From: dkelvey at hotmail.com To: cctalk at classiccmp.org Subject: RE: S-100 power supply transformer Date: Sun, 17 May 2009 18:15:27 -0700 ---------------------------------------- ---snip--- Hi What you want is an 8 volt and a 13.5 volt AC. You forget that the DC voltage is created from the peaks not the RMS average. You have to account for diode drop as well. That is how I got the numbers. I assumed full wave rectification and .6v per diode. Dwight

From: steven.alan.canning at verizon.net To: cctech at classiccmp.org; cctalk at classiccmp.org Subject: Re: S-100 power supply transformer Date: Mon, 18 May 2009 23:36:52 -0700 Dwight, I think a little explanation is in order, otherwise your info borders on incorrect. You need to increase the CURRENT rating of the transformer by 1.4, but this only applies to a full wave rectifier topology and not a full wave BRIDGE topology. reference: The holy book " The art of electronics " page 47 Best regards, Steven for Hi I forgot to mention. You need to derate the transformer as well. The problem is that the current flows when the voltage is the highest ( or close to it ). The math is a little complicated but it means you need to increase the rating of the transformer by at least 1.414. If you can't find a transformer with the right windings, I recommend looking at toroidal transformers. These can easily have turns added to increase or buck the voltage of the secondary. Dwight

From: steven.alan.canning at verizon.net To: cctech at classiccmp.org; cctalk at classiccmp.org Subject: Re: S-100 power supply transformer Date: Mon, 18 May 2009 23:36:52 -0700 Dwight, I think a little explanation is in order, otherwise your info borders on incorrect. You need to increase the CURRENT rating of the transformer by 1.4, but this only applies to a full wave rectifier topology and not a

wave BRIDGE topology. reference: The holy book " The art of electronics " page 47 Best regards, Steven for Hi I forgot to mention. You need to derate the transformer as well. The problem is that the current flows when the voltage is the highest ( or close to it ). The math is a little complicated but it means you need to increase the rating of the transformer by at least 1.414. If you can't find a transformer with the right windings, I recommend looking at toroidal transformers. These can easily have turns added to increase or buck the voltage of the secondary. Dwight

from the rectifier (IDC) is: IAC = KFF x IDC where KFF is the form factor.

From: steven.alan.canning at verizon.net To: cctalk at classiccmp.org Subject: Re: S-100 power supply transformer Date: Tue, 19 May 2009 14:34:12 -0700 I respectfully disagree. ? In a center-tapped full-wave rectifier circuit the output voltage is half what you get if you use a bridge rectifier. It

not the most efficient circuit in terms of transformer design, because

half of the secondary is used only half the time. Thus the current through the winding during that time is twice what it would be for a true

circuit. Heating in the windings, using Ohm?s law, is I^2R, so you have

times the heating half the time, or twice the average heating of an equivalent full-wave bridge rectifier circuit. You would have to choose a transformer with a current rating 1.4 times as large as compared with the ( better ) bridge circuit; besides costing more, the resulting supply

