"Hello World" CMoy preflight check
Sep 14, 2008 at 1:36 AM Thread Starter Post #1 of 23

ashmedai

100+ Head-Fier
Joined
Aug 9, 2008
Posts
298
Likes
0
This is my first audio circuit & first soldered circuit, so please help me make sure my plans & part choices make sense. Hard enough to get your first attempt to go smoothly without any flaws in the design!
biggrin.gif


In order: design planning, parts list, questions, schematic.

Design Planning

I want a wallwart-only Cmoy design. To that end, I took the TLE2426 NR circuit from Tangent's site, and added back in the power switch and power LED. The amplification section is unchanged.

In the original power circuit, there are two 220uF caps in series, or an effective 235uF capacitance between the Vcc+/- rails. This is roughly equivalent to the capacitance in the adapted 2426 circuit I'm using. Therefore I assume the original range suggested for the power caps holds.

There are no voltage dividers in this circuit other than the gain control, which can bloody well be off by a few percentage points provided there isn't a major difference between the left and right channel. Assuming 1% metal film resistors and a gain of 11, R3 can be off by at most 10Ω and R4 by at most 100Ω. Note that any deviation by R3 and R4 in the same direction is canceled to the extent that the percentage overlaps, since only their ratio matters. Assuming worst-case opposing deviation, this gives a del+ gain of 11.202 and a del- gain of 10.802, or a difference of 0.400 given a gain of 11. Practically, a worst case deviation with 1% resistors will yield a 3.6% difference in gain between channels - but this is using conditions that are sufficiently improbable that I am willing to give it a go and fix it later if it comes up (a full 1% deviation is rated to occur once in 10000 resistors for the parts chosen, deviations further have a 50% chance of cancelling to some extent).

I plan to make the following substitutions:

1x C1: 220µF --> 470µF. Better power storage.
2x C2: 0.1µF --> 1.0µF. Reduces f 3dB- to ~1.6Hz, improves distortion.

I chose a 2V red LED and assume a target current of 1-2 mA. Therefore RLED should be 11-22 kΩ.

1x RLED: 11-22 kΩ --> Radio Shack 15kΩ or 2x10kΩ. Tolerance completely unimportant.
2x R2: 100kΩ. Tolerance affects f 3dB- matching between channels, but already using 5% caps which will have much larger impact on RC constant error. Tolerance reasonably unimportant.
2x R3: 1kΩ. Affects gain, but tends to minimize errors.
2x R4: 10kΩ. Affects gain, but tends to minimize errors.

Parts List

OPA2134PA op-amp for amplification (Mouser 595-OPA2134PA, 2/chip)
DIP socket for potential rolling (Mouser 575-113308)
TLE2426CP rail splitter, NR version (Mouser 595-TLE2426CP)
24V AC/DC Adapter (Mouser 412-124013)
2.5MM DC jack (Mouser 163-4024)
ALPS RK097 10kΩ potentiometer/switch (Courtesy of Tangent)
Volume Knob (Radio Shack 274-0403)
Power LED (Mouser 859-LTL-1CHE)
Vishay-Dale 1kΩ 1/4 watt 1%/100ppm (Mouser 71-RN60D-F-1.0K)
Vishay-Dale 10kΩ 1/4 watt 1%/100ppm (Mouser 71-RN60D-F-10K)
Vishay-Dale 100kΩ 1/4 watt 1%/100ppm (Mouser 71-RN60D-F-100K)
Vishay-Dale 15kΩ 1/2 watt 1% (Mouser 71-CMF5515K000FHEK)
Vishay/Sprague 470µF/50V/20%/85C electrolytic cap for C1, 12.5x20.0 mm, need one (Mouser 75-515D477M050DG6AE3)
Vishay/BC 1.0µF/100V/10%/85C metallized polyester film for CNR, 5.0x11.0x17.5 mm, need one (Mouser 594-2222-373-25105)
Vishay/Roederstein 1.0µF/250V/5%/100C metallized polypropylene film cap for C2, 10.5x18.5x26.5 mm, need two (Mouser 75-MKP1840510254)
3.5MM stereo jacks (Mouser 502-35RAPC4BH3)
*case*

Questions

Is the power circuit good for this purpose?

