ESR Quest

[trem]

I have been looking at ESR Meters for a few years.
Any of you guys want to venture an opinion on this one.?
Thank You
AnaTek Corporation - Blue ESR Assembled

[Tom Phillips]

Check out the past discussions here at MEF too.
There are lots of threads discussing ESR meters listed at https://www.google.com/search?as_q=E...ts=&gws_rd=ssl

[Alan0354]

No opinion on the meter. But if people worry about high ESR of the big caps, it is a standard practice to parallel smaller value caps to lower the ESR. Larger value caps inherently tend to have higher ESR loss. So above certain freq, the cap become inductive after resonance. If you parallel a smaller value cap, the smaller cap still below the resonance frequency and hold the impedance low.

I always wonder why guitar amp doesn't do that. But lately I have been on DIYaudio, people in audiophile definitely doing this. So if you have a 40uF cap, put a 1uF or a 2.2uF in parallel. Then if you are anal about it, put a 0.1uF non polar cap. That should hold the impedance low to MHz range. Now you can get any cheap caps.

JMHO

[trem]

So THAT is the wisdom behind those metal-film (or whatever) cap i see across the reservoir caps....seems like a lot of Voxy builds use that.?
Anyway.....thanks for the info and suggestions.
I will do a search.

[Jazz P Bass]

High value capacitors take care of the big stuff.

Smaller values take care of the little stuff.

[Tom Phillips]

And you still want to use decent high value caps. Paralleling a small film cap can fix certain problems. However, that approach is not a cure all. It won't make up completely for a high ESR electrolytic.

[trem]

And you still want to use decent high value caps. Paralleling a small film cap can fix certain problems. However, that approach is not a cure all. It won't make up completely for a high ESR electrolytic.

.........Between Nichicon, Xicon, IC, and sometimes JJ, F&T, and some "generic" caps sold by CE.....I have never had a problem (that I know of) with a new Elytic Cap.
Like with paralleling a 0.10uF cap across a 50uF cap...what is that fixing exactly.?
How does ESR manifest itself in the Sound/Tone of a guitar amp.?
If you remove the smaller value cap from the Power Supply, what are you likely to hear.?
Thank You

[Tom Phillips]

..If you remove the smaller value cap from the Power Supply, what are you likely to hear.?...

In an otherwise well built guitar amp I say that it is highly unlikely that you will hear any difference at all.

[Alan0354]

This is an article from Illinois Cap. http://www.illinoiscapacitor.com/pdf...factor_ESR.pdf. The graph in page two seems to indicate that ESR is the lowest impedance the cap can be and it is at the resonance point.

At frequency below the resonance point, the impedance of the cap is governed by Xc=1/(2 pi f C). This means as long as you can match a smaller cap with lower ESR that has higher resonance frequency than the resonance of of big cap, the impedance will continue to drop below the ESR of the big cap. Then use a even smaller cap to take care of the resonance point of the middle cap.

[nsubulysses]

I have that meter and have found it useful. It allows you to spot high ESR cathode bypass caps quickly and easily. I am glad to have it.

But, FWIW, a scenario that often comes up is a cap is borderline high ESR, then you replace it and the amp performs just the same. Not sure if this means you are needlessly replacing capacitors, or if it is maybe good because over time the cap will go higher ESR and you are pre-empting it (george bush-style). Or as Tom Phillips said perhaps a slightly high ESR cap is not a big deal in an otherwise good guitar amp. For a modern amp you probably just replace the inexpensive cap and move on, but it could lead to some head scratching when you have a vintage unit that seems to work fine but the caps test poorly.

One possibly troublying aspect of the meter is the graph only goes to 250V, so I think you are unable to truly test power supply caps. You can kinda guesstimate it, but I always wondered if there was a better meter that tested higher voltage caps, and why the Blue ESR meter cannot test high voltage caps.

[Alan0354]

Also, it's a question whether ESR is really that important in tube amps. Tube amps have high impedance and low current circuits. In the preamp section, the +B usually going through RC network of say 2.2K and a 20uF cap as filter. If you calculate a 20uF at 1KHz, the impedance is 8ohm. This is still way higher than ESR, and it's still dominate the impedance. Maybe that's the reason you don't see people dying to search for low ESR cap or parallel caps. I can understand for SS amp, ESR becomes important as the circuit is very low impedance, people use 33000uF caps then ESR becomes dominant. Maybe that's the reason the old caps works just great even the ESR is bad.

[trem]

nsubulysses -
You raise a very good question. I believe there is a meter called The Capacitor Wizard that can test caps in circuit, and at High Voltage. I THINK it is about 250-300 bux.

But maybe Alan0354 (and others) raise the most relevant point.....in a Tube Guitar amp, how important is one of these meters really.?
You can parallel a filter cap with another one of like value, and see if your troubles go away...or just replace all the inexpensive by-pass caps, etc etc.
I guess that is why I have never bought one...guys with Much More knowledge/experience than myself usually put a pretty low priority on an ESR Meter.
I Dunno.........

[Chuck H]

JM2C on ESR...

I'll start with Alan's observation that paralleling a smaller value film cap across a larger value power filter isn't done on guitar amps. Well, I saw Dan Torres do it in 1989. Say what you will about Dan, but he has always tried to innovate and he's had at least one good idea for every five that get publicly bashed. How many of us can say that?

