PINT Problem

[Heady]

Quote:

Originally Posted by tangent As it says in the documentation, 10 to 100 ohms.

I have to buy the resistors in lots of 50 for each value, granted they are not expensive. But 10 - 100 ohms is a wide range. I am asking what is a good value to start with? And what increments to work on if the first value don't work. I don't want to have a box full (fuller?) of unused parts.

Quote:

Originally Posted by tangent Since the resistors are outside the feedback loop now, you can only do it with the left and right channel resistors. With the resistors inside the loop, I believe you could have done it with the ground channel resistor as well.

If I un-mini3-fy the built, I replace all 3 inductors with the resistors. With the min3-fy built, I replace only the L/R inductors.

Sorry if I am being so confused. I thought it was only my ground opamp which was oscillating. I am not good with electronic knowledge, my training was in body parts not electronics.

 

[amb]

Quote:

Originally Posted by Heady If I un-mini3-fy the built, I replace all 3 inductors with the resistors. With the min3-fy built, I replace only the L/R inductors. Sorry if I am being so confused. I thought it was only my ground opamp which was oscillating. I am not good with electronic knowledge, my training was in body parts not electronics.

Yes, this is because with the ground ferrite taken out of the feedback loop, if you change it into a resistor it will increase the output impedance of the ground opamp, and that would adversely affect the stereo separation.

Changing just the L/R ferrites to resistors (or switching to "stronger" ferrites) may be enough to tame an unstable ground opamp, if the oscillation is caused by capacitive load at the output. This is because the ground opamp still "sees" the L/R opamps' outputs via the load, so they all interact in some fashion. However, if any of your opamps oscillate even without anything plugged into the headphone jack, playing with ferrites or resistors probably won't help at all. I would start looking at other causes of problems, such as inadequate power supply bypassing. You may also want to try lowering the rail splitter resistor values from 47KΩ to 20KΩ.

 

[tangent]

Quote:

Originally Posted by Heady But 10 - 100 ohms is a wide range. I am asking what is a good value to start with?

It's in the documentation: start high and work your way down.

 

[FallenAngel]

Hey guys, some good news and some bad news, I replaced the ground chip with an AD45048 and the offset stablalized quite a bit. It is now around 5mv OG->OL and about 4mv OG->OR at full volume.

The bad news is the current draw is pretty huge. It starts off at around 80mA and slowly starts climbing from there. I waited a few minutes until it reached 110mA and turned it off. Not sure exactly what that means and would appreciate a little help understanding what could be wrong.

Thanks

 

[Filburt]

Quote:

Originally Posted by FallenAngel Hey guys, some good news and some bad news, I replaced the ground chip with an AD45048 and the offset stablalized quite a bit. It is now around 5mv OG->OL and about 4mv OG->OR at full volume. The bad news is the current draw is pretty huge. It starts off at around 80mA and slowly starts climbing from there. I waited a few minutes until it reached 110mA and turned it off. Not sure exactly what that means and would appreciate a little help understanding what could be wrong. Thanks

It means oscillation.

 

[Filburt]

Quote:

Originally Posted by FallenAngel Damn, that sucks! Just to confirm though... would the amp really still play music with a fried ground chip and would it really have such precise DC offsets on both channels?

Yes, it could still play sound through it. I know from experience

 

[FallenAngel]

Uhm... Just to ask the obvious next question... how do I fix oscillation in the PINT? I read tangent's "Working with Cranky Op-Amps" but unfortunatly I don't have an engineering background and some of the solutions seem like they are written in a different language all together.

 

[Heady]

Quote:

Originally Posted by tangent It's in the documentation: start high and work your way down.

I was afraid of that. Looks like I will have to built up a collection of 1026 resistors in those values. Sigh!

On another note, amb has mentioned improving the bypassing to stop oscillations. Can one just solder a polyester cap to the V+ and V- legs of the opamp? And what values if this is possible? Thanks amb.

 

[n_maher]

Quote:

Originally Posted by FallenAngel Uhm... Just to ask the obvious next question... how do I fix oscillation in the PINT? I read tangent's "Working with Cranky Op-Amps" but unfortunatly I don't have an engineering background and some of the solutions seem like they are written in a different language all together.

Unfortunately there's really only one answer, and it's the obvious one. You have to track down what is causing the oscillation. Is that easy? Perhaps no, but that's the reality of the situation. There's only so much that people here are going to be able to help even if you post pictures and so on. As I discovered in the past with one of the PPA's that I helped a friend build sometimes solutions are not easily found and require both learning and rebuilding. So, hit the books my friend. Try to figure out what might be causing the oscillation and start trying to eliminate it.