Is the 20-40 mA stability limit on the rail splitter an issue of any sort in this application?

Is there any preferred cap type for CNR? I assume the criteria are similar to signal caps, although perhaps a bit more loose (i.e. the nonlinearity of ceramic/tantalum caps might not be a major problem in this spot, shaving a couple bucks off the build cost).

Are my cap choices good for audio? Especially, what's up with all this Vishay/* branding (since Vishay seems commonly recommended for resistors, electrolytics, AND film caps)? Am I fine skimping and using a polyester cap for CNR (I notice the rail splitter is usually used in more complex circuits, where this is less of a concern, so CNR hasn't been experimented with much)?

Do I need to worry about insulated audio jacks since I'm using wall power? Tangent's notes on the subject tend to make me think yes, but I only see RCA jacks. I'd rather use 1/8" jacks. I could always find a non-metal enclosure if that's the best solution (an altoids tin is right out with just the cap choices, so...).

Are there any other tweaks I should look at before I place my Mouser order?

Is this cool or what?
biggrin.gif



I'm working on circuit board layout & enclosure options now.

Schematic

dccmoyschematicpj7.jpg
 
Sep 14, 2008 at 2:29 AM Post #2 of 23
I would suggest a different power supply section (I was to lazy to add the NR pin so feel free to DIY that), A regulated AC adapter (or maybe adding a regulator to the PSU section) and a Isolated DC jack if you are planning on using a metal enclosure.
 
Sep 14, 2008 at 2:38 AM Post #3 of 23
Comments:
  • You may need more than 1 DIP socket since the TLE you chose is also a DIP-8 package. Not sure why you are using this instead of the TO92 package.
  • Isolation: you might want to isolate both the power jack and the 3.5mm jacks from the case. A metal case is useful for the shielding, so I think isolated jacks are a better solution than going with a non-metal case.
    Power jack: 163-4303-EX
    3.5mm jack: 161-7300-EX (I like these better than the Switchcraft anyhow).
  • The Vishay branding is because they have been buying suppliers left and right for the last decade or longer. When people say they are using Vishay, they need to be a bit more specific.
  • Not sure what phones you are planning on using, but the stock gain of 11 is excessive with almost all phones. You might get some alternate R3s, such as 2.2K, etc.
 
Sep 14, 2008 at 3:01 AM Post #4 of 23
I'm not terribly attached to metal if helps not to use it. I think most of the cheap options are plastic anyway.
biggrin.gif


I'm having trouble finding isolated DC jacks and regulated AC/DC adapters, so it might need to be plastic. Should I still add the extra caps between V+/V- and Vgnd?

What about just cribbing the PIMETA power section? I'd need more caps and the diodes, but I'm not exactly worrying about weight for this application. I do want to keep it relatively cheap/simple since I'm new at this (also it's a CMoy).

pimetapowersectionvg8.jpg
 
Sep 14, 2008 at 3:10 AM Post #5 of 23
Quote:

Originally Posted by Pars /img/forum/go_quote.gif
[*]You may need more than 1 DIP socket since the TLE you chose is also a DIP-8 package. Not sure why you are using this instead of the TO92 package.


DIP-8 has NR, which I figured was good for some laughs. I figured on only getting a socket for the op-amp since the TLE will be a permanent part of the circuit while I might swap the op-amps. Will I still want a socket for the TLE? I imagine it might be easier to solder than the IC.

Quote:

Originally Posted by Pars /img/forum/go_quote.gif
Isolation: you might want to isolate both the power jack and the 3.5mm jacks from the case. A metal case is useful for the shielding, so I think isolated jacks are a better solution than going with a non-metal case.
Power jack: 163-4303-EX
3.5mm jack: 161-7300-EX (I like these better than the Switchcraft anyhow).



A very good point about the shielding, especially since it will be on a desk near monitors and various other computer bits instead of in a pocket. Thanks for the part numbers, I've swapped those out. It's also nice that the phono jack is cheaper.
biggrin.gif


Does this solve the power section, or should I still look at adding more caps in on the right and/or using the PIMETA power layout?

Quote:

Originally Posted by Pars /img/forum/go_quote.gif
The Vishay branding is because they have been buying suppliers left and right for the last decade or longer. When people say they are using Vishay, they need to be a bit more specific.


