Print

[WB2VVV]

Dunno if that includes Huawei power supplies or not.   The German made B26 amp still uses a single 3kw Huawei 50 vdc switching supply.  He was buying 25 power supplies at a time.   They are cheap, and 100% reliable.
[/quote]

No one has said whether they are QUIET.

For me Huawei is a non-starter in any case, based on the way they pirated Lucent base station technology, and then showed up trying to sell exact copy base stations that looked precisely like Lucent but instead the badge said Huawei. They used to come and visit me when I was a 3rd level officer at a Fortune 10 cellular company. My boss made me see them out of courtesy, even though we both knew that we would never do any business with them.

There are many other power supplies out there, and I don't need to violate my ethics...

[KM3F]

Could consideration be possible to locate the noisy power supply in an Acoustic lined wood container and still have enough air flow and located away from the operating position?

[WB2VVV]

As it turns out I have had an alternative 48 VDC switching power supply on my spares shelf that I bought used and very inexpensively back when I was building my MOSFET amplifier deck several years ago. It was the same kind of chassis like the Eltek Flatpack2, that you had to slide into a frame to make the AC and DC connections. Back then I took one look at the back of it, and those obscure push-in connections, and bought the Mean Well power supply which had sturdy screw terminals.

Well with this talk of Flatpack chassis power supplies that were intended to go in a frame, I just pulled the rest of the bubble wrap off it, and it's an HP ESP120 used for server farms I suppose. The tag says 2950 Watts, 51.4 VDC, 57 Amps. It has all kinds of approval stamps including CE and TUV. This time I took a picture of the tag with my phone and googled the part numbers. It turns out that N4GA suggests that the HP ESP120 is one of the best of these power supplies out there for solid state amplifiers, and that it is also inexpensive on the secondary market. On his website at N4GA.com he provides the pinout of the obscure connector, and he also provides info on how to go inside and turn down the fan speed to make the power supply QUIET. He has been using them this way for some time and is happy with them. Now I have to figure out what to do about those connections. I am thinking that I will mount a sturdy screw terminal strip to the top of the chassis just above the connections, and solder some short heavy insulated solid wire between the obscure connections and the screw terminals - after I go inside and perform his mod on the fans to make them QUIET. It sure seems like I can wedge a heavy gauge solid wire between each pair of connector leafs and solder it in place so wire stays put.

Anyone ever use an HP ESP120 power supply?

[WB2VVV]

Could consideration be possible to locate the noisy power supply in an Acoustic lined wood container and still have enough air flow and located away from the operating position?

I guess anything is possible depending upon one's effort and priorities, and this might make sense so thanks. I briefly considered building an acoustic baffle lined with acoustic absorption foam around my current noisy power supply, and then moved to putting a noise gate in the microphone line. I have a couple sheets of Sonex foam left over from another project, and what would be necessary is a Helmholtz Resonator tuned to the whine of the fans and set up to guide that frequency noise into a damped cavity, and the Sonex foam could be the damping part. Tuning the resonator is a lot more complex. My current noisy power supply is mounted on rubber shock mounts hanging beneath my 2 inch thick solid wood operating bench-top, with the rig and amplifier deck on top. I didn't expect the fans to be as loud as they are, and I expected that the sound would be deflected downwards away from the operating position, shielded from the operating position above by 80 pounds of solid wood bench-top. In fact it is much louder if I bend down and listen under the bench-top where the bench-top probably acts as a sounding board in a piano. Unfortunately these fans have a loud whine to them, and they only operate at their maximum speed. The fan noise was clearly audible between words on the other end of an SSB QSO. Now with the noise gate it's not - but after several years I am asking myself why have so much fan noise in the first place for low duty cycle SSB and CW analog modes...

It is always my opinion with low voltage DC power supplies to keep the DC power leads as short and thick as possible. As to moving the power supply away from the operating position: more distance = more voltage drop. My shack is a one room building in a field at the base of my HF tower, so there's truly no place to hide from the noise.

[AD0AR]

Here's one that I haven't tried, but I have used  a variant of this for 12v applications.  No noise because it is fanless. 
  It does require 240V to achieve the full output of 31.5 A with an adjustable output voltage ranging from 50%-120%, i.e. 24-57 vdc, and 96% efficient!!!

https://www.meanwellusa.com/upload/pdf/UHP-1500/UHP-1500-spec.pdf

Kinda expensive at $448, but available in the USA.

https://power.sager.com/uhp-1500-48-5787756.html

It is also potted for vibration resistance and also has a MINIMUM MTBF rating (mean time between failures of 535,000 hours or about 61 years)
Set it and forget it NO NOISE at all! 

[W9IQ]

Do take note of the aluminum mounting plate (heatsink) requirements necessary to operate the UHP-1500 supply within its derating curve if you are trying to avoid the use of forced air cooling.

