[stvnharr]

I now have had my balanced e-Pot Max operational for just over one month.
Some things have cropped up that are a bit of a mystery to me. So I would like to share them. Perhaps others have had these same issues. But I noticed there has not been much written on the forum.

1. I have had the secondary controller drop out of working at times. This has happened several times, since the get go.

It is always the same situation too - big climaxes in Mahler symphonies on Channel Classics sacds.
These may not be common listening material for folks here, but they are for me.
I have listened to these for years. I never had this issue when I ran single ended volume controllers. I have used a Tortuga controller for 8 years now.
Now at the big climaxes, the left channel, controlled by the secondary board, just goes nearly silent when it all happens.
I have tried different impedances. I have tried different volume levels. It has happened even at low volume level, and the climax volume only comes up to moderate.
It is only on these recordings too. I can play other music, loud, and there is no issue.
Things I have tried thus far:
1. swapping LDR modules between controller boards.
2. reseating the serial data links with Pro Gold contact enhancer.
3. trying to time the climax and lower the volume - tricky.
4. listening at low level so that climax does not get loud.

In order to restore operation I have to recalibrate and power cycle. So I take about a 4 minute pause.
This is not the end of the world, but it is annoying.

Does anyone have any thoughts or ideas here? I would write directly to Mort, but he is busy packing. And there is a whole Tortuga community here.

2. Once I learned about the adjustable impedance feature and began having the above issue, I began to try different impedance settings. I learned through trial and error that to make a new impedance I had to do a power cycle after calibrating the new impedance. And when I tried to do an impedance change on the fly I could not do so. I also had to do a power cycle. So there is no trying different impedances on the fly.
This is pretty minor. But it is not mentioned in the documentation, or at least I could not find it.
However, I did read that one should do impedance calibration changes with interconnects disconnected, or having the controller out of the system. Is that still true?

3. Because of the adjustable impedance feature I was able to find a really good impedance setting for my system, 10k. This is half of the single ended impedance setting. The controllers came with a 50k default setting. The sound level was much as I had with the 60k resistor module in the old single ended days. So thanks to adjustable impedance I was able to get back the wonderful sound of the single ended 20k default setting that I wrote about on my Diyaudio LDR listening impressions thread. This sound is what determined me to go balanced and simplify my system cabling. And this new balanced operation sound is even better than the 20k single ended sound. So this is all a very good thing, other than issue No. 1.

Thank you for reading this longish post. Issue No. 1 is concerning, and a complete mystery to me.

[stvnharr]

Instead of adding to yesterday's post I will just make a new one with some updates.
Yesterday I had another dropout occurrence. This time it happened on a jazz disc on a snare drum hit. Volume was medium and the drum hit was not a big smack, just a common drum hit.
In order to restore function, I had to do TWO calibration/power cycles, and all was well thereafter.

With that in mind, earlier today I decided to set all the Impedance settings to 10k, other than No. 1 which cannot be changed. I have found that 10k is the best impedance value. I did two recalibration/power cycles for each setting. Thus, if I get another dropout I can hopefully quickly change to another impedance number and resume play quickly. I can always do the recalibrations later.

I also made a small change in the wiring. I have 2 output connectors on each channel. One for the main amp and one for a subwoofer. The main amp connections are to the pins on the end of the controller board. The subwoofer connections are to the holes on the board.

Now to see how it goes.

[krustykat]

I had the same problem with the LDR300x.v3. when I tested it during a tour.   If I recall correctly, Morten had tracked it down to an issue with the power management on the board.  Needle drops and big dynamic sounds would cause the same problem.  Best to reach out to him. 

[tortugaranger]

I've received a small number of similar reports regarding the V3. They are not limited to balanced configurations so the fact that your setup is balanced is probably not a key factor. I have tried everything to reproduce this behavior here with the equipment I have but have not been successful in doing so. The fact that I can't reproduce the behavior has prevented me from confirming my suspicions as to what is going on.

Several weeks ago I was contacted by a customer in France who reported similar behavior. A certain passage on a certain track would cause the controller to reset. Basically it turned off the V3. Afterwards the V3 would turn back on normally. It didn't require recalibration to resume working normally. Same customer with otherwise same hardware then switched from his phono preamp source to his DAC source. When running with the DAC as the source he could not make the same behavior happen even with the same music. I think it's fair to conclude that this behavior, while triggered by a specific musical passage from certain music sources yet not others, has nothing to do with the music per se.

My working theory is this has to do with the fact that the V3 utilizes solid state analog switches for selecting inputs instead of conventional electromechanical relays. These analog switches are powered with a split voltage power supply which operates around +/- 4.7 volts. In order for these analog switches to operate correctly the incoming audio signal from the connected source (phono preamp or DAC etc. ) needs to be at the same ground reference as the analog switch. An audio input that is running at say +/-2 volts peak to peak relative to ground should be no problem provided that ground reference is the same as the ground reference feeding the analog switch. If the audio signal instead has a significantly DC offset plus has a large magnitude AC peak (as with certain musical passages), it's conceivable that the audio signal could exceed the +/-4.7 volts supply within the V3, disrupting the power and causing the microcontroller to reset - i.e. turning off the controller.

