CMoy using Joshatdot PCB - With Parts List
Oct 18, 2007 at 10:12 AM Post #31 of 115
Here's a question I've been sort of hinting at throughout the thread...

If the proportion between the resistors controls gain, and the input capacitors eliminate incoming DC offset (if these caps are in place), then what does the value of the resistors control?

There are a lot of different resistor values that end up with a gain of 6, and I understand that lower-valued resistors mean less noise, but is there a reason that I would want higher-valued resistors?
 
Oct 18, 2007 at 3:21 PM Post #32 of 115
The resisters balance the input bias currents. If they aren't balanced, you get small oscillations as the opamp reacts to it and then tries to balance one leg to the other. The faster the opamp, the more likely this will become an issue. Also, if you roll in a bipolar opamp, the bias and offset currents are higher, making DC offset a significant issue.

For the adapter, it looks expensive coming from some international place. You could just use any old adapter (I use Brown Dog, but I think Tangent sells them individually instead of 5 at a time) and run a cap from V+ to V- on the adapter. You can use 399-4264-ND. Keep in mind of the voltage also, if you decide to go to 24V or higher.

For small builds, and 1 or 2 items of things, I prefer Mouser over Digikey...unless you plan to stock up.
 
Oct 18, 2007 at 4:03 PM Post #33 of 115
Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif
Here's a question I've been sort of hinting at throughout the thread...

If the proportion between the resistors controls gain, and the input capacitors eliminate incoming DC offset (if these caps are in place), then what does the value of the resistors control?

There are a lot of different resistor values that end up with a gain of 6, and I understand that lower-valued resistors mean less noise, but is there a reason that I would want higher-valued resistors?


As holland noted, the choice of resistors somes down to balancing the inputs of the opamp. With input coupling capacitors, and the need to make R1 (different references for some amps, but the resistor normally from the input to ground) 10 times the pot value or better, you are forced into rather high resistor values for the feedback loop. For example, the original PINT needed somewhere in the neighborhood of 768K or so for the feedback resistor, resulting in noise. In the mini3-ified PINT, these were pulled down into <10K range IIRC.

The input capacitor's effect is more of insulating you from the DC variability of the pot in parallel with R1 (i.e. 0K-10K || 100K = ~0K - 9.xK input resistance). With input capacitors, from the opamp input perspective, the impedence looking out from the non-inverting input is 100K fixed.
 
Oct 19, 2007 at 11:51 AM Post #34 of 115
Thanks guys, that makes it a little clearer. Without input capacitors, is there some way to calculate the minimum resistor values I'll need for a particular op-amp to avoid oscillation? Or is the 100K resistor at R2 essentially the majority of the resistance I'll need?

I probably shouldn't be so paranoid, but I definitely don't want to fry a pair of headphones for the sake of some extra sound quality.
tongue.gif


I think that for the first build, for the sake of simplicity, I'll try using the OPA2228P. I've read that this chip needs gains of 5 or more for stability, so a gain of 6 should be okay. The calculated DC offset is below 1 mV, so I'm assuming it should be okay there as well.
 
Oct 19, 2007 at 4:40 PM Post #35 of 115
Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif
Without input capacitors, is there some way to calculate the minimum resistor values I'll need for a particular op-amp to avoid oscillation? Or is the 100K resistor at R2 essentially the majority of the resistance I'll need?


Use the Mini3 values or values I and others posted. Those are values without the input cap. 100K for R2 is standard as you need a minimum of 10x the pot. In place of C2, place a resistor of which has been posted. Mini3 value is 330. My personal CMoy value is 825. There were other values listed as well. The R in place of C2 is calculated as the parallel value of R3 and R4.
 
Oct 19, 2007 at 4:56 PM Post #36 of 115
Quote:

Originally Posted by holland /img/forum/go_quote.gif
In place of C2, place a resistor of which has been posted. Mini3 value is 330. The R in place of C2 is calculated as the parallel value of R3 and R4.


Gotcha... I must have misinterpreted. So, if I chose to use R3=330 and R4=1.5K, I would need another set of 330 resistors (or more accurately, ~270) to use as C2, not the 1K as previously mentioned.
 
Oct 23, 2007 at 5:58 AM Post #37 of 115
Does anyone stock the OPA2107? I can't find it anywhere...

Here are the op-amps I'm considering rolling with the current setup. Please tell me if I risk damaging anything by using these with the setup listed in the first post (
smily_headphones1.gif
):

Burr-Brown OPA2132P
Burr-Brown OPA2228P (better than 2227?)
Burr-Brown OPA2107AP

National LM4562NA

Am I missing any great DIP-8 chips, for example from Analog Devices? (AD823 maybe?)
 
Oct 27, 2007 at 8:50 AM Post #40 of 115
I just submitted my Digi-Key order, so hopefully everything will arrive soon.
biggrin.gif


And now I've submitted my order with amb. Exciting!
580smile.gif


Does anyone have some spare solder or flux that I could borrow/purchase? Anything better than Radio Shack's offerings will do. I've heard that the flux can help quite a bit with less-than-perfect soldering irons and 2% silver solder.
 
