TOPIC:

Hi Devteam,

Just testing a balanced Flex here and 1 channel has 35dB more distortion. Any thoughts on what could cause this?


I swapped cables, rebooted, changed presets, and nothing improved. Is there anything I could be doing wrong that would cause this?

Thanks,
Mike

Hi Mike

Something is odd here.. Are you 100% sure that all filters are disabled on the outputs and not something odd of routing? Could be a simple measurement artifact when you have crossovers enabled or something else. 
Not sure what else to suggest but starting a support ticket with our team to have the official channel of support. Here is the portal: support.minidsp.com 

DevTeam

Everything is bypassed on all channels. All levels are the same. I’m using the same cable and same input channel on my analyzer for all channels. Just moving the TRS connector to each output and enabling the 1khz sine wave in the generator. 

Could you please describe a simple measurement setup how to check if the device shows the distortion?
Hopefully this can be done with some software using the PC only, rather than having special equipment.

You need an audio analyzer to take accurate measurements like this. 

For reference, this case is being investigated by our team/engineers that indeed pulled the APX555 logs for this unit showing that all channels passed QC with all the correct levels. 35dB would obviously fail all the THD limits. Something else is going on here... 
Before making too many assumptions that the design is flawed or that all units are like this, let's maybe investigate what is the issue with this specific unit and we'll get back to you guys? 

35dB difference is "NOT" normal, it's NOT part of the design.. It's a hardware failure and once we see the unit we'll understand what is going on. 
At this stage, our team is somehow guessing some failure on a connector or shipping damage. Once we get the unit we will once again know better than theories.  

DevTeam

I tested a 2nd unit. Here’s the THD+N results from all 4 channels. 

 

Mike, thank you for the update.
From your measurements, there seem to be some variations between the channels.
The variations in your measurement may be larger than the +/-3 dB specified by miniDSP, but I think that performance is quite good for all channels.

It would be important to know why one channel had bad THD+N in the original device.
Did you receive any explanations?  I want to be sure that my device does not have a similar problem.
 

This unit is a completely different flex. Not the one that was returned for repair. Regarding the returned unit, MiniDSP has had it back for 9 days now. And no response from them or returned emails. 

Yes this is the 2nd unit in a row now that doesn’t meet spec. I guess I haven’t gotten the golden sample that ASR got for the review yet. 

It also should be noted that all my measurements are with the DSP bypassed. With a simple 100hz LR4 crossover, and 4 bands of PEQ, the distortion went up by 7dB on the best channel. 

Update after 24hrs burn in. Channels 1&2, and 3&4. And this is with the DSP bypassed as well. 

 

I’ll also add that this is not a unit that was a replacement for the bad one. It was a completely different order from a different person than ordered the last one. I still have the return of the damaged one coming at some point. In which I’ll be able to make a 3rd unit measurement. But something tells me that one will be dialled in good. 

Is anyone buying this unit with the intention of not utilizing the DSP section at all, and bypassing it? Because the measurements I shared are with the DSP bypassed. And same with the measurements ASR made. But is this the real world use case with these units? To bypass the DSP section? 

I’ve discovered an increase of 7dB of harmonic distortion with a simple 100hz LR4 high pass filter, and 4 PEQ bands between 150-600hz. I haven’t tested the FIR or Dirac yet to see if the performance degrades even further with higher load on the DSP. 

I suspect that the increased distortion with the DSP utilized is due to an increase in conducted EMI from the DSP to the DAC with higher load on the DSP. Rather than the algorithms on the SHARC being inferior. 

Further investigation will need to be conducted. 

Devteam,
Looks like the new Flex unit is having a few teething problems.  I would like to see the following:
1. I am sure you can solve these problems. I would not place an order Before you solve these problems.
2. On completion of solving these problems can you inform us here on this forum.
3. Also can you inform us EXACTLY when the Flex unit with DIGITAL will be available.

Thank you.

Mivera1 wrote:

It also should be noted that all my measurements are with the DSP bypassed.

To be clear, by 'bypassed' I assume you just mean not using any of the DSP features, rather than modifying the unit in some way to bypass the DSP chip? It will always be in the signal path otherwise I believe, since it is used for volume control.

Have you made measurements on other hardware such that you're certain there are no issues with you measurement equipment or technique?

Bypassing means enabling the bypass button in the PEQ, and using no crossover on all channels. Or simply leaving the PEQ flat so the DSP chip isn’t being utilized at all. Like how Amir measured his unit on ASR. Is that how most people will use their Flex in the real world?

Are you suggesting that the 115 SINAD I measured from the best channel when the unit was new was due to user error with my analyzer? Explain how that is possible?