Review and Measurements of Topping DX3Pro DAC and Headphone Amp

[kprog]

I would also like to know why they are broken!
I want to buy one but .....
If it is due to excess temperature, I can make some openings in the upper part of the box to evacuate heat.

 

[finneybear]

Are you afraid that the reverse current, in our case the current from the potential of 100 mV, can damage 4493? Can we install diodes instead of capacitors to prevent this?

Not very sure what "reverse current" you meant but probably you were right. I searched internet for a while and found this similar to what 4493's output looks like:

This is a differential SCF. It's basically a sample & hold circuit. The OP is responsible for keeping the cap CA voltage stable.
So the output load impedance is very important. It has to be high enough to keep the OP current output minimal.
If you have reverse current at the output then it will disrupt the charge of CA.

 

[simonchretien]

Whos gonna be the first to drill some holes?

 

[finneybear]

I'd rather have the hottest girl, but I can't afford such an expense, I'm affraid.

I measured the AK4490 surface temperature with a (K-Type) thermocouple connected to a digital multimeter.
DAC looks like this:

LOL. Then probably you should get a black one instead? Black is hot yet runs cool.

Obviously you have the magic to run things hotter than they should have been.

 

[finneybear]

I've got parts rolling in.

Kemet KO-CAP 100uf that fit perfectly for the DAC output.
Some transformers to play with

More caps coming in. I follow up to Finnybear's advice. Still waiting for the polymer radials and the Panasonics for the digital input.

It's spring festival here.. I tried to persuade many shops to still send me some with the last open shipping company. but it's difficult

I really want to see whether the cap swap will improve the stability issue. When those DC-DC converters and regulators look OK, those power caps are bare minimal. Each 4493 can consume as much as 0.25W which is about 5X of mobile ESS chips. The power supply to 4493 is a concern to me now.

 

[777]

Yea babe, we have the music now

 

[finneybear]

Re tube amps:

SE transformers have gapped cores to be able handle DC.

Push-pull transformers operate in full diff mode, with same voltage at two ends of the winding, resulting in net zero voltage across transformer.

Small signal transformers will saturate with as little as 40mV offset.

You really need to know what you're talking about when dishing out advice. I'm not the one who's confused. Your post on mono mode is simply wrong too. Even voltage out DACs have theoretical benefits when used properly. Go read a few datasheets and application notes to refresh your knowledge please. Not trying to be rude, but you're simply wrong on a number of things.

Unfortunately that I do have the datasheet for this permalloy transformer. It can easily withstand 10mA input/output current.
Now, just take 4493 full swing output at 5V, the transformer has a 10K impedance : 5V / 10K = 0.5mA. So 0.5mA bias can saturate a 10mA transformer?
I kept telling you that the transformer has 10K impedance and 4493 will emit very little DC but now you are even saying 40mV offset will saturate the transformer? LOL

Transformer has long been used for DC isolation. I only brought up tube amp as an example. Dont want to get too technical. In fact, I had designed a 300B SET amp with a group of people. One guy has been doing output transformers for over 10 years.

Those permalloy transformers were originally designed as input transformers. Input transformers can not take 3V or 5V DC input?

 

[finneybear]

Yea babe, we have the music now

My God! What are you doing here? LOL

Actually you do not need bi-polar here. Silmic will work just fine.

 

[777]

My God! What are you doing here? LOL

Actually you do not need bi-polar here. Silmic will work just fine.

I don't want "just fine" . Believe me, deserves bi-polar. I finally have the wanted bass, a lot of bass shades.

 

[MacDac]

LOL. Then probably you should get a black one instead? Black is hot yet runs cool.

View attachment 20963

Obviously you have the magic to run things hotter than they should have been.

Well, I used to utilize that magic for picking up girls some 30 years ago, but unfortunately it's not the case anymore
What I have left for my autumn years is disassembling broken Chinese devices and blowing wind on them. Now, call it life, if you will...

Btw, DAC is working flawlessly now, regardless of ambient temperature. Moreover, after a minor recap it sounds really great. And I don't mean the whizzing sound of the fan

 

[Cidious]

What kind of transformers are those?

The ones in the photo are 10k:10k Jensen clones. I've got a pair even better quality coming in. I got the taobao recommendation from Pavel a bit earlier in this thread. The shop owner is really cool ! And delivers quality and service. He started out cloning buy now his own versions are better quality and after spring festival he has some new products coming. I'll give them a swing.

I worked taobao hard to get the capacitors needed during this spring festival time. And it seems like I've got most of them incoming. The only thing i couldn't get was a 100uf polymer cap for the 3.3V dac power supply. I've got a 220uf though... might squeeze that one in, give it a try if it fits.

I'll replace all the tantalums with polymers kemet co-caps. I found the right sizes and values. And I'll replace the 10uf at the opamp with 33uf ko-cap too.

 

[Pavel]

The ones in the photo are 10k:10k Jensen clones.

