What if you prefer the tants over film or electrolytic?
Warm Audio WA-84
- Thread starter Emmathom
- Start date
Help Support GroupDIY Audio Forum:
Hi Dominic, I thought there are better caps for the signal path than tantalum... I would never think of tantalum in my mic tube pre's or in any small signal path > Muse BP for low volatge or PIO/Film for HV
For instance, J. Hiraga advises and uses Muse BP in his low signal path boards (pre RIAA, Dac, etc)
For instance, J. Hiraga advises and uses Muse BP in his low signal path boards (pre RIAA, Dac, etc)
@Emmathom
I know many members here express their disdain of Tantalum caps. Apparently prone to failure etc.
But at the same time, a lot of Pro gear has Tantalums in its audio path (or not) and it still sounds good you know.
AKG C414’s, D 224, Neumann U87’s, Neve modules and many more.
Not a capital sin to have a Tant in.
Try it if you really want to change it. If you have a matched pair of mics, record audio with both mics close together. Mod one, re-record in same position and A-B both audio files. Compare - try to be objective.
M
Attachments
Last edited:
..I'm pretty convinced that tants are important players when it comes to making a sontec clone sound like an actual sontec..
It's not a lot, but it's certainly there
/Jakob E.
It's not a lot, but it's certainly there
/Jakob E.
Admiral-dk
Well-known member
I will try to rate my choices here - but beware there are a number of exeptions - depending on application :
Audio :
1) Ceramic NP0/C0G - If the size of the case makes sense + can get them.
2) Electrolytes if I need big values and they are followed by something where leakage isn't critical - usually OpAmps.
3) Filmcapacitors where I connect the Anode of a Tube to the next stage (passive or active). Manually mounted.
As most of my Production stuff is guitar pedals - NO Filmcapacitors, as they usually degrade really Badly in a Reflow-Oven.
4) Tantalum - when I need a big value and the above isn't an Option - like leakage + they will see about a third of their rated DC Voltage over them.
I would NEVER use a Tantalum if it will see zero Vdc. - as it will distort badly and sound ******.
Power Supply :
1) Electrolytes.
2) Unless the size is small enough to consider Ceramic SMDs.
3) Tantalum - if it makes sense - this can be in relation to price / avaliable parts (defenitely a reason during Covid and for some time after).
Oh - and in repair situations - I NEVER change a good Capacitor ...!!!.... no matter the type. If the value and ESR is good and I don't have any leakage that give me problems. And just to be clear : I change a lot more Filmcapacitors than I change Elelctrolytes in old Tube gear - simply because they leak current into the next stage, and this can destroy the Tube that comes after .... Of cause old Oil- or Paper- capacitors are a lot worse (like what you find in old Teisco Amps).
So why do I recomend not changing the Tantalun in your Microphone - there isn't any room for one that will be an improvement. The Electrolytes you keep on suggesting might work as well as the original Tantalum in this application - but I have a hard time seeing them as an improvement. The Transformer might not care much about any DC Current through it in this application or it might react just as bad as those old Tradania wonder ones did.
In this regard I recomend listening to Jakob E. as he has much, much more experience than me, in regard to Line and Microphone Level Transformers - like how they sound and if they distort in a good or bad way
I also admit that sometimes it better if you make your own experiences - simply to learn - but keep an open mind, like did it truely make a difference or is it just because people tell you it will do so
Per
ps I didn't recomend 1µF 50V electrolytes in your application - it was an example from my previous work.
Audio :
1) Ceramic NP0/C0G - If the size of the case makes sense + can get them.
2) Electrolytes if I need big values and they are followed by something where leakage isn't critical - usually OpAmps.
3) Filmcapacitors where I connect the Anode of a Tube to the next stage (passive or active). Manually mounted.
As most of my Production stuff is guitar pedals - NO Filmcapacitors, as they usually degrade really Badly in a Reflow-Oven.
4) Tantalum - when I need a big value and the above isn't an Option - like leakage + they will see about a third of their rated DC Voltage over them.
I would NEVER use a Tantalum if it will see zero Vdc. - as it will distort badly and sound ******.
Power Supply :
1) Electrolytes.
2) Unless the size is small enough to consider Ceramic SMDs.
3) Tantalum - if it makes sense - this can be in relation to price / avaliable parts (defenitely a reason during Covid and for some time after).
Oh - and in repair situations - I NEVER change a good Capacitor ...!!!.... no matter the type. If the value and ESR is good and I don't have any leakage that give me problems. And just to be clear : I change a lot more Filmcapacitors than I change Elelctrolytes in old Tube gear - simply because they leak current into the next stage, and this can destroy the Tube that comes after .... Of cause old Oil- or Paper- capacitors are a lot worse (like what you find in old Teisco Amps).
