[amb]

Quote:

Originally Posted by phobus I was re-wiring my M3, and put everything back together, but now... With the Ground opamp out, the voltages measure fine. Once I put the opamp in, the voltage between pins 4 and 7 read 6.5 volts, and the power supply gets REALLY hot (its not the opamp, I tried other ones).... This doesn't happen with the L/R opamps so I'm guessing I blew something out in the ground channel. Is there anything to check for, or should I start replacing components? Thanks

Don't start replacing anything just yet. The symptom sounds like some MOSFETs are simply over-biased and passing too much current.

With the power off, first try turning the BIAS trimpots of all channels to the fully-counterclockwise position and the BIASOP trimpots to the middle of their travel, remove the L/R channel opamps, then re-do the initial setup procedure found on the M³ website. See if that is successful.

 

[adhoc]

Quote:

Originally Posted by sean-xenos I second what guzzler wrote. Especially with smaller, temperature sensitive components, it doesn't make sense to cut after soldering.

imo most importantly, cutting before soldering reduces the chance of a fractured solder joint. ime the mechanical shock (think 'snip!') of cutting a long lead off a completed joint is sometimes too much. ymmv.

 

[Voodoochile]

Nice looking work, rreynol.

 

[Teerawit]

nevermind

 

[phobus]

Quote:

Originally Posted by amb Don't start replacing anything just yet. The symptom sounds like some MOSFETs are simply over-biased and passing too much current. With the power off, first try turning the BIAS trimpots of all channels to the fully-counterclockwise position and the BIASOP trimpots to the middle of their travel, remove the L/R channel opamps, then re-do the initial setup procedure found on the M³ website. See if that is successful.

Unfortunately its still showing those symptoms... Any ideas?

 

[amb]

Quote:

Originally Posted by phobus Unfortunately its still showing those symptoms... Any ideas?

Well then it's time to test the board for problems. Since you say that the L/R channels are ok, and only the ground channel is exhibiting a problem, then it should be easy to use the L/R side as reference. With the power off (and all rail caps completely discharged), check each MOSFET in your DMM's ohms mode between the drain and source pins to see if there are any shorts. Check the Q2G transistor between the B-E junction, B-C junction and the C-E junctions and compare against Q2L and Q2R. Check the Q3G and Q4G JFETs similarly. Basically, see if you could find any obvious difference between the ground channel and the other channels (except where they are supposed to be different, see the schematic).

Also, it would be a good idea to examine the board very carefully to see if there are any bad solder joints, solder bridges, shorts, charred parts, etc.

 

[poolorpond12]

I have started trouble shooting my amp and get +25v and +12v at the rail splitter. This corresponds to the same voltages on pins 4 and 7 at the opamp. The steps power supplies about 36v unloaded. I have been sneaking away from the new baby to build this and have not had much time to spend with this. Any thoughts?

 

[n_maher]

Quote:

Originally Posted by poolorpond12 I have started trouble shooting my amp and get +25v and +12v at the rail splitter. This corresponds to the same voltages on pins 4 and 7 at the opamp. The steps power supplies about 36v unloaded. I have been sneaking away from the new baby to build this and have not had much time to spend with this. Any thoughts?

First, turn the STEPS down, a lot! As I learned (the hard way) there is absolutely no need to run with that much voltage. It serves virtually no purpose and can only complicate things.

Second, where exactly are you taking the measurements from? You say at the rail splitter but do you really have the board flipped upside down and you're measuring the TLE pins? If so, BE CAREFUL. That's an easy place to short something. Also, measure the resitance from pin to pin on the TLE and report back, it could be that you already fried it and it's internally shorted.

Finally, from the sound of it you've got a short somewhere on your board. The only way that I can see to get positive volts into the negative rail is to have a short. After spending as much time pouring over my PPA that's really all I can come up with (I experienced similar issues as you with that amp). If you've got a flatbed scanner scan the bottom side of the board and look at it very closely, or post the image and let us look at it.

HTH,

Nate

 

[amb]

Quote:

Originally Posted by poolorpond12 I have started trouble shooting my amp and get +25v and +12v at the rail splitter. This corresponds to the same voltages on pins 4 and 7 at the opamp. The steps power supplies about 36v unloaded. I have been sneaking away from the new baby to build this and have not had much time to spend with this. Any thoughts?

I think you're probably just testing with the probes at the wrong points for one of your readings. 12V is about half of 25V, so it seems to be splitting correctly.

To verify this, keep your meter's black probe on IG, then touch the red probe on pin 7 of the opamp. You should get +12V. If you have a digital meter, then keep the black probe at IG, and touch the red probe on pin 4 of the opamp, you should get about -12V. If you have a meter with "needle pointer", then put the red probe on IG and the black probe on pin 4, and you should get +12V. By the way, I don't really recommend using a needle-type meter for measuring these circuits, because they tend to have much lower input impedance and would load the circuit you're trying to measure. Use a digital multimeter instead.

 

[poolorpond12]

Sorry guys. I was being hasty and should not have even wasted your time. I kept measuring from pins 4 and 7 to the common lead on the rail splitter, not the IG. Everything checks out.

The power supply is only putting out 26v, not 36v.

 

[silvervarg]

My M³ build has been delayed by tons of things, but I finally got to the point where I want to ask for troubleshooting help.

With no OP-amps in I measure:
rails 22.6V, sockets (pin 7-4) 21.4V and pins 7/4 to IG about half that.
Everything seems fine so far.

Once I put in an OP at the ground channel (using NE5534 during tests) the voltages drops dramatically.
Rails 11.5V, sockets (pin 7-4) 10.1V and pin 7/4 to IG about half that.

So, any idea what is happening?
It seems my rails are dropping to close to half the voltage with an op-amp in place.

