Recapping Questions

DraKnight13

Active Member
Hello,
I am getting into getting some work on some of my tube amp (namely an Eico HF-87 atm). I have couple of questions.
  1. Does the wattage of a resistor matter at all (some are 1/4W, some are 1/2W)? Can I just replace with 1W resistor?
  2. Using a multimeter, some of the range on the capacitor is puzzling to me (some are dead on like the label, but some is much higher). I have cap that rated at 50uF measured at 70uF, sometime 2 or 3 times the rated label. Am I measuring them wrong or this is normal????? I am trying to keep them as original as possible, so anything that isn't out of spec I don't want to replace. If one of the pair went out then I will go ahead and replace them all.
Thank you!
 
Register to hide this ad
1. It is usually safe to increase wattage, but not to decrease wattage. The exception to that is in those instances where a resistor is intended to be fusible, but I believe that's a lot more common in SS units, and pretty uncommon in a tube chassis.

2. I've seen that as well, and if the cap is intended as a filter instead of a coupler, extra capacitance is not usually a problem. That's a big jump, however, and well over the typical ±20% tolerance.
 
Thanks for the answer.
I just want to clarify, if it is coupling cap, I should just replace them? It is higher than the 20% tolerance. I will order orange drop cap since it seems to be popular, is there anything other cap I should be buying?
 
Replace all of the electrolytic capacitors in the signal path with film. Your ears will thank you.

Replace all of the electrolytic capacitors in the power supply with electrolytics.

An electrolytic capacitor may have a higher value than is specified. Many run higher than values shown on the printed label or the schematic. The specifications on even new parts are typically -20% to +80%. It's inherent in how the parts are made.

Replace the carbon-composite resistors with metal film, as this has the lowest noise in an affordable package. (Wirewound has no noise, but is very expensive.)
 
Most resistors in tube gear are 1/2w unless noted. The schematic will tell you where they are different but the higher wattage ones are bigger. You can up the wattage if you want but its generally not needed.

and yeah, cap tolerances are pretty much the wide side of a barn. Pay little mind to what they read, replace with basically the same marking on the label, +/- 10% as needed to get values that can be had now.
 
I already order a metal film resistor kit, hopefully it will give me a baseline of what I need. If I need more I can order them individually. The power supply part I think I can get. My question is the choice of film capacitor (most post I see favor the orange drop) is there better alternative (I know they are subjective). Also film capacitor can have higher value than rated?
 
Hello,
I am getting into getting some work on some of my tube amp (namely an Eico HF-87 atm). I have couple of questions.
  1. Does the wattage of a resistor matter at all (some are 1/4W, some are 1/2W)? Can I just replace with 1W resistor?
  2. Using a multimeter, some of the range on the capacitor is puzzling to me (some are dead on like the label, but some is much higher). I have cap that rated at 50uF measured at 70uF, sometime 2 or 3 times the rated label. Am I measuring them wrong or this is normal????? I am trying to keep them as original as possible, so anything that isn't out of spec I don't want to replace. If one of the pair went out then I will go ahead and replace them all.
Thank you!
The 87 is an excellent sounding amp. I believe that the 50uf cap is the cathode bypass cap. If true you should replace it with a hi quality 100uf x50vdc like Elna Cerafine. I usually also bypass the electrolytic with a film cap. I bought some 3uf x 50vdc PIO cap a long time ago and they work great. The reason to do this is because the EL34 tubes are cathode biased which means the cathode bypass caps are part of the signal chain.
If your coupling caps are measuring high or out of specs it usually means they are leaky and definitely need to be replaced. I like to use a PIO, paper in oil , like .1 x 600vdc vitaminQ , which sounds very nice in this amp.
Also the voltage doubler caps should be increased in value to around 500uf. It is very hard to get can caps to replace the original units so you will probably need to mound the new ones on clamps or you can epoxy glue them to the chassis, which works well and looks cleaner.
 
Arbitrarily increasing the value for capacitors in the signal path can have unwanted effects. If the capacitor is a filter, you're changing the frequency response.

If the original was +/- 20% and you're slightly increasing the nominal value by, say, 5%, you're still well inside the original tolerance.

Here's some words I wrote on the subject of the shield marking on film capacitors.

Film capacitors deliver a performance difference depending upon how the outer foil and inner foil are connected, and the printed label on the case is often done at random and has no value whatsoever. (See the video below listed.)

It is better to connect the outside of the capacitor foil to the low-impedance side (previous stage's plate) to act as a shield. The grid side is very high impedance, so it's best to pull the signal from the interior of the capacitor as it is better shielded.

