I just posted a schematic in the Gallery for a regulated power supply using the LR8 adjustable 3-terminal regulator. The supply is fully overload protected with adjustable foldback current limiting. <A HREF="http://antiqueradios.com/photogallery/LR8N3PSq.gif" TARGET=_blank>http://antiqueradios.com/photogallery/LR8N3PSq.gif</A> <P>The supply will provide at least 100 mA output over a voltage range of about 10 to 300 volts. Output current capacity can be increased by adding a second TIP50 connected in a Darlington configuration. This may require a reduction in the value of R103, to prevent the current-limiting circuit from activating.<P>Parts cost, including everything except the transformer and housing, should be under $15. The LR8N3 is available from Mouser (catalog #689-LR8N3) <A HREF="http://www.mouser.com" TARGET=_blank>http://www.mouser.com</A> for $0.82, in a TO-92 (small transistor) package. They also have the TIP50 (catalog #511-TIP50) for $0.80.<P>The VTL5C8 optoisolator is available from Allied Electronics (catalog #980-0720) <A HREF="http://www.alliedelec.com" TARGET=_blank>http://www.alliedelec.com</A> for $3.20. It's the most expensive part other than the transformer. Allied has the TIP50 (catalog #568-2055) for $0.51.<P>If you use a junk-box transformer, make sure you don't exceed 250 volts on the secondary. The LR8 is rated for a maximum input of 450 volts. The TIP50 is rated for 400 volts.<P>R102 is wired to provide maximum resistance when rotated fully clockwise. This corresponds to maximum output voltage. R103 is wired to provide minimum resistance when rotated fully clockwise, corresponding to maximum output current.<P>------------------<BR>73 de Leigh W3NLB | | Leigh@AtwaterKent.Info<BR> <A HREF="http://www.AtwaterKent.info" TARGET=_blank>http://www.AtwaterKent.info</A>

Sure puts the old tube circuits to shame.<P>Darlington? I'd have thought parallelled.<P>Do you happen to know the output impedance vs. frequency, and the ripple level?<P>------------------<BR>

Hi Alan,<P>You need the Darlington configuration to increase the effective beta. The output current of the LR8 is limited to about 10 ma, and it drops as the voltage differential increases. The minimum beta of the TIP50 is 30, so connecting two of them in Darlington mode gives you 900.<P>Don't know the output impedance offhand. Ripple attenuation on the LR8 is ~60-65 db, varies a bit depending on the load current. The data sheet is available on the Supertex website. The test circuit they use is 65 volts DC with 20 volts rms 60 Hz superimposed, and a 1 mfd cap on the output. Certainly a worst-case scenario.<P>I think the input filter on this thing should be increased to around 100 mfd if you were to use it at max output voltage. The ripple is a bit high with 20 mfd.<P>------------------<BR>73 de Leigh W3NLB | | Leigh@AtwaterKent.Info<BR> <A HREF="http://www.AtwaterKent.info" TARGET=_blank>http://www.AtwaterKent.info</A>

Hi Leigh<P> Great job. I printed it.<P> I would increase the input filter cap if high current is drawn at higher voltage. Filter cap value doesn't matter if only a few ma. are drawn.<P>------------------<BR>Norm

Thanks, Norm.<P>I think I'll redraw it with a larger input filter cap, maybe 60 mfd. I was trying to keep the cost down, but it's only a couple more bucks.<P>The ripple rejection on the regulator is so high that you could drop the input cap to 10 mfd and still get a clean output. 10 volts ripple on the input produces less than 10 millivolts ripple on the output, and that's with only a 1 mfd filter cap on the output side.<P>------------------<BR>73 de Leigh W3NLB | | Leigh@AtwaterKent.Info<BR> <A HREF="http://www.AtwaterKent.info" TARGET=_blank>http://www.AtwaterKent.info</A>