So why do I recomend not changing the Tantalun in your Microphone - there isn't any room for one that will be an improvement. The Electrolytes you keep on suggesting might work as well as the original Tantalum in this application - but I have a hard time seeing them as an improvement. The Transformer might not care much about any DC Current through it in this application or it might react just as bad as those old Tradania wonder ones did.
In this regard I recomend listening to Jakob E. as he has much, much more experience than me, in regard to Line and Microphone Level Transformers - like how they sound and if they distort in a good or bad way
I also admit that sometimes it better if you make your own experiences - simply to learn - but keep an open mind, like did it truely make a difference or is it just because people tell you it will do so
Per
ps I didn't recomend 1µF 50V electrolytes in your application - it was an example from my previous work.
Last edited:
Ok ok Per... some rules are really gold ones like "I never change a good capacitor" and you're rightI will try to rate my choices here - but beware there are a number of exeptions - depending on application :
Audio :
1) Ceramic NP0/C0G - If the size of the case makes sense + can get them.
2) Electrolytes if I need big values and they are followed by something where leakage isn't critical - usually OpAmps.
3) Filmcapacitors where I connect the Anode of a Tube to the next stage (passive or active). Manually mounted.
As most of my Production stuff is guitar pedals - NO Filmcapacitors, as they usually degrade really Badly in a Reflow-Oven.
4) Tantalum - when I need a big value and the above isn't an Option - like leakage + they will see about a third of their rated DC Voltage over them.
I would NEVER use a Tantalum if it will see zero Vdc. - as it will distort badly and sound ******.
Power Supply :
1) Electrolytes.
2) Unless the size is small enough to consider Ceramic SMDs.
3) Tantalum - if it makes sense - this can be in relation to price / avaliable parts (defenitely a reason during Covid and for some time after).
Oh - and in repair situations - I NEVER change a good Capacitor ...!!!.... no matter the type. If the value and ESR is good and I don't have any leakage that give me problems. And just to be clear : I change a lot more Filmcapacitors than I change Elelctrolytes in old Tube gear - simply because they leak current into the next stage, and this can destroy the Tube that comes after .... Of cause old Oil- or Paper- capacitors are a lot worse (like what you find in old Teisco Amps).
So why do I recomend not changing the Tantalun in your Microphone - there isn't any room for one that will be an improvement. The Electrolytes you keep on suggesting might work as well as the original Tantalum in this application - but I have a hard time seeing them as an improvement. The Transformer might not care much about any DC Current through it in this application or it might react just as bad as those old Tradania wonder ones did.
In this regard I recomend listening to Jakob E. as he has much, much more experience than me, in regard to Line and Microphone Level Transformers - like how they sound and if they distort in a good or bad way
I also admit that sometimes it better if you make your own experiences - simply to learn - but keep an open mind, like did it truely make a difference or is it just because people tell you it will do so
Per
ps I didn't recomend 1µF 50V electrolytes in your application - it was an example from my previous work.
I'm gonna leave this mic as it is...
FYI, in my mic pre, in the signal path between tube 1 and 2 I use Wima MKP10 or PIO russian (values are 150nF & 470nF)
MKP10 are a good value for money and prices rise (to hell) if you tend to very good (excellent) film caps like Janzten, EVO Oil, etc (let's forget about Duelund, Audyn, Clarity Caps, and many more out of my purse)
A high-end cap review :
https://www.humblehomemadehifi.com/Cap.html
Last edited:
Hi gentlemen,
One of my Warm Audio WA84 pair works poorly or rather alternately... a few times it works well and delivers the same sound and level as the other, but more often it generates a big roaring sound, like an storm in the distance when connected to +48v... The circuitry is poorly assembled and so are the solders. I can't read the ref. on the fet because it's half engraved half written Warm Audio is definitely not a serious brand (WA84 pair is sold at 700€ !)
I think it could be the fet... or a capacitor but not the transformer (no other parts left then).
I found a paper saying the fet is BF245 and another one which says you can replace BF245 by MPF102 or J113 (both in my stock but no BF245 that's why I'm asking)
Regards
One of my Warm Audio WA84 pair works poorly or rather alternately... a few times it works well and delivers the same sound and level as the other, but more often it generates a big roaring sound, like an storm in the distance when connected to +48v... The circuitry is poorly assembled and so are the solders. I can't read the ref. on the fet because it's half engraved half written
Warm Audio is definitely not a serious brand (WA84 pair is sold at 700€ !)I think it could be the fet... or a capacitor but not the transformer (no other parts left then).
I found a paper saying the fet is BF245 and another one which says you can replace BF245 by MPF102 or J113 (both in my stock but no BF245 that's why I'm asking)
Regards
Attachments
Last edited:
RuudNL
Well-known member
Geoff Tanner (Neve) always warned not to replace tantalum capacitors with aluminium ones.
