Recapping = massive improvement & a great fun!

[Svart]

Hey BR,

your post kinda got overlooked.. did you say that the console was recapped 8 years ago? If so I doubt that it would need recapping if quality caps were used! It may need to be ON for a period of time before being actually used to let the caps reform themselves though..

 

[Svart]

fucanay,

Signal path is easy. Its a cap that is actually IN the signal path. audio comes from a source and into one lead on the cap, out the other lead and into another part.

You can do some not so fun math equations and figure out the optimal size for the cap, or you can just up the value to a next higher value, going lower might form a "HPF" and going up might "fix" any HPF like areas. This is what they were talking about.

My console used everything from 22uf to 220uf in the audio path.. so i just jellybeaned it with 220uf for everything. much cheaper to buy a huge bag of the same thing than a handful of a few different values and I had plenty of room for the larger caps so I won on both sides.

:thumb:

 

[ulysses]

This is kind of a wide subject, but there seems to be a number of people asking pretty much the same question, so let me toss a couple of ideas onto the pile.

I would say it's really not a great idea to recap a console without first having a look at which caps are doing what. This is especially true if you're planning on changing capacitor values.

When people talk about recapping a piece of gear, they're talking about replacing the aluminum electrolytic capacitors. There's generally no need to replace the ceramic, film, or mica caps because they don't dry up and lose capacitance with age like electrolytics do. Tantalum electrolytics generally only fail if they're shorted or reverse-biased (which they do very quickly and spectacularly) so you generally don't need to replace those either.

Invest in a good desoldering iron. I hope the OP is charging enough for recapping this console to pay for a desoldering station if he doesn't already have one. You can find something decent used on Ebay for a few hundred bucks. Then practice desoldering on some old piece of junk before you touch the good gear.

When recapping or otherwise modifying a console, it's really important that you recap the power supply first, and then the monitor section. Then proceed to the master section, and do the channel strips last. This will allow you to hear the changes you're making. Don't copy a mod across the whole desk until you've looked at how it performs and listened to how it sounds in one spot.

Electrolytics are typically used in three primary applications: Power supply filtering and decoupling, signal coupling, and filter tuning.

All of the power supply filter and decoupling capacitors will be connected between a power supply voltage point and ground. These are not really "in the signal path" (except in the case of single-ended Class A amps like Neve 1073). They can generally be cheaper caps, but it's still a good idea to use 105 degree caps (or even 125 degree caps) because they'll last longer than 85 degree caps. Use better caps the closer you are to the actual audio circuit (the raw filtering in the main power supply rectifier circuit generally are very large and not top-grade parts). Of course it's up to you how much money you want to spend. Power supply decoupling caps (the ones that are right near audio circuits, usually with a small resistor between them and the main power rail) can usually be increased in capacitance by a factor of 2 to 20. The only limit is physical size, cost, and the initial inrush current surge at power up (if you put too much capacitance across the power supply, it'll blow the fuse when you first turn the power on). Other than that, just make sure you observe polarity and voltage rating. I tend to use panasonic caps, including the various TS series (TSHA), NHG, FC, and the new FM stuff. I will probably shift more toward the FM caps because they're lead-free, they're the lowest-impedance caps you can find, and Digikey stocks them.

Signal coupling caps are directly in the signal path, used to feed audio signal from one circuit section to another while blocking any DC offset that may be superimposed on the signal. This is where you want to use the best capacitors. It's better to use good design to eliminate the need for electrolytic coupling caps, either by eliminating DC offsets or by contolling circuit impedances at key coupling nodes so that smaller (film) capacitors may be used. But when you're recapping somebody else's design, you don't generally have that option. Replacing old caps with new ones can greatly improve frequency response, both at the bottom and top of the audio range. Increasing the capacitance can improve low-frequency response and phase performance, but beware that sometimes the low-frequency cutoff is being controlled for a reason. And just about any decent electrolytic cap available today is going to have much lower ESR and much lower inductance than the old cap ever had, which can open up the top end a lot. Of course, this might reveal ugly things you didn't want to know about in the HF response of the circuit. Sometimes the crummy old caps sound better. But in general, you can try making coupling caps 2-10x larger and see if you like it. New caps are more compact than old caps, so this is pretty easy to do and sometimes you can't avoid it because they just don't make some of those really small values anymore.