be bulkier and heavier ?. from the good book ? ? The art of Electronics ? The disadvantage of a bridge is two diode voltage drops. Diodes are a lot cheaper than transformers though. Best regards, Steven ------------------------------------------------------------------------- Hi This is the result of when the current actually flows. It only flows when the voltage is near the peak. This changes the average power dissipated in the secondary winding ( the VA rating ). Altough, the 1.414 is the worst value and the real value is somewhere between 1 and 1.414. It make no difference what type of rectification is used, fullwave bridge or center tap. Watts are watts and there is no way around it. It is true that the load current rating is halved for the secondary turns for the center tapped fullwave rectification but the VA rating of the transformer remains the same. I'm not sure what the book you reference says. I do know how power is calculated and it is always the vectored volts times amps. The current flows out of the transformer when the output voltage is close to the peak. This means that the rms current of the secondary needs to be compensated to account for when the current actually flows. If your transformer has a secondary rated at 10 amps that can only feed a rectifier and capacitor filter of 7.07 amps. That would change if feeding a inductive filter because of when the current flows. If it was a center tapped with two diodes, it would be 14.14 amps instead of 20 amps. Still, the VA rating of the transformer does change. Dwight ---------------------------------------- full > wave BRIDGE topology. > > reference: The holy book " The art of electronics " page 47 > > Best regards, Steven > > >> ---------------------------------------- >>> From: lynchaj at yahoo.com >>> To: cctalk at classiccmp.org >>> Subject: S-100 power supply transformer >>> Date: Sun, 17 May 2009 16:51:44 -0400 >>> >>> Hi! Does anyone know of an available transformer suitable for making an >>> S-100 power supply? I need 120VAC input and 10VAC output and +/- 18VAC >>> outputs. >>> >>> The best I solution I can find is two separate 80VA transformers; one >>> parallel 10VAC output and another 36VCT output which can be configured > for >>> +/-18VAC outputs. >>> >> ---snip--- >> >> Hi >> What you want is an 8 volt and a 13.5 volt AC. You forget >> that the DC voltage is created from the peaks not the RMS >> average. You have to account for diode drop as well. >> That is how I got the numbers. I assumed full wave rectification >> and .6v per diode. >> Dwight >> > > Hi > I forgot to mention. You need to derate the transformer > as well. The problem is that the current flows when the > voltage is the highest ( or close to it ). The math is a little > complicated but it means you need to increase the rating of > the transformer by at least 1.414. > If you can't find a transformer with the right windings, > I recommend looking at toroidal transformers. These can > easily have turns added to increase or buck the voltage > of the secondary. > Dwight

---------------------------------------- What voltages were they talking about for these types of numbers? The 1.414 number is the best case. allowing some for the diode and as you save, expected ripple. Dwight

On 17 May 2009 at 21:32, Steve Stutman wrote: While I can understand that using a 78xx/79xx vintage regulator might be important for the sake of accuracy, I'd be sort tempted to use modern monolithic switching regulators if one wanted to preserve the S100 "on board" regulation. No more "scorched PCB" or big honking heatsinks--and less conversion of amps into heat. --Chuck

David Griffith wrote: What did you do with the power feeds on the backplane? Got any pics?

Alexandre Souza wrote: The problem is compatibility. Imagine that each S-100 board you would connect on your system has to be modified. And every board that will be sold, lended or given away, will ALSO have to be modified. Why not just change the regulator for a newer one? They are so cheap nowadays :o)

from cards that have been converted plug in here. I don't have a lot

I suspect that it's a combination of high-power use and the quality of the TO-220 packaging back then. I recall that on some MITS boards, (was it the 4K DRAM board?) bridging the regulator with a power resistor was done to keep the temperature down to that of a slow bonfire.

-----Oorspronkelijk bericht----- Van: cctalk-bounces at classiccmp.org [mailto:cctalk-bounces at classiccmp.org] Namens Teo Zenios Verzonden: maandag 18 mei 2009 12:02 Aan: General Discussion: On-Topic and Off-Topic Posts Onderwerp: Problem with an old Tandy CM-4 monitor A while ago I got a Tandy 1000 system with a CM-4 monitor. The system worked fine for 20 minutes and then the monitor died. So far I havn't seen anything that looks burnt on the inside, but the tube doesn't light up when you provide power and there is a crackling noise going on. Is there a common issue with these old CGA monitors, and is this fixable without gutting a working one?

The problem with the onboard regulation is that many of those original regulators, though they might be working, are actually no longer any good. Not sure if it's just age, or it's a lifetime of certain ones running too near max, but in all the ones I've replaced, I've cut the amperage draw of the board way down, sometimes in half. If they're going to be replaced *anyway*, then makes sense just to remove them and their associated, equally suspect tantalums/aecs altogether. In my IMSAI, I feed each board power from a dedicated, pluggable line running to it... ground is still done through the connector.