Additional paper capacitor? - UK Vintage Radio Repair and Restoration Discussion Forum

WaveyDipole · 13th Jun 2020, 12:55 pm

I have recently acquired a Kriesler radio and started work on it. On studying the radio and circuit, I noticed a couple of anomalies where the radio layout does not appear match the circuit. I am still trying to understand one of them but would like to ask about the presence of a "0.009μF" (9nF) paper capacitor in the tuning circuit. Firstly the value is unusual which suggests that it was deliberately selected, but secondly AFAIK paper capacitors are not stable enough to be included in tuning circuits? I have circled the cap in question on the attached circuit. The 435pF Mica capacitor C17 is still present but has the paper 0.009μF capacitor in series with it. What would be the purpose of such a modification?

The ciircuit here does not have particularly high volatage to warrant belt and braces type isolation and although the capacitance of the Mica 435pF is modified slightly to 415pF, there is already a 20pF trimmer at the other end of the coil for adjustment purposes. I am not sure whether to replace this with a 0.01μF or omit it altogether?

GrimJosef · 13th Jun 2020, 1:08 pm

I am guessing, but could it be there as some attempt at temperature compensation ?

Cheers,

GJ

Radio Wrangler · 13th Jun 2020, 1:59 pm

From the point of view of the tuned circuit, it is in series with both the inductor and the variable capacitor. It therefore reduces the total C seen by the inductor and therefore increases the oscillator frequency above the frequency tuned by the RF tuned circuit. Usually called a padder capacitor.

By playing about with values for the L and the padder a superhet can be made to track with an error that crosses twice through zero within the tuning range and only needs a variable capacitor with the same values and laws for RF and LO sections.

Notice that the set has separate padders for the two wavebands.

They could have been fitted in series with the hot ends of the coils, but the physical arrangement might work better one way than the other.

Paper capacitors are not brilliant, but small value ones have been used in tuned circuits. The 'Micamold' capacitors in the RF decks of AR88 weren't mica at all, they were paper, faking it!

David

turretslug · 13th Jun 2020, 2:39 pm

Is this a set with reasonable tuning scale resolution? It may be that someone had a touch of the Friday afternoon syndrome when calculating the padder value, or stray capacitance (e.g. oscillator coil self-capacitance) may have been higher than anticipated, and the series 9nF was a quick-and-easy production correction. 415 or 435pF padder suggests a relatively low frequency band on a typical superhet, or at least around MW region, so the oscillator coil might have quite a few turns and winding capacitance.

Some superhets have three-point tracking, particularly LF up-converting sets where the oscillator's end-to-end ratio is a small fraction that of the RF circuit span and the padder value is low, critical and prone to stray error. The extra adjustment is a trimmer across the padder and is adjusted towards the LF end of the band, the coil inductance around mid-span and the "usual" trimmer towards the usual HF end.

WaveyDipole · 13th Jun 2020, 5:21 pm

I am not sure how to answer the question regarding the scale resolution, but physically the scale is about the size of the one you find in a DAC90. It is certainly no Eddystone. However, the radio does have a short-wave band which would perhaps require a more accurate resolution than your typical long/medium wave broadcast band.

Having done a search on the term "padding capacitor" it seems that this arrangement would not be unusual in some radios and might possibly be related to three point tracking. Whatever the case, if what is stated here is true, then its going to be a tricky proposition to change the caps.

https://antiqueradios.com/forums/vie...p?f=6&t=299219

The tolerance of these paper caps is ±15% so would something within that tolerance range be a suitable replacement?

turretslug · 13th Jun 2020, 5:52 pm

I was thinking along the lines that if the scale had sufficiently good resolution, then aligning the set for consistent performance with the wrong value of padder would result in noticeable discrepancy at some point in the scale and the manufacturer might have been sufficiently guilt-stricken that they had to do something about it! Otherwise, if it only had a marker every 100kHz or so, they might have got away with it. I believe (and I stand to be corrected here) that there was a run of R1155s that had an incorrect padder fitted on one band, this set having sufficient scale resolution to make the problem obvious.