- Glenn W9IQ

[WB2VVV]

As it turns out I have had an alternative 48 VDC switching power supply on my spares shelf that I bought used and very inexpensively back when I was building my MOSFET amplifier deck several years ago. It was the same kind of chassis like the Eltek Flatpack2, that you had to slide into a frame to make the AC and DC connections. Back then I took one look at the back of it, and those obscure push-in connections, and bought the Mean Well power supply which had sturdy screw terminals.

Well with this talk of Flatpack chassis power supplies that were intended to go in a frame, I just pulled the rest of the bubble wrap off it, and it's an HP ESP120 used for server farms I suppose. The tag says 2950 Watts, 51.4 VDC, 57 Amps. It has all kinds of approval stamps including CE and TUV. This time I took a picture of the tag with my phone and googled the part numbers. It turns out that N4GA suggests that the HP ESP120 is one of the best of these power supplies out there for solid state amplifiers, and that it is also inexpensive on the secondary market. On his website at N4GA.com he provides the pinout of the obscure connector, and he also provides info on how to go inside and turn down the fan speed to make the power supply QUIET. He has been using them this way for some time and is happy with them. Now I have to figure out what to do about those connections. I am thinking that I will mount a sturdy screw terminal strip to the top of the chassis just above the connections, and solder some short heavy insulated solid wire between the obscure connections and the screw terminals - after I go inside and perform his mod on the fans to make them QUIET. It sure seems like I can wedge a heavy gauge solid wire between each pair of connector leafs and solder it in place so wire stays put.

Anyone ever use an HP ESP120 power supply?

N4GA was absolutely correct about the HP ESP120 "flatpack style" power supplies! Using his website at www.N4GA.com I was able to read the pinout of the HP ESP120 I have had on my shelf for a number of years.

I soldered short pigtail leads to the push-in connector blades and terminated them to a nice heavy duty (60A) screw terminal block that I epoxied to the top of the cabinet. Then I ran the 240 VAC connections on to a double pole switch that I mounted on top of the cabinet above the fans, after removing the simple cosmetic plastic baffle and blade "pull" lever which revealed a mounting plate with screw holes ready for a switchbox. I installed a line cord to a 240 VAC plug from the switchbox. This power supply provides 2950 Watts at 51.7 VDC, quite more than I need.

I followed W4GA's steps to open the cabinet and install voltage dropping resistors in each of the four (4) red fan leads that go to the four fans, to slow down the very loud fans and make them QUIET. They are simple 2 wire fans at two openings and are double stacked push-push in layout. After installing heat shrink tubing over the over-rated 3W resistors I had soldered into the leads, I took the time and blew out a small amount of dust and then buttoned the cabinet up. BTW the simple plated steel cabinet has very sharp edges you will find on the inside edges and I got two nicks on my hands to prove it.

The moment of truth came when I plugged the power supply into 240 VAC and flipped the newly installed switch on. Right away the fans started to purr at a reasonable audible volume, yet with good airflow, and then 10 seconds later the green LED lit up. All systems go! I checked the DC voltage and it was right as it should be per the tag, 51.7 VDC. After a while I flipped the switch off and the fans continued to run as the yellow LED took over for the green one - until power was gone and the LED went off along with total "Silence of the Fans".

This is a long skinny "flatpack" style cabinet 21 inches long, though it ended up 25 inches long including the gentle bends of my pigtails at one end and the toggle of the double pole switch at the other end. It has basic metal screens over the fan intakes to keep out fingers, and that is just under my newly installed power switch.

https://flic.kr/p/2nDC4Et

I am now planning to de-install my current Meanwell power supply, and check out its fan architecture as those too are full on all of the time, and can probably be slowed down too to avoid their bodacious noise.

I set the HP ESP120 power supply aside for the bigger LDMOS amplifier on my bench...

[AI5BC]

4. Durability: maybe it is my bad luck, but so far I have lost the canbus connection on two Elteks due to surges in electrical storms.

No sir not bad luck, poor product. Many utilities like TXU, FPL, AEP, Con Edison, and several wireless carriers are replacing their Eltek Dc power plants. All have the same piss poor performance and isolation issues making them subject to lightning damage. They are the Astron Power Supplies of commercial world, a POS only a ham would use.     

[AJ4SN]

I am also using the MeanWell 48 VDC 1500 Watt switching supply for my amplifier. I put it in a cheap plastic tool box, added some packing material and drilled a few holes for ventilation. It sits on the floor below the amp. It still makes noise, but it is acceptable. Roughly speaking it's about as noisy as my friends Mercury III amp. You might give that a try. You could certainly do a better job than I did with my junkbox approach.

73,

Stan

[WB2VVV]

I am also using the MeanWell 48 VDC 1500 Watt switching supply for my amplifier. I put it in a cheap plastic tool box, added some packing material and drilled a few holes for ventilation. It sits on the floor below the amp. It still makes noise, but it is acceptable. Roughly speaking it's about as noisy as my friends Mercury III amp. You might give that a try. You could certainly do a better job than I did with my junkbox approach.