Here's a tester you can easily build, which not only will identify the shield end but it proves that manufacturers randomly label capacitors:
As far as, "does it matter?", the answer, summarized, is "that depends".

The shield will help minimize noise because of the high-impedance/low-impedance distinction.

There's a lot of noise in the amplifier: internal noise from diode Qrr broadcast and tank circuits combined with external noise from SMPS, lighting, and stray RF, some of which comes in on the input cables. The high-impedance grid is very, very good at rectifying such noise. This will generally not be audible, except AM radio is sometimes demodulated by the high-gain phono stage, but it will reduce the amplifier bandwidth because the signal is being amplified. Just like oscillation is amplified and uses bandwidth, even if one can't hear it.

Tubes are very, very good amplifiers and will happily amplify inaudible signals from the hundreds of kHz to the many tens of MHz. So one must go looking for the noise with an oscilloscope to see the effects.

My view is build for minimal noise wherever possible because it adds up. So lower-noise metal film resistors, snubbers on low-Qrr diodes, snubbers on transformers, etc.
 
The 87 is an excellent sounding amp. I believe that the 50uf cap is the cathode bypass cap. If true you should replace it with a hi quality 100uf x50vdc like Elna Cerafine. I usually also bypass the electrolytic with a film cap. I bought some 3uf x 50vdc PIO cap a long time ago and they work great. The reason to do this is because the EL34 tubes are cathode biased which means the cathode bypass caps are part of the signal chain.
If your coupling caps are measuring high or out of specs it usually means they are leaky and definitely need to be replaced. I like to use a PIO, paper in oil , like .1 x 600vdc vitaminQ , which sounds very nice in this amp.
Also the voltage doubler caps should be increased in value to around 500uf. It is very hard to get can caps to replace the original units so you will probably need to mound the new ones on clamps or you can epoxy glue them to the chassis, which works well and looks cleaner.
Thanks I did increase all the value of the voltage doubler cap (560uF). I use a combination of glue and clamps. It doesn't look too bad I don't think. What does the 3uF 50VDC cap replace in the amp????? (if it possible can you tell me the caap number base on the schematic). For the 50uF cap, I will check out and buy the Elna cap as you recommended.
I bought this HF-87 on impulse (it is complete other than the voltage doubler cap damaged), also rust everywhere on the transformer (trying to clean it out but doesn't look that great). I want to see if I can do this before I brought my brother pair of HF-60 to the chopping block.
 
Thanks I did increase all the value of the voltage doubler cap (560uF). I use a combination of glue and clamps. It doesn't look too bad I don't think. What does the 3uF 50VDC cap replace in the amp????? (if it possible can you tell me the caap number base on the schematic). For the 50uF cap, I will check out and buy the Elna cap as you recommended.
In my 87 i used the 3uf cap in parallel with the 100uf x 50vdc Elna Cerafine for the cathode bypass capacitor.
 
I forgot to mention that the grounding scheme on the 87 is sub par, since they will use some of the rivets that hold down components as ground points. You should make a star ground or bus ground for best results. The eico was a kit and it seems that the designers did not want to make things overly complicated for a first time builder and so they used convenient grounding points
 
Arbitrarily increasing the value for capacitors in the signal path can have unwanted effects. If the capacitor is a filter, you're changing the frequency response.

If the original was +/- 20% and you're slightly increasing the nominal value by, say, 5%, you're still well inside the original tolerance.

Here's some words I wrote on the subject of the shield marking on film capacitors.

Film capacitors deliver a performance difference depending upon how the outer foil and inner foil are connected, and the printed label on the case is often done at random and has no value whatsoever. (See the video below listed.)

It is better to connect the outside of the capacitor foil to the low-impedance side (previous stage's plate) to act as a shield. The grid side is very high impedance, so it's best to pull the signal from the interior of the capacitor as it is better shielded.

Here's a tester you can easily build, which not only will identify the shield end but it proves that manufacturers randomly label capacitors:
As far as, "does it matter?", the answer, summarized, is "that depends".

The shield will help minimize noise because of the high-impedance/low-impedance distinction.

There's a lot of noise in the amplifier: internal noise from diode Qrr broadcast and tank circuits combined with external noise from SMPS, lighting, and stray RF, some of which comes in on the input cables. The high-impedance grid is very, very good at rectifying such noise. This will generally not be audible, except AM radio is sometimes demodulated by the high-gain phono stage, but it will reduce the amplifier bandwidth because the signal is being amplified. Just like oscillation is amplified and uses bandwidth, even if one can't hear it.

Tubes are very, very good amplifiers and will happily amplify inaudible signals from the hundreds of kHz to the many tens of MHz. So one must go looking for the noise with an oscilloscope to see the effects.