"Those tants aren't there without a reason", he used to say...
- Joined
- Mar 11, 2014
- Messages
- 260
If you have a unit from the 70's or 80's were all the electrolytics in the signal path are tantalums the reason is that they performed better than aluminium back then. That's not the case anymore BUT they still have som benefits so if you have a circuit with a mixture you should probably keep the tantalums. And of course, if you want the same sound as the original unit keep the tantalums.
Disco Volante
Well-known member
It's my impression that tantalums got their bad rep from being used as decoupling on power rails, where their tendency to fail short would be, well, bad...
I've also seen quite a few PSU electrolytics blow up in older gear though... Usually covers the entire machine in gunk and kills the rectofyer, if not the PT too..
I've also seen quite a few PSU electrolytics blow up in older gear though... Usually covers the entire machine in gunk and kills the rectofyer, if not the PT too..
Admiral-dk
Well-known member
Well you got one right +
As an Apprentice [1984], I got the job of figuring why a new product (5MB HDD, Controller + Interface to 4 computers) had Random Data Errors.
EMC problems was suspected and sure enough it was. First major Hardware improvement was to add about 50 pcs. 100µF 6V3 Tantal Capacitor @ the supply pins on IC's. This worked fine until we got an upgraded PCB Layout (rather needed too). Shortly after we ran out of those Tantalums and had to quickly source some more from a different supplier. Not many minutes after the first completed PCB from the Production was hooked up for a longer test run - they had many very LOUD Bangs and the whole PCB caught Fire (or so it appeared). They test a few more and the only difference was how long it took before the same thing happened - WHY ????
Turned out that the other Manufactor had marked the Negative Pin instead of the Positive Pin - and used a stylized Negative Sign, that depending on where in World you're from might be read as a Minus, a Divisor (Fraction Sign) .... But most likely will be read as nice + Sign .... wich they had done in Production
I know that you risk something similar if a Tantalum Capacitor has the correct Polarity, and is getting too close to Max Voltage too ...!
And missed one important one for Audio : If you use a Tantalum in the Signal path and there isn't a propper DC Voltage across it - you will get some very nasty Distortions (unfortunately not the Musically pleasant ones) as a former Boss of mine discovered during his Apprentice period - when the Company he worked in @ the time made a new Measuring Device and had a Hell of a time trying to find out why it measured AC values all over the place - but the correct.
After they change those capacitors to something else (can't remember the type) they suddenly had a perfect new Instrument ...!!!
Per
As an Apprentice [1984], I got the job of figuring why a new product (5MB HDD, Controller + Interface to 4 computers) had Random Data Errors.
EMC problems was suspected and sure enough it was. First major Hardware improvement was to add about 50 pcs. 100µF 6V3 Tantal Capacitor @ the supply pins on IC's. This worked fine until we got an upgraded PCB Layout (rather needed too). Shortly after we ran out of those Tantalums and had to quickly source some more from a different supplier. Not many minutes after the first completed PCB from the Production was hooked up for a longer test run - they had many very LOUD Bangs and the whole PCB caught Fire (or so it appeared). They test a few more and the only difference was how long it took before the same thing happened - WHY ????
Turned out that the other Manufactor had marked the Negative Pin instead of the Positive Pin - and used a stylized Negative Sign, that depending on where in World you're from might be read as a Minus, a Divisor (Fraction Sign) .... But most likely will be read as nice + Sign .... wich they had done in Production
I know that you risk something similar if a Tantalum Capacitor has the correct Polarity, and is getting too close to Max Voltage too ...!
And missed one important one for Audio : If you use a Tantalum in the Signal path and there isn't a propper DC Voltage across it - you will get some very nasty Distortions (unfortunately not the Musically pleasant ones) as a former Boss of mine discovered during his Apprentice period - when the Company he worked in @ the time made a new Measuring Device and had a Hell of a time trying to find out why it measured AC values all over the place - but the correct.
After they change those capacitors to something else (can't remember the type) they suddenly had a perfect new Instrument ...!!!
Per
I've shared this story before. Back in the 80s CBS granted me a CX (vinyl noise reduction) license so I could make a CX record playback decoder kit. In the course of studying the licensee package I saw the encoder design schematic and it used a tantalum capacitor in the side chain for the dominant encode/decode time constant. Tantalum caps are known to have high dielectric absorption (DA). In the specific application of a side chain being charged and discharged with different attack/decay impedances, DA could alter the attack/release envelope. To make my playback decoder as accurate as possible I used a tantalum capacitor in the CX kit decoder side chain too.
This was one of the very few times I have used a tantalum capacitor on purpose. Back in the day they were valued for low impedance at HF, and good density (high capacitance in small packages). They are also notorious fire starters when used for PS decoupling.