You can calculate the corner frequency of a hi-pass or low-pass filter if you know the value of the capacitor and the resistance connected to it:
f=1/(2*pi*r*c) where f is frequency of the -3dB point, pi is 3.1415926, r is the circuit resistance in ohms, and c is capacitance in farads (1uF=0.000001F). Determining the resistance seen by the capacitor can be tricky, but the general idea is that you imagine removing the capacitor and replacing it with an ohm meter that measures with an AC signal (don't actually do that, just imagine it). Remember that power supply rails are considered ground points as far as audio circuits are concerned. So are the inverting inputs of op amps. This is really a bigger topic than can be covered here, but keep staring at schematics until it starts to make sense. A ballpark corner frequency for a coupling capacitor should be around 1Hz, give or take. Remember that if you have six of these 1Hz hi-pass filters scattered throughout a channel, your net result is a hi-pass filter at 32Hz. Imagine where you would be if you set the corner at 20Hz. Same goes for low-pass filters. You might think you only need bandwidth out to 20kHz, but remember what happens when you stack a bunch of circuit stages, either within one device or when you connect several pieces of gear in series. This is why it's very typical to see a HF corner frequency up around a couple of hundred kHz or more. You can start to see why the tiny bit of inductance in a capacitor can be a sonic issue, especially in a console that includes dozens of amplifier stages in series. This is not complicated math, it's simple algebra and you're welcome to use a calculator. You wouldn't expect to do this kind of work without knowing how to solder, and the math is every bit as integral a part of electronic tomfoolery as soldering.

Filter tuning capacitors, such as in equalizers and hi-pass filters, have to be a specific value or the EQ frequency will change. Leave those values alone. Except of course, if the original is 25uF then you'll probably have to go with 22uF because 25 is no longer a standard value. Luckily, EQs generally use film caps for frequency-critical areas. Most electrolytics have such a wide tolerance they wouldn't be very good in an EQ anyway.

One final word - we keep talking about how some caps can "generally" be increased in value by a pretty large margin. This applies to typical audio gear, which is built to a price point, and was built at a time when large caps were bulky and expensive. You can bet that in many cases, the designer would have liked to use a larger, more expensive cap to begin with. Especially if they had access to the much higher-performance caps available today. But remember that this might not apply to every manufacturer. Some use large, high-quality caps to begin with. For example, I manufacture a compressor that just recently hit the market. It doesn't have any electrolytic caps in the audio circuit, but the power supply caps are ridiculously oversized and very good caps to begin with. so in 15 years or so when somebody goes to recap one of my compressors, I hope they don't blindly follow my advice and increase all the values by a factor of 20. Make sense?

 

[BR]

Thanks svart.
I'll have the console on for at least a day before trying anything funny.
take care
Gil

 

[Viitalahde]

Yep - Justin has many good points there. :thumb:

Definately no use to change poly/mica/ceramic caps, it's only the electrolythics and failed tantalums.

In the case of this desk, 47uF is used pretty much everywhere, except for some rail filtering. The voltage rating is important - originals are 25v so I keep them 25v (though AC coupling in a bi-polar powered desk COULD take smaller caps too).

Caps are smaller (physically) today. It's very clear to see - those FC's I buy come at 120uF smallest at 25v rating. They also have a 2,5mm pitching, which kinda sucks, but you'd need to go to enormous sizes/voltage ratings to get a 5mm pitching. This seems to apply to any high quality electrolythic I see in the Farnell cataloque.

