Hi Jan Peter,
Great this effort!
But that time schedule seems rather optimistic. Such a PSU is a major challenge, even for experienced and skilled engineers. Apart from designing, testing for reliability and EMC compliance is a time consuming business. So I guess end of this year is more realistic?
Cheers
Great this effort!
But that time schedule seems rather optimistic. Such a PSU is a major challenge, even for experienced and skilled engineers. Apart from designing, testing for reliability and EMC compliance is a time consuming business. So I guess end of this year is more realistic? Cheers
We have very skilled engineers...
For pre-compliance EMC measurements we have a tools in house. And an official test-house is 15min driving from our place.
Of course I am not talking about products available from stock!
In about a few months we will know when we will have something available.
For pre-compliance EMC measurements we have a tools in house. And an official test-house is 15min driving from our place.
Of course I am not talking about products available from stock!
In about a few months we will know when we will have something available.
Not that I have nothing against PFC.
It is only that it undoubtfully adds complexity (and cost) to the system, as well as reduced efficiency under some conditions. My experience is that there is no sonic problem with the increased noise due to harmonics with PF < 1. Most well-known and well-reputated PA audio manufacturers have eliminated PFC from their designs due to reduced reliability, etc. But I understand that electric companies prefer that your current consumption is as resistive as possible and that they try to convice us of the advantages of PFC.
Regulations are another thing and there is nothing to argument against them... I remember that catalytic converters were also made mandatory for cars and that doesn't mean that they are any good for performance Any car lover will tell you they are rubbish (taking apart the environmental considerations) ...anyway, don't need to argue more about this. I am getting passionate
Good luck with your new project, Jan Peter! We are all sure that you will end up with a very good product.
Best regards,
Pierre
It is only that it undoubtfully adds complexity (and cost) to the system, as well as reduced efficiency under some conditions. My experience is that there is no sonic problem with the increased noise due to harmonics with PF < 1. Most well-known and well-reputated PA audio manufacturers have eliminated PFC from their designs due to reduced reliability, etc. But I understand that electric companies prefer that your current consumption is as resistive as possible and that they try to convice us of the advantages of PFC.
Regulations are another thing and there is nothing to argument against them... I remember that catalytic converters were also made mandatory for cars and that doesn't mean that they are any good for performance
Any car lover will tell you they are rubbish (taking apart the environmental considerations) ...anyway, don't need to argue more about this. I am getting passionate Good luck with your new project, Jan Peter! We are all sure that you will end up with a very good product.
Best regards,
Pierre
ssanmor said:
Hello all,
Just hooked up my SPS80 (coldamp) SMPS in a test setup.
Connected a 60W 100V lightbulb between + and - rail and hooked up a scope.
SMPS starts up very quick, 0 to 57-58V in about 200ms, startup with and without the load is the same. SMPS is also stable without load, probably as there is some load on the SMPS board itself. No funny noises at startup.
After that, I hooked up some extra cerafine caps at the output (63V 10000uF). This works very nice as well. Startup becomes only a little bit slower but output of the SMPS seems to stay very stable (with the lightbulb load of about 80Watt), no oscillations.
I think adding 10000uF per rail is likely close to max as that 10000uF is charged up in about 200ms from 0 to 60V, this is a slope of 300V/sec which corresponds with a current of about 3A to charge up the Cerafines in 200ms. Add to that the 2A that will flow into the caps onboard of the SMPS and you have a current of 5A that flows during startup. Corresponding with 600W at 60V rail voltage. A bit more caps and it might trip the overcurrent protection.
No time to hook it up to some UcD modules as I leave for a business trip of a week tomorrow.
Will keep you updated.
Gertjan
ssanmor said:
Yes, there is a 20mA load per rail, including the LEDs
Thanks, yes, I have seen the LEDs glowing
When I'm back, I will further test out the SPS80 to make sure it has no issues with the extra 10000uF of the Cerafines.
To do this I plan to use resistor dummy loads and suddenly change the load by switching in some additional resistors. This will allow me to see how the supply reacts to sudden load changes, hopefully no oscillations/ringing on the rails.
Thanks and best regards
Gertjan
ssanmor said:
Yes, there is a 20mA load per rail, including the LEDs
I see indeed a 10k resistor in series with the leds and an additional bleeder of 3.9K, looks like a 2Watt resistor, correct? It has to burn about 1W so may get a bit hot. I also see two caps C11 and C10 that are parellel (one of them per rail) to the two 3300uF caps (per rail). These two caps are 100nF. Actually I think those 100nF caps maybe responsible for some ringing at something like 3-4Mhz in combination with the two 3300uF caps. It may help to have a snubber there as my scope seems to indicate a bit of swithing induced ringing.
Attached is a scope picture of the AC component of the + rail voltage measured at the cerafines (about 10cm away from the SMPS) and loaded with the lightbulb (60W 100V bulb fed with about 120V, connected between + and - rail).
Some ringing can be seen. HF residues are not very large, the scale is 25mV/div, reducing the ringing by adding a snubber maybe nice. The ringing is also there when the cerafines are not used, so I think it originates in the SMPS itself.
Best regards
Gertjan
Attachments
ghemink said:
Finally hooked up an SPS80 SMPS to two UcD400 modules that I use to drive the woofers of the left and right speaker (4 Ohm load for each UcD). The dipole woofers are equalized for almost flat response. This means that low frequencies are more amplified than with normal woofers. I added an ELNA Cerafine 10000uF to each power rail. I monitored the power rails while playing music with heavy LF synthesized bass.
Conclusion, this supply works great. Output voltage is very stable, close to 58V and hardly drops under heavy load. Of course you can see some ripple on the supply rails depending on the music content as of course the impedance of the supply is not 0.
