Soundcraft 200B mixer recap question

[snk]

Hello,

- First, I hope this thread is not totally off-topic (it's about a mixer, not a stompbox), if so, please mods feel free to say so, move or delete it-

I have an old Soundcraft 200B mixer from 1980. It works fine, doesn't hum or buzz, and is not very noisy.
My concern is that I find the bass range a bit weird : while the treble can get quite harsh when using the eq, the bass range doesn't have as much presence as, for instance, a more modern mixer like an A&H GL2400.

As the mixer is rather old (40+ years), I thought that some electrolytic capacitors may be defective or off-value, leading to a worse performance.
I have searched the web, and found many topics talking about fully recapping or modding the 200B, but I wouldn't want to make a full recap "just for the sake of it" (there are 400+ electrolytics in the mixer, and good chances to make more harm than good when desoldering it). Also, most mods had the goal to reduce hum and buzz, and mine is fine as far as buzz is concerned.


So, my questions are the following :

1- Do you think that recapping some areas (maybe with increased caps values) would bring back some bass range, or is mainly a signal/noise ratio improvement (or fairy dust : "i modded something, so it may be better")? Or do you maybe think that the mixer is working like it should (like a venerable low-end mixer from the 80s)?

2- is there a way to check the capacitors value without removing them, in order to check which ones seem faulty, and which ones seem ok?

3- Which areas of the mixer/circuit should I try to recap first (i'm guessing the power supply, the coupling/decoupling capacitors, the power rail), without doing a full recap? Is recapping the power supply and the master section (without touching the tracks modules) prone to noticeable improvements, or is it a "all or nothing" scenario?

[snk]

Some info :

user manual with schematics :
https://www.soundcraft.com/zh/product_documents/soundcraft-200b-ug-zm2310-pdf

A picture of the mixer :


Some mods websites :
http://eli-audio.blogspot.com/2013/09/the-200b-saga-continues-part-2-master.html
http://eli-audio.blogspot.com/2013/08/soundcraft-200b-modifications-ones-that.html
http://eli-audio.blogspot.com/2013/09/the-200b-saga-continues-part-3-sweep-eq.html
http://eli-audio.blogspot.com/2013/09/the-200b-saga-continues-part-3-sweep-eq.html

[PRR]

Don't trust ears. Set up to measure frequency response. Repeatably.

That is not one box but dozens of sub-circuits, many of them with insert jacks. Measure response stage-by-stage an see if any stage looks goofy.

Power supply *may* be the least profitable mod because you have no buzz and all the chips have serious supply rejection. (Yes, when you do a 99-track mixdown, rail cross-talk may matter; for now just get one channel sweet.)

[snk]

Thank you for the advice, PRR.

Don't trust ears. Set up to measure frequency response. Repeatably.

Yes, I am going to proceed to some tests. I have used it the whole afternoon, and I enjoy it. But I find it less "detailled" and "punchy" than the A&H GL2400 which i'm also used to use.

That is not one box but dozens of sub-circuits, many of them with insert jacks. Measure response stage-by-stage an see if any stage looks goofy.

I noticed that :
- the AUX4 return is slightly distorted, and sometimes it produces some random noise even when nothing is played through.
- I have to be careful not to push the input gain, as the input stage is not very forgiving (it produces ugly clipping, not "warmth"). But I think it's the expected behaviour so i can live with that.

Power supply *may* be the least profitable mod because you have no buzz and all the chips have serious supply rejection.

Ok (Too bad : it's the easiest servicing to do  ).
I have read a lot that these Soundcraft 200B could be noisy, but here it isn't the case 

[snk]

Ok, I made some tests using pink noise (and a sine osc as well, later).
The results aren't that bad at all...

The frequency range is quite linear on all channels, as well as the aux returns (and the sends). There is a slight bass cut from 50Hz, and some channels are a bit louder than others (but I have more gain than needed, so it's easy to compensate for).
Sending a sine oscillator to the channels shows some harmonics, but no noticeable distortion.


The overall noisefloor is ok. Not mindblowing by today standards, but quite fine for a 40 years old mixer : with all channels muted and aux returns set to minimum, i get a -80dB noisefloor.
With gain at noon and faders at unity gain on every channel as well as on aux, the noisefloor is around -60dB (with peaks at 50, 100, and 150Hz : fundamental and harmonics from electric signal).


