I have a question about adding more capacitors to my MingDa tube (MC368-b90) amplifier. In the power supply, it uses two 330uf/450v caps in series to filter the output from the rectifiers. I'd like to parallel this rail of caps with a rail of two 470uf/450v caps in series to increase the capacitance of the power supply. Is this a safe mod? I understand that this increase in capacitance may result in more peak currents drawn from the rectifiers thus I may have to replace them with ones that have a higher current rating. Besides this, are there any potential problems that I need to watch out for?
Thanks in advance for any helps,
Frank
Thanks in advance for any helps,
Frank
What are you hoping to achieve by adding the capacitance?
You may run into problems of ringing in your power transformer. You may also generate more radio frequency interference which will be difficult to get rids of once created.
Are the rectifiers snubbered.
Overall this may prove to be a bad idea. There is a very good case to be made for minimising power supply capacitance. Some sort of choke (even a very small one) might be a better addition to the power supply.
Shoog
You may run into problems of ringing in your power transformer. You may also generate more radio frequency interference which will be difficult to get rids of once created.
Are the rectifiers snubbered.
Overall this may prove to be a bad idea. There is a very good case to be made for minimising power supply capacitance. Some sort of choke (even a very small one) might be a better addition to the power supply.
Shoog
Thanks for the quick response. I got some Rifa PEH200 470uf/450v caps intending to replace the two stock AeroM 330uf/450v as I think the former are better in terms of lower ESR. I heard that more capacitance in the power supply can help the bass response as it would provide more peak current, so I thought why not just parallel it. This is not a good idea after all?
What did you mean by "snubbered rectifiers"? The stock ones are IN5408x8, and I plan to replace them by the "ultra fast and soft recovery" ones.
Frank
What did you mean by "snubbered rectifiers"? The stock ones are IN5408x8, and I plan to replace them by the "ultra fast and soft recovery" ones.
Frank
I would echo Shoog's caution against using too much capacitance after the rectifiers. You have more to lose than gain, IMHO, for the reasons he states.
The usual snubber used with silicon rectifier diodes is a 0.01uF capacitor of 1,000v AC rating across each diode. It helps a great deal to suppress the spikes associated with most SS rectifiers, even fast-recovery diodes.
The usual snubber used with silicon rectifier diodes is a 0.01uF capacitor of 1,000v AC rating across each diode. It helps a great deal to suppress the spikes associated with most SS rectifiers, even fast-recovery diodes.
Lowering ESR can be beneficial if the caps are really poor, however lowering ESR beyond a certain point may bring penalties as well. It was found with Gainclones that a little bit of ESR helped to damp ringing in the power supply which was clearly audible.
Sometimes we second guess designers who know infinately more than we do as hobbyists.
Shoog
Shoog said:
Shoog said:
I agree. But I am still not sure if the amps in question fulfill such design thoroughness criteria and associated development phase measurements, or they just sport cap values that came handy economically and were in stock. I would just listen and compare if I was in the thread starter's shoes. I would just take out the original ones with care not to destroy them, and they can be fitted back.
I would totally agree with this statement. Try it and see.
The only thing to be aware of is that if it does introduce ringing, it is very likely on initial switch on things will sound 'better'. This is because the top end will sound more extended and lively, which will probably tighten up the bottom end. Only extended listening will reveal that the sound is more fatiguing and harder to listen to. This is what intermodulation distortion tends to sound like.
Just be careful and don't jump to conclusions.
Shoog
Just as I was about to change the stock caps by Rifa caps, I noticed that the Rifa caps that I got (Rifa PEH200 470uF/450VDC) have date code 0209, meaning they were made in September 2002 according to the information given by Rifa web page, which also says that these caps have a shelf life of four years. So the caps that I have are more than one year past their shelf life. Are they still safe to use? What kind of checks that I can do to make sure that they are still usable?
Thanks in advance,
Frank
Thanks in advance,
Frank
Old unused electrolytic caps usually are fine to use provided you test and "reform" them first. Chemical changes take place in the electrolyte when they sit dormant, and this process needs to be reversed, prior to placing them in service.
There are a number of web sites that have detailed explanations of the process and why it is necessary. It's actually quite easy to do, and in fact it's easy to build a simple reforming circuit out of "junk box" parts, a couple meters and a variac. I did this years ago and routinely do this to all electrolytics that I use, old or new, since it's also a testing process.
While your caps may seem old to you, I'd consider these fairly new, since I've reformed dozens of NOS caps with date codes as far back as the 1970s, and most of them come up just fine!
Google the subject, or here's a link to get you started:
http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm
As always, when working with high voltage, stay alert and work safely!
There are a number of web sites that have detailed explanations of the process and why it is necessary. It's actually quite easy to do, and in fact it's easy to build a simple reforming circuit out of "junk box" parts, a couple meters and a variac. I did this years ago and routinely do this to all electrolytics that I use, old or new, since it's also a testing process.
While your caps may seem old to you, I'd consider these fairly new, since I've reformed dozens of NOS caps with date codes as far back as the 1970s, and most of them come up just fine!
Google the subject, or here's a link to get you started:
http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm
As always, when working with high voltage, stay alert and work safely!
Frank,
I have some of those very same Rifa capacitors with the same date code. I got them from eBay out of Florida. I had to buy some M5x10 pan head screws for them. A four year shelf life is a riduclously conservitive short amount of time. Rifa makes a good capacitor. Go ahead and use them. Just bring the voltage up slow for the first time, then they'll be fine.
