[nikongod]

This thread is a continuation of the discussion that started in the post pics of your builds thread around this date.
 
 
 
 
 
 
 
 
Quote:

Can we move this to a new thread so we don't clutter up this one?
 
Quote:

     Quote:

Another test you can do if you can measure the current consumption is to
measure it with no signal and compare it to the sum of all the currents from
the specs in the data sheets. If it's pretty close then your good.
If it's way higher than it should be then you have a problem.
Either oscillations or a short.
My BUF634's don't get more than warm in my ground channels.
 
Quote:

Quote:

Good job on the A47.
Do you have the BUF634 running high bandwidth or low?
In wide band mode, the BUF634 heats up a little.
In low band mode it stays cool.
 
I use OPA633 buffers in one of my amps and they run hot
all the time...
 

 

I believe I'm in low bandwidth mode. I snipped the BW lead on the BUF634. I'm still going with the theory that the tin is acting as ground and for some reason this is causing the BUF634 to heat up and get really hot. How much heat tolerance does the BUF634 have? I do have a spare just in case it bites the dust.
 

 

 

Excuse me for my ignorance. How would I measure the the current consumption? Also, I have downloaded the datasheet -- where would I find it's current consumption rating? Another thing to note -- before I added in some insulation I noticed this loud hum/buzz type noise that got progressively louder. I had to unplug my Grados to make sure they didn't get damaged. In my initial testing without the tin enclosure I didn't get this. So the heat and loud hum were the symptoms. Now with the insulation I'll get intermittent heat depending on if I fiddle with the thing too much. Perhaps, I should build a case that isn't made out of metal to insure I have no heat problems in the future. Any recommendations for the A47?
 

 

 
I dont think the huddler software makes it easy to move posts or split threads. Its really designed for SEO before friendlessness towards any of its users, moderators/admins included.
 
The best bet is to copy all of this into a new thread yourselves and continue on from there, but since Im too lazy to do that I'l post here
 
BUF-634 has a lot of protection circuitry built in. They will not overheat, although I do believe that a lot of what people describe as the "buf-634 sound" is the sound of a warm buf-634 in self-preservation mode. 
 
When the buf634 gets too hot it limits current, without regards to the sound. This is a real problem in the dip8 packages where they have very little heat dissipation, and people like to run them hot. In the TO220 packages they can be heatsinked effectively and taah-daah! My limited experience with a TO220 buf-634 (on an original millet hybrid, before the head-fi revisions) was QUITE favorable. 

 

[Avro_Arrow]

Thanks Nikongod...
 
 
What you need to measure is the idling current for your amp.
To know what it should be you need the know the quiescent
current for the chips you are using. BUF634 is 20mA worst case
in wide band mode. OPA2227 is 4.2mA (per amplifier) worst case.
If you add them all up and add a little for wiggle room, you should
see no more than 55mA being used when the amp is idle.
To test this with your amp, set you meter on the 200mA DC setting.
With the amp turned off and nothing connected, place one probe
on one of the switch terminals on you volume control and the other
probe on the other switch terminal. This will short out the switch and
turn the amp on. You will get a reading on the meter, it will either
be positive or negative depending which probe you put on which side
of the switch...anyway the polarity doesn't matter, only the value.
It should be around 50mA more or less, but not more than 55mA.

 

[holland]

ammeter between the battery and your amp.  the reading should be close to the quiescent current.  this is a good way to check for oscillation too, without a scope.  the current draw will bounce wildly.  if you're not class a biased, your quiescent current should be close to the spec of the opamps and buffers combined (sum them all up), otherwise sum up the biases.
 
look for quiescent current in the spec sheets.
 
you can also lift the pins and measure current between (v+ and opamp/buffer) and (v- and opamp/buffer), or drop a resistor in there and measure the voltage drop.  that shouldn't be necessary now.  go between the battery and the amp.
 
if you're concerned about heat, put a piece of aluminum on top of the opamp so when the lid closes it touches the case.  use heatsink tape (what is used for older CPUs) to tape the lid to the aluminum and the aluminum to the buffer(s).  that will help pull some of the heat out instead of building it up in a small vent space.  you can even link all the buffers together.