Nate

 

[FallenAngel]

Heady: There are already caps on V+ and V-, they are C8.

Would lowering this value help oscillation?

 

[Heady]

Quote:

Originally Posted by FallenAngel Heady: There are already caps on V+ and V-, they are C8. Would lowering this value help oscillation?

Hi, I have built two PINTs following the list of parts exactly and both do not work correctly. If you read the thread, some post back, amb suggested that if everything is correctly built and the opamps still oscillate, he thinks that bypassing caps may need to be increased in values. I am asking him if polyester caps instead of ceramics can be used and what values to use.

Basically, I can discern two main lines of opinion here re: PINT and they are documented in this thread. One is the straight PINT and the other is the mini3-fied PINT.

Tangent provides his opinion on the straight PINT while amb seems to be the authority on the mini3-fied PINT.

I have a friend who has built 3 of the PINTs and have had no problems. I built two with the same parts and both of mine don't work. So it must be me.

I have put in considerable time and money on this project and while I am enjoying the challenge and learning, unfortunately, I have nothing working yet. And I have built several cmoys, Pimetas and others in the past, so I am not a newest newbie. This amp is not consistently reliable as is and Tangent has said so.

 

[tangent]

Quote:

Originally Posted by FallenAngel how do I fix oscillation in the PINT?

I think you're looking for easy solutions. If there were an easy solution, I'd have just designed the amp with it included. You're at the point where easy solutions have failed, and you will have to either get down to engineering, or start throwing wild guesses at the problem, or abandon the project. Them's your choices.

Quote:

Originally Posted by Heady Looks like I will have to built up a collection of 1026 resistors in those values. Sigh!

No, that's not whta it means. It means that if you want to buy just one value that will fix the problem, use the highest recommended value: 100 ohms. Going lower is to improve performance, at the risk of a less effective fix.

Quote:

Can one just solder a polyester cap to the V+ and V- legs of the opamp?

Yes, but if the problem truly is inadequate bypassing -- and it certainly may not be! -- then you should stick with ceramic caps in the standard positions. Also, "improving the bypassing" doesn't necessarily mean making the value larger. In fact, oftentimes, a lower value for the bypass cap will do a better job, because it will have a higher self-resonance frequency. I'd try 0.01uF before I tried something higher than 0.1uF.

 

[Heady]

Quote:

Originally Posted by tangent No, that's not whta it means. It means that if you want to buy just one value that will fix the problem, use the highest recommended value: 100 ohms. Going lower is to improve performance, at the risk of a less effective fix. Yes, but if the problem truly is inadequate bypassing -- and it certainly may not be! -- then you should stick with ceramic caps in the standard positions. Also, "improving the bypassing" doesn't necessarily mean making the value larger. In fact, oftentimes, a lower value for the bypass cap will do a better job, because it will have a higher self-resonance frequency. I'd try 0.01uF before I tried something higher than 0.1uF.

Thanks Tangent for clarifying these points. Ok, doesn't looks so bad now. I also got some TS922s (the only ones I could find from your list of alternates) to substitute the opamps with if all these don't work. I hope they are not as cranky as the AD8397s.

 

[FallenAngel]

Hey guys, I'm just curious, what Caps are you using for C4?

I'm using 470 µF / 16V from Tangent's parts list and I'm running off 2 9V rechargables which are actually around 16.5V. Would this be a problem now and how much of an issue would this be when I plug in my 24.0V TREAD to recharge them?

I never even noticed at first when I ordered the initial parts, so now I have 3 of those and 3 more of the 680 µF 10V and no idea what to do with them.

Just in case, I just ordered some ELNA Duorex II 25V 1000uF from www.PartsConnexion.com, great canadian store, even though prices in USD, it's cheap and fast shipping to me.

 

[blueworm]

The value of the power cap C4 in the pint needs to higher than the supply voltage because its is the actual voltage the cap see's.

This is not the case on the pimeta for example, since in normal conditions the power caps only see half the supply voltage even so it is recomended that the caps are rated above the max voltage, just in case the railspliter fails.

In my first pint I used a panasonic FC 330 uF 25v.
In my mini3'fied pint I used a elna cerafine 470 uF 25v.
In both cases pints are supplied by 2x 9v.