Yes, I am also well acquainted with this manufacturer. I bought more than 60 transformers from him. But you bought the wrong model! jensen clone sounds bad. you had to buy exactly what I indicated. This model is called R3.
https://www.audiosciencereview.com/...dx3pro-dac-and-headphone-amp.4967/post-142624

 

[pejakovic1]

Well a mean well 15V 1.66A didn't help, the device shut down in the middle of the game. Was keeping it on A-C too. So this is not the solution.

 

[yue]

Well a mean well 15V 1.66A didn't help, the device shut down in the middle of the game. Was keeping it on A-C too. So this is not the solution.

There's something seriously wrong. I encountered various issue as well.
you can send your device to @amirm. He is interested in purchasing a dead unit to investigate the issue.

@amirm, you should put a big warning in your review article.

 

[pejakovic1]

There's something seriously wrong. I encountered various issue as well.
you can send your device to @amirm. He is interested in purchasing a dead unit to investigate the issue.

@amirm, you should put a big warning in your review article.

Mine isnt dead, it just shut down, i turned it back on with remote after.

 

[splkn]

Yea babe, we have the music now

It was necessary after soldering four capacitors and four shunts to try. Now it is too late, do not otpaivat back this hedgehog. And on power 1612 what capacitors put? We must also pick up a good ceramic capacitor. In datashit on 1612 write:
"Use low-ESR, ceramic bypass capacitor"

 

[yue]

Mine isnt dead, it just shut down, i turned it back on with remote after.

Mine doesn't shutdown. it just stop producing sound in the middle of a track. Apart from that all UI behaves like normal. Rebooting doesn't fix it. One need to Unplug and Replug the power cord to make it work again.

There're also a couple of times it can't play DSD files and need unplug/replug power cord to fix the issue.

 

[simonchretien]

There's no way anyone can recommend this, the amount of dead units and random problems is insane on this device.
I cant even send mine back because the ebay seller has a 20% restocking fee on returns (didnt notice when buying), plus the shipping is way too expensive. I'll have to pray my 200$ bucks device keeps working.
I mean I knew I was buying a chineese product directly, never imagined any product could be this bad.
They should stop selling the unit until they can fix their crap.

 

[Sangram]

Unfortunately that I do have the datasheet for this permalloy transformer. It can easily withstand 10mA input/output current.
Now, just take 4493 full swing output at 5V, the transformer has a 10K impedance : 5V / 10K = 0.5mA.

Once again, you are basically wrong.

10k impedance only applies to AC currents. DC currents are determined by the winding DCR, not the impedance. You do know the difference between impedance and resistance, don't you? If you can confuse these two, I'm not any of your advice is sensible.

Since you claim to have 'Designed' SET amplifiers (and I'm being generous and as courteous as possible here), you will know the transformer is different from a PP transformer. They are built to handle a large amount of DC current without saturating, by gapping the core. This also leads to needing a lot more copper and iron for the same amount of output power, which is why SE transformers are much heavier than PP ones. But you knew this already since you 'designed' such an amp, yes?

To go back to the example here, I am looking at a Lundahl 1690 datasheet that states primary DCR is 150 ohms. With 3.3V across the winding, you will have 20mA of steady state current. At 1.65V as per your 'advice' of omitting cap and connecting single ended DAC output directly to transformer winding, the core is already saturated at 10mA. All I/O application datasheets will tell you that DC current is a no-no as even if a small amount of current reduced output capability and increases distortion.

 

[finneybear]

Once again, you are basically wrong.

10k impedance only applies to AC currents. DC currents are determined by the winding DCR, not the impedance. You do know the difference between impedance and resistance, don't you? If you can confuse these two, I'm not any of your advice is sensible.

Since you claim to have 'Designed' SET amplifiers (and I'm being generous and as courteous as possible here), you will know the transformer is different from a PP transformer. They are built to handle a large amount of DC current without saturating, by gapping the core. This also leads to needing a lot more copper and iron for the same amount of output power, which is why SE transformers are much heavier than PP ones. But you knew this already since you 'designed' such an amp, yes?

To go back to the example here, I am looking at a Lundahl 1690 datasheet that states primary DCR is 150 ohms. With 3.3V across the winding, you will have 20mA of steady state current. At 1.65V as per your 'advice' of omitting cap and connecting single ended DAC output directly to transformer winding, the core is already saturated at 10mA. All I/O application datasheets will tell you that DC current is a no-no as even if a small amount of current reduced output capability and increases distortion.

Ta-Da! I left this gotchya on purpose to see whether you would bring up the impedance vs resistance thing.
At the same time, I also reminded you that those were input transformers. Unfortunately you still fell into the trap! LOL

The truth is, check out this Jensen transformer datasheet:

http://www.jensen-transformers.com/wp-content/uploads/2014/08/jt-11p-1.pdf

The resistance is 1.5K. Enough is enough?