With SW tuning and typical IF, the padder ends up usually being low nF in range (say 1.5-4nF depending on span), so strays aren't as significant whereas MW padder is often mid-hundreds of pF and LW even lower, so discrepancies are more of a problem. The Eddystone 670A here showed discrepancy on its relatively-highly resolving LW scale, so I replaced the original 150pF padder with 120 + 15pF in parallel with a trimmer for three-point tracking to minimise error. The caps used often seem to be 1 to 2% tolerance but the 9nF capacitor in question has such a large value compared to its series counterpart that tolerance is less critical. I suppose that you could use 390pF or 400pF with a trimmer in parallel with that to replace the original 435 + 9,000pF series combination.

WaveyDipole · 13th Jun 2020, 6:22 pm

There are no markers of that kind, only numbers from 16, 19, 25, 31, 40, 50 in metres spaced in a non-linear fashion, so in that respect, there is no real precision, only that the pointer should align with a number at the given wavelength, which if my calculations are correct would correspond with 6-18.7MHz.

Thanks for you suggestion with regards to the trimmer in parallel. That does sound like a reasonable solution. I will check my box of bits for something suitable.

WaveyDipole · 19th Jun 2020, 11:37 am

Ok, this is embarrasing!

It would appear that the caps were not in series after all. I could see that the ends were twisted together, but they were also tied to the chassis underneath the coil which I didn't spot at first. The caps are, in fact, C17 and C18. C18 which was a 0.009 waxie, is shown as having a value of 0.008 in the BOM list, so that now makes sense.

I replaced this and C21 (both waxies) with new caps closely matching in value.

13th Jun 2020, 12:55 pm	#1
WaveyDipole Nonode Join Date: Mar 2013 Location: Leicester, UK Posts: 2,535	Additional paper capacitor? I have recently acquired a Kriesler radio and started work on it. On studying the radio and circuit, I noticed a couple of anomalies where the radio layout does not appear match the circuit. I am still trying to understand one of them but would like to ask about the presence of a "0.009μF" (9nF) paper capacitor in the tuning circuit. Firstly the value is unusual which suggests that it was deliberately selected, but secondly AFAIK paper capacitors are not stable enough to be included in tuning circuits? I have circled the cap in question on the attached circuit. The 435pF Mica capacitor C17 is still present but has the paper 0.009μF capacitor in series with it. What would be the purpose of such a modification? The ciircuit here does not have particularly high volatage to warrant belt and braces type isolation and although the capacitance of the Mica 435pF is modified slightly to 415pF, there is already a 20pF trimmer at the other end of the coil for adjustment purposes. I am not sure whether to replace this with a 0.01μF or omit it altogether? Attached Thumbnails

13th Jun 2020, 1:08 pm	#2
GrimJosef Dekatron Join Date: Sep 2007 Location: Oxfordshire, UK. Posts: 4,311	Re: Additional paper capacitor? I am guessing, but could it be there as some attempt at temperature compensation ? Cheers, GJ __________________ http://www.ampregen.com

13th Jun 2020, 1:59 pm	#3
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,876	Re: Additional paper capacitor? From the point of view of the tuned circuit, it is in series with both the inductor and the variable capacitor. It therefore reduces the total C seen by the inductor and therefore increases the oscillator frequency above the frequency tuned by the RF tuned circuit. Usually called a padder capacitor. By playing about with values for the L and the padder a superhet can be made to track with an error that crosses twice through zero within the tuning range and only needs a variable capacitor with the same values and laws for RF and LO sections. Notice that the set has separate padders for the two wavebands. They could have been fitted in series with the hot ends of the coils, but the physical arrangement might work better one way than the other. Paper capacitors are not brilliant, but small value ones have been used in tuned circuits. The 'Micamold' capacitors in the RF decks of AR88 weren't mica at all, they were paper, faking it! David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