Oh, the monitoring section is going to get a nice re-cap. There already are some differences between left and right channel, but even through that mess, the one-channel recap difference came out loud and clear. :thumb:

 

[t-wurst]

Hey Jaakko!
I own the smaller D&R 4000 series. Im sure its never been racapped although its from 1980 as far as i remember. the surface is very dusty cause it stands at least one year unconnected and uncovered in the cellar. So i allready have to take it apart and have a closer look to all potis and faders and i have to remove the frontplates to clean them. Would you recommend to recap it, starting with the main section? After reading your post it seems to make a huge sonic difference. And maybe you can give me a hint wich ones are to change and share some tips´n´tricks u discover during your "recapping fun". I dont think there´s much difference between the 8000 and 4000 series. :grin:

best regards, toby

 

[Viitalahde]

Yep, I'll tell you about my experiences. I would probably start with the master section if I were you, in this case we started from channel 2 because it was clearly problematic.

As what to change.. I'd say anything. For example, out of the ~45 caps in the channel, about 40 are directly in the signal path and while you have it open, it hurts no-one to change the rest either (psu filtering and such) while you're there. :thumb:

Just ordered a huge bunch of Panasonic FC's and NHG's. :shock:

 

[jensenmann]

@Stephan: welcome to the Lab!
Usually you can choose the next bigger value without problems. Try to get caps with at least 25% higher voltage rating than the supplyvoltage, in your case 35V is the deal. I did some recapping with Panasonic FCs and I´m really satisfied. Very HiFi-ish result. Those ones you can get at www.rsonline.de . If you want to compare different brands get yourself some caps (105Degree types) from www.reichelt.de, the cheapest place in germany, and compare them to the FCs. The above mentioned Nichicons are hard to source in Germany. Another highly regarded brand are Evox Rifa Caps from rsonline, though i never tested them.
Don´t hesitate to contact me via pm and later when you´ve done your tests report us your listening expieriences here.
:sam:

@crazylama and snatchman
Steffen uses two M3500

:sam: :sam: :sam:

 

[Gus]

something to keep in mind. Older >15years ago electros were often much higher than the marked value, A older 22uf could measure 33 or more uf when new. New electros are often just under the rated value a NEW 10uf will often measure 9.9uf in spec if rated 10%.

 

[Snatchman]

[quote author="jensenmann"]@Stephan: welcome to the Lab!
Usually you can choose the next bigger value without problems. Try to get caps with at least 25% higher voltage rating than the supplyvoltage, in your case 35V is the deal. I did some recapping with Panasonic FCs and I´m really satisfied. Very HiFi-ish result. Those ones you can get at www.rsonline.de . If you want to compare different brands get yourself some caps (105Degree types) from www.reichelt.de, the cheapest place in germany, and compare them to the FCs. The above mentioned Nichicons are hard to source in Germany. Another highly regarded brand are Evox Rifa Caps from rsonline, though i never tested them.
Don´t hesitate to contact me via pm and later when you´ve done your tests report us your listening expieriences here.
:sam:

@crazylama and snatchman
Steffen uses two M3500

:sam: :sam: :sam:[/quote]Thanks for the reply jensenmann! I think I read somewhere when I first bought the M-2600, Tascam stated that the M-2600 was the replacement" for 3500 series! I could be wrong tho. Thanks

 

[ulysses]

[quote author="Viitalahde"]In the case of this desk, 47uF is used pretty much everywhere, except for some rail filtering. The voltage rating is important - originals are 25v so I keep them 25v (though AC coupling in a bi-polar powered desk COULD take smaller caps too).

Caps are smaller (physically) today. It's very clear to see - those FC's I buy come at 120uF smallest at 25v rating. They also have a 2,5mm pitching, which kinda sucks, but you'd need to go to enormous sizes/voltage ratings to get a 5mm pitching. This seems to apply to any high quality electrolythic I see in the Farnell cataloque.[/quote]

In most cases I would choose a cap with the correct pin spacing, even if it means going to a significantly higher capacitance value. It's very good to keep your caps seated firmly against the circuitboard so they can't vibrate with signal, and that's tough to do when you're stretching pins to reach widely-spaced holes.

