Meanwell IRM-60-24

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The switching power supply Meanwell IRM-60-24 is widly used to power power audio amps expecially the chip one such as LM3886.

My question is about how the IRM-60-24 behaves with high capacitive load used to bypass the power amplifier.

I performed a test with a "generic" switching power supply 24 V 2 A loaded with several thousand of uF and the result was a "wild" oscillation at the output.

What about the IRM-60-24 loaded with e.g 1000 uF or more?
 
Last edited:
With Mains-to-DC converters, no bulk-capacitance should be used. The required peak-current delivery comes from the power-supply itself - so the supply should be rated to give the peak current your load (example: speakers) demand, with a definite margin.

The point-of-load capacitance should be only about 100µF. The IRM-60 does not appear to specify maximum capacitance, but similar modules I have used (for industrial purposes) can only tolerate 470µF, at most.

If your design actually requires large capacitances to function, a 50Hz transformer & rectifier is more suitable.
 
I performed a test with a "generic" switching power supply 24 V 2 A loaded with several thousand of uF and the result was a "wild" oscillation at the output.

IRM-60-MEAN WELL Switching Power Supply Manufacturer

What about the IRM-60-24 loaded with e.g 1000 uF or more?

Ask them... Do not expect a response or, if they do respond, any meaningful answer. I'm not inclined to try and reverse engineer it for you.

Sorry to be harsh but I have little to no sympathy for people who arbitrarily slap anything off the end of a Flyback converter, or indeed any other SMPS employing feedback, because they are upset with the output ripple and noise figures.

Basically, in the case of a Continuous Flyback, you send the feedback loop second order at crossover and it becomes unstable. There are other effects but that is the initial one.
 
With Mains-to-DC converters, no bulk-capacitance should be used. The required peak-current delivery comes from the power-supply itself - so the supply should be rated to give the peak current your load (example: speakers) demand, with a definite margin.

The point-of-load capacitance should be only about 100µF. The IRM-60 does not appear to specify maximum capacitance, but similar modules I have used (for industrial purposes) can only tolerate 470µF, at most

I have little to no sympathy for [cut] who arbitrarily slap anything off the end of a Flyback converter, or indeed any other SMPS employing feedback, because they are upset with the output ripple and noise figures.

Basically, in the case of a Continuous Flyback, you send the feedback loop second order at crossover and it becomes unstable
Many thanks for the replies.

I asked the previous question while I was studing the highly regarded Modulus-86 (which uses a LM3886) and the SMPS-86 (which uses two Meanwell IRM-60-24) from Neurochrome.

Neurochrome states:

"Who [cut] have low-efficiency 4 Ω speakers, should consider building the SMPS-86 + Modulus-86 as mono blocks"

Modulus-86 specs:

65 W (4 Ω) [cut] output power when using the Power-86 and recommended transformer.

Although there is no reference to SMPS-86, I suspect that there is not a big difference in output power when a SMPS-86 is used.

Simple math:

65 W (4 Ω) means 22.8 Vp

Ipeak = 22.8 / 4 = 5.7 A

Iave = 5.7 / pi = 1.81 A

We see that we need an average current of 1.81 A and a peak current of 5.7 A for each rail for a mono channel.

While the average 1.8 A is well within the rating of Meanwell IRM-60-24 (24 V 2.5 A 60 W) the 5.7 A peak is well beyond its rating.

To overcome this (and to fullfill the bypass requirements of LM3886) the SMPS-86 and the Modulus-86 are equipped with (very) high uF electrolitic capacitors on the supply rails.

From the SMPS-86 description:

"LC filtered output to minimize noise" and "This project contains two surface mount components: The inductors used in the output filter."

I suspect that the inductor on the Meanwell IRM-60-24 output is used to filter the output to minimize noise as well to insure stability of the switching module.

Two more questions:

a) is the inductor necessary to insure stability of the switching module?

b) is the inductor value a math calculation or a trial-and-error result for a specific switching power supply?
 