Again, this is a working theory only. I can't prove this. But the fact that the problem disappears by switching to another source and the fact that most users do not experience this problem with most sources, suggests to me that certain source devices are incompatible with these analog switches. Again, exactly why and how is unclear but I suspect it has to do with the nature of the audio signal levels.

The V3 is a direct coupled device meaning there is no input coupling capacitor that isolates a possible DC offset on the input from the AC of audio signal. If I could duplicate this behavior, I would add an input coupling capacitor and observe if the behavior stopped. This would confirm the working theory. 

[stvnharr]

Thank You Mort. Very much appreciate this.

This is interesting news about automatic resetting of the controller. Perhaps I have always been rather impatient.

The very first couple of times this happened I thought it was the overcurrent protection circuit in my Class D amplifier going off again because I was listening too loud. It was just that it was the "wrong" channel. Right channel overcurrent protection circuit was the one that always was more sensitive, not the left channel. And now I had the left channel going out. And it did not reset itself after the 20 seconds of protection shutdown. So the next time this happened I immediately checked the amplifier to see if the bright red protection circuit led was glowing. And it was not. So then I realized it was the volume controller cutting out.
I believe that I did notice an automatic reset of the controller without doing anything, once. But that was after a few minutes of being down and investigating. Then I tried power cycling, to no effect. Then I tried changing to a different impedance number, to no effect. But when I tried recalibration, it worked. Yet, the other evening I had to recalibrate twice to get full operation.
So if there is a next time, I will just wait, and see what happens.

[stvnharr]

And now for a brief update following the changes that I did yesterday...
I have noticed a difference and it is not particularly subtle.
There is a difference in sound. It is now much stronger.
Let me clarify a bit.

For the first 3 weeks of balanced operation with the 50k impedance setting I found the common listening level to be in the low to mid 90's on the display. This is comparable to the display readings of 120's, with the top setting being 127, that was the common listening level with the 60k resistor module in single ended. And not only were the display readings comparable, also the sound characteristics were comparable.
Then I started to change the impedance settings to 25k and 20k, but I did not notice any differences at all.
But when I changed the impedance to 10k one week ago I found the common listening level to be in the low to mid 80's on the display. This was comparable to the common listening level displays when I ran the 20k LDR module in single ended. And the sound characteristics were also comparable.

Now that I have done the double calibration/power cycle for the 10k impedance level, I have noticed that the common listening level is now in the low to mid 70's, and with the same sound characteristics. It is not hard to read the display. Must be something to the double cycle.

Still have to play another Channel Classics Mahler recording, and eventually will do so.

[tortugaranger]

And now for a brief update following the changes that I did yesterday...
I have noticed a difference and it is not particularly subtle.
There is a difference in sound. It is now much stronger.
Let me clarify a bit.

For the first 3 weeks of balanced operation with the 50k impedance setting I found the common listening level to be in the low to mid 90's on the display. This is comparable to the display readings of 120's, with the top setting being 127, that was the common listening level with the 60k resistor module in single ended. And not only were the display readings comparable, also the sound characteristics were comparable.
Then I started to change the impedance settings to 25k and 20k, but I did not notice any differences at all.
But when I changed the impedance to 10k one week ago I found the common listening level to be in the low to mid 80's on the display. This was comparable to the common listening level displays when I ran the 20k LDR module in single ended. And the sound characteristics were also comparable.

Now that I have done the double calibration/power cycle for the 10k impedance level, I have noticed that the common listening level is now in the low to mid 70's, and with the same sound characteristics. It is not hard to read the display. Must be something to the double cycle.

Still have to play another Channel Classics Mahler recording, and eventually will do so.

Couple of comments....

First, when used in balanced configurations the stated impedance level is nominal and not actual in that being balanced the effective impedance when set to 10k is actually 20k.  This is one of those single-ended vs balanced differences that I deliberately decided to not reflect in the display values because of the way the V3 design works. In balanced configurations there are actually 2 V3 boards each set at say 10k which is technically true for each board, yet the effective overall balanced impedance is 20k.

Second, when making any change to the impedance level, even if only changing it from say 10k to 11k, this requires all 400 (4 LDRs x 100 steps/LDR) of the calibration values to be set up at the new level. The first cal pass through at this new level requires some estimating. The second cal pass at the same level will improve the precision of the calibration data set because it uses some of the information from prior pass to inform the next pass. Bottom line, run 2 cal cycles when changing to a new impedance setting #/level for the first time.

[stvnharr]

Prior to yesterday, and the evening before that, I had only been doing the recalibration once, and with mixed results.
In about my 30th reading of all things balanced in the documentation section, I did pick up doing the second run for recalibration. Then when I HAD to do it, it just confirmed what I had read.
And now you have explained it very well.

