Article 1083 of rec.antiques.radio+phono: Path: news.umbc.edu!hookup!news.kei.com!yeshua.marcam.com!usc!howland.reston.ans.net!news.sprintlink.net!bga.com!bga.com!nobody From: vancleef@bga.com (Henry van Cleef) Newsgroups: rec.antiques.radio+phono Subject: Re: 400 MFd capacitor **TEST RESULTS** Date: 20 Oct 1994 11:13:36 -0500 Organization: Real/Time Communications - Bob Gustwick and Associates Lines: 62 Message-ID: <3864vg$2qh@ivy.bga.com> References: <385n3o$g8k@fnnews.fnal.gov> NNTP-Posting-Host: ivy.bga.com In article <385n3o$g8k@fnnews.fnal.gov>, Dans Cockatoo Ranch wrote: > >After reading all the speculation about what happens when you install a 400 Mfd >filter capacitor in an AC/DC set designed for 40 to 50 Mfd in the first >section I decided to measure it. Here are the results: > >Unit Under Test: >Crosley Model 11-114U >Manufactured 1951 >Five tubes 12BA6, 12BE6, 12AT6, 50C5, 35W4 > > 50 Mfd FILTER 450 Mfd FILTER >STARTUP PEAK CURRENT: <400 MA <420 MA >OPERATING PEAK CURRENT: 400 MA 420 MA >OPERATING RMS CURRENT: 115 MA 119 MA >OPERATING PEAK VOLTAGE: 144 V PEAK 138 V PEAK >OPERATING DC VOLTAGE: 133 VDC 134 VDC >RIPPLE PEAK TO PEAK: 18 VPP 2 VPP > >-Measured current is the plate current of the 35W4. >-Measured voltage is the cathode voltage of the 35W4. >-Volume turned to minimum, no station tuned. >-Peak voltages and currents measured with a Tektronix 2235 oscilloscope. >-Peak currents measured using a 1 ohm shunt in the plate circuit of the 35W4. >-RMS and DC voltages and currents measured with a Fluke 8060A true RMS multimeter. > >The normal first filter capacity of this set is 50 Mfd. For the large capacity >measurements a 400 Mfd capacitor was placed in parallel with the 50 Mfd >capacitor in the set giving a total of 450 Mfd. > >On turn on the plate current started at zero and slowly rose to the operating >current for both capacities. No startup transients were observed. Other than >a greatly reduced ripple voltage in the first filter section there is little >difference in the operating characteristics between the normal and large >capacitors. >The operating peak voltage actually drops slightly due to the decreased ripple >and the operating DC voltage rises slightly for the same reason. The final >result is that it makes little difference if you install a much larger filter >capacitor in the set. The changes in the operating characteristics are >minimal. Daniel, did you measure the plate-cathode voltage on the 35W4? The tube is spec'ed to have 18 volts drop at 200 ma. DC average should be on the order of 60-70 ma. for that set, not around 115 ma. The reduction in peak voltage on the cap is as expected, but I'll bet you have a huge current spike when the tube is conducting, and the Fluke may not be accurate with that waveform. I'd be much happier if that circuit had 15 ohms between the 35W4 cathode and the big barrel cap, and I'll bet it only drops the DC by about a volt. Also, a .05 or a .1 across the pilot lamp wiring to bypass some of the transients might be a nice idea. What kills 35Z5's and 50L6's is mechanical shaking of the heater from the magnetic fields induced by heater current. -- *********************************************************** Hank van Cleef vancleef@bga.com vancleef@tmn.com *********************************************************** Article 1089 of rec.antiques.radio+phono: Path: news.umbc.edu!eff!news.kei.com!uhog.mit.edu!sgiblab!swrinde!elroy.jpl.nasa.gov!lll-winken.llnl.gov!fnnews.fnal.gov!usenet From: schoo@fnal.gov (Dans Cockatoo Ranch) Newsgroups: rec.antiques.radio+phono Subject: Re: 400 MFd capacitor **TEST RESULTS** Date: 20 Oct 1994 17:12:42 