Last edited:
Many thanks for the replies.

I asked the previous question while I was studing the highly regarded Modulus-86 (which uses a LM3886) and the SMPS-86 (which uses two Meanwell IRM-60-24) from Neurochrome.

Two more questions:

a) is the inductor necessary to insure stability of the switching module?

b) is the inductor value a math calculation or a trial-and-error result for a specific switching power supply?

Switch-mode power supply for audio amplifiers.
IRM-60-MEAN WELL Switching Power Supply Manufacturer

Yes the inductor will act to isolate the feedback loop within the SMPS. Unfortunately you are still grasping at straws.

The report suggests 2400mV in 500uS 40% load transient, 0.8A. Design crossover is 4.4KHz. Output impedance is 3R at crossover. Pick an inductor with an impedance of 6R at 2.2KHz.

I was studying the highly regarded Modulus-86. Neurochrome states:...

Good Luck. Drops Mic.
 
Switch-mode power supply for audio amplifiers.
IRM-60-MEAN WELL Switching Power Supply Manufacturer

Yes the inductor will act to isolate the feedback loop within the SMPS. Unfortunately you are still grasping at straws.

The report suggests 2400mV in 500uS 40% load transient, 0.8A. Design crossover is 4.4KHz. Output impedance is 3R at crossover. Pick an inductor with an impedance of 6R at 2.2KHz.



Good Luck. Drops Mic.
Thank you for the link to the documents (I had already read).

Unfortunately I'm not able to find the figures you state (2400 mV etc).
Which report are you refering to?
Can you please post the link of the report?

You didn't answer my question (which one 1 or 2?):

is the inductor value

1) a math calculation

or

2) a trial-and-error result for a specific switching power supply
 
I thought I had dropped the Mic..

Thank you for not linking me to the documents you have already read. Fortunately I managed to find them.

I posted a link to the report. Presumably you have not bothered to click REPORT and use it.

I have answered your question.

I assume you know how to tie your own shoe laces.

Perhaps you need some advice in respect of updog.

..
 

Attachments

  • Screenshot at 2017-12-16 20:26:28.png
    Screenshot at 2017-12-16 20:26:28.png
    7.6 KB · Views: 578
I thought I had dropped the Mic..

Thank you for not linking me to the documents you have already read. Fortunately I managed to find them.

I posted a link to the report. Presumably you have not bothered to click REPORT and use it.

I have answered your question.

I assume you know how to tie your own shoe laces.

Perhaps you need some advice in respect of updog.

..
Dear MorbidFractal,
thank you for your kind reply and your patience.

1) I had already read the documents at the link YOU sent me and NOW I repeat below
Switch-mode power supply for audio amplifiers.
IRM-60-MEAN WELL Switching Power Supply Manufacturer

2) I have NOT "bothered to click REPORT and use it" because the word "report" (in your post # 5) is not a link (at the best of my knowledge)

3) Attached to this post you can find the PDF downloaded from the link (you posted)
IRM-60-MEAN WELL Switching Power Supply Manufacturer. I cannot find any "REPORT"

Please check the image below to verify my statements 1) 2) & 3).

Until I can read the report I cannot make sense about:
"The report suggests 2400mV in 500uS 40% load transient, 0.8A. Design crossover is 4.4KHz. Output impedance is 3R at crossover. Pick an inductor with an impedance of 6R at 2.2KHz."

What about report?

Best regards
 

Attachments

  • MeanWell-SMPS.png
    MeanWell-SMPS.png
    96.6 KB · Views: 580
  • IRM-60-SPEC.PDF
    538.5 KB · Views: 70
Last edited:
Old School
Joined 2008
Paid Member
diy_audio_fo:

A couple of friendly questions: If the rated amperage of the Meanwell RM-60-24 is 2.5 Amps (see the second page of the PDF you posted above), is it realistic to assume that 2500 mA, which is essentially 2400 mA, as MicroFractal indicates, can sustain peaks of 5.7 Amps?

