SMSL SP400 Review (headphone amp)

[Guy.Cohen]

One of my amplifiers is SE class A based on 4 6336 output tubes.
I made it by myself about 50W.
Everything is extremely hot and no fault. No capacitors, resistors issues and if course no valves issues

 

[Roland68]

OK, so replacement of all these parts every 5-6 years is a good engineering?
I have a friend who had this amplifier. It was also blow output transistors while pushing Apogee.
It was always on the floor with no ventilation issues, just always has failure

But you can't blame that on the power amplifier.
If you buy a sports car with a sophisticated engine, you don't have to be surprised if it breaks down after just 100,000 km because you only had the oil changed.
Capacitors, diodes, transistors, relays, etc. are wearing parts. Most people don't realize this because these parts can last 20-50 years under normal to low loads.
The Krell KSA250 has 700 watts of power loss, you complain about the heat with the SP400 at around 5-10 watts.
With such a large power amplifier, regular overhaul makes sense because if there is a defect, a lot more can break, including the speakers, for example. In this case there is also a big difference whether the power amplifier runs 24/7 or just for a few hours on the weekend.
But that's off topic now.

If one of the power supplies on the SP400 actually fails, they can be easily replaced and cost around 15-20 €/$, no problem.

One of my amplifiers is SE class A based on 4 6336 output tubes.
I made it by myself about 50W.
Everything is extremely hot and no fault. No capacitors, resistors issues and if course no valves issues

And again this indication "Hot", that's everything from to. Electron tubes can reach 250° without any problem, and that's ok. How warm are your capacitors really and how many hours of operation have they already been running?
Buy an infrared thermometer laser gun, costs around € 12-20 and can measure at least up to 350-400°. How much time have you wasted worrying about how warm your devices are?

Some of our high energy products run for several thousand hours at around 1000°, which is what I call "hot".

 

[Guy.Cohen]

From the beginning I said that for me the warmth of the SMPS of the SMSL seems too much for only half Amper consumption, I was not too worry and I was wondered if someone has a failure SP400, just in case it is a real problem.

Yes, it is common, specially for western luxury cars, watches etc to be serviced event short periods.
Price of being comfort (and I am not suggesting buying Lexus).

If you will read the datasheet of 6336 you will notice each valve has filament consumption of 2.5A.
I have 4 6336, and puts about 50W on 8Oohm, so it is really hot.
Each valve section draw about 125mA, so totally of about 1 Amper for all 8 sections. B+ is 400V.
What I am saying each monoblock dissipates about 500W, all together with the front end and driver.
It is extremely how but all capacitors are 105 Celsius.

 

[Guy.Cohen]

I had a thermometer of UNI-T of my friend which is infrared but you always need to put this factor of emissivity I was always not sure what is the correct temperature.
This factor you need to enter has a lot of impact on the temperature

Do you have a suggestion for a good thermometer that can just read simply ?

 

[Roland68]

From the beginning I said that for me the warmth of the SMPS of the SMSL seems too much for only half Amper consumption, I was not too worry and I was wondered if someone has a failure SP400, just in case it is a real problem.

Yes, it is common, specially for western luxury cars, watches etc to be serviced event short periods.
Price of being comfort (and I am not suggesting buying Lexus).

If you will read the datasheet of 6336 you will notice each valve has filament consumption of 2.5A.
I have 4 6336, and puts about 50W on 8Oohm, so it is really hot.
Each valve section draw about 125mA, so totally of about 1 Amper for all 8 sections. B+ is 400V.
What I am saying each monoblock dissipates about 500W, all together with the front end and driver.
It is extremely how but all capacitors are 105 Celsius.

I wasn't talking about luxury cars, but about sophisticated sports car engines, which is something completely different.
We used to repair, pump and manufacture power tubes for radio and TV stations, so I know completely different performance classes.

But we're talking past each other, your capacitors in the tube power amplifier can have an operating temperature of 40° or 90°, without measuring you don't know anything. You have different heat zones with such a device, which should be taken into account when setting it up. Sometimes they lose 20-40° of component temperature with 1 cm more distance from the tubes.

I had a thermometer of UNI-T of my friend which is infrared but you always need to put this factor of emissivity I was always not sure what is the correct temperature.
This factor you need to enter has a lot of impact on the temperature

Do you have a suggestion for a good thermometer that can just read simply ?

Just look on Amazon and the reviews. I've been using this for my needs (kitchen, hobby, etc.) for a few years:
Eventek Laser Infrared Thermometer, -50 to +550°C
I have already compared it a few times with the calibrated industrial device (up to 1750°) in our company, and it always differed by less than 1°.

 

[Guy.Cohen]

Thank for the thermometer link, it is cheap, I will buy it.
I mentioned luxury cars as they usually have more sophisticated engine but I guess you meant really sophisticated, not Lexus, Cadillac or maybe even BMW.
However it is true that cars from the west maybe more comfortable and more sophisticated have more maintenance needs than those coming from the east.

 

[Guy.Cohen]

I wasn't talking about luxury cars, but about sophisticated sports car engines, which is something completely different.
We used to repair, pump and manufacture power tubes for radio and TV stations, so I know completely different performance classes.

But we're talking past each other, your capacitors in the tube power amplifier can have an operating temperature of 40° or 90°, without measuring you don't know anything. You have different heat zones with such a device, which should be taken into account when setting it up. Sometimes they lose 20-40° of component temperature with 1 cm more distance from the tubes.


Just look on Amazon and the reviews. I've been using this for my needs (kitchen, hobby, etc.) for a few years:
Eventek Laser Infrared Thermometer, -50 to +550°C
I have already compared it a few times with the calibrated industrial device (up to 1750°) in our company, and it always differed by less than 1°.