My view is build for minimal noise wherever possible because it adds up. So lower-noise metal film resistors, snubbers on low-Qrr diodes, snubbers on transformers, etc.
I will try to read the material you provide.
I am looking for more tool atm (oscilloscope and signal generator and analyzer). Right now I just try to replace what is out of specs. It is a bit overwhelming but I am gettting the hang of it.
 
If you have a local ham radio club, someone there likely has a scope and would help you mark the shield end of those capacitors. For the dozen you'd need to check it would take five minutes.

It's mostly a mechanical issue—desolder, insert, solder—so the first time you do it has the learning curve.
 
You can ID the outer foil with an audio amp too. It will hum louder when you hold the cap and have the center lead on the outer foil.
 
I forgot to mention that the grounding scheme on the 87 is sub par, since they will use some of the rivets that hold down components as ground points. You should make a star ground or bus ground for best results. The eico was a kit and it seems that the designers did not want to make things overly complicated for a first time builder and so they used convenient grounding points
that went way over my head cause I lack the background knowledge on this. I will note this and come back cause this seem to need modification (moving wires around). I know what grounding mean, and it seems like bus ground (from what I glean from internet) was grounding them all to one point to reduce noise?
 
A buss is a fancy way of saying a long heavy wire that serves as the tie point for all the grounds. Its connected to the chassis in one spot. I used a hunk of 14awg scrap from some Romex scrap. Star grounds are where you use a central tie point to chassis with wires out to sections as needed. It looks somewhat like a spiderweb.

Common chassis grounding can work and be quiet but the chances of it being noisy are higher than a well done buss or star system. Loose rivets or corrosion between the tab and the chassis can cause you all sorts of gremlins though.
 
that went way over my head cause I lack the background knowledge on this. I will note this and come back cause this seem to need modification (moving wires around). I know what grounding mean, and it seems like bus ground (from what I glean from internet) was grounding them all to one point to reduce noise?
Since the HF87 is around 60 years old, there is most likely some corrosion developing between the metal contact held by rivets and screws. And if the fastener has gotten loose over time it can cause intermittent noises and other problems. At the least you can make all the power supply caps running off the same ground. If you are doing a complete rebuild it just make more sense to install a ground bus or star ground at this point when you are changing grounding resistors and caps to ground. Also redoing the ground on the input and output jacks to a single point or attach to the end of the ground bus. There are many ways to retrofit a ground scheme like having a bus bar that is isolated by standoffs and then grounded via insulated stranded wire to the star ground. Much written about audio grounding since you can have hum and other noises if it is not done correctly.
 
In my 87 i used the 3uf cap in parallel with the 100uf x 50vdc Elna Cerafine for the cathode bypass capacitor.
So basically I replace the 50uF with the Elna 100uF, then attach the 3uF cap in parallel with that????
A buss is a fancy way of saying a long heavy wire that serves as the tie point for all the grounds. Its connected to the chassis in one spot. I used a hunk of 14awg scrap from some Romex scrap. Star grounds are where you use a central tie point to chassis with wires out to sections as needed. It looks somewhat like a spiderweb.

Common chassis grounding can work and be quiet but the chances of it being noisy are higher than a well done buss or star system. Loose rivets or corrosion between the tab and the chassis can cause you all sorts of gremlins though.
For these grounding, does every point need to be ground to the same point (every resistor, cap, power supply etc...) If it just a matter of running extra wire I don't think it is too complicated for me. The problem is wire organization I guess.
 
The 87 is an excellent sounding amp. I believe that the 50uf cap is the cathode bypass cap. If true you should replace it with a hi quality 100uf x50vdc like Elna Cerafine. I usually also bypass the electrolytic with a film cap. I bought some 3uf x 50vdc PIO cap a long time ago and they work great. The reason to do this is because the EL34 tubes are cathode biased which means the cathode bypass caps are part of the signal chain.
If your coupling caps are measuring high or out of specs it usually means they are leaky and definitely need to be replaced. I like to use a PIO, paper in oil , like .1 x 600vdc vitaminQ , which sounds very nice in this amp.
Also the voltage doubler caps should be increased in value to around 500uf. It is very hard to get can caps to replace the original units so you will probably need to mound the new ones on clamps or you can epoxy glue them to the chassis, which works well and looks cleaner.
Sorry for so many question. I saw on Mouser the Elna 100uF 50VDC, but they are Silmic II. Would that work? or where should I get them? Also for the PIO cap, where is a good source cause they are all over the place with so many different options (I believe these are vintage nos capacitor?)
 
Back
Top Bottom