 

[Avro_Arrow]

Oh, I just see you said you snipped off the bandwidth pin on the BUF634.
You are in low bandwidth mode. Current is only 2mA in this mode.
 
Total current should be 37mA, not 55mA.

 

[Mullet]

 
 
 

BUF-634 has a lot of protection circuitry built in. They will not overheat, although I do believe that a lot of what people describe as the "buf-634 sound" is the sound of a warm buf-634 in self-preservation mode. 
 
When the buf634 gets too hot it limits current, without regards to the sound. This is a real problem in the dip8 packages where they have very little heat dissipation, and people like to run them hot. In the TO220 packages they can be heatsinked effectively and taah-daah! My limited experience with a TO220 buf-634 (on an original millet hybrid, before the head-fi revisions) was QUITE favorable. 

 
 
Thanks nikongod for moving the thread. In hindsight 20/20 I should have just left the build thread with a few pics and then created a separate thread. Anyways, now we're here.
 
It's good to know that the BUF634 won't overheat and self-preservation mode is kicking in. I'm using the BUF634T, which comes in the TO220 package. I'm not sure how I'd heatsink this puppy with the casing I'm using... perhaps holland's solution below might be sufficient.
 

 
Quote:

Thanks Nikongod...
 
 
What you need to measure is the idling current for your amp.
To know what it should be you need the know the quiescent
current for the chips you are using. BUF634 is 20mA worst case
in wide band mode. OPA2227 is 4.2mA (per amplifier) worst case.
If you add them all up and add a little for wiggle room, you should
see no more than 55mA being used when the amp is idle.
To test this with your amp, set you meter on the 200mA DC setting.
With the amp turned off and nothing connected, place one probe
on one of the switch terminals on you volume control and the other
probe on the other switch terminal. This will short out the switch and
turn the amp on. You will get a reading on the meter, it will either
be positive or negative depending which probe you put on which side
of the switch...anyway the polarity doesn't matter, only the value.
It should be around 50mA more or less, but not more than 55mA.

 
I have a Fluke 12 DMM. Just from doing a little digging, I don't think it can measure amperage. It looks like it measures for VDC, VAC, capacitance, resistance, and continuity.
 
Also if I could measure amperage on my DMM, where would I put the probes on the Alps RK097? I'm assuming it would be where the power is routed through the switch. Not the signal terminals.
 
Quote:

ammeter between the battery and your amp.  the reading should be close to the quiescent current.  this is a good way to check for oscillation too, without a scope.  the current draw will bounce wildly.  if you're not class a biased, your quiescent current should be close to the spec of the opamps and buffers combined (sum them all up), otherwise sum up the biases.
 
look for quiescent current in the spec sheets.
 
you can also lift the pins and measure current between (v+ and opamp/buffer) and (v- and opamp/buffer), or drop a resistor in there and measure the voltage drop.  that shouldn't be necessary now.  go between the battery and the amp.
 
if you're concerned about heat, put a piece of aluminum on top of the opamp so when the lid closes it touches the case.  use heatsink tape (what is used for older CPUs) to tape the lid to the aluminum and the aluminum to the buffer(s).  that will help pull some of the heat out instead of building it up in a small vent space.  you can even link all the buffers together.

I'm not using a battery. I'm using a wallwart. I'm assuming it's still the same. I'm going to try and follow Avro's instructions and see if his idea works for measuring amperage. Regarding heat dissipation, your idea about the aluminum sounds like a good idea. 
 
 
Finally I decided to measure DC offset and I'm getting a reading of -2 mV for the left out and -1 mV for the right out. Does this seem correct?

 

[nikongod]

Regarding heat-sinking the TO-220 packages, try gluing them to a penny. There are a few other truly cheap DIY solutions floating around the web. If memory serves tangent posted a pretty well done comparison of a few of them. Whatever you do, be mindful of the fact that the screw-tab is electrically connected to V-. 
 
Quote:

 
I have a Fluke 12 DMM. Just from doing a little digging, I don't think it can measure amperage. It looks like it measures for VDC, VAC, capacitance, resistance, and continuity.
 