GMT Organization: Fermilab Lines: 48 Distribution: world Message-ID: <3868ea$nmp@fnnews.fnal.gov> NNTP-Posting-Host: eeddps.fnal.gov X-Newsreader: In article <3864vg$2qh@ivy.bga.com> vancleef@bga.com (Henry van Cleef) writes: >Dans Cockatoo Ranch wrote: >> 50 Mfd FILTER 450 Mfd FILTER >>STARTUP PEAK CURRENT: <400 MA <420 MA >>OPERATING PEAK CURRENT: 400 MA 420 MA >>OPERATING RMS CURRENT: 115 MA 119 MA >>OPERATING PEAK VOLTAGE: 144 V PEAK 138 V PEAK >>OPERATING DC VOLTAGE: 133 VDC 134 VDC >>RIPPLE PEAK TO PEAK: 18 VPP 2 VPP >> >>-Measured current is the plate current of the 35W4. >>-Measured voltage is the cathode voltage of the 35W4. >>-Volume turned to minimum, no station tuned. >>-Peak voltages and currents measured with a Tektronix 2235 oscilloscope. >>-Peak currents measured using a 1 ohm shunt in the plate circuit of the 35W4. >>-RMS and DC voltages and currents measured with a Fluke 8060A true RMS multimeter. >Daniel, did you measure the plate-cathode voltage on the 35W4? The >tube is spec'ed to have 18 volts drop at 200 ma. No, I did not happen to measure that value. >DC average should be on the order of 60-70 ma. for that set, not around >115 ma. The reduction in peak voltage on the cap is as expected, but >I'll bet you have a huge current spike when the tube is conducting, and >the Fluke may not be accurate with that waveform. Yes there is a huge current spike. The 400 ma number is the peak plate current measured with the 'scope using a 1 ohm shunt. The Fluke is very accurate and designed to give the true RMS values of waveforms just like this one. I could confirm by measuring the DC out from the filter. The comparison is what is important in that it shows that despite the much larger capacity, the plate current (peak or RMS take your choice) is not effected very much. > >I'd be much happier if that circuit had 15 ohms between the 35W4 >cathode and the big barrel cap, and I'll bet it only drops the DC by >about a volt. Well, with a peak plate current of 420ma and a 15 ohm resistor this would produce a drop of 6.3 volts peak. Not enough to bother anything significantly. From the numbers I really didn't see the need for it but if it makes you happy I'll allow you that pleasure. V Daniel Schoo (o o) Electronics Design Engineer ( V ) Fermilab, Batavia, Illinois, USA .......m.m......Dan's Cockatoo Ranch vvv Article 1114 of rec.antiques.radio+phono: Path: news.umbc.edu!hookup!news.kei.com!sol.ctr.columbia.edu!howland.reston.ans.net!EU.net!uunet!news.tek.com!tekgp4.cse.tek.com!dank From: dank@tekgp4.cse.tek.com (Daniel G Knierim) Newsgroups: rec.antiques.radio+phono Subject: Re: 400 MFd capacitor **TEST RESULTS** & Resistor dissipation Date: 21 Oct 1994 08:02:06 GMT Organization: Tektronix, Inc., Beaverton, OR. Lines: 42 Message-ID: <387shu$o4q@tekadm1.cse.tek.com> References: <385n3o$g8k@fnnews.fnal.gov> <3864vg$2qh@ivy.bga.com> Reply-To: dank@mdhost.cse.tek.com NNTP-Posting-Host: tekgp4.cse.tek.com In article <3864vg$2qh@ivy.bga.com> vancleef@bga.com (Henry van Cleef) writes: >In article <385n3o$g8k@fnnews.fnal.gov>, >Dans Cockatoo Ranch wrote: >>OPERATING PEAK CURRENT: 400 MA 420 MA >>OPERATING RMS CURRENT: 115 MA 119 MA >DC average should be on the order of 60-70 ma for that set, not around 115 ma. Ah, now I see. Let's clarify something. The DC average current supplied by a half-wave rectifier is necessarily less than the RMS current flowing through it. The DC average *is* around 60-70 ma, while the RMS is around double that, or 115-119 ma. The RMS current thru a half-wave rectifier is *always* higher than the DC average current thru the rectifier. Schematics are much more likely to quote