Actually it looks like you can get 5mm Panasonic caps in the 180 to 330uF range, depending on whether you choose 25 or 35V, FC or FM. That's not what I would call "enormous" by any means. Increasing the value of signal coupling and supply decoupling caps by a factor of 10 or 20 is pretty common when recapping a typical board from the 1980s. In the Digikey catalog, the 300uF 25V Panasonic FM is $109 for 1000 pieces, cheaper than the equivalent FC. The FMs are better caps with significantly lower impedance. I imagine they'll eventually replace the FC series entirely.

By the way, I don't know where you're shopping, but I see 25V FC values as low as 10uF in the Digikey catalog.

 

[leigh]

[quote author="Svart"]Signal path is easy. Its a cap that is actually IN the signal path. audio comes from a source and into one lead on the cap, out the other lead and into another part.[/quote]

Could you clarify: would a cap that shunts an audio source to ground be part of the "signal path"? It might be kind of a semantic argument, because that cap wouldn't be in the "path" between the in and the out of the circuit. However, it certainly would affect the sound...

Leigh

 

[fucanay]

[quote author="leigh"][quote author="Svart"]Signal path is easy. Its a cap that is actually IN the signal path. audio comes from a source and into one lead on the cap, out the other lead and into another part.[/quote]

Could you clarify: would a cap that shunts an audio source to ground be part of the "signal path"? It might be kind of a semantic argument, because that cap wouldn't be in the "path" between the in and the out of the circuit. However, it certainly would affect the sound...

Leigh[/quote]

Yeah, I don't quite understand either, but that's due to not understanding schematics all that much. I've only been at this for a few months now, so I'm still trying to grasp a lot of things at once.

It would be cool to see a schematic tutorial based on either a preamp or compressor circuit that has a little of everything in it. I don't plan to make much more than those type projects, so it would be really helpful to me as a newbie. It could be something like a series of schematic pictures with a section drawn in red and then an explanation of what that part of the circuit is doing. I dunno, I guess I just learn better that way.

On a different note, I've just reinstalled 8 input channels and the buss/master channel in an old Yamaha MC1604 console. All electrolytics were replaced and on 4 channels I replaced the JRC4558 chips with MC33078 chips. I haven't yet compared the two different chips, but I did a preliminary test on the channels with the new chips. Now, I may be being dramatic here, but I would say the way the channels were before was like listening through a wall and now I've just came into the room. It's crazy how different and great it now sounds. Very exciting for a newbie like me to have real success on my first outing.

And of course, it's all thanks to the fine folks here. I wish I had a beer for all of you but these will have to do. :guinness: :guinness: :sam: :sam:

Matt

 

[Viitalahde]

[quote author="ulysses"]It's very good to keep your caps seated firmly against the circuitboard so they can't vibrate with signal, and that's tough to do when you're stretching pins to reach widely-spaced holes.[/quote]

..Or I could do a little L-shaped angle for the other leg.

By the way, I don't know where you're shopping, but I see 25V FC values as low as 10uF in the Digikey catalog.

I do my shopping at Farnell. Damn, I always thought they had a good selection. :?

 

[ulysses]

[quote author="Viitalahde"]..Or I could do a little L-shaped angle for the other leg.[/quote]

You still end up with the cap sitting on top of the lead, even if the lead is resting nicely against the circuitboard. When you have caps with the correct lead spacing, the leads comeout of the bottom of the cap and go directly into their holes, allowing the bottom of the cap to rest firmly against the board (if you're careful while soldering them). This is the sort of thing I'm more anal about than I probably need to be, but attention to detail is what separates the men from the boys when it comes to squeezing maximum audio performance from the realities of physics and economics.

 

[Viitalahde]

[quote author="ulysses"]You still end up with the cap sitting on top of the lead, even if the lead is resting nicely against the circuitboard.[/quote]

Well, that's true. Never thought it that way.

This is the sort of thing I'm more anal about than I probably need to be

We all have our own anal factors. :grin: I like trimming bipolar supplies to, say, +/- 15.0000000000000001 volts.

