Can I use this supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have a sub plate amp that is not repairable. The power supply however is still working well. It has a SMPS that outputs 90V. The power supply was rated to support 200W continuous audio output. 2 of the taps from the transformer are ran through diodes facing the same way I am guesing for current increase. I figure this could be split into +-90V by adding 2 diodes in the reversed direction.

Is there a 100V single supply amp or a 100V dual supply amp that is not too complicated to build that will deliver ~200W @ 8ohms

So far all I have found was this:
http://www.ti.com/lit/ds/snas396d/snas396d.pdf

I would hate to junk something that I could have fun with. The supply also has 2 aux supplies 1 at +-35V 2A and one at +35v 500ma

Thanks
 
AX tech editor
Joined 2002
Paid Member
Did you measure the supply, is it 90V single ended? That might support 200W into 4 ohms (see Andrews post above).
Building an amp that supports +/-90V is a very difficult project, I would stay away from that!
But using the 90V for a +/-45V amp looks doable, though the circuit complexity increases due to the single ended supply.
Not an easy choice.

Jan
 
So I took some measurements and the SMPS is putting out a square wave with some overshoot and the rising and falling edge. A DMM reads ~48V ac on both taps when half rectified it reads 107VDC no load and ~77VDC with a 100W load. I suppose something is dooable but I will need an amp that can support some serious rails. Any final ideas or feedback on the part that I listed? This is mainly for fun but I do have 9 of these supplies in my parts box. They were cut off from damaged Klipsch Sub 10 amplifiers.
 
Half Bridge

Here is a half bridge implementation. You will have to modify the supply by changing the components. Use the same diodes that is currently used.

An other option is that you clip open the transformer, remove the aux windings that you don't want and then modify the winding that you do want to be center tapped. On a switch mode PSU there are not many windings and is pretty simple to modify.
 

Attachments

  • fun.jpg
    fun.jpg
    25.2 KB · Views: 152
You set you had a few SMPS - haul out one transformer and analyse it, make some mods and then replace it. The primary is normally wound first around the bobbin. nIf the secondary that you want to modify is 107V and you want it to be 50 - 0 - 50 then just count the windings, halve the number and wind two from the same wire and into the same direction as before. The secondary could have as few as 10 turns on it.
 
AX tech editor
Joined 2002
Paid Member
So I took some measurements and the SMPS is putting out a square wave with some overshoot and the rising and falling edge. A DMM reads ~48V ac on both taps when half rectified it reads 107VDC no load and ~77VDC with a 100W load. I suppose something is dooable but I will need an amp that can support some serious rails. Any final ideas or feedback on the part that I listed? This is mainly for fun but I do have 9 of these supplies in my parts box. They were cut off from damaged Klipsch Sub 10 amplifiers.

The DMM is calibrated for RMS of sine waves so will underestimate if you use it on square waves. That's why the big difference between the DMM and the rectified voltage.

Jan
 
Most DMMs use averaging. Take lots of samples and average the resulting voltage.

A square wave using averaging gives a good result.
A sawtooth or triangle wave using averaging gives a good result.
A sinewave using averaging gives a good average. BUT it is a bad result as far as informing us of the energy contained in the waveform. The manufacturer scales (multiplies) the average result to give a good equivalent to the heating effect (rms) of the waveform.

But that scaling from average to sinewave destroys the usefulness of the triangle and squarewave results. If you know the scaling factor used by the manufacturer, you can apply that in reverse to find the true average of those straight edged waveforms.

The scaling factor is roughly sqrt(2). Can anyone confirm the actual scaling factor?
 
Last edited:
Understanding how they work allows one to decide if they can be used for non sinewave measurements.

Trying to identify the extent of ripple on a DC voltage is just such a use.

All my DMMs, both hand held and bench give a good approximation of ripple voltage on a supply rail without reading the rms effect of the dominant DC voltage.

Using a scope confirms that the DMM mVac reading is very approximately 1/3rd of the peak to peak voltage shown on a scope.
 
Last edited:
AX tech editor
Joined 2002
Paid Member
Understanding how they work allows one to decide if they can be used for non sinewave measurements.

Trying to identify the extent of ripple on a DC voltage is just such a use.

All my DMMs, both hand held and bench give a good approximation of ripple voltage on a supply rail without reading the rms effect of the dominant DC voltage.

Using a scope confirms that the DMM mVac reading is very approximately 1/3rd of the peak to peak voltage shown on a scope.

Andrew, they give 'a good approximation' in as far as the ripple comes closer to a sine wave. You still confuse 'the way they work' with the calibration. It doesn't matter how they work. They are calibrated to show the correct RMS value of a sine wave with a certain peak value. They are calibrated to show 0.707 of the peak value as RMS value. If you measure anything else than a sinewave they are off because a non-sinewave does not have the 0.707 ratio between peak value and RMS value.

Of course they can read the AC on a DC supply - coupling cap! That said, many DMMs have the option to include the DC component at the flick of a switch.

Giving 1/3rd of the ripple peak to peak value is to be expected. The DMM 'measures' the peak value which is 1/2 of the peak-to-peak and then indicates 0.707 of that, which leads to - surprise! - 1/2 * 0.707 = 0.35.
That does not mean that the RMS value of the ripple is 1/3rd of the peak-to-peak value, although that is what the meter is telling you.

Jan
 
It has been a long time since I have torn one of these transformers apart. If my memory is correct the windings are staggered on the ER35 bobbin. Re-winding would be a pain. I suppose this would be a good learnign expierence on how this SMPS actually works and the transformer ratios.
 
I learned a few more things. The main supply operates in a push pull method. The transofrmer then has a couple different tapps. It has a low current +-35V winding that is dropped down to +-15v with linear regulators(they do get pretty hot). The second winding measures between the suplies unloaded are around 90-100V. Would it be easier to just add a DC/DC buck to drop it down to 48V?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.