Review and Measurements of Topping DX3Pro DAC and Headphone Amp

[777]

The most probably it's about "inrush current". An NTC(thermistor) inside the power adapter will solve that.

 

[Sangram]

Good response from Topping, finally.

I would really like to see possibility of firmware upgrade kit with new microcontroller, or a replacement board which can either be cross-shipped or paid for at cost by certified owners (on proof of purchase). Soldering the QFN package is possible by visit to PC technician. I would really like to see a fix for disabling the 0dB mode from remote without a second press or some kind of confirmation. That is my big issue, since I am connected to very loud speakers and have some consideration for neighbors.

Now my unit has never actually had this standby issue (knock on wood). Mine is a very early production unit - 1808xxx, runs off a very basic linear supply and has never been connected to Windows except for initial testing, which is also the only time the included power supply was ever used. It means I have never had the XMOS update released by Topping. It has also been connected to a pair of headphones exactly once. However sometimes the display will go off when the source powers down. Pressing power button awakens it.

 

[TP-Wong]

The most probably it's about "inrush current". An NTC(thermistor) inside the power adapter will solve that.

So we did not rule out the possibility of adapter failure causing all this. And when we call back defect units, we will ask our users to send back whole set including power adapter.

 

[Sangram]

The most probably it's about "inrush current". An NTC(thermistor) inside the power adapter will solve that.

I am not sure about that, because all voltages inside the DACs have regulators, unless the design is marginal it is unlikely the regulators will fail (which is the only reason for downstream problems).

I am guessing a failure of the ST mcu. After all, it's ST. Everything they make is prone to blowing up at some point. There is a jtag on the PCB so waking it up should be possible with a basic kit

 

[Klenfo]

It is a pity that 15v are used here, not 12v. There are a lot of good 12v psu and easy to find. But where to get a good 15v psu?

 

[Yuno]

So we did not rule out the possibility of adapter failure causing all this. And when we call back defect units, we will ask our users to send back whole set including power adapter.

I doubt adapter has anything to do with this. There have been multiple users here who purchased mean well aftermarket power supply to test out if it helps out, and I believe nothing has changed.
My issue is a bit different than the one you are talking about, but I feel it is connected somehow: my unit goes into suspend mode by itself every once in a while (every 2-3 weeks), happened exactly 4 times since I purchased it. I have no issues with waking it up by pressing power on button (so far). Few weeks ago I bought mean well GS25A15-P1J (15V 1.66A power adapter) and exactly 2 days ago experienced another self-shutdown. So in my case changing power supply did not help.

 

[TP-Wong]

Good response from Topping, finally.

I would really like to see possibility of firmware upgrade kit with new microcontroller, or a replacement board which can either be cross-shipped or paid for at cost by certified owners (on proof of purchase). Soldering the QFN package is possible by visit to PC technician. I would really like to see a fix for disabling the 0dB mode from remote without a second press or some kind of confirmation. That is my big issue, since I am connected to very loud speakers and have some consideration for neighbors.

Now my unit has never actually had this standby issue (knock on wood). Mine is a very early production unit - 1808xxx, runs off a very basic linear supply and has never been connected to Windows except for initial testing, which is also the only time the included power supply was ever used. It means I have never had the XMOS update released by Topping. It has also been connected to a pair of headphones exactly once. However sometimes the display will go off when the source powers down. Pressing power button awakens it.

This is what we are considering. More and more things were done by MCU, but firmware of our MCU is not able to update, this is a potential risk.

 

[TP-Wong]

I doubt adapter has anything to do with this. There have been multiple users here who purchased mean well aftermarket power supply to test out if it helps out, and I believe nothing has changed.
My issue is a bit different than the one you are talking about, but I feel it is connected somehow: my unit goes into suspend mode by itself every once in a while (every 2-3 weeks), happened exactly 4 times since I purchased it. I have no issues with waking it up by pressing power on button (so far). Few weeks ago I bought mean well GS25A15-P1J (15V 1.66A power adapter) and exactly 2 days ago experienced another self-shutdown. So in my case changing power supply did not help.

Since it could not turn on, that means the adapter or built-in power circuit defect. Your situation proved that your adapter is not defect. Finding out why that happen make more sense, that's why we need defect items back.

 

[simonchretien]

TP-Wong, any lead as to why the units could randomly enter standby mode ? Not being able to get out of standby is something, but for most people it's just shutting down itself for no reason.

 

[TP-Wong]

It is a pity that 15v are used here, not 12v. There are a lot of good 12v psu and easy to find. But where to get a good 15v psu?

We did not do long test for DX3 Pro working on DC12V so I'm not sure it is suitable or not. But sure it works well with our LPS P50 which is coming soon, same size with D50 but double height of it, providing 1xDC15V, 2xDC5V output. We bring the P50+DX3 Pro combination to Canjam Shanghai for show last year.

 

[TP-Wong]

TP-Wong, any lead as to why the units could randomly enter standby mode ? Not being able to get out of standby is something, but for most people it's just shutting down it.

There must be some problem in DX3 Pro or power adapter,

I don't know where, but I will send out result and solution here when we have. Some users here had changed adapters but no improvement, so I think the problem is more likely in DX3 Pro.