 

[Jazper]

Quote:

Originally Posted by silvervarg My M³ build has been delayed by tons of things, but I finally got to the point where I want to ask for troubleshooting help. With no OP-amps in I measure: rails 22.6V, sockets (pin 7-4) 21.4V and pins 7/4 to IG about half that. Everything seems fine so far. Once I put in an OP at the ground channel (using NE5534 during tests) the voltages drops dramatically. Rails 11.5V, sockets (pin 7-4) 10.1V and pin 7/4 to IG about half that. So, any idea what is happening? It seems my rails are dropping to close to half the voltage with an op-amp in place.

What amount of current is your power supply capable of? I'd guess you're overloading it.

do you realise that the NE5534 is a bipolar opamp and will not work in the m3? Try a TL071 if you want a test chip that's cheap.

 

[GregVDS]

Hi,

Long time I did not ask something. Too busy at work, but I now I just finished the initial setup. I still have to receive the boxes for STEPS and M³ and mount everything in them. I so don't have heard the amp in itself, for the crossfeed filter, switch and rca are still to be connected.

I hope to have the boxes soon, so I can take precise quotes and modify lastly if required the front panel express fronts and backs. I think I will not be able to make a rectangular 34 x 72 mm hole, nor round holes of 24mm of diameter, required for headphone jack and speakon connectors I will use for power between PSU and AMP.

I have some questions:

I made the setup, following scrupulousely all the steps described. First I noticed the power voltage drop somewhat each time another opamp was introduced (not that much, don't remember the figures but I wrote them down for record). Is it normal? As I understand, the STEPS give a somewhat higher voltage without load, and less and less the load increasing, right?

Through R5X (X = G, L, R) I first put 0.5V adjusting BIASOPX, then adjusted BIASX for 80mV through R9X. Coming back to chaeck to R5, I read a lower value, around 0.491V, depending on the channel (G being hoter, and working for longer than R, and this last one longer than L). I thought this is ok, but re-adjusted BIASOPX for 0.497V through R5X. What do you think of this.

After the half-hour warm up, I pushed the quiescent current through R9X to 100mV PSU giving 24.77V, for the amp will drive a Grado RS-325i (32 ohms, sensitivity not that high, so volume put higher, having in mind also an Audio-Technica ATH-EW9, rated 28ohms). I had to readjust some time, the temperature counter-effect droping somewhat the quiescent current. Should I play reading-adjusting a long time or there is no point to be maniac here?

What differences between R5's and R9's of reading is allowed (I mean to what point I should refine and quest for the same values for G, L and R)?

I achieve almost nothing mV between IG and OG, maybe 0.1mV, 1.5~1.6mV for L and 2.3mV for R. Is it ok, keeping in mind my source will be an iPod giving around 3mV of DC offset, passing through a Modified Linkwitz filter (dropping of DC offset here or not?)?

Many thanks again for any advice.

All the best,

GregVDS

ps, I promise, when everything will be finished I will manage to post some pics, but I still to manage to reduce them enough for posting.

 

[amb]

Quote:

Originally Posted by silvervarg Once I put in an OP at the ground channel (using NE5534 during tests) the voltages drops dramatically. Rails 11.5V, sockets (pin 7-4) 10.1V and pin 7/4 to IG about half that.

The NE5534 is not unity-gain stable without a special compensation cap between pins 8 and 5, and the ground channel runs at unity gain. It is probably oscillating like mad, increasing the current draw dramatically. When opamps oscillate on the M³ it has the potential of damaging Q5+ and Q5-.

Try using one of the recommended opamps, and if the problem persists, replace Q5+/Q5-.

 

[amb]

Quote:

Originally Posted by GregVDS I made the setup, following scrupulousely all the steps described. First I noticed the power voltage drop somewhat each time another opamp was introduced (not that much, don't remember the figures but I wrote them down for record). Is it normal? As I understand, the STEPS give a somewhat higher voltage without load, and less and less the load increasing, right?

Yes, a small amount of drop (fraction of a volt) with each additional opamp is ok.

Quote:

Through R5X (X = G, L, R) I first put 0.5V adjusting BIASOPX, then adjusted BIASX for 80mV through R9X. Coming back to chaeck to R5, I read a lower value, around 0.491V, depending on the channel (G being hoter, and working for longer than R, and this last one longer than L). I thought this is ok, but re-adjusted BIASOPX for 0.497V through R5X. What do you think of this.

That's fine. A variation in the third decimal place is too small to matter.

Quote:

After the half-hour warm up, I pushed the quiescent current through R9X to 100mV PSU giving 24.77V, for the amp will drive a Grado RS-325i (32 ohms, sensitivity not that high, so volume put higher, having in mind also an Audio-Technica ATH-EW9, rated 28ohms). I had to readjust some time, the temperature counter-effect droping somewhat the quiescent current. Should I play reading-adjusting a long time or there is no point to be maniac here?

Just get somewhere close to what you want is fine. It's more important to make it consistent between channels than to hit a specific target precisely. When you put everything into a case, the thermal rise on the MOSFETs will increase somewhat and the quiescent currents will settle at a different point again.

Quote:

What differences between R5's and R9's of reading is allowed (I mean to what point I should refine and quest for the same values for G, L and R)?

See above. It also depends on how picky you are.

Quote:

I achieve almost nothing mV between IG and OG, maybe 0.1mV, 1.5~1.6mV for L and 2.3mV for R. Is it ok, keeping in mind my source will be an iPod giving around 3mV of DC offset, passing through a Modified Linkwitz filter (dropping of DC offset here or not?)?

As far as the amp itself is concerned, your offset readings are fine. The crossfeed circuit will reduce the effect of the source-induced offset somewhat (because it is lossy), but the overall offset will be whatever the amp "sees" at its input, multiplied by the gain of the amp.