 

[peeceebee]

[quote author="leigh"][quote author="Svart"]Signal path is easy. Its a cap that is actually IN the signal path. audio comes from a source and into one lead on the cap, out the other lead and into another part.[/quote]

Could you clarify: would a cap that shunts an audio source to ground be part of the "signal path"? It might be kind of a semantic argument, because that cap wouldn't be in the "path" between the in and the out of the circuit. However, it certainly would affect the sound...

Leigh[/quote]

EVERY component in a circuit effects the sound...
BUT some more obviously than others-

 

[ulysses]

[quote author="Svart"]Could you clarify: would a cap that shunts an audio source to ground be part of the "signal path"? It might be kind of a semantic argument, because that cap wouldn't be in the "path" between the in and the out of the circuit. However, it certainly would affect the sound...[/quote]
It depends on the circuit. In a single-ended class A amplifier, such as a Neve BA283 as found in the 1073 and 1272 modules, there is no inherent power supply rejection and the power supply itself can be considered to be in series with the audio path. This means that the decoupling caps at each amp are extremely important. Other circuits, such as an API style or your basic IC op amp have lots of power supply rejection, but decoupling caps near the amp can still have a big impact on the circuit performance in terms of bass response, headroom, and op amp stability.
Regarding your specific question about caps that short audio to ground - That doesn't happen very often with large electrolytic caps. You'll see it with emitter bypass caps (or cathode bypass in tube circuits) which increases gain by removing the degenerative feedback signal from the transistor (or tube). But any time you put a large cap from a signal-carrying node to ground, it stops being a signal-carrying node. But the quality of the capacitance can be significant to the audio performance, since the audio circuit is relying on there NOT being signal in that spot, in order to maintain the correct DC bias.

You're more often going to see tiny ceramic capacitors from signal to ground, which would be low-pass filters to remove unwanted RF noise and such.

 

[DerEber]

Dear forum Members,
Thak you all!!!!
I just had my first recapping experiences and indeed....... Great fun!

First I changed all the caps of one of my V272 to cheap BC-caps.
This was not that exiting at the first moment. I plugged in and compared the recapped to one original pice and I was a little dissappointed about the results. The chage in sound was very very obvious, but as the original pices soundes old and dull, they still sounded like a pice of good gear.
(who want´s all this highs?).
The BC-cap one sounded in one way more direct with more highs, but.....
I don´t know..... it reminded me a bit on "Sony HiFi" type of sound. But definately not very "interesting"compared to the original.
Letting the device "on" for some hours, it changed to good!!!
But then I did another one with RIFA-caps jensemann recommended to give a try.
Think it is always a good sign if you just cannot stop talking in that microphone
Ok what to say..... the definately kick out the BC´s!!!!
Dear forum Members,
Thak you all!!!!
I just had my first recapping experiences and indeed....... Great fun!

First I changed all the caps of one of my V272 to cheap BC-caps.
This was not that exiting at the first moments. I plugged in and compared the recapped to one original pice and I was a little dissappointed about the results. The chage in sound was very very obvious, but as the original pices soundes old and dull, they still sounded like a pice of good gear.
(who want´s all this highs?).
The BC-cap one sounded in one way more direct with more highs, but.....
I don´t know..... it reminded me a bit on "Sony HiFi" type of sound. But definately not very "interesting"compared to the original.
Letting the device "on" for some hours, it changed to good!!!
But then I did another one with RIFA-caps jensemann recommended to give a try. (thanks!!!!)
Think it is always a good sign if you just cannot stop talking in that microphone
Ok what to say..... the definately kick out the BC´s!!!!
As I had very good experience with the old one, using it for "overbright" Keyboards I will perserve 2 originals.
But still two 272 left
Lookig forward to try other caps too!!!!!!!


greetings,
Stephan

 

[Esset]

read this thread and gave me some ideas.

I've got an old Alesis X2 mixer, buildt around 1995 never recapped. Examined the inside and have sourced the parts I need. Only got one question.

Nichicon PW seems to be recommended when recapping, but the Rubycon YXF seems to be the recommendation when building say the GreenPre. Would there be any difference in sound? If so in what way? If anybody know this without having tried it? Or if somebody have tried it, which did you prefer?