Question for SSLtech +/- 18, 48v split supply

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
G

Guest

Guest
Hi everybody,

I'm excited - finally I've got all my parts together (almost) to build my little PS. I'm hoping someone might have an answer to this question about ssltech's power suppy schem, for which I am ver greatful!

Here it is: My toroid will be 2x18v (20.9v no load), instead of 2x20 on the schem. I'm assuming this doesn't mean much to the +/-18v supplies, as the regulators are fixed. I did start thinking however that I might have have to do something with with the resistance on adjust of the LM 317. Then I realized that I'm not even sure how he's getting 48v out of that.

Here's a link to the schematic:

http://www.beatbazar.com/guests/ssltech/kps-1/index.htm

And here goes:

Are C5 and R1 setting the input voltage on ic3?
What will I recalculate for IC3, once I understand it, with respect to the new primary voltages?

Thanks folks, and sorry to bug.


Kelly
 
Kelly, your link doesn't show up.

But without seeing the schematic, whatever is feeding the LM317 needs to be sufficiently higher than the desired output voltage. The output voltage itself, when the input voltage condition is met, will be set by the resistive divider values and the internal reference of the 317.

If the input voltage to the regulator isn't a minimum amont higher at all times than the output, there's no change in the output-determining network that will make the thing regulate to the desired voltage.

Brad
 
Hi Kelly,

Your 18-0-18V toroid will be fine. Once rectified and smoothed, the DC voltage applied to the voltage regulators will be above the output voltage required, so no problem there.

R1 (8k2) and R2 (240R) set the output voltage for the LM317T. These values are found using a bit of simple maths (or using a pot to get the correct value, measuring it with a DMM, and the replacing with the nearest-value fixed resistor)

But I'm sure Keith did the maths...

:wink:

Check the LM317T product page and datasheet for more info:

LM317 Homepage

Mark
 
thanks a lot, fellas...you guys even helped without the link!

truly, I am a moron. if I'd thought to read down the page before asking i would have found the explanation for the multiplier (to 48v). Thanks a lot, and my apologies.

kelly
 
Hah! And the link shows now too. There is the full-wave voltage doubler that I hadn't seen before, that I mentioned in the voltage multiplier thread. Drawn differently the bridge of diodes wasn't as recognizable. I love his reference to it as a Cockcroft-Walton.

You are probably ok with the lower V transfomer, but it does cut down the margin a bit, especially if the mains input sags. Fortunately we can usually count on the tranformers having a fairly much higher output voltage below rated load, so perhaps that is the case here as well. After you put it togther be sure to measure the voltage at C5 when the PS is feeding its rated loads.
 
In my research of the multiplier, I found this:

http://ejksdesktop.homelinux.com/cockroftwalton.html

In the second part, there's a crockett multiplier that sounds truly menacing. I want to build that, but I would electrocute myself. At least no more dumb questions on the board :razz:

Again, thanks all.

kelly
 
I just did this one up on veroboard, using the cheap torroid($5) from the other thread, and 7815 and 7915. Took all of 10 minutes, and works spiffy (left out the phantom section, as I didn't need it for this one ).

Funny, I've had this schem in a filing cabinet for months, and the day I use it, someone posts about it.

Regards

ju
 
Ahh high voltage. I used to love the smell of ozone.

Yes, if the 18V windings were really putting out an 18V rms perfect sinusoid, and the diodes having a typical silicon forward drop, the transformer voltage is marginal: the output looks like 48.5V d.c. with the ripple dipping it to 48.3V, with a roughly 10mA load. This is less than the 317 needs to regulate to 48V out.

OTOH, as I say the actual tranny output will be higher most likely, so you may be ok.

If you are a bit deficient you could reduce the programmed output voltage a bit and few microphones would be the wiser, nor would you likely be visited at 3AM by the phantom power police. The alternative of having a bunch of mains hum appearing in your preamp output would be decidedly worse.
 
The Phantom Power Police! I want that job. Voltmeter in a holter...citation book...

:razz:
 
I'm thinking of adding a few protection diodes to the SSL tech power supply.
this is pretty common practice and as far as I know this doesn't change the noise level... I could be wrong, I've used this supply for guitar effects and a diy synthesizer and never had any noise problems.

I'm curious if this does add any noise/ripple to the supply?


This would be implemented as an IN400x diode from the out to in points across each of the LM78xx regulators... the same as D1 which protects the LM317.
http://www.epanorama.net/links/psu_linear.html#dualpolaritylinearpsu

Thanks
Sleeper
 
Hi Sleeper,

The diode connected in reverse-sense to the Input and Output pins helps prevent the situation that would arise if the input voltage became lower then the output voltage- e.g. if the input supply fused and the fed circuit had large reservoir/smoothing caps. A higher voltage at the Output pin than the Input pin can damage the regulator. The diodes aren't normally used within equipment- most manufacturers couldn't warrant this "expense" for a once-in-the-lifetime-of-the-equipment chance happening.

But DIY-ing means that you can go to whatever lengths you want to make your gear roadworthy/dependable.

Check this TI 78xx datasheet, page 7, for more diode-protection info. And no, as long as you use a silicon diode (i.e. low leakage and low capacitance), the extra protection diodes won't increase any noise. Just be careful you install them the right way round!!

http://www.sunmark.com/datasheets/7805.pdf

The diagram at the top of page 7 is essentially the schematic you linked to above.

Mark
 
Hi Mark,
I was mostly thinking about the noise here.
and as DIY'ing also involves playing around with components and putting those alligator clips to work, it seems I've blown a couple of regulators on the ssl supply during testing of various pieces of gear.

I do extremely rigorous testing on all my DIY gear, involving things like putting chips in backwards, or making circuit boards that mistakenly have the power supply footprint for a 5532 where a 5534 should go, etc. :roll: :wink:

until I get a good bench supply, anything I can manage to keep the psu intact during these procedures is well worth the effort.

Thanks for the tips
Sleeper
 
I'm about to start building a solid state pultec clone which needs +15v/-15v and 24v. As starting point for the power supply I want to use Ssltech's power supply. I will replace the LM 7918/7818 with LM7915/7815. The question now is: can I replace the LM317 for a LM 7824 and if so, what value's should R1 and R2 be?

Thanks in advance,

Radiance
 
Hi radiance,

It would depend on the current consumption you'd require for the +24V rail. Keith's PSU board uses a voltage multiplier circuit for the +48V rail, and this is fine because P48 has a very low current requirement.

What is the +24V rail for? What kind of current do you need? Is it for Steffen's Pultec?

Mark
 
Sorry that I'm late to the party on this one, but my Valiant deputy Mark Burnley seems to have answered in my stead.

bcarso, this is a Cockroft-Walton, -right? -Or am I mistaken? -(happens a lot these days!)

I've been out of the country with little access to the net for the last week or so, and prior to that I've been working crazy hours... another week to go and I'll be back to more regular visits.

Cheers all!

Keef
 
How about this??

SSLTech_power_supplyEDIT.jpg


And skip C3 and C4 and make C5 a 1000uF cap..
 
Keef, it's a variant of the Cockcroft-Walton which is a half-wave form: see for example the left side schematic of this: http://www.wenzel.com/pdffiles/voltmult.pdf

The full-wave takes advantage of both sides of the power trafo so the ripple freq is doubled and the impedance lowered. Also the net secondary d.c. current from the trafo is zero.

C & W were using these to make atom-smashing proton beams, and they didn't need much current.

Brad
 
Back
Top