Ok, folks!
I'll go for the 5532 (Sakis' and Wahad's suggestion) opamp using Molly's suggestion of a 10uF cap seried with a 3.9 ohms 1/4W resistor across pins 4 and 8 in order to avoid oscillations??
Any other tip?
The OPA2604 will be fine in typical "Baxandall" tone control. So will the 5532 and many others too.
Any wide bandwidth device has the potential for stability issues but I don't think this application is one of them
This is what can happen, and yes this is an OPA604 (the single version of the 2604) but don't let that put you off here. I think its a good choice for your amp.
http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
Fit a socket and try them both. Another common but excellent sounding device (which many will disagree with) is the TL072.
Any wide bandwidth device has the potential for stability issues but I don't think this application is one of them
This is what can happen, and yes this is an OPA604 (the single version of the 2604) but don't let that put you off here. I think its a good choice for your amp.http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
Fit a socket and try them both. Another common but excellent sounding device (which many will disagree with) is the TL072.
Yes, positive end to pin 8 and negative to pin 4. It doesn't matter which order the R and C connect when in series.
Many, many, MANY THANKS!! Your linked thread, Molly, will go to my library/bookmarks right now!
Since these devices don't cost that much, I can buy several different ones to test myself on a good 8-pin socket:
TL072 - 4MHz 16V/us (€0.25)
NE5532 - 10MHz 9V/us (€0.50)
OPA2604 - 20MHz 25V/us (€4.10)
LM4562NA - 55MHz 20V/us (€2.20)
JRC4558DA - 3MHz 1.7V/us (the one to be replaced)
And if know a few more recommendations, please tell me. I'm testing them anyway and can use at anytime in the future!
Last edited:
Hi all,
I was investigating over the Internet and saw about the practice of putting capacitors across rectifier diodes to reduce the RF Noise generated by the rectifiers diodes themselves.
Picture below is a modd from an Onkyo receiver from the Internet.
Couldn't it be a further action to reduce from the beginning any kind of noise going into the circuit?
And since the transformer in the Kenwood KA-405 has +-31VAC from its secondary to the board and +-43VDC out the original GP25D rectifiers, could I replace these with faster/soft recovery ones (fuses are 6.3A) with a ceramic 47nF in parallel for each diode?? Any idea for such faster/soft recovery rectifiers?
Picture below is from our Kenwood KA-405's original rectifiers:
I see there are already two 10nF ceramic disc caps. So they may be already doing this job, but I don't know if at least the rectifiers could be replaced with faster ones with 33nF or 47nF in parallel. This would prevent also any noise coming from trafo as well?
Inicially I was thinking the Schottky SB560 could be used here, but they are rated 60V 5A (fuses are 6.3A)
I know it's a bunch of required info coming at one time, but when it's understood all the work for reducing RF noise and prevent oscillations, a lot of problem are avoid from the beginning.
Thanks again!
I was investigating over the Internet and saw about the practice of putting capacitors across rectifier diodes to reduce the RF Noise generated by the rectifiers diodes themselves.
Picture below is a modd from an Onkyo receiver from the Internet.
An externally hosted image should be here but it was not working when we last tested it.
Couldn't it be a further action to reduce from the beginning any kind of noise going into the circuit?
And since the transformer in the Kenwood KA-405 has +-31VAC from its secondary to the board and +-43VDC out the original GP25D rectifiers, could I replace these with faster/soft recovery ones (fuses are 6.3A) with a ceramic 47nF in parallel for each diode?? Any idea for such faster/soft recovery rectifiers?
Picture below is from our Kenwood KA-405's original rectifiers:
An externally hosted image should be here but it was not working when we last tested it.
I see there are already two 10nF ceramic disc caps. So they may be already doing this job, but I don't know if at least the rectifiers could be replaced with faster ones with 33nF or 47nF in parallel. This would prevent also any noise coming from trafo as well?
Inicially I was thinking the Schottky SB560 could be used here, but they are rated 60V 5A (fuses are 6.3A)
I know it's a bunch of required info coming at one time, but when it's understood all the work for reducing RF noise and prevent oscillations, a lot of problem are avoid from the beginning.
Thanks again!
Last edited:
12Pf won't do much (the picture)
Commutation noise from rectifiers is very real (it actually causes a burst of RF as the diode comes out of conduction interacting with the inductance of the transformer).
You can add caps across the rectifiers (around 0.01uf and of suitable voltage) or just as effective, one cap in series with a resistor (say 0.1uf and 1 ohm) directly across the secondary.
The voltage rating for any of these has to equal the transformer peak to peak voltage plus a good safety margin. A 40 volt AC winding would need 160 volt rated caps.
Commutation noise from rectifiers is very real (it actually causes a burst of RF as the diode comes out of conduction interacting with the inductance of the transformer).
You can add caps across the rectifiers (around 0.01uf and of suitable voltage) or just as effective, one cap in series with a resistor (say 0.1uf and 1 ohm) directly across the secondary.
The voltage rating for any of these has to equal the transformer peak to peak voltage plus a good safety margin. A 40 volt AC winding would need 160 volt rated caps.
12Pf won't do much (the picture)
Commutation noise from rectifiers is very real (it actually causes a burst of RF as the diode comes out of conduction interacting with the inductance of the transformer).
You can add caps across the rectifiers (around 0.01uf and of suitable voltage) or just as effective, one cap in series with a resistor (say 0.1uf and 1 ohm) directly across the secondary.
The voltage rating for any of these has to equal the transformer peak to peak voltage plus a good safety margin. A 40 volt AC winding would need 160 volt rated caps.
Correct all the way through !!