Oct 27, 2007 at 3:29 PM Post #41 of 115
Radio Shack solder is fine. People have their preferences, but it's fine. There's not much to solder. Radio Shack flux is fine too.

I have spare Kester 63/37 solder and 44 flux, but it's for me. No sharing.
smily_headphones1.gif


Why silver solder? Flux isn't going to help if your iron isn't up to snuff as you need higher temps to work with silver. Radio Shack silver solder flows easier though.

So...you mentioned AMB, you're not building a CMoy any more but a Mini3? Or are you building both?
 
Oct 28, 2007 at 4:23 AM Post #42 of 115
How did I miss this thread! I guess I been working too much, and not surfing HF enuff
eggosmile.gif


I don't have any opinions on most everything here, but I do for the RLED value. The LED in my BOM (350-1560-ND) has a forward voltage of 3.5v, current draw of 10mA (or 20mA if you look at DK's Catalog PDF).

LED Calculator says if source voltage is 18v (dual 9s) and LED forward voltage is 3.5 and current draw is 10mA, the RLED should be 1.5k ohm. If source voltage is 9v, then the RLED is 560 ohm. All that is to achieve 12.5 mcl brightness @ 10mA current.

In my CMoy's I've always use dual 9s, 18v, and with the LED in my BOM 2k is plenty bright. I tried a 10k with that LED, and it was not visable at all. And with a 4.75k it was very dim.
 
Oct 28, 2007 at 4:49 AM Post #43 of 115
Quote:

Originally Posted by holland /img/forum/go_quote.gif
Why silver solder? Flux isn't going to help if your iron isn't up to snuff as you need higher temps to work with silver. Radio Shack silver solder flows easier though.

So...you mentioned AMB, you're not building a CMoy any more but a Mini3? Or are you building both?


I'd planned to use RS's silver bearing solder because it comes in a nice small gauge. My iron isn't the best (a Weller I bought for ~$18 at Lowe's), so my solder tends to blob.

I needed a few Alps potentiometers, since that's what Josh's board is designed to accept. I also bought my railsplitters, a few DC input jacks, and an AD8397 SOIC for another time.

Quote:

Originally Posted by Joshatdot /img/forum/go_quote.gif
I don't have any opinions on most everything here, but I do for the RLED value. The LED in my BOM (350-1560-ND) has a forward voltage of 3.5v, current draw of 10mA (or 20mA if you look at DK's Catalog PDF).

LED Calculator says if source voltage is 18v (dual 9s) and LED forward voltage is 3.5 and current draw is 10mA, the RLED should be 1.5k ohm. If source voltage is 9v, then the RLED is 560 ohm. All that is to achieve 12.5 mcl brightness @ 10mA current.

In my CMoy's I've always use dual 9s, 18v, and with the LED in my BOM 2k is plenty bright. I tried a 10k with that LED, and it was not visable at all. And with a 4.75k it was very dim.


Thanks, I'll definitely keep that in mind. I bought extras of a few resistor values and I have a few left over from my first build, so I should have a few choices. I've heard that placing a higher-valued resistor in the RLED position can add battery life, since the LED takes some power to operate.

I noticed that the blue LED you specified in your parts list was $2.17, while the ones Tangent had in his list (160-1708) were $0.09. Is there a difference between them? I see that his has a lower candela rating...
 
Oct 28, 2007 at 7:35 AM Post #44 of 115
Quote:

Originally Posted by holland /img/forum/go_quote.gif
Radio Shack solder is fine. People have their preferences, but it's fine. There's not much to solder. Radio Shack flux is fine too.


Solder has a wide range of melting temperatures, and some kinds are eutectic. It's definitely not a case of "solder is solder."

I tried a couple kinds of Radio Shack solder...I found it pretty difficult with all of them to make a nice clean shiny joint. I've tried other kinds that were plain awful to work with.

Then I got some Welborne Labs solder...wow, that stuff is nice. It flows well, sets nicely and always has a nice shiny appearance--even using a dirt-cheap 15w iron. It cost me around $5 for 60 feet of it (the website says 30 ft for that price, and they sent me a bag with 60 ft)! Even 30 ft would be a far better value than RS solder, and it's much better anyway. I recommend it, you can't go wrong.
 
Oct 28, 2007 at 7:56 AM Post #45 of 115
With a bright blue LED, those resistor values will give you a flashlight. At 18 V the LED is very bright with a 10 k resistor.

It's much harder to get a good result with silver solder. I'd say stick to some "normal" cheap solder. Do some test soldering on surplus parts so you get a feel for it. I think it's totally unimortant to use silver solder. I doubt that anyone in the world could point out silver solder vs normal solder in a blind test. It's more important to get a metal to metal contact from component to component or board.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top