Second, NeuroChrome indicates that the Modulus-86 should be used with the "recommended transformer." Have you checked the rating of NeuroChrome's "recommended transformer" to get a sense of the VA they expect for a power supply for the Modulus-86? It would appear to me that the reference to a "transformer" for their preferred power supply suggests that NeuroChrome expects the Modulus-86 to be powered by a linear power supply, notwithstanding that it is a switching amplifier.

Cheers
 
If I understand correctly the OP wants to filter the output of a supply capable of 2.5A continuous such that burst currents, above about 6A, will not trigger the supply's current limiting. They want to do this with additional filtering without rendering the supply unstable.

It strikes me that we are down in the realms of achieving PMPO^5 performance. I do not doubt that the additional filtering will reduce noise and might improve sound but why not just use a power supply that is rated to cope with the amplifier's requirements.

Anyway. I did roll some, valid, sums but ended up at the doorstep of despair. It's just chasing someone else's meaningless but 'highly regarded' dream.

Code:
Version 4
SHEET 1 1016 680
WIRE 112 128 80 128
WIRE 224 128 192 128
WIRE 320 128 224 128
WIRE 416 128 320 128
WIRE 512 128 416 128
WIRE 656 128 512 128
WIRE 688 128 656 128
WIRE 224 160 224 128
WIRE 320 160 320 128
WIRE 416 160 416 128
WIRE 512 160 512 128
WIRE 320 256 320 224
WIRE 416 256 416 224
WIRE 416 256 320 256
WIRE 512 256 512 224
WIRE 512 256 416 256
WIRE 80 288 80 128
WIRE 224 288 224 224
WIRE 512 288 512 256
WIRE 656 288 656 128
WIRE 80 400 80 368
WIRE 224 400 224 368
WIRE 224 400 80 400
WIRE 512 400 512 368
WIRE 512 400 224 400
WIRE 656 400 656 368
WIRE 656 400 512 400
WIRE 80 432 80 400
FLAG 80 432 0
FLAG 688 128 VOUT
IOPIN 688 128 Out
SYMBOL ind 96 144 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L1
SYMATTR Value 1m
SYMBOL cap 208 160 R0
SYMATTR InstName C1
SYMATTR Value 22000µ
SYMBOL cap 304 160 R0
SYMATTR InstName C2
SYMATTR Value 22000µ
SYMBOL cap 400 160 R0
SYMATTR InstName C3
SYMATTR Value 22000µ
SYMBOL cap 496 160 R0
SYMATTR InstName C4
SYMATTR Value 22000µ
SYMBOL res 208 272 R0
SYMATTR InstName R1
SYMATTR Value 10m
SYMBOL res 496 272 R0
SYMATTR InstName R2
SYMATTR Value 200m
SYMBOL voltage 80 272 R0
WINDOW 0 36 47 Left 2
WINDOW 3 35 67 Left 2
SYMATTR InstName VIN
SYMATTR Value 24V
SYMBOL current 656 288 R0
WINDOW 0 -280 -206 Left 2
WINDOW 3 -283 -186 Left 2
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName I1
SYMATTR Value SINE(0 6 100 100m 0 0 0.5)
TEXT 120 432 Left 2 !.tran 0 150m 95m

Whilst the sums, I am not going to reproduce them here, are approximate the values that result when inserted in the model appear to be close. Unfortunately the values are... huge and that's all you get for a half cycle 100Hz 6A demand. Yes the power supply is below current limit but I really see no actual point to the exercise.
 

Attachments

  • Screenshot at 2017-12-17 13:39:12.png
    Screenshot at 2017-12-17 13:39:12.png
    40.5 KB · Views: 381
I assume you know how to tie your own shoe laces.
No, I ain't.