Re: ESR... This is just my own speculation based on observation with very little concrete test results to support my positions. Take it with a pinch of salt

There is plenty of reason to suspect that even operational ESR parameters have an influence on tone and performance. One reason is that the power supply for the preamp doesn't so much use the capacitors for additional filtering as decoupling. Because of the high gain and high current nature of guitar amps (VERY high gain for some) any signal riding on the rail has the potential to affect the tone of the amp for better or worse. The higher the ESR the greater the effect. If your power supply is set up correctly (as is the case with most vintage amps) the decoupling caps prevent anti phase anomalies. A high ESR cap (as was typical in the day OR as they were when the amps came to be revered) will allow for more negative feedback in the signal chain. This can have a positive or negative affect depending on the specific model and/or design. The Amp Garage has had a couple of discussions on this issue WRT getting the tone of the TrainWreck designs right (as it seems all clones are considered inferior ). Still, some validity to the issue seems to have been realized. On the other end of it there is the possibility of positive feedback in a poorly designed power supply. This is the case with many contemporary Fender products. It doesn't help that many of the early models used lousy caps. The HR series amps, for example, are well known to suffer instability due to decoupling failure. This is directly related to the ESR of the decoupling capacitors. I would think that these amps in particular could benefit from paralleling film caps across the preamp filters.

Main power filters probably won't benefit a lot from the parallel caps since their main job is to filter and act as reservoirs. The frequencies they target in their duty is well below the knee of any smaller film cap. There may be voltage spike and diode switching issues that might be lessened though. I haven't tested for this.

Dai once experimented with small value resistors in series with the power supply filters and his observation was that there was an audible change in tone with different resistors. I don't recall that there was a general good or bad consensus though. It was probably fourteen years ago.

So, depending on the amps tonal or stability requirements the ESR can have a noteworthy affect on the tone of an amp. That's my own speculation on the matter.

Sorry, I can't comment on ESR meters because I've never owned one

[Alan0354]

We can talk theory until the face turn blue, but I know by now guitar amp is a kind of it's own!!! Things that's not supposed to matter DO matter in guitar amps. So I guess people just have to test and see.

As for me, my motivation in this discussion is to find ways to be CHEAP as I born cheap!!! I have to be convinced before I am going to spend more money to buy FT caps. So far, I use all cheap 600V caps from Weber and I am happy camper. Sound is so subjective, I am happy with my amp with all cheap parts including the filter caps.

I did experience oscillation before. It was feedback on the +B from later stages back to the first stage of the high gain channel. But it's not because of ESR, I don't have enough caps. I had 10uF. So when I change to bigger cap, it's cure. Now I have 3 separate RC filter for 4 gain stages, not a hint of oscillation. Only squeal I had was from the OT that I had a thread on that here. So I stay with my cheap Weber caps and no intention in buying an ESR meter.

BTW, on the subject of caps, Don't use tantalum caps for signal or cathode bypass. They have 1/f noise. I had random popping sound before and I chased down to the tantalum cap I use to bypass the cathode on the first stage. I switch to electrolytic, it still give me a slight swooshing sound!! I ended up parallel a 1uF ceramic cap to get the best result.

[R.G.]

It's important for understanding to separate the ESR from the ESL (Equivalent Series Inductance).

ESR is, well, resistance. Although the conceptually ideal cap inside every real cap has an impedance that goes down to infinitesimally small with increasing frequency, ESR makes it stop going down and level out at the ESR value, so the impedance of the cap no longer decreases with increasing frequency.

ESL is worse. At frequencies above the point where the inductive impedance is bigger than the capacitive impedance, the overal impedance of the cap starts going UP. Above that frequency, the cap acts like an inductor. The physically bigger the cap, the lower this frequency.

The standard approach to 'fixing' this comes from RF practice, where caps with smaller capacitance values but also smaller ESR and ESL are paralleled with bigger caps so they shunt the increasing ESL of the bigger caps and still have a low, and decreasing impedance as frequency rises.

Audiophiles have adopted this as part of the never ending turd polishing that goes on in the audio freako culture. The grain of truth here is that for really big electros, you ... might ... get some advantage down in the audio range by one set of paralleling if you are good at measuring and designing an appropriate shunt cap value. But there is frankly little advantage in having paralleled caps good up to 100MHZ in a guitar amp, and precious little in other audio stuff. Maybe the only real advantage is where the caps are good enough to prevent self-oscillation of linear ICs. But that is not a consideration to someone who thinks that using teflon circuit boards is necessary and thinks they can hear the chemical composition of their speaker wires.

But it's ESL that makes the paralleling of smaller values useful.

ESR can be helped - where this is really needed, which isn't all that often, and mostly in switching power supplies - by paralleling many nearly-equal capacitors to get the same capacitance but lower ESL. This does lower ESL, too, by the same ratio. This approach is much more for curing the time domain response, not the frequency domain response.

No opinion on the meter. But if people worry about high ESR of the big caps, it is a standard practice to parallel smaller value caps to lower the ESR. Larger value caps inherently tend to have higher ESR loss. So above certain freq, the cap become inductive after resonance. If you parallel a smaller value cap, the smaller cap still below the resonance frequency and hold the impedance low.

I always wonder why guitar amp doesn't do that. But lately I have been on DIYaudio, people in audiophile definitely doing this. So if you have a 40uF cap, put a 1uF or a 2.2uF in parallel. Then if you are anal about it, put a 0.1uF non polar cap. That should hold the impedance low to MHz range. Now you can get any cheap caps.