the DC average current used by (various stages of) a set, and rectifier specs typically state the continuous DC current that can be supplied, but what we need to use in calculating resistor dissipation is the RMS. In another article (r.e. a Zenith T-O), Henry writes: >I used the RMS value of 100 ma., which is the maximum continuous DC >output from many rectifier tubes, and which is more than the average >3-way set will actually draw. RMS, or "root mean square," is the heat >value of an AC current, and the big issue with resistors is heat. 100 ma may be the maximum continuous DC output for many rectifiers, and more than the average 3-way set will actually draw (again, for DC average current). But the RMS current will be much higher than the DC average, typically twice or more. Since power dissipation in a resistor is proportional to I^2, that leads to power that is around four times higher in a resistor in series with a rectifier than you would get if the same resistor were placed after the filter cap! In the specific case of a 3-way set, the filaments of the other tubes draw at least 50 ma DC from the rectifier, and the plates draw some more current, so the DC load on the rectifier is typically around 80 ma or so. This leads to an RMS current thru the rectifier well over 100 ma. - Dan Knierim These opinions aren't worth the standard disclaimer form they're printed on! Article 1117 of rec.antiques.radio+phono: Path: news.umbc.edu!haven.umd.edu!ames!lll-winken.llnl.gov!uwm.edu!fnnews.fnal.gov!usenet From: schoo@fnal.gov (Dans Cockatoo Ranch) Newsgroups: rec.antiques.radio+phono Subject: Re: 400 MFd capacitor, the final chapter Date: 21 Oct 1994 11:50:24 GMT Organization: Fermilab Lines: 33 Distribution: world Message-ID: <3889u0$kct@fnnews.fnal.gov> NNTP-Posting-Host: eeddps.fnal.gov X-Newsreader: >In article <3864vg$2qh@ivy.bga.com> vancleef@bga.com (Henry van Cleef) writes: >In article <385n3o$g8k@fnnews.fnal.gov>, >Dans Cockatoo Ranch wrote: >> >>OPERATING PEAK CURRENT: 400 MA 420 MA >>OPERATING RMS CURRENT: 115 MA 119 MA > >DC average should be on the order of 60-70 ma for that set, not around 115 ma. Ok now I'll throw another rock on the pile. I measured the plate current again last night and this time I used the DC amps setting on the meter. The meter read 50 ma DC current which is close to the number Hank quoted. All of the other current measurements made with the Fluke were made using the AC setting so that I could get the RMS values. The peak and RMS currents were the main point of discussion as to the effect of putting a large capacitor in the first filter stage of an AC DC radio. For your viewing pleasure I also measured the total RMS current draw of the radio, 240 ma RMS. The radio is spec'ed at 270 so it's pretty close to normal. BTW when current and voltage numbers are given in a schematic they usually also report what kind of a meter was used to measure them. If you use a different type of meter I.E. a high tech Fluke instead of the Simpson 260 that they used, the readings are likely to vary. The Fluke can read a complex waveform and accurately display an RMS value. A Simpson VOM on the other hand assumes a sinusoid, rectifies it and applies a fixed conversion constant to display the RMS value. This is fine for most things as long as everybody measures the circuit with the same type of meter, they all get the same results. If you use a meter that differs from the one that was used by the manufacturer you should be aware that the readings may differ. V Daniel Schoo (o o) Electronics Design Engineer ( V ) Fermilab, Batavia, Illinois, USA .......m.m......Dan's Cockatoo Ranch vvv