 

[Sangram]

This is what we are considering. More and more things were done by MCU, but firmware of our MCU is not able to update, this is a potential risk.

Hi

I can get a new chip soldered. More advanced users will be able to program it using USB-I2C using the onboard header if you provide the code to access the MCU - but I think that's too advanced, so maybe just sending replacement chip with instructions on removing and replacing it. Not a very difficult job, any tech with a hot air station and good hands can do it in about 5 minutes, and it's cheaper than sending the whole unit back to factory.

Good Luck. The DX3 is a good sounding unit in spite for few basic mistakes in design and is a good deal at the price. Maybe next time get a good analog designer for the output stage, or ask AKM to lend a hand. The pop noise can quite easily be cured, so the relay activation won't be as much of a problem.

Re 12V - it should be fine and put little less strain on the regulators, but there might not be enough headroom for the 10V rails. 13V or greater should be fine as most regs are fine with 3V across them. I have my LPS set to 14V and it works great.

 

[777]

TP-Wong, did you change the capacitors type at some point ? Maybe another one with lower ESR ? I was add two capacitors of 220uf/10v with ESR=0.14ohm in the analog power supply and....bang. Never could start without a voltage source with over 2.6 Amps. So, probably the cause is an over-current somewhere.

 

[TP-Wong]

Hi

I can get a new chip soldered. More advanced users will be able to program it using USB-I2C using the onboard header if you provide the code to access the MCU - but I think that's too advanced, so maybe just sending replacement chip with instructions on removing and replacing it. Not a very difficult job, any tech with a hot air station and good hands can do it in about 5 minutes, and it's cheaper than sending the whole unit back to factory.

Good Luck. The DX3 is a good sounding unit in spite for few basic mistakes in design and is a good deal at the price. Maybe next time get a good analog designer for the output stage, or ask AKM to lend a hand. The pop noise can quite easily be cured, so the relay activation won't be as much of a problem.

Re 12V - it should be fine and put little less strain on the regulators, but there might not be enough headroom for the 10V rails. 13V or greater should be fine as most regs are fine with 3V across them. I have my LPS set to 14V and it works great.

Soldering or programming MCU is already very hard for normal users, so it is better to make the MCU easier to change, or add USB port to make firmware update possible. We discussed a lot in our product meeting.

 

[TP-Wong]

TP-Wong, did you change the capacitors type at some point ? Maybe another one with lower ESR ? I was add two capacitors of 220uf/10v with ESR=0.14ohm in the analog power supply and....bang. Never could start without a voltage source with over 2.6 Amps. So, probably the cause is an over-current somewhere.

Sorry we already got result so I need to edit this post.

Turn on>>>>Over current in OPA2140>>>>Over load of TPS54331>>>>Shut down. This is a cycle and happen in very short time, so you can only see"not able to turn on"

 

[pejakovic1]

Over current, should be one possibility lead to defect: Over current somewhere>>>>Over load or over heat of TPS54331>>>>Dead.

Most users that reported completely dead units were using < 40 ohm headphones. If I'm not wrong, testing the unit with low impedance might be worth the shot.

 

[kukocz]

We did not do long test for DX3 Pro working on DC12V so I'm not sure it is suitable or not. But sure it works well with our LPS P50 which is coming soon, same size with D50 but double height of it, providing 1xDC15V, 2xDC5V output. We bring the P50+DX3 Pro combination to Canjam Shanghai for show last year.

Could you reveal a little more info about P50 - release date, price range? It could be really interesting LPS.

What about fixed line out mode adjustable via remote. Have you came with any idea how to fix it (i.e. firmware update)?

 

[dmac6419]

See that's why i like this site,straight English i can understand,not like all them other so called audiophile sites,just honest reviews,nothing but the facts mam.

 

[ChuckT]

During my modding of the DX3 board, one earphone channel went dead (could be me as I replaced the headphone opamp capacitor), however, the usb also become non functional. However, bluetooth and dac output is perfectly functional and I am getting very good sound from it. In fact, I took the Topping Dx3 to a very expensive dac shootout. The top winning dac is the Lampizator R2R tube out dac which cost upward of 6k usd. The DX3 dac uses the original wall wart and did not seat at the bottom. The bottom dac was the Audio GD R2R which I also own (very flat sounding without body). Even if the DX3 was not up to the top dac level, it was certainly a listenable among the very expensive dac.

Now few days ago I use the linear power supply for my Intel NUC which output 19V into the DX3 and I can hear further improvement on the overall sound. The bass become more punchy. It has be running few days now with 19V and no problem so far.
PS. I have request a replacement board from the manufacturer for a replacement board (which I will pay for) . No reply from them yet.

 

[tired_guru]

Maybe topping should spend more time on testing ? Maybe better/bigger/test plans/scenarios/testers crew is needed ? Products measure and sound great but my feeling is they simply put new models to the customers too early thus many problems during first batches.

Hopefully new upcoming D70 will be checked out enough to avoid this kind of situation. In this price range this would be not cool at all having similar issues to dx3pro.

This is not only great benefit to the customers but also big cost saving solution for the manufacturer (all these gurantee responsibilities for flawed units, wasted time for shipping back+extra costs). It's better to run small amount of units to testers around the world first, and after like 2-3 months collect them back to see what & why happened and do another iteration of the unit - either software or hardware design.