All better now!

I also read about disconnecting interconnects when doing recalibration. I did the recalibrations yesterday when I had the unit on the work table after changing a connector on the rear panel. In the past I had been doing recalibrations while still all connected.

These seem like small issues when reading about them. But they can be important, very important for best operation.

[tortugaranger]

I also read about disconnecting interconnects when doing recalibration. I did the recalibrations yesterday when I had the unit on the work table after changing a connector on the rear panel. In the past I had been doing recalibrations while still all connected.

With the current V3 model you shouldn't have to disconnect the inputs/outputs to run calibration. All audio I/O gets disconnected internally within the V3 during cal. That doesn't guarantee that  you won't get some hum or noise coming out of your amp if it's on. Amps tend to not be too happy about having nothing connected to them.

Rather than disconnect all interconnects, better to simply have your amp powered down before initiating a cal cycle.

[stvnharr]

Here is another, and hopefully final, update.
When I began to have the dropouts, I then also started to do recalibrations. I didn't do them properly and then began to have balanced sound issues with the left channel, secondary controller, always a bit louder than the right channel, primary controller.
So yesterday I tried a new way of recalibration. I ran the recalibration cycle twice. I then power cycled and selected the next impedance and recalibrated twice, etc.
This method has given me perfect balance of sound now. And it is actually better than ever as I am now using the proper impedance setting for best sound.
I can't do anything about the random dropouts, but now I know all about recalibration. So everything now is All Good.

[tortugaranger]

Here is another, and hopefully final, update.
When I began to have the dropouts, I then also started to do recalibrations. I didn't do them properly and then began to have balanced sound issues with the left channel, secondary controller, always a bit louder than the right channel, primary controller.
So yesterday I tried a new way of recalibration. I ran the recalibration cycle twice. I then power cycled and selected the next impedance and recalibrated twice, etc.
This method has given me perfect balance of sound now. And it is actually better than ever as I am now using the proper impedance setting for best sound.
I can't do anything about the random dropouts, but now I know all about recalibration. So everything now is All Good.

Your last note got me thinking about a change that was made earlier this year to the V3 firmware that handles how the calibration data is stored, read, and used. Cycling power after running one or more calibration cycles may in fact be needed to get the firmware to reinitiate properly with the latest cal data set. This is something I intend to look into once I'm on the other side of the move to Oregon.

[stvnharr]

All I can add is that when I power cycled after one calibration and then did a second calibration, it resulted in sound balance issues with the secondary controller being louder. It was that way with all 4 impedance settings that I did.

When I power cycled after doing two calibrations, I got perfect sound balance.

[stvnharr]

Update 2 weeks after I recalibrated the controller.
It has been sonic bliss since I did the recalibration.

Then today I got curious about the discs that had given me the issues, Channel Classics SACD Mahler symphonies. So I played one. Nothing has changed. At the big climaxes the secondary board goes out. This happens when I am playing the sacd layer. I tried different impedances and different attenuation levels with the same results.
Because I could isolate where on the disc that the dropout would occur, I could switch to the cd layer and re-play the same passage. There were no issues with the climaxes on the cd layer.
I never had issues with the sacd layer on these same discs and passages when I used a single ended controller board.
SO, hard to really know if the issue is something on the sacd layer or something with the secondary board. It is always the secondary board that drops out, never the primary board.

I found that I could change impedance on the fly and sometimes the sound would return in full. At other times this did not work. To regain full sound with an impedance change, a power cycle would be needed. So then there is a short pause of a few seconds.
However, I did learn that the secondary board would restore itself without recalibration. So at least that is a plus.

More investigations will need to be done. The single ended success was with different, earlier version controllers, to what I have now.

[stvnharr]

New Update on this issue:
Given that I had previously listened, in single ended, to the discs that now cause issues with balanced control, I reconstructed the single ended volume control with each of the new V3 boards. I was able to play the discs without issue. Or I should say that I chose a disc passage that would not play in balanced control without the secondary board dropping out, and this passage played without issue with each of the V3 boards configured for single ended play.

The next thing I did was to change the Orchard Audio Active RCA to XLR converter, as I have 2 of them. The same disc passage had the same result in balanced control with the second converter, with the secondary dropping out.

Thus, in order to play these discs I will need to reconfigure the balanced control box to be able to play single ended with the primary board. This is not a real big thing as it involves drilling just 4 holes for rca connectors. I already have an open holes for the rca input connectors and have already installed these connectors.
When I wish to play a disc that is known to not play with balanced control, I can change the cabling and play in single ended. I invested too much in the balanced control option to just go back to full time single ended play with the more complicated cabling and 2 active converters on a full time basis.
Given that I pretty much know the few discs that have the issue, it is a fairly minor compromise. It is about the best that I can come up with at the present time. Seems a bit strange, but it is what it is.

[stvnharr]

In reading the V4 update above, it looks promising for balanced operation in the future.