Abletec 53v Dual Rail 450W Supply for $20

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Founder of XSA-Labs
Joined 2012
Paid Member
Abletec and IRS2092 in a proper setup

I finally had a chance to mount my IRS2092 (200w) and the Abletec onto a proper amp setup. Listening to it now on my XL-S speakers. Sound is excellent - very clean and bass authority is very good.

The pre-made wiring harness for the Abletec makes this very much more convenient.

535905d1457305186-abletec-53v-dual-rail-450w-supply-20-irs2092-abletec-setup.png
 

Attachments

  • IRS2092-Abletec-setup.PNG
    IRS2092-Abletec-setup.PNG
    692.1 KB · Views: 1,380
The 5V rail on the PSU could be useful for adding a bluetooth module for anyone looking to build a high power wireless setup.

Sure, except there is no "5V" supply on the unit. That would just be all too convenient! Instead you will find a 5.6V output rated for up to 1A. That voltage is a bit too high for some electronics that expect a true 5V DC supply, e.g. power from USB (max spec for USB 5V power is 5.25V).

But with an LDO regulator like this:
Exar SPX29300T-L-5-0 5V fixed LDO regulator
you should be able to make a nice 5V supply out of it. 5V@1A is just about enough power to run a Raspberry Pi without much connected to it. Perfect for onboard DSP...
 
Alternatively, use the 7V rail at 1.6A with one of these buck or boost regulators to power a Pi with a bunch of stuff connected, or miniDSP, or...

10pcs Mini 360 DC DC 4 75V 23V to 1v 17V Buck Converter Step Down Power Supply | eBay

10pcs MT3608 2V 24V 2A DC DC Step Up Power Apply Module Booster Power Modulemax | eBay

I suppose you could even use 2 buck or boost regulators, and parallel the two auxiliary rails to get even more power if you needed it.

EDIT: Just be sure to put a schottky diode in series with each DC-DC output if you parallel them.
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
I was going to build an all active 6 channel system with these as the woofer amps, and then probably a class AB for mids, or the TDA7498 (100w) units for the mids, and Sanwu TPA3118 PBTL (30w) for the tweeters. The miniDSP can be powered with 5v to 24v (it has a built in regulator). Most class D amps with BT also have a 5v regulator as the amps can take anything up to 24v typically. Although I wonder how clean the 5.5v is, I may be better off with a linear regulated supply for that critical part as it acts as the preamp.
 
Indeed. I suppose the point I was trying to make is just that the ~18W available on the auxiliary rails is useful for a whole lot more than just heating up some resistors to get the high voltage rails up and running.

Sure is! In one application I have a 12VDC "quiet" fan connected to the 7.5V aux supply. This keeps it spinning VERY quietly but still plenty of airflow to exchange air through the amp chassis.

The resistors are there to draw the minimum current needed for reliable operation of the SMPS, not just to generate heat. If you can come up with better uses for the supplies that will always draw the minimum required current feel free to eliminate the resistors.
 
I'm a total newbie to class D amps, and solid state (I'm a tube amp guy), so bear with me on this. I'm using 3 of these supplies to run three IRS2092 amps, in a DIY variable frequency 3 phase motor controller, for a DIY turntable project. I don't understand how the proposed bleeder resistors could work for V1 and V2. Straighten me out on this, please.

In the first post in this thread it was stated that the minimum current requirements for a stable power supply for V1 and V2 (+ and - 53V) is 100ma, V3 (7.5V) is 50ma, and V4 (5.6V) is 30ma. It was suggested that the use of 150 ohm 1 watt resistors would work, in other posts it was suggested that 150 ohm 5 watt resistors be used.

I'm confused here, according to ohms law, a 150 ohm resistor on V1 and V2 would draw about 353ma, and burn almost 19 watts. Of course that would quickly fry a 150 ohm 5 watt resistor.

For V3 (7.5V) it does work out to 150 ohms, and the power consumption is .375 watts, so a 1 watt resistor would be fine there. If you wanted V4 (5.6V) to draw 30ma, it would technically require a 180 ohm resistor, and would draw .156 watts, so the 150 ohm 1 watt resistor would work there as well. I'm pretty sure the 5.6V supply would be happy at 37ma.

Are you guys just relying on the amp to draw at least 100ma on V1 and V2 even at idle, so eliminating the need for bleeder resistors on V1 and V2? BTW, I do have power to spare, about 18.5 watts continuous needed per amp. If I could gain a little stability, and sacrifice a little power, by using bleeder resistors there, it would be a good thing. I was thinking of using 420 ohm 15 watt resistors there. That would draw about 126ma, and burn about 6.7 watts. Lets say the amps are about 85% efficient (don't really know, they came with absolutely no documentation), that would be a total power draw of about 28.5 watts on the supplies with the resistors. Any advice?

twystd
 
I'm powering a L15D-Pro, IRS2092 amp which is driving a single 12" sub. When I play a 20Hz tone, the power supply shuts down before the sub even breaks a sweat. If I had to guess, I'd say the sub is seeing maybe 60W, which is fine for every day listening, however I bought 4 of these supplies thinking that they would be comparable to my current linier PS, but in a smaller / lighter package. I usually power the same amp and sub with a 450W, linear power supply which drives my subwoofer to ridiculous levels.