Electronics Borg!
eek.gif


So is Vishay/* good in general, or is it just the Vishay/Dale parts that I hear about more often that are of interest?


Quote:

Originally Posted by Pars /img/forum/go_quote.gif
Not sure what phones you are planning on using, but the stock gain of 11 is excessive with almost all phones. You might get some alternate R3s, such as 2.2K, etc.


Thanks for the heads up, I'll do that. It's for my Sony MDR-V6 (63 ohms impedence at 1 kHz, 106 dB/mW sensitivity), which aren't as friendly about sources as my ATH-A900s.

I thought about using a 10k pot for R4 - but keeping it internal-only and tuning it with the power off until the volume range is appropriate to the particular phones. I read a few things about variable gain being a tricky proposition, but I think most of these were looking at either switching between several settings or having an externally accessible pot for online gain adjustment. To my knowledge, there's no electrical difference between a pot and a static resistor...so am I fine just plunking down a 10k pot on the board for R4 provided there's no external access and I confine it to offline tuning? If this is bad, why?

Decided the gain pot is more trouble than it's worth, and PCTS I can always use a little practice with desoldering.
biggrin.gif
 
Sep 14, 2008 at 3:13 AM Post #6 of 23
You choose another power supply. 100mA current spec of the power supply you listed above is bit weak. Any adaptor rated at 12V~18V/300mA or above will work well, and everyone seem to have one or two laying arround.

I do not think TLE makes any noticable difference at sound or stability of the amp.

If you use metal enclosure, you need to isolate power jack and/or 3.5mm jack from the enclosure for they share common ground. They must be isolated from each other.

Gain of 11 is excessive at most cases. Use 2.2K or 3.3K for R3. Visahy RN series are recommended not because they are something special, but because they are readily available, look good and not too expensive.
 
Sep 14, 2008 at 3:35 AM Post #8 of 23
Quote:

Originally Posted by Seaside /img/forum/go_quote.gif
You choose another power supply. 100mA current spec of the power supply you listed above is bit weak.


Tangent's site says that the TLE goes unstable at 20-40 mA, and that 100mA is more than plenty for the circuit. Notably, I'm modifying it to take 24V...but if it really needs more than 100mA, I probably need to ditch the TLE.

Quote:

Originally Posted by Seaside /img/forum/go_quote.gif
Any adaptor rated at 12V~18V/300mA or above will work well, and everyone seem to have one or two laying arround.


I looked, oddly I couldn't find any that weren't actually in use.
eek.gif


(Now that I say that, I actually found a 9V/0.7A adapter...was hoping for a bit more voltage, though?)

Quote:

Originally Posted by Seaside /img/forum/go_quote.gif
I do not think TLE makes any noticable difference at sound or stability of the amp.


Hmm...well I got the idea to try the TLE instead from this article:

Virtual Ground Circuits

Should I just go with the standard power circuit instead? It would eliminate the current limit, assuming the stability issue can be neglected. In this case, I would probably go for some 0.1% resistors for the divider and keep 470uF caps.

Quote:

Originally Posted by Seaside /img/forum/go_quote.gif
Visahy RN series are recommended not because they are something special, but because they are readily available, look good and not too expensive.


Yeah, I figured they were just the cheapest ones which are well-behaved. That's the main call "Black Gate" caps used to have, after all - cheapest readily-available part that does the job properly.
biggrin.gif


Quote:

Originally Posted by MisterX /img/forum/go_quote.gif
Regulated AC adapter = https://www.alliedelec.com/Search/Pr...px?SKU=8791080


Meep! Um, how badly does it need to be regulated? Figure $30ish shipped...that's more or less the entire project budget for a CMoy usually.

Quote:

Originally Posted by MisterX /img/forum/go_quote.gif
Isolated DC jack = 163-1100-EX


Thanks. Should I care about the 12V rating on the jack?
 
Sep 14, 2008 at 4:19 AM Post #10 of 23
Okay, I found a 19V/1.6A supply in an old parts box. I have no clue what this thing went to, therefore it's fair game. Any way to tell if it's regulated? 6.5/3.0mm jack, which is fun.

potranset7.jpg


Looking into the TREAD circuit...