Victor
I have some of those very same Rifa capacitors with the same date code. I got them from eBay out of Florida. I had to buy some M5x10 pan head screws for them. A four year shelf life is a riduclously conservitive short amount of time. Rifa makes a good capacitor. Go ahead and use them. Just bring the voltage up slow for the first time, then they'll be fine.
Victor
Thanks for the information. I feel comfortable to use the caps now. While thinking about how best to fit them into my amp, I came upon an idea. Instead of putting two caps in series with a pair of dividing resistors as is normally done when V is over the rated voltage of C1:
----------- V
| |
C1 R1
| |
---------
| |
C2 R2
| |
--------- ground
how about just use one cap:
----------- V
| |
C1 R1
| |
---------
|
R2
|
--------- ground
This way, by making R1 bigger than R2, I can actually make the voltage across C1 and R1 closer to the rated voltage of C1. Would this work?
-Frank
----------- V
| |
C1 R1
| |
---------
| |
C2 R2
| |
--------- ground
how about just use one cap:
----------- V
| |
C1 R1
| |
---------
|
R2
|
--------- ground
This way, by making R1 bigger than R2, I can actually make the voltage across C1 and R1 closer to the rated voltage of C1. Would this work?
-Frank
From a strictly voltage point of view across the capacitor, yes it would control the DC. But otherwise it is not a good thing to do. This would make the power supply "spongy" and raise it's internal inpedance. Regulation and base would suffer greatly. I would only do this if I were stranded on a desert island with only one capacitor left. Capeesh?
Victor
Hi Victor: I was guessing something must be wrong with that scheme of mine as I did not see anything like that anywhere, but could not figure out why. You said it would make the power supply "spongy", why? I would've thought otherwise. With two caps in series, the overall capacitance is less. By getting rid of one, I thought I'd have increased the stored energy, thus making the supply more stable...
Frank
Frank
Originally asked by f1802llk:
The quick answer is that by adding a series resistor you current limit the supply.
If you think about how the capacitor works it becomes clear. Under no load conditions, the capacitor charges to full voltage and remains charged. But as the output of the supply is loaded down, the capacitor is tasked to supply current. It's job is to fill in the voids between the half sine pulses coming from the rectifier. This happens at a 120 cycle rate for full wave recitfication. (60 cycle for ½ wave)
By adding a resistor in series with it the peak current available is reduced because the time constant is increased. The time it takes for one charge/discharge cycle to happen. The higher the resistance, the longer the time constant becomes. This results as a reduced charge/discharge level between cycles, which translates to less power available to the load and a softer supply. Further, usful power will be wasted in that series resistor.
Generally one time constant period is required to charge a capacitor to 63.2% of Es. This can be found by: t (in seconds) = R (in ohms) x C (in farads) Five time constant periods are required for C to be considered full charged. The opposite is true for discharge. (Decreasing to 36.8% in one time period.)
Here's a simple analogy:
Think of the capacitor as if it were a two stage water pump. First filling then pumping out the water through a hose. By adding a resistor it's like reducing the hose diameter which will limit the water (current) flow over a given length of time.
At a quick glance this might seem true, but it isn't. The added resistance, and time constant, makes it not so. Actually, a larger capacitor also produces a longer time constant for a given charge level. For this reason a point of diminishing return is reached as the value increases beyond a certain capacitance when all things are considered. Although I do not think that point has been reached with the values discussed here.
To prove it to yourself, you might use simple ohm's law to select two resistors and try it. You will not damage anything in so doing, but I'll wager you won't like the resulting performance.
Hope this makes things a little clearer.
Victor
The quick answer is that by adding a series resistor you current limit the supply.
If you think about how the capacitor works it becomes clear. Under no load conditions, the capacitor charges to full voltage and remains charged. But as the output of the supply is loaded down, the capacitor is tasked to supply current. It's job is to fill in the voids between the half sine pulses coming from the rectifier. This happens at a 120 cycle rate for full wave recitfication. (60 cycle for ½ wave)
By adding a resistor in series with it the peak current available is reduced because the time constant is increased. The time it takes for one charge/discharge cycle to happen. The higher the resistance, the longer the time constant becomes. This results as a reduced charge/discharge level between cycles, which translates to less power available to the load and a softer supply. Further, usful power will be wasted in that series resistor.
Generally one time constant period is required to charge a capacitor to 63.2% of Es. This can be found by: t (in seconds) = R (in ohms) x C (in farads) Five time constant periods are required for C to be considered full charged. The opposite is true for discharge. (Decreasing to 36.8% in one time period.)
Here's a simple analogy:
Think of the capacitor as if it were a two stage water pump. First filling then pumping out the water through a hose. By adding a resistor it's like reducing the hose diameter which will limit the water (current) flow over a given length of time.
At a quick glance this might seem true, but it isn't. The added resistance, and time constant, makes it not so. Actually, a larger capacitor also produces a longer time constant for a given charge level. For this reason a point of diminishing return is reached as the value increases beyond a certain capacitance when all things are considered. Although I do not think that point has been reached with the values discussed here.
To prove it to yourself, you might use simple ohm's law to select two resistors and try it. You will not damage anything in so doing, but I'll wager you won't like the resulting performance.
Hope this makes things a little clearer.
Victor
Victor's explanation is great. The way I think of it is that when there is a sudden increase in the demand on current, it should be filled virtually instantaneously by the reservoir cap of the PS. That requires the series resistance of the cap to be as low as possible.
- Status
- Not open for further replies.