73,

Stan

That's a very interesting approach Stan, and one that I would consider especially if I was concerned about dust intake into the power supply as it would be easy to incorporate a removable and washable dust intake filter with an airflow plenum, and a loosely screened exhaust port, into a secondary enclosure. BTW, that might be something that interests you since you're sitting your secondary enclosure on the floor. However, my MeanWell 1500-48 is currently mounted on rubber isolation shock mounts/hangers - hanging up under my heavy wood bench table-top, far away from dust that settles to the floor, and it has a big current shunt hanging off the + DC terminal for my DC ammeter - plus a big 2 pole 240 VAC primary switch is mounted under bench beside the power supply.

When I get a chance I'm going to unmount my MeanWell power supply and open it up now that it is well outside of its warranty, and see if I can simply slow the OEM fans down to a less audible medium level airflow, like I did with my HP ESP120 3kW server power supply.

[HAMHOCK75]

I was studying a switching power supply I used with the Icom 2KL. This is how they continuously control the fan speed via a thermistor. Not sure how loud the fan is compared to what you have now though so could not make any recommendation.

[WB2VVV]

Using one simple voltage dropping resistor in the lead to each fan has been working just fine. I made sure the resistor value was selected using a resistance substitution box and was based on getting a quieter but still substantial airflow thru the switching supply. I don't operate any digital modes so don't need any higher duty cycle than SSB/CW.

With the mechanical fan noise issue resolved, I noticed some electrical noise on some parts of the 10 meter band (now that the 10 meter band has frequently been open) and wasn't sure if that noise was getting into my system from the switching power supply DC output to the solid state amplifier. As a test I bought a circuit board filter from VK Amps that was designed to filter off switching power supply noise. The board arrived and looks very well made and well designed for this purpose. I installed this board in a diecast metal project box using 6 metal standoffs and star washers to deliver a good connection between the ground sections of the PC board and the metal project box. I used a short pair of leads with ring terminals on the OUTPUT end of the board to connect to the RF Deck DC Input, and a pair of 45A/60V rated Binding Posts on the INPUT end of the board to accept the ring terminal leads coming from the switching power supply. That way I could easily install the electrical noise filter in-line with the DC power leads coming from the switching power supply to the RF Deck, to see if it made an improvement.

Here are pictures of the electrical noise filter board in the box:

https://flic.kr/p/2nWFcGZ

https://flic.kr/p/2nWFB8V

Yes, this dramatically reduced the electrical noise I was hearing, so this simple solution worked for me! It can also be easily removed, or moved to another noisy power supply...

[KD6VXI]

Using one simple voltage dropping resistor in the lead to each fan has been working just fine. I made sure the resistor value was selected using a resistance substitution box and was based on getting a quieter but still substantial airflow thru the switching supply. I don't operate any digital modes so don't need any higher duty cycle than SSB/CW.

With the mechanical fan noise issue resolved, I noticed some electrical noise on some parts of the 10 meter band (now that the 10 meter band has frequently been open) and wasn't sure if that noise was getting into my system from the switching power supply DC output to the solid state amplifier. As a test I bought a circuit board filter from VK Amps that was designed to filter off switching power supply noise. The board arrived and looks very well made and well designed for this purpose. I installed this board in a diecast metal project box using 6 metal standoffs and star washers to deliver a good connection between the ground sections of the PC board and the metal project box. I used a short pair of leads with ring terminals on the OUTPUT end of the board to connect to the RF Deck DC Input, and a pair of 45A/60V rated Binding Posts on the INPUT end of the board to accept the ring terminal leads coming from the switching power supply. That way I could easily install the electrical noise filter in-line with the DC power leads coming from the switching power supply to the RF Deck, to see if it made an improvement.

Here are pictures of the electrical noise filter board in the box:

https://flic.kr/p/2nWFcGZ

https://flic.kr/p/2nWFB8V

Yes, this dramatically reduced the electrical noise I was hearing, so this simple solution worked for me! It can also be easily removed, or moved to another noisy power supply...

I get it that the board worked for you.

But I have to wonder, since the noise coming from a switch mode supply is common mode, why did they use one turn on the negative lead and multiple on the positive?


--Shane
WP2ASS / ex KD6VXI

[VK6HP]

Shane,

The filter is symmetrical - it's just not obvious on the picture supplied.  See below for details and alternative view:

https://www.vk-amps.com/products/a-50a-2-stage-dc-line-noise-filter-circuit

73, Peter.

[VE7RF]

Shane,

The filter is symmetrical - it's just not obvious on the picture supplied.  See below for details and alternative view:

https://www.vk-amps.com/products/a-50a-2-stage-dc-line-noise-filter-circuit

73, Peter.

That looks like an excellent filter.  What ga wire are they using, to handle 50 amps ccs ?   I see they also make a single stage 20+50 amp filter.