Wait, is that thing set up to take AC in? Well, not what I want to do for my first adventure in soldering - I'd rather build a few things before I fry myself playing with AC. But fun for another time. Or am I interpreting the schematic incorrectly?

With DC-input, would this be the way to modify it? Modification based on "Choosing a Power Source" here. I notice this leaves no caps over 22uF...is that okay given it's using wall power, or did I screw it up?
biggrin.gif


treadqh0.jpg
 
Sep 14, 2008 at 5:06 AM Post #11 of 23
Yes, the power supply above is regulated and DC. You can use it with your Cmoy with no problem.

It's simple amp, don't worry too much about it. Your selection is ok. Make one first, and dip yourself into complicated circuit... later.
smily_headphones1.gif


**
As for your question...
TLE gets instable at 20~40mA has nothing to do with power supply.
TLE is rail spliter that can handle +80mA/-80mA. It splits the input voltage into two exactly the same voltages with half the value of input voltage, and its max current would be 80mA each. If 18V DC is input voltage, TLE will split it to +9V and -9V up to 80mA each. But what if the split voltages are uneven due to the problem at other part of the amp? TLE will try to fight this difference, but it is not able to handle if the current difference b/w two rails is more than 20~40mA and become instable. That's what the notion means. Normaly, you don't have to worry about this if all other parts of the amp are correctly selected and assembled.
 
Sep 14, 2008 at 5:30 AM Post #12 of 23
Cool, I'll start looking for an appropriately sized jack to fit it then. How do you tell it's regulated, for future reference? I couldn't find a spec sheet on it anywhere, so I assume that's not how you found out (unless your google-fu is just very strong
biggrin.gif
).

Thanks for the clarification on the TLE. As you concluded, I had the impression it was the total current it could handle and not the differential.

Changes from OP at this point:

Better phono jacks.
19V regulated supply found
Adding 100uF electrolytic caps between V+/Vgnd and V-/Vgnd at right of virtual ground circuit.

Need to find Isolated 6.5/3.0 mm DC jack.

Note that RLED now needs to be 8.5k to 17k because of the changed supply voltage.

Revised schematic:

dccmoyschematicmodifiedll5.jpg


Found that the power supply apparently went to an LCD. Off to search my dead LCD pile...maybe I'll find a jack. No 6.5/3.0s on Mouser that I could see, so hopefully it won't be problematic. I might hunt for a metal enclosure with plastic end-plates similar to one of the Hammond options for the Mini3 if the only DC jack I can find for it isn't isolated.

Bingo...it went to a crappy Rosewill LCD I'm not using anymore. I wonder if I should just desolder the jack and junk it for parts...
biggrin.gif
biggrin.gif


Digikey has one: http://search.digikey.com/scripts/Dk...ame=CP-048H-ND
 
Sep 14, 2008 at 5:57 AM Post #13 of 23
That is a switchmode power supply and that means it is about as useful as an unregulated power supply is. Err, actually. If it is not isolated it even less useful because it will just short circuit the TLE
 
Sep 14, 2008 at 6:13 AM Post #14 of 23
Doh. Well, I guess that saves me getting the unusual jack. How are you guys figuring this out, though? Better spec sheets? Something about the symbols on it? The fact that it was for an LCD?

Honestly, I don't want to blow $30 on a supply for this thing if I can avoid it. I'd rather take the money and put it into my next project once I can solder parts without setting my hair on fire. So...

Is that circuit from #10 a correct regulator given a DC source? Would I connect it to the V+/V-/Vgnd points at the output of the power section (i.e. vgnd >> regulator >> amp)? I assume that's where it will do the most good. Would I take out the 100uF caps in this case? I figure if I can get the regulator working, that's about another $3 in parts instead of $30. But boy howdy am I going to get a lot of practice soldering down ICs...somehow I think that I'll be appreciating the ESD-safe iron.
 
Sep 14, 2008 at 3:09 PM Post #15 of 23
I gave you the wrong DC jack part number... use the one MisterX provided instead as this one is isolated. For wallwarts, Jameco's Reliapro series are quite good and cheap linear regulated. The Elpac that MisterX provided the info for would be good as well.

The Tread circuit will take either AC or DC. You could eliminate the rectifier bridge, etc. for DC only if you wanted.
 

Users who are viewing this thread

Back
Top