Hi Mike,<P>The TIP50 is a common transistor made by ON Semiconductor (used to be Motorola) <A HREF="http://www.onsemi.com" TARGET=_blank>http://www.onsemi.com</A> and several other manufacturers. Summary page at <A HREF="http://www.onsemi.com/site/products/summary/0,4450,TIP50,00.html" TARGET=_blank>http://www.onsemi.com/site/products/summary/0,4450,TIP50,00.html</A> , and the datasheet can be downloaded from that page.<P>The LR8N3 is made by Supertex <A HREF="http://www.supertex.com" TARGET=_blank>http://www.supertex.com</A> . Enter LR8 in their Search box and it will bring up nine pages, including a product summary page and the data sheet link. The N3 is the designator for the TO-92 package.<P>The optoisolator is made by Perkin-Elmer <A HREF="http://www.perkinelmer.com/" TARGET=_blank>http://www.perkinelmer.com/</A> . The VTL5C8 datasheet is available at <A HREF="http://optoelectronics.perkinelmer.com/catalog/Product.aspx?ProductID=VTL5C8" TARGET=_blank>http://optoelectronics.perkinelmer.com/catalog/Product.aspx?ProductID=VTL5C8</A> . Note that the digit 8 suffix is VERY IMPORTANT. Most of the optoisolators are low-voltage devices. This is the only part in the product line with the appropriate voltage rating for this application.<P>I would think all of these parts would be available in the UK. All three manufacturers are active in the global market. If you can't get them or appropriate subs, email me and I can buy some here and ship them to you. It will probably take a while in transit, but at least you'll have them.<P>BTW, it would probably be a good idea to increase the value of the input filter capacitor from 20 mfd to 60 or so. I'm going to revise the drawing to reflect that.<P>Good luck with the project, and thanks for your interest.<P>------------------<BR>73 de Leigh W3NLB | | Leigh@AtwaterKent.Info<BR> <A HREF="http://www.AtwaterKent.info" TARGET=_blank>http://www.AtwaterKent.info</A>

Is there any reason why this power supply could not be directly wired into an old time radio to replace the rectifier tube and the need for a choke and/or speaker inductance coil, assuming the voltage and current and ripple requirements are met? (and maybe keeping the filiment warm on the rectifier tube for authenticitlys sake)<P>------------------<BR>

Hi Bear,<P>You could do that. Receivers don't normally require a regulated supply, but there's no reason you can't use one. It does get around the choke / field coil problem. And the current limiting provides a measure of safety should one of the old components fail.<P>------------------<BR>73 de Leigh W3NLB | | Leigh@AtwaterKent.Info<BR> <A HREF="http://www.AtwaterKent.info" TARGET=_blank>http://www.AtwaterKent.info</A>

Well, the cost seems very reasonable, especially if one can make use of the existing power transformer. I have been stymied in trying to find an appropriate field coil speaker for my Zenith 6S222 and just using a resister and a pm speaker results in too much hum.<P>------------------<BR>

Hi Bear,<P>My only concern would be the raw HV coming out of the rectifier. The LR8 regulator is good for 450 volts, but the TIP50 is only rated for 400 volts. I made a note in the original post to not exceed 250 volts rms on the transformer secondary. If your transformer has higher voltage, you would need to reduce it by some means.<P>------------------<BR>73 de Leigh W3NLB | | Leigh@AtwaterKent.Info<BR> <A HREF="http://www.AtwaterKent.info" TARGET=_blank>http://www.AtwaterKent.info</A>

grampibear, you should be able to reduce the hum in your Zenith by increasing the filter caps and maybe adding a choke to replace the FC.<BR><P>------------------<BR>

Hi Lou. I haven't been able to find a choke even close to the 1250 ohm (resistance) field coil. I did increase the filter capacitance, but fear that increasing it too much will reduce the plate voltage too much. The closest choke I have able to find so far measures about 350 ohms. Haven't tried this one yet though.<P>------------------<BR>

Your right, you won't find a choke with that high a resistance but you can use a combination of choke and resistance. Increasing the output cap will not lower your voltage but raise it. Try using the choke right after the rectifier,then a cap, a resistor and another cap. <P>------------------<BR>

I raised this old post.....does anyone have an archived copy of the schematic ? It's no longer in the gallery.

Easy enough to implement the LR8N with only a pass transistor (done), but I'd like to see how he used the optoisolator and had adjustable current limiting.

Alan is looking for his copy. Id like to see it also.

Carl

Thank you kindly.

First of all, thanks to all previous posters for the fruitful discussion. I am planning to build a HV bench power supply too, so I decided to present yet another schematic for your critique.

I started from the idea that using LR8 is nice (indeed, I found many positive feedback about this device on Internet), however, I wanted to address the following issues of the TIP50-based version.

First, the high-voltage optocoupler is quite costly, so I wanted to avoid using it. Second, the current sensing resistor degrades load regulation, so I started looking for alternative solutions for current limiting. National's datasheet for LM317 gave me an idea that I employed in the new design.

Third, I wanted to increase the maximum current to 0.5A. This is addresed by using two power MOSFETs in parallel as pass elements.

I am attaching the design that came out. I didn't breadboard it yet, so please note that it's totally untested. I would be most grateful for any comments.