Another very good point of this supply is that it is not very sensitive to a difference in current drawn from + and - rail. When the UcD modules are on but without any input signal, the current drawn from the + rail is less than that from the - rail. With my other KT6 SMPS, this resulted in a significantly larger + rail voltage than - rail voltage (a kind of power supply pumping effect), I had to add additional resistors from + rail to GND to prevent that the + rail became to high. With the SPS80 this is not needed anymore. Very good, very happy with this supply. I will run it for a couple of days (not continuosly of course) to see how it behaves in the long run.
Best regards
Gertjan
ackcheng said:
It sure looks like a promising SMPS and it is quite compact, at least in comparions with the KT6. With the additional ELNAs, HF switching noise is very strongly reduced. Just some switching noise is left with frequency components around 10-20mHz, I think those components will not do any harm as I`m quite sure that the LC filter on the UcD amps will filter them out, will hook up a scope there and check it sometime. Sound is great, however, only used now for the woofers, need to hookup some mid and tweeter amps in the future.
Best regards
Gertjan
Thanks for the review, Gertjan.
We are making some improvements in the SPS80 that will become standard, such as bigger mosfets in TO247 format (with them we have tested the SPS80 up to 1KW _continuous_ with proper cooling) , a thermal protection, etc.
It is nice to know that with added capacitance the measured noise is reduced. However, customers must know that even with the standard capacitors, it was already very low, causing no problems at all when connected to amplifiers, such as noises, whistles, etc. This applies to its utilization with a variety of Class-D modules (of course also no problem with linear amplifiers).
There are also no interferences with nearby tuners, etc, in fact I have one of them with 5 BP4078 modules just by my TV and receiver
If you like how it sound for bass, just wait to hear it with a full-range amplifier ;-)
Best regards,
Sergio
We are making some improvements in the SPS80 that will become standard, such as bigger mosfets in TO247 format (with them we have tested the SPS80 up to 1KW _continuous_ with proper cooling) , a thermal protection, etc.
It is nice to know that with added capacitance the measured noise is reduced. However, customers must know that even with the standard capacitors, it was already very low, causing no problems at all when connected to amplifiers, such as noises, whistles, etc. This applies to its utilization with a variety of Class-D modules (of course also no problem with linear amplifiers).
There are also no interferences with nearby tuners, etc, in fact I have one of them with 5 BP4078 modules just by my TV and receiver
If you like how it sound for bass, just wait to hear it with a full-range amplifier ;-)
Best regards,
Sergio
ssanmor said:Thanks for the review, Gertjan.
We are making some improvements in the SPS80 that will become standard, such as bigger mosfets in TO247 format (with them we have tested the SPS80 up to 1KW _continuous_ with proper cooling) , a thermal protection, etc.
It is nice to know that with added capacitance the measured noise is reduced. However, customers must know that even with the standard capacitors, it was already very low, causing no problems at all when connected to amplifiers, such as noises, whistles, etc. This applies to its utilization with a variety of Class-D modules (of course also no problem with linear amplifiers).
There are also no interferences with nearby tuners, etc, in fact I have one of them with 5 BP4078 modules just by my TV and receiver
If you like how it sound for bass, just wait to hear it with a full-range amplifier ;-)
Best regards,
Sergio
Hi Sergio,
Thanks for the additional remarks, Yes, I agree, noise even without the extra caps is low. I just happen to have a digital scope with 12 bit resolution and a spectrum analyzer so I can measure very low noise levels (hanging the scope on the power rails, using AC coupling and a 200mV measurement range). I`m kind of a freak and try to get the noise as low as possible (within reasonable limits and with only small extra investments, although the ELNAs are expensive but I had them lying around so they are basically at 0 cost for me to use).
In case of UcD however, the noise will be heavily filtered out by the LC filter on the UcD, I think the L was 300nH and the C is 470uF, so that would give a quite low roll-off point so my extra filtering is probably overkill.
In fact, I like the current SPS80 so much that I`m thinking of getting additional modules (the smaller version) to power my mid range and tweeter amps seperate from the woofer amps. Very decadent, 6 channels with each their own SMPS, not cheap but I like the regulation that the SMPS gives.
Best regards
Gertjan
The thing is that the SPS80 PSU is extremely efficient.
Just an example: if you connect 2 or even 4 typical class-D modules they will get much warmer than the supply even with no signal. In fact the supply remains cold.
Pushing two of our BP4078 400W channels hard with a single SPS80, all screwed to a very small alluminium case, we have never seen a temperature rise of, say, 15ºC above ambient in the supply.
The fan output is more for the amplifiers than for the supply itself, and it is rarely used.
Obviously, when we did our 1KW test, we screwed the supply to a small plate and used a low speed CPU fan to blow some air at the top of the supply, mainly to cool the inrush limiter and transformer a bit. See attached photo: it was delivering around 990W. After some minutes you couldn't touch the loads heatsink, but the PSU plate was 20-25ºC above ambient
Just an example: if you connect 2 or even 4 typical class-D modules they will get much warmer than the supply even with no signal. In fact the supply remains cold.
Pushing two of our BP4078 400W channels hard with a single SPS80, all screwed to a very small alluminium case, we have never seen a temperature rise of, say, 15ºC above ambient in the supply.
The fan output is more for the amplifiers than for the supply itself, and it is rarely used.
Obviously, when we did our 1KW test, we screwed the supply to a small plate and used a low speed CPU fan to blow some air at the top of the supply, mainly to cool the inrush limiter and transformer a bit. See attached photo: it was delivering around 990W. After some minutes you couldn't touch the loads heatsink, but the PSU plate was 20-25ºC above ambient
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.