Also if I could measure amperage on my DMM, where would I put the probes on the Alps RK097? I'm assuming it would be where the power is routed through the switch. Not the signal terminals.

Measure the voltage drop across a known resistance. 
 
If you know the target current (37mA was posted above) you could select a resistor for 0.5-1V of drop...
V=I*R (volts=current*resistance, respect units or they will make you sad) one of the few formulae worth remembering.
0.5-1V is a good target because you can read the meter with very good resolution, and the drop should not cause problems in most circuits.
It will require a bunch of measuring & a bit of simple calculating, but thats part of the fun of this.
 
Regarding your voltage offset:
less than 2mv is very good. Its hard to consistently do much better than that.

 

[Avro_Arrow]

As you mentioned, your Fluke 12 does not do amperage.
But no problem. We can just do as Nikongod suggested
and use our own shunt and measure the voltage drop
across it. Either a 1 ohm resistor or a 10 ohm resistor
will make it pretty easy. Just temporarily solder the resistor
across the switch on the Alps RK097 and measure the
voltage across it. For the 1 ohm resistor, mV = mA. So if
you get a reading of 0.345 for example, then the amp is
using 345 mA. If the reading is 0.035 then you are using
35 mA. Pretty easy. For the 10 ohm resistor we just have
to divide the reading by 10. Now, 0.345 would be 34.5 mA.
The resistor for the shunt does not have to be anything
special. The same size and type as the other resistors
in the amp is fine. The switch on the Alps RK097 are the
last two terminals where the power passes through it.

 

[Mullet]

Cool. Could I use an unused 1k Ohm resistor that I have from a CMOY project? I'm assuming I'd divide by 1000 instead of 10 or 1.

 

[nikongod]

V=I*R Such a simple equation, and so powerful.
 
I=37mA=0.037A
R=1Kohm=1000ohm
V=??? (Its a lot)
 
The idea with this is to drop as little voltage across the resistor as possible while still having enough voltage to read.
 
Do you have a few 47 ohm resistors? 3 47 ohm reistors in parallel should be good. Measure the actual impedance before you hook the resistors up to the amp (it should be about 12 ohms) and calculate away after you measure the voltage across it.

 

[Mullet]

I'm going to have to go out to RatShack and get myself an assortment of resistors. Thus far I've only bought what I need regarding DIY projects.
 
So to understand this fully and to put it in laymens english so others might understand the procedure...
 
Take 3 47ohm resistors, solder them together. Measure the resistance of all three resistors. Save this number for later. Now solder them to the power terminals of my switch ie Alps RK097. Then grab my multimeter. Set it to the VDC position. Apply power to the amp. Put the red test lead on one terminal. Put the black lead on the other. Measure the voltage when it settles on a value. It will drop .5 to 1V. Take this value and divide it by the resistance. This will give you the current.
 
I think this is what I'm supposed to do.
 
Alternatively, it seems like Avro's method is a bit more simple if using a 1 ohm or 10 ohm resistor.
 
As a side note, I've decided to tear apart the amp and make the hole larger where the dc jack comes in. Then I need to figure out how make a grommet or something that will shield the jack from the tin. Ultimately, this is what is going to solve the issue with heat, hum, etc. After I really figure out a better solution for the jack touching the tin then I'll make an effort to measure the power consumption and see what results I get.

 

[holland]

Quote:

I'm going to have to go out to RatShack and get myself an assortment of resistors. Thus far I've only bought what I need regarding DIY projects.
 
So to understand this fully and to put it in laymens english so others might understand the procedure...
 
Take 3 47ohm resistors, solder them together. Measure the resistance of all three resistors. Save this number for later. Now solder them to the power terminals of my switch ie Alps RK097. Then grab my multimeter. Set it to the VDC position. Apply power to the amp. Put the red test lead on one terminal. Put the black lead on the other. Measure the voltage when it settles on a value. It will drop .5 to 1V. Take this value and divide it by the resistance. This will give you the current.
 
I think this is what I'm supposed to do.
 
Alternatively, it seems like Avro's method is a bit more simple if using a 1 ohm or 10 ohm resistor.
 