13th Jun 2020, 2:39 pm	#4
turretslug Dekatron Join Date: Nov 2011 Location: Surrey, UK. Posts: 4,396	Re: Additional paper capacitor? Is this a set with reasonable tuning scale resolution? It may be that someone had a touch of the Friday afternoon syndrome when calculating the padder value, or stray capacitance (e.g. oscillator coil self-capacitance) may have been higher than anticipated, and the series 9nF was a quick-and-easy production correction. 415 or 435pF padder suggests a relatively low frequency band on a typical superhet, or at least around MW region, so the oscillator coil might have quite a few turns and winding capacitance. Some superhets have three-point tracking, particularly LF up-converting sets where the oscillator's end-to-end ratio is a small fraction that of the RF circuit span and the padder value is low, critical and prone to stray error. The extra adjustment is a trimmer across the padder and is adjusted towards the LF end of the band, the coil inductance around mid-span and the "usual" trimmer towards the usual HF end.

13th Jun 2020, 5:21 pm	#5
WaveyDipole Nonode Join Date: Mar 2013 Location: Leicester, UK Posts: 2,535	Re: Additional paper capacitor? I am not sure how to answer the question regarding the scale resolution, but physically the scale is about the size of the one you find in a DAC90. It is certainly no Eddystone. However, the radio does have a short-wave band which would perhaps require a more accurate resolution than your typical long/medium wave broadcast band. Having done a search on the term "padding capacitor" it seems that this arrangement would not be unusual in some radios and might possibly be related to three point tracking. Whatever the case, if what is stated here is true, then its going to be a tricky proposition to change the caps. https://antiqueradios.com/forums/vie...p?f=6&t=299219 The tolerance of these paper caps is ±15% so would something within that tolerance range be a suitable replacement?

13th Jun 2020, 5:52 pm	#6
turretslug Dekatron Join Date: Nov 2011 Location: Surrey, UK. Posts: 4,396	Re: Additional paper capacitor? I was thinking along the lines that if the scale had sufficiently good resolution, then aligning the set for consistent performance with the wrong value of padder would result in noticeable discrepancy at some point in the scale and the manufacturer might have been sufficiently guilt-stricken that they had to do something about it! Otherwise, if it only had a marker every 100kHz or so, they might have got away with it. I believe (and I stand to be corrected here) that there was a run of R1155s that had an incorrect padder fitted on one band, this set having sufficient scale resolution to make the problem obvious. With SW tuning and typical IF, the padder ends up usually being low nF in range (say 1.5-4nF depending on span), so strays aren't as significant whereas MW padder is often mid-hundreds of pF and LW even lower, so discrepancies are more of a problem. The Eddystone 670A here showed discrepancy on its relatively-highly resolving LW scale, so I replaced the original 150pF padder with 120 + 15pF in parallel with a trimmer for three-point tracking to minimise error. The caps used often seem to be 1 to 2% tolerance but the 9nF capacitor in question has such a large value compared to its series counterpart that tolerance is less critical. I suppose that you could use 390pF or 400pF with a trimmer in parallel with that to replace the original 435 + 9,000pF series combination.

13th Jun 2020, 6:22 pm	#7
WaveyDipole Nonode Join Date: Mar 2013 Location: Leicester, UK Posts: 2,535	Re: Additional paper capacitor? There are no markers of that kind, only numbers from 16, 19, 25, 31, 40, 50 in metres spaced in a non-linear fashion, so in that respect, there is no real precision, only that the pointer should align with a number at the given wavelength, which if my calculations are correct would correspond with 6-18.7MHz. Thanks for you suggestion with regards to the trimmer in parallel. That does sound like a reasonable solution. I will check my box of bits for something suitable.

19th Jun 2020, 11:37 am	#8
WaveyDipole Nonode Join Date: Mar 2013 Location: Leicester, UK Posts: 2,535	Re: Additional paper capacitor? Ok, this is embarrasing! It would appear that the caps were not in series after all. I could see that the ends were twisted together, but they were also tied to the chassis underneath the coil which I didn't spot at first. The caps are, in fact, C17 and C18. C18 which was a 0.009 waxie, is shown as having a value of 0.008 in the BOM list, so that now makes sense. I replaced this and C21 (both waxies) with new caps closely matching in value.