Hi Roland,

I bought UNI-T infrared thermometer model UT301A+.
The measurements shows case temperature of 42.2 Celsius (bottom).
I guess internally it might reaches 50 degrees. Not terrible but for me it seems high. No good reason for that.

There is an emissivity parameter, that I put 0.95. I am not sure about that value, you know the metal of the SP400 is plated black, so I think 0.95 is fine..
What you are using for this parameter in your thermometer?

Thank,
Guy

 

[Snoopy]

If you really think the SP400 should run cooler, it shouldn’t be that difficult to open it up and mod the case . Maybe remove the glass covering the top. Drill some holes in the case as exhaustion holes or put a USB fan on top.

But even 50 degrees assuming you are not running air conditioning in your room shouldn’t be a issue at all.

 

[Guy.Cohen]

I don't think this 50 degrees is an issue. I think it is too much for such a device that should be more efficient. It is surprisingly mainly the SMPS that is hot.

 

[Snoopy]

I don't think this 50 degrees is an issue. I think it is too much for such a device that should be more efficient. It is surprisingly mainly the SMPS that is hot.

I imagine the sealed case with the glass on top are the perfect isolated box to let not much heat out.

 

[Roland68]

Hi Roland,

I bought UNI-T infrared thermometer model UT301A+.
The measurements shows case temperature of 42.2 Celsius (bottom).
I guess internally it might reaches 50 degrees. Not terrible but for me it seems high. No good reason for that.

There is an emissivity parameter, that I put 0.95. I am not sure about that value, you know the metal of the SP400 is plated black, so I think 0.95 is fine..
What you are using for this parameter in your thermometer?

Thank,
Guy

0.95 is ok.
At what ambient temperature did you measure the device?
What is the temperature on the top of the device?

 

[Guy.Cohen]

0.95 is ok.
At what ambient temperature did you measure the device?
What is the temperature on the top of the device?

I didn't open the device, what I have measured is from outside, near the SMPS. It is at a room temperature but here, now, it is about 25 degrees Celsius

 

[Roland68]

I didn't open the device, what I have measured is from outside, near the SMPS. It is at a room temperature but here, now, it is about 25 degrees Celsius

I asked for the temperature of the top of the device, not inside.
You said 42.2 Celsius bottom.

 

[Guy.Cohen]

I asked for the temperature of the top of the device, not inside.
You said 42.2 Celsius bottom.

Ahh... I checked. Not important, top or bottom, or side, but near the SMPS area it is about 42 Celsius. Could be 2 Celsius variation.

 

[Guy.Cohen]

I asked for the temperature of the top of the device, not inside.
You said 42.2 Celsius bottom.

I bought a 5k thermistor for my Agilent DMM in order to measure temperature also by coupling. More or less the same, 40.1 Celsius at bottom after half an hour of coupling. Gives little better than emissivity of 1 in the UNI-T.
Like a black hole .

 

[Guy.Cohen]

One more: I looked at Topping Audio L70 (for instance). Written DNR of 146db.
I don't think there exists a source with such a DNR. I guess they meant that it is theoretical calculation. They took the maximum swing (RMS) and divide by the measured noise of 0.3uV.
This measured noise is measured by pre amplification and then RTI of the amplifier.
About the noise they are mentioning the procedure but for the dynamic range just claim 146db.
I think SMSL related measurements are more realistic (or direct).

 

[Roland68]

I bought a 5k thermistor for my Agilent DMM in order to measure temperature also by coupling. More or less the same, 40.1 Celsius at bottom after half an hour of coupling. Gives little better than emissivity of 1 in the UNI-T.
Like a black hole .

40° at 25° ambient temperature is completely OK. The internal temperature will only differ by a few degrees and will not be 50°.
I have already written something about the "heat" hysteria, as well as about the stable working temperature, "influence of temperature on the working point", temperature fluctuations and excessive cooling, e.g. due to uncontrolled convection, etc.
Personally, I think it is a bad idea to cut cooling openings in devices that run in a good and stable temperature range, but everyone can decide for themselves.

 

[Guy.Cohen]

At the beginning of my posts I asked if someone has a failure in this device because of the heat. That was my original question.
I don't think the temperature is not acceptable, it will not prolong the life of the internal capacitors for sure but it is still acceptable. I am not going to drill holes at the plastic top or at bottom.
It is a surprise for me that SMPS of relatively low consumption device reaches such temperature. The efficiency of SMPS is normally very high and for such small amount to current the PS needs to be cool, like my DO300EX, that really run cool (and had only one PS). This is the designer choice and maybe the situation is improved in their new H400.

 

[Roland68]

At the beginning of my posts I asked if someone has a failure in this device because of the heat. That was my original question.
I don't think the temperature is not acceptable, it will not prolong the life of the internal capacitors for sure but it is still acceptable. I am not going to drill holes at the plastic top or at bottom.
It is a surprise for me that SMPS of relatively low consumption device reaches such temperature. The efficiency of SMPS is normally very high and for such small amount to current the PS needs to be cool, like my DO300EX, that really run cool (and had only one PS). This is the designer choice and maybe the situation is improved in their new H400.

With the SP400 with this THX circuit I assume that the amplifier section is biased. This is why the heat does not increase significantly during operation under load.
I actually think the DO300EX is a bit too cool. It heats up even less during use than the DO400.

 

[Lucian_Petrasescu]

Does anyone know of an alternative way to remote control the SP400?
The original remote is trash and the down button now barely works.
The SMSL RC-8C remote does not work with the SP400.
Are there other remotes or an Android app or something?