As a side note, I've decided to tear apart the amp and make the hole larger where the dc jack comes in. Then I need to figure out how make a grommet or something that will shield the jack from the tin. Ultimately, this is what is going to solve the issue with heat, hum, etc. After I really figure out a better solution for the jack touching the tin then I'll make an effort to measure the power consumption and see what results I get.

Get a 1 ohm or 0.47 ohm power resistor (5W, 10W, etc.), since you are buying.  Wirewound is fine.
 
nikongod's suggestion was really about if you had it on hand or not.
 
it would be good for you to find electronics/surplus stores in your area.  i buy most of my immediate need stuff in places like that, and other things like standoffs are way cheaper.
 
you shouldn't need a grommet if the hole is big enough.  the DC jack will hold the plug in place.  you are using a pcb mounted jack, I presume.  I don't understand how making a tiny hole a little bit bigger is going to solve your problems.  What jack did you use, as you listed hum.
 
Edit: I just checked the prices of resistors at radio shack, expensive.  The wirewound is $2.  Forget it.  They have 10 ohm 1/2 W carbon film 5-pack for $1.00  Get that instead.

 

[Mullet]

holland -- Thanks for the suggestions on buying local. I'm sure the Rat Shack will have everything I need or at least I hope -- they are more like cellphone stores nowadays.
 
Regarding the hole and possible grommet -- I'm getting a spark when the metal dc jack touches the mint tin. Then the dang BUF634 gets hot. I probably failed to mention this before -- at least the part about the spark. BTW I used a metal panel mount dc jack like this...
http://bit.ly/dVU4Kn
 
It's rated for 13V and 2A, but I'm sending in 24V. Could this be my issue -- too much voltage for the dc jack?

 

[holland]

Quote:

holland -- Thanks for the suggestions on buying local. I'm sure the Rat Shack will have everything I need or at least I hope -- they are more like cellphone stores nowadays.
 
Regarding the hole and possible grommet -- I'm getting a spark when the metal dc jack touches the mint tin. Then the dang BUF634 gets hot. I probably failed to mention this before -- at least the part about the spark. BTW I used a metal panel mount dc jack like this...
http://bit.ly/dVU4Kn
 
It's rated for 13V and 2A, but I'm sending in 24V. Could this be my issue -- too much voltage for the dc jack?

 
No, it's not too much voltage, but you're shorting your circuit out, I think.  What input and output jacks are you using?  What is your pot?  The issue with virtual ground, well, actually any general circuit, is that you want to control ground current flow.  In general, I never connect with the case unless I have to.  With virtual ground, your headphone out and input grounds are going to be sitting at a higher potential than earth.  The "ground" on the wallwart is your negative rail (due to the splitter).
 
I didn't see your original posts so I don't have much history about what it is you are using or have done.  This is a picture from AMB's site, regarding what to isolate and what not.  The same applies to you since you.  Here, the chassis (your tin) is at the virtual ground potential as opposed to your negative rail.
 
I'm surprised your TLE2426 hasn't fried.
 

 

 

[holland]

Use this jack

 

[Mullet]

Here's a pic of my case design and amp in general. I'm using switchcraft 3.5mm jacks. An Alps RK097 pot, and that DC jack from mouser.
 
I'm still a newb when it comes to understanding how electricity works. So the TLE2426 splits my 24v supply into +12v and -12v. If I understand you correctlyl the negative -12v is coming from the wallwart and then the +12v comes from the TLE2426. The virtual ground is at 0v -- right? I'm still not clear on what "earth" is and how my headphone input and output grounds relate to "higher potential". I'm just trying to wrap my head around all of this and it's tad bit dizzying. 
 
Here is a pic of what I came up with in this build...
 

 
As you can see I tried using masking tape to isolate the jack. I guess it didn't work very well. How would I know if I fried my TLE2426? I'm assuming nothing would work then or the amp would sound really bad. Another hunch is that I would have full 24 volts on both the positive and negative rails. Just a note -- these issues with spark, hum, heat from the BUF634 are all intermittent at best. Since this picture was taken I've removed the tape and I'm looking for ideas on what to do next to isolate the jack.