I'm a blind/dumb guy ;-)
When I read the following page
IRM-60-MEAN WELL Switching Power Supply Manufacturer
I saw only the content of the page and NOT the frame (around the content) with the tab "REPORT"

Now back to the technical stuff

You write:
"The report suggests 2400mV in 500uS 40% load transient, 0.8A. Design crossover is 4.4KHz. Output impedance is 3R at crossover. Pick an inductor with an impedance of 6R at 2.2KHz"

In the report I find the following figures:
"TRANSIENT RECOVERY TIME
V1:2400 mVp <500us
I/P:230 VAC O/P:40% LOAD CHANGE

Now I can obtain:

Out Imp = 2400 mV / 0.8 A = 3 R

I (politely) ask:

Where 0.8 A come from? (40% of 2.5 A (max out current) is 1 A)
and Design crossover frequency 4.4 KHz?

I assume that for 6R and 2.2 KHz you considered a X2 "safety factor"

If I understand correctly the OP wants to filter the output of a supply capable of 2.5A continuous such that burst currents, above about 6A, will not trigger the supply's current limiting. They want to do this with additional filtering without rendering the supply unstable
You nailed the point :)
My application is not an audio amp but your understanding is perfect!
I proposed the Modulus-86/Meanwell-SMPS just as an example.

Thanks a lot for your valuable simulation.
 
Last edited:
Thank for rising interesting points.

A couple of friendly questions: If the rated amperage of the Meanwell RM-60-24 is 2.5 Amps (see the second page of the PDF you posted above), is it realistic to assume that 2500 mA[cut]can sustain peaks of 5.7 Amps?

I think so if:

- peaks are "narrow" enogh
- added (output) capacitor are "large" enough
Second, NeuroChrome indicates that the Modulus-86 should be used with the "recommended transformer." Have you checked the rating of NeuroChrome's "recommended transformer" to get a sense of the VA they expect for a power supply for the Modulus-86? It would appear to me that the reference to a "transformer" for their preferred power supply suggests that NeuroChrome expects the Modulus-86 to be powered by a linear power supply, notwithstanding that it is a switching amplifier.

Although my application is not an audio amp I proposed the Modulus-86/Meanwell-SMPS just as an example.

Switch-mode power supply for audio amplifiers.
"Who [cut] have low-efficiency 4 Ω speakers, should consider building the SMPS-86 + Modulus-86 as mono blocks"

From the page: "Power Supply Design"
LM3886 chip power amplifier power supply design.

Crest Factor (dB) Power Transformer VA Rating
3 (sine wave) 138 VA
6 31.8 W 98.9 VA

The SMPS-86 has the following specs (there are two combined Meanwell IRM-60-24 module):

24 + 24 V 2.5 A 120 W
 
Now I can obtain:

Out Imp = 2400 mV / 0.8 A = 3 R

I (politely) ask:

Where 0.8 A come from? (40% of 2.5 A (max out current) is 1 A) and Design crossover frequency 4.4 KHz?

Obviously I could try to claim that the mistake was deliberate to make sure you were paying attention but... My Bad. Somehow my brain fixated on 2A. You are right 2.5A 2400mV/1A 2R4

The loop will be first order at crossover. In effect an LR or RC section. Recovery is specified to a percentage of the excursion either based on 2.2RC or 5RC. Crossover becomes.

2.2/500u = 4400Hz

or

5.0/500u = 10000Hz

The Mean Well example you are using is a flyback converter. Elsewhere, and I am not going to try and find it again, its switching frequency is specified as being 67KHz. It will operate with continuous transformer current and therefore suffer from a Right Half Plane Zero, RHPZ, which will vary with load current.

It is possible to determine where that RHPZ will be but wet rule of thumb is that the loop is closed at 1/10th of the switching frequency in order to avoid it. 6.7KHz but you also have to allow for component tolerances. That's why I chose 2.2 as opposed to 5.

The loop will use Pole-Zero compensation. A DC pole with, ideally, the zero placed at half of the crossover frequency of 4400Hz or 2200Hz. That's where the 'safety factor' comes from.

Otherwise the choice is somewhat arbitrary. You can see that in my model I used 1mH in order to chase a solution without having to use Farad sized capacitors.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.