I've got the 5.5v output loaded with a Raspberry Pi and the 7v loaded by a bluetooth module... I also tried loading those same outputs with 1/2 watt, 200 Ohm resistors, but no change. I've tried 2 of the 4 that I have but they both perform horribly :(

Am I overlooking something?
 
I decided to measure the output myself. Hopefully someone can correct me if my testing method was flawed or skewed in any way. I'm learning as I go :)

I don't have much for test equipment, I just wanted to see what voltage I could get into a resistive load... I built a 2.4‎Ω dummy load with 5 paralled 12‎Ω / 100W resistors mounted to a fair sized heatsink (that's all the resistors I had on hand but I figure that should be more than sufficient for this purpose). The test amp was my modified JLAmp (It's an IRS2092 based class D amplifier that is 2‎Ω stable and can put out well over 400W into a 2‎.4Ω load.

I tested with both 60Hz and 120Hz tones and slowly increased the gain (measuring output voltage) until the power supply cut out... With a 120Hz tone the power supply cut out @ about 22v (That's about 200W into this load) With 60Hz tone cut out at only 12v (about 60W)... This test was done with 200Hz resistors loading the 5v & 7.5v outputs. With the same setup, but using my linear power supply (+/-42v) I ran out of gain before before I was able to max out my power supply (29v into 2.4Ω and 350W)

Can anyone else out there using this supply share their experience? Like I said, I bought 4 of these supplies and I'm hoping I'm doing something wrong here
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Have you tried the test on one of the other 3 amps to see if the same thing happens? It does seem like it is cutting out early but one thing to note is the rated spec for high power transients is very short - perhaps your amp has a sharp spike in current draw? Does your amp have a very large capacitor on its input rails (like one would use for smoothing a transformer input)? SMPS don't like capacitive loads. Try reducing the cap size on the IRS2092 amp? Maybe two 330uF 100v caps can work?
 
Have you tried the test on one of the other 3 amps to see if the same thing happens? It does seem like it is cutting out early but one thing to note is the rated spec for high power transients is very short - perhaps your amp has a sharp spike in current draw? Does your amp have a very large capacitor on its input rails (like one would use for smoothing a transformer input)? SMPS don't like capacitive loads. Try reducing the cap size on the IRS2092 amp? Maybe two 330uF 100v caps can work?

The JLAmp only has 120uF on the supply rails. I'm sure the LJM amp has more. I'll try that one out anyway and see if/how it behaves differently. I'll also re-configure the load to make 4Ω and test the $20 IRS2092 amp. I'll try this out shortly and keep you posted.

Thanks for the advice xrk971
 
Let's see... using V=IR we have:
R=2.4 Ohms
V=22 Vrms
what's the current?

Rearranging, Irms=Vrms/R (resistive load so no phase angle to worry about!)

Irms=22/2.4=9.16A (rms)
Ipk=1.41*Irms
Ipk=12.8A peak

Perhaps the PS is not so happy when there is this much current demand?

Looking at the spec sheet, this PS is rated for 1.38A continuous, 4.13A peak (up to 5 min) and 16.5A for 10milliseconds (so forget about that one, because your frequencies are low). I think that this is per rail, so double these number. This would put the peak rms current at about 8.4Arms. Your test is drawing another Ampere or so above this.

I think that you are exceeding the current draw that the supplies are capable of, causing them to shut down. Not sure exactly why this is happening at a much lower voltage with a 60Hz input.

I would try to repeat your test with a higher impedance load so that current demand is much lower, e.g. 8R. Then you should be able to reach the full rated power (or that is my hunch). If you are driving a 2-ohm nominal driver with this PS, then this is NOT a good choice because much more current is needed. Is that why you chose the 2.4 ohm test load by any chance? Also, keep in mind that a real driver represents a reactive load and this can jack up the current demand at some frequencies.

I have a stock of these that I bought for various projects. I have been using two of them to power a pair of active loudspeakers for the past two years and have had zero issues although I am probably not pushing the supplies to their limits. I have seen other complaints about these PS not living up to their rated outputs, so I am very interested to see how this particular testing proceeds since there is actually some numbers behind the claim. Please test again and post the results. Thanks.
 
Last edited:
Indeed. The numbers add up and this makes perfect sense... I just tested @ 4Ω and I was able to push 28.5v @ 120Hz / 30v @ 240Hz but again the low frequency capabilities was much lower, giving up @ only 18v. Longer sine-waves, I suppose?.

This really clears things up and I suppose it should have been obvious that I'd run into current issues with such a low impedance... 8Ω, stereo seems to be the best matchup this power supply.

Thanks to both of you
 
Indeed. The numbers add up and this makes perfect sense... I just tested @ 4Ω and I was able to push 28.5v @ 120Hz / 30v @ 240Hz but again the low frequency capabilities was much lower, giving up @ only 18v. Longer sine-waves, I suppose?.

This really clears things up and I suppose it should have been obvious that I'd run into current issues with such a low impedance... 8Ω, stereo seems to be the best matchup this power supply.

Thanks to both of you

Hmmm, that's a little concerning. You are still only putting around 200W or a little more into the load. If the power can't meet spec for some reason perhaps this is why the PS is found as surplus for cheap?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.