I guess I'd be happy if I could reduce the noisefloor a bit in the bass range, but as the console is in good working order, I won't go for a full recap.

[Rob Strand]

The response looks OK but there is a slight downward tilt on the low-end.     On the schematic there seems to be a high-pass filter in the signal chain.  I haven't analysed the frequencies or looked at the circuit close enough to see if there is any low-cut remaining when the filters are switched out.  You could check your switch settings in this area.

Perhaps the Bass and Treble EQ frequencies aren't to your liking?   From what I can see the later models add a frequency control to the Bass and Treble EQ.    It's a simple matter to change the bass and treble caps to tweak the tone control frequencies.

[snk]

Thank you, Rob.

I think you're right : I need to spend more time with it. I may be too used to my GL2400 (I love the eq on this console), making me find the Soundcraft 200B harsher.
I have also found some advices on forums about finding the sweet spot in the gain structure, as the headroom might not be as forgiving as i expected 

The 200B does not have a bunch of head-room, so watch the gain structure like a hawk.
If you find that you have to run the busses or the main outs faders lower than -5 to get the level under control, the board is going to start compressing like crazy and not in a good way.

I spent some time yesterday mixing a 14 tracks mix with the 4 aux running effects, and with a careful gain stage, it sounds ok. Well, it sounds even better if i gently bump the bass around 80Hz and roll off the highs a bit afterwards, but it's usable.

I have a channel with a slightly sticky fader, so I might take it as an enticement to remove the module, clean the fader and see what else I can do.

From what i have read, it seems that replacing a ceramic cap by a Wima can improve the high shelving eq...

The one thing which seems to be a consensus is that the High Frequency shelving EQ can get a little harsh and it is possible to change a capacitor to improve this.

Also, it seems that increasing the values of the coupling and decoupling caps can bring back some low end...

- There are (3x) 47uF/25v coupling caps (C15, C25, C29) which I will replace with 220uF/35v Nichicon PW series.
- The power rails have (4x) decoupling caps (C16, C17, C30, C31), 47uF/25v which I will replace with 470uF/25v Panasonic FM series.(ou 220µF/25V)

If it's working fine, do you think that changing the PSU capacitors (for newer caps with bigger values) would improve the noisefloor a bit, or would it be useless?

[ElectricDruid]

Also, it seems that increasing the values of the coupling and decoupling caps can bring back some low end...

- There are (3x) 47uF/25v coupling caps (C15, C25, C29) which I will replace with 220uF/35v Nichicon PW series.

I'd be very surprised if 47uF wasn't large enough to get sufficient bass response. Ok, I'm used to effects circuits, where the resistances tend to be higher than in mixers (where you tend to use lower values to lower noise and not worry about power consumption so much), but 47uF is *huge*, even with a low resistance.
We'd need to see a schematic to be sure, since without it, I'm just guessing what they're doing, but 47uF is a big value.

- The power rails have (4x) decoupling caps (C16, C17, C30, C31), 47uF/25v which I will replace with 470uF/25v Panasonic FM series.(ou 220µF/25V)

If it's working fine, do you think that changing the PSU capacitors (for newer caps with bigger values) would improve the noisefloor a bit, or would it be useless?

In general I'm against random re-capping, but this I might actually try. Because it's only a few caps, so it's not a huge job. And because decent power smoothing is pretty much always better (although you might have to think about how the rest of the circuit reacts to having the rails come up a lot slower than previously. Hence I might not multiply the factory values by ten without further research!) And because 40 years is a long time for a cap that's been filtering mains hum all its life.

That said, I replaced the power smoothing caps in my Pro-One synth (early eighties, so similar era). Made no audible difference whatsoever. Still, at least now I know they're not about to die on me. For a synth full of expensive CEM chips, that's worth quite a bit.

[snk]

Hi, Electric Druid.
Thank you for your reply.

That said, I replaced the power smoothing caps in my Pro-One synth (early eighties, so similar era).

The Pro-One is a wonderful synth indeed. Mine is getting a lot of use 
I just had to service the keyboard and change a logic chip once, a couple years ago.

I'd be very surprised if 47uF wasn't large enough to get sufficient bass response. Ok, I'm used to effect circuits, where the resistances tend to be higher than in mixers (where you tend to use lower values to lower noise and not worry about power consumption so much), but 47uF is *huge*, even with a low resistance.

