op amps) The opa2604 sounded better than the opa2227. I believe it may also be $1.00 cheaper and I know there is a single version for the ground channel as well (Ijust didn't have it).
 
Sep 29, 2005 at 7:15 PM Post #8 of 61
Quote:

Originally Posted by darkisz
Is there a problem placing the TREAD right next to the Pimeta board in the case?


It's unlikely. Just don't do something silly like running the audio input wires next to the DC input wire.
 
Oct 3, 2005 at 5:59 PM Post #9 of 61
My parts came in, w00t!

Now it's time for more Q and A
icon10.gif


On the TREAD with the new wallwart, I'm getting 27v in, and around 26v out. I know this is too high, so I need to adjust with VSET. I've been turning the screw clockwise and it seems to be dropping the output voltage, however, it takes forever. I've tightened it about 50 turns, and I've only gotten the voltage down to about 25.7v, and it fluctuates +\- .05v. Is this normal? Might I be doing something wrong?

I double checked the resistor values for R1/R2. They are 120/2k, which should put my voltage in the 22-27v range, according to the graph on the schematic.

I guess I just need to keep twisting!
 
Oct 3, 2005 at 6:08 PM Post #10 of 61
Quote:

Originally Posted by darkisz
My parts came in, w00t!

Now it's time for more Q and A
icon10.gif


On the TREAD with the new wallwart, I'm getting 27v in, and around 26v out. I know this is too high, so I need to adjust with VSET. I've been turning the screw clockwise and it seems to be dropping the output voltage, however, it takes forever. I've tightened it about 50 turns, and I've only gotten the voltage down to about 25.7v, and it fluctuates +\- .05v. Is this normal? Might I be doing something wrong?

I double checked the resistor values for R1/R2. They are 120/2k, which should put my voltage in the 22-27v range, according to the graph on the schematic.

I guess I just need to keep twisting!




How many turns is your vset? 50 turns sounds awfully high (make sure you arent continuing to turn it after you hear the soft click). The mouser part number from Tangents part list is only a 25 turn (nom) pot. It sounds like something isnt right but all I can say is make sure all parts are placed correctly and of the correct value. I'm sure someone else will chime in with better advice.
 
Oct 3, 2005 at 6:12 PM Post #11 of 61
Part of what you're observating is most likely due to the fact that wall warts are generally rated "under load". With just the TREAD there's hardly a load on the power supply so it's outputting higher voltage than expected. I don't know how much (or how to tell) the output will drop once you attach it to your PIMETA but you can certainly measure it to see if it's acceptable.

HTH,

Nate
 
Oct 3, 2005 at 6:58 PM Post #12 of 61
VSET: 1 KΩ Bourns 3296W trimmer , from Tangent's kit.

I was, um, wearing headphones
rolleyes.gif
, so I probably missed any clicks. Don't tell me I killed it?

At least I had good results trimming my pot shaft
tongue.gif




Time to pay some attention to the Pimeta board. There are many parts yet to be mounted.
 
Oct 3, 2005 at 11:13 PM Post #13 of 61
Well, the board is finished. I just need to drop in the IC's and it'll be ready for a test drive. If that goes well, I'll start casing it up.

I would still like any reassurances that I'm not going to blow my 25W caps if the TREAD is actually putting out 25.7W. I trust n_maher in that once it has some load, the volts will come down, but if anyone else has info otherwise, now is the time to save me from frying anything
wink.gif


Pics, too.







Comments and input are always welcome.

Thanks to everyone that has helped me out so far!
 
Oct 3, 2005 at 11:20 PM Post #14 of 61
if you screwed the pot, you could always remove the pot and replace R2 with a 1.6kOhm Resistor then solder a wire like the gray line. That would net you about 23V which shouldn't make much difference in terms of the amp's performance. Alternatively, for better precision, get another 83ohm resistor and solder them in series with the 1.6k so that it adds up to 1683(around this value), that would give you very close to 24V
smily_headphones1.gif

pcb-v1.gif
 
Oct 3, 2005 at 11:33 PM Post #15 of 61
Quote:

Originally Posted by darkisz
VSET: 1 KΩ Bourns 3296W trimmer , from Tangent's kit.

I was, um, wearing headphones
rolleyes.gif
, so I probably missed any clicks. Don't tell me I killed it?

At least I had good results trimming my pot shaft
tongue.gif




Time to pay some attention to the Pimeta board. There are many parts yet to be mounted.



Well the click is there to tell you when you've reached the end of its travel in that particular direction (either at 0 or 1K ohms). I also recall that when I set up my TREAD, I was able to dial it down to the desired output voltage without having it loaded. It probably wouldn't be a good ideal to hook it up to the Pimeta and see if is the case of loaded/unloaded not allowing you to get it down to 24V (since you are probably using 25V caps). However, if your caps are rated for 25.X volts you should try hooking it up and then measuring vout.


Edit: Technically, there is a load on the output side (the LED and RLED).
 

Users who are viewing this thread

Back
Top