For years now, it has been popuar "audiophile" practice to use Ultra Fast Recovery diode rectifiers rather than standard types. More recently, there have been nay-sayers telling us that these diodes add more noise than they prevent. Evidence has been posted on the forum too. The guy posting recently showed that standard diodes with 100nF film caps did a better job of silencing the audible spikes of rectifier switching than even soft recovery diodes.
Those ceramic caps were once required by the US FCC to snub hash switching noise that leaked back into the mains power line and were not necessary to reduce spikes on the amplifier power rails, so you see a variety of values there, depending on the compliance necessary for the model. Now they are used as part of many optimistic ways to "upgrade" amplifiers with more exotic and expensive parts. My understanding is that soft recovery diodes do reduce power supply noise in the audio range but actually increase the energy of RFI. Capacitors can control both, on the mains and the rails, without using ultra fast diodes and at lower cost.
Frankly, I don't know what the example in your pics using 12pF caps is for, perhaps UHF equipment?
So if your amplifier has standard diodes and suitable value capacitors fitted, I would say that will give you as good or better results than many fast diode tweaks you see parroted on the net.
Yeah!
Actually, for me it could also be a bigger effort, but very worthy at the end. As I have the unit already open for refurbishing, I rather do everything in one time and close it for the next 20 years. And the joy of learning all this! Yes, I prefer the first option (adding the 100nF to each rectifier).
But 100nF won't hurt, right? I see people saying to use from 22nF up to 47nF!
So, you'd recomment only adding the 100nF caps across the rectifiers and keeping these, right?
Thanks, guys! Been learning a lot about preventing both RF noise and oscillations here!
---------------------------
P.S.: And tomorrow, Monday 13th, I'll go buying all the components.
So the amplifier must be running on Tuesday if everything goes easy!
Last edited:
100nf (0.1uf) is OK although I would still consider the single 0.1uf cap and resistor across the (or each) secondary as perhaps the best solution. Reason for saying that is because a cap on its own can (in theory) provoke any transients to cause "ringing" when they interact with the inductance of the transformer.
Like anything, the fine detail can be sweated ad-infinitum and in practice any of the above mentioned tactics will work just fine. Personally though... I would go for the single cap and resistor.
Like anything, the fine detail can be sweated ad-infinitum and in practice any of the above mentioned tactics will work just fine. Personally though... I would go for the single cap and resistor.
It always hurts me to see someone 'optimizing' a nice, old-school unit like this to death.
IMHO, it is ok and usually required to replace electrolytics due to aging. Of course one can also try to replace opamps with more recent, pin-compatible substitutes. Technology has moved on there over the years.
But why throw out a phono board or a headphone jack to get improvements on a scale that mostly goes unnoticed due to the limitations of the remaining device chain.
If you want everything tweaked and up-to-date, build or buy a recent amplifier that has all the bells and whistles.
If you like the 'classic' unit you have for some reason, try to modify it as non-intrusive as possible (new caps, maybe better connectors that fit into the original places, better potentiometers, ...)
Rundmaus
IMHO, it is ok and usually required to replace electrolytics due to aging. Of course one can also try to replace opamps with more recent, pin-compatible substitutes. Technology has moved on there over the years.
But why throw out a phono board or a headphone jack to get improvements on a scale that mostly goes unnoticed due to the limitations of the remaining device chain.
If you want everything tweaked and up-to-date, build or buy a recent amplifier that has all the bells and whistles.
If you like the 'classic' unit you have for some reason, try to modify it as non-intrusive as possible (new caps, maybe better connectors that fit into the original places, better potentiometers, ...)
Rundmaus
Ok, the transformer has + - + outs (2 reds and 1 common black), so it means that, for each red wire, I can add a seried 100nF and 1 ohm resistor across them.
The 100nF might be something like 200V (or higher) rated. And for the 1 ohm resistor something like 1W (or higher)??
Here below a detail from the 1980 Kenwood brochure showing the KA-405:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
You are absolutelly right!
But I'm not "optimizing this unit to death".
I just don't use turntables, so the Line In (under the Opamp) are more important to me here.
And all this work on preventing RF noises and oscillations refers exclusively to the new Opamp that will take place over the older 4558.
Last edited:
Actually, FET opamps are more immune to RF hash than bipolar types like the 4558.
For the transformer, it sounds like the black lead is the centre tap of the windings and this will go to the amplifier ground. So the suppression is just one cap and resistor in series across the two red leads. Thats all you need. A 1 watt carbon or metal film and 200 volt cap is perfect.
For the transformer, it sounds like the black lead is the centre tap of the windings and this will go to the amplifier ground. So the suppression is just one cap and resistor in series across the two red leads. Thats all you need. A 1 watt carbon or metal film and 200 volt cap is perfect.
Thank you, Molly! Very precious info.
And Rundmaus and folks, I'm not destroying any nice Kenwood old-school amp here as I described above. It's just recapping it and replacing the Opamp without affecting the original sound. After all, the amp needs a deep work as MANY solder joints are broken, cold and dry. And all the pots are dirty and yelding "cracks" and "pops" noise sound. This is a labor of love actually.
Ja! Bedankt!Sorry, didn't want to offend you. Sounds better already
But when refurbishing the whole unit, why not paying a few hours and some components and redo the phono section too? Leading to a nice, classic amp in fully working condition...
Good luck with the work!
Rundmaus
Actually I was going to redo the whole Phono input as well putting fresh new electrolytics and metal film 1% resistors. But then the Opamp replacement, as being more difficult to work on, took over it. By now!
Later, I'll do the Phono input, too. It's a much easier job!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.