Yes, I was quite surprised by the value increase suggestions I found on Gearspace and elsewhere (by a factor of 10 most of the time  ).
But well, I don't have enough knowledge to know if increasing the original value by 10 would be that much better than increasing it by 2 (and this is why i'm asking here, because i don't want to follow some random advice found once on internet)...

Well, to be fair, I spent quite some time searching on the net. Some discussions about changing every opamp, every electrolytic cap, several resistors and several ceramic caps made my jaw drop : it was like trying to change that mixer into something it's not, and sometimes, some forum posters seemed a bit fooled by their ears (finding a night and day difference between two opamps one day, and changing their mind the other day). But on Gearspace I also found several posts by Jim Williams, who seems an experienced and trustful man, and might be a fine guideline. I can provide links to the topics if needed.

In general I'm against random re-capping, but this I might actually try. Because it's only a few caps, so it's not a huge job. And because decent power smoothing is pretty much always better (although you might have to think about how the rest of the circuit reacts to having the rails come up a lot slower than previously. Hence I might not multiply the factory values by ten without further research!) And because 40 years is a long time for a cap that's been filtering mains hum all its life.

Yes, this is what I have in mind, too. There are only a couple capacitors, the circuit is easy to reach, and, like you wrote, 40 years is quite a lot... I don't really know what to expect from it, but I think it can't be harmful?
I have seen advices to double the capacitors values in the power supply, which seems quite wise from what I can tell?

We'd need to see a schematic to be sure, since without it, I'm just guessing what they're doing, but 47uF is a big value.

The schematics should be in the last part of the user manual posted above (page 52 and onwards) : https://www.soundcraft.com/zh/product_documents/soundcraft-200b-ug-zm2310-pdf

[Rob Strand]

From what i have read, it seems that replacing a ceramic cap by a Wima can improve the high shelving eq...

For ceramic caps less than say 1nF they are often NPO types or have very liittle impact on the audio range in the first  place.   Ceramics in these position usually have no effect on the sound.

When you get to ceramic cap values >= 1nF they often aren't NPO.  If the cap passes audio signal then it's often a good idea to upgrade those to polymer film caps.    No need to change supply decoupling caps or other non-audio parts though.   Throughout the circuit I can see a few 2.2nF and 3.3nF caps which might be ceramics.

Also, it seems that increasing the values of the coupling and decoupling caps can bring back some low end...

- There are (3x) 47uF/25v coupling caps (C15, C25, C29) which I will replace with 220uF/35v Nichicon PW series.

I'd be suspicious the coupling caps are bad more than the value is too low.   Bad electrolytic coupling caps can do some weird things.  You could measure the capacitance value with a cap meter and check the cap's ESR with an ESR meter.   If they look off then the that would confirm the caps are bad.

In general I'm not in favour of blindly replacing electrolytic caps.   It's a fix that's based on hope more than evidence.  It's best to confirm you have a problem in that area first.   If you do find bad caps then that would be cause to suspect other caps and perhaps justify replacing a number of them.    Eventually electrolytic caps go bad it's just a matter of when.

If it's working fine, do you think that changing the PSU capacitors (for newer caps with bigger values) would improve the noisefloor a bit, or would it be useless?

It could go either way.   As a test I sometimes just put a large cap across the existing ones to see if it changes anything.  If I can hear or measure a change then that's a reason to suspect something is wrong.

[ElectricDruid]

As a test I sometimes just put a large cap across the existing ones to see if it changes anything.  If I can hear or measure a change then that's a reason to suspect something is wrong.

Yeah, this is a good idea. It's easy to tack-solder a big cap in place to see if it helps.

[blackieNYC]

I can't quite make out the markings on the EQ knobs, but if you put in a tone that is the center of the bass frequency band, and you turn up that band, you should get the gain that the knob may indicate. If the center freq is 100Hz, and the gain knob is labeled "+15dB" at its maximum, see if you indeed get 15dB of gain with a 100hz tone.  If you get less than 15dB of gain, like 8dB, the series caps in the signal path need to be replaced.  At least in the EQ, but probably the rest of the signal path needs it too.  Because the caps are all the same age and probably manufacturer's series.   

[snk]

Thank you, Blackie, it's a very simple way to consider things and I didn't even think about it 
I just checked, and I think I found a possible cause why I found the bass a bit "hollow" and the high sometimes "harsh" : it may be because the shelving slope is very smooth (and goes up to 500Hz -thus explaining "un-focused bass sound"), and also because the high shelf boosts more than the bass shelf (thus explaining "harsh" treble when boosting).
With the treble boost/cut, I reach the claimed "+/-15dB".
With the bass boost/cut, I get in the "+/-10~12dB" ball park.



Do you think that it means the capacitors need to be replaced?

I certainly can live with +/-10dB (it's already a lot!), it's mostly a matter of "knowing my tools" (and I didn't expect to see so gentle slopes, going so high in the medium range).

[snk]

It's interesting to compare the analysis between the Soundcraft 200B and the A&H GL2400 :

- The 200B gives a +8~10dB boost, starting at 1KhZ, with a plateau shelving around 100Hz.
- The GL2400 produces a +12dB boost, starting at 300Hz, with a plateau shelving around 50Hz.

Given these results, the 200B doesn't seem so much off-specs (i mean, the boost is supposed to be +15dB, but the more recent GL2400 doesn't give me a +15dB boost either. The Soundcraft is "just" 2-3dB below the GL2400).
Also, we can see the slight low cut on the 200B, but it's around 20Hz, and at 50Hz it's still very close to unity gain (matching the pink noise at input).



I think that what I didn't like at first was the bass boost soft slope, going up to 1khz (and the treble boost being more pronounced than the bass boost), but now that I know my tools, I can mix with it and enjoy the way it works (as long as it works as it's supposed to be, which seems to be the case).

I can change a couple capacitors on the channel modules, and every caps in the power supply if it can improve the noisefloor or make the console more reliable on the long run, but I know I won't go the hard way and replace hundreds of capacitors blindly just for the sake of it : i'd rather use it to make music 

[Rob Strand]

- The 200B gives a +8~10dB boost, starting at 1KhZ, with a plateau shelving around 100Hz.
- The GL2400 produces a +12dB boost, starting at 300Hz, with a plateau shelving around 50Hz.

For apples to apples comparison you will find with the GL2400 set to +8dB to +10dB the 50Hz will move up and the 300Hz will move down.

Scanning over the schematics both units use the same Bass pot value but the Bass caps on the 200B are 3 or 5 times smaller than the GL2400, depending if you compare mono or stereo input.    To me the 200B bass frequencies are quite high and overly warm whereas the GL2400 is more "normal".   

FWIW, the treble frequencies on the 200B are lower than the GL2400.  That would explain why the 200B treble control seems harsh.

If you want to bend the 200B closer to the GL2400 you would need to increase the bass caps and decrease the treble caps.

I can change a couple capacitors on the channel modules, and every caps in the power supply if it can improve the noisefloor or make the console more reliable on the long run, but I know I won't go the hard way and replace hundreds of capacitors blindly just for the sake of it : i'd rather use it to make music

You can replace the PSU caps with values double the size to fight hum a bit.   Increasing the cap voltage ratings will help increase the life.   However, you need to find (good) caps which match the hole pitch and the available height.

You might find the noise comes from a design limitation.  For example large resistor values adding noise.  You can also get hum from the PCB layout and the way the connectors bundle the grounds and connect to the chassis.  Even the proximity of mains wiring and the transformer to the PCB can make an impact.    You might be able to make small improvements but you can also get into a situation where you have to severely hack-up the unit to make improvements.   If that's the goal then there's nothing wrong with doing that, even though you could argue it's making it into something it is not.   It can also be a lot of work.   I guess this is where other people's work on the web can narrow things down a bit to reduce to amount of redesign.

[Rob Strand]

Yeah, this is a good idea. It's easy to tack-solder a big cap in place to see if it helps.

Sometimes I just poke short lead caps from my junk box on the back of the board by hand.   You have to be a bit careful since you will expect some consequences of the cap charging up while the unit is on.   Also you don't want to charge the cap up to +15V and slap it across another point which won't handle it - you might want to discharge it in between.    I suppose the problem with "quick" checks like this is it needs lot more thought about what you are doing.

[snk]

Thank you, Rob!
First, I will spend some time mixing with the 200B to see how I like playing with the eq.
Then, I'll order some caps for the PSU.
Then, when I will remove one channel module to clean the stuck fader, I'll check the channel board, and consider changing a couple caps if needed