JR
k brown
Well-known member
. . . and I'll take the smell of a popped electro over the stink of a fried tant any day!
Hi
I disassembled my broken WA84. As I already said it suffers from noise & cracking. The mechanical aspect of this microphone is crap. The "tube" that receives the capsule connection peak is just soldered between the 2 PCBs and it is painful to unsolder / resolder. The PCB itself is not quality either and a kind of "mud" recovers the components and melts with the solder when you try to re-solder the components.
In short, I redid all the soldering (re-melt them) and removed the 1uF tantalum in series to measure it. It is good. I then changed the jfet (J112 installed despite the fact that it should be a BF245 according to some web review) for an MPF102. No improvement ...
One thing to note is that by dint of unsolder / resolder the PCBs they start to get damaged...
So I decided, next time working on the microphone again, to change all the components (except the OPT). On this occasion, and against all odds, I will replace the tantalums with chemical but quality ones (Panasonic FC or Elna Silmic II). According to the KM84 schematic, all 5 capacitors (except polystyren & Wima MKP 100nF) are 4,7uF and 50v is ok for coupling as for filtering. In another hand, Banzai (https://www.banzaimusic.com/Tantalum/) does sell tantalums (can we rely on their quality ?)
The question is: should the J112 be replaced by another reference? and if so, which one do you recommend? On the original schem. it is said 2N3891 or J113... or similar (?). I got a stock (~30) J113...
Regards
I disassembled my broken WA84. As I already said it suffers from noise & cracking. The mechanical aspect of this microphone is crap. The "tube" that receives the capsule connection peak is just soldered between the 2 PCBs and it is painful to unsolder / resolder. The PCB itself is not quality either and a kind of "mud" recovers the components and melts with the solder when you try to re-solder the components.
In short, I redid all the soldering (re-melt them) and removed the 1uF tantalum in series to measure it. It is good. I then changed the jfet (J112 installed despite the fact that it should be a BF245 according to some web review) for an MPF102. No improvement ...
One thing to note is that by dint of unsolder / resolder the PCBs they start to get damaged...
So I decided, next time working on the microphone again, to change all the components (except the OPT). On this occasion, and against all odds, I will replace the tantalums with chemical but quality ones (Panasonic FC or Elna Silmic II). According to the KM84 schematic, all 5 capacitors (except polystyren & Wima MKP 100nF) are 4,7uF and 50v is ok for coupling as for filtering. In another hand, Banzai (https://www.banzaimusic.com/Tantalum/) does sell tantalums (can we rely on their quality ?)
The question is: should the J112 be replaced by another reference? and if so, which one do you recommend? On the original schem. it is said 2N3891 or J113... or similar (?). I got a stock (~30) J113...
Regards
Attachments
I know your preference for silver solder, but the elevated temperatures required foar this type of solder might damage the PCB. I'd propose to use Sn63Pb37 solder instead, which has a much lower melting point. I have no reason to assume that this solder will affect sonic qualities of the microphone, though the audiophile community may think differently.
I wouldn't be afraid of using regular electrolytic capacitors. Tantalums are by no means a guarantee of good sound. In a blind test with several KM84 clone microphones, my KM84 clones with regular elcaps were better liked than the one with tantalum caps. Btw, a 1uF coupling cap (C4) is probably the best choice, just as in the original KM84 circuit. For an explanation why 1uF is the optimum, I'd like to refer to my website here.
J113, 2N3819 and BF245 should all work, though they may result in slightly different gains due to different gm and parasitic capacitance. Just give the J113 a try if you already have it.
That "mud" you are referring to could be conformal coating to protect the electronics against moisture. Nothing bad about that, except that it might be difficult to remove the coating.
I wouldn't be afraid of using regular electrolytic capacitors. Tantalums are by no means a guarantee of good sound. In a blind test with several KM84 clone microphones, my KM84 clones with regular elcaps were better liked than the one with tantalum caps. Btw, a 1uF coupling cap (C4) is probably the best choice, just as in the original KM84 circuit. For an explanation why 1uF is the optimum, I'd like to refer to my website here.
J113, 2N3819 and BF245 should all work, though they may result in slightly different gains due to different gm and parasitic capacitance. Just give the J113 a try if you already have it.
That "mud" you are referring to could be conformal coating to protect the electronics against moisture. Nothing bad about that, except that it might be difficult to remove the coating.
Just say that with some expérience and a temperature iron set to ~305° I just touch the part less than 1.5 second to melt it right. I like to solder at 300° to 350° but fast doing
This WA circuit is indeed crap (for 700€ a pair = a shame)
This WA circuit is indeed crap (for 700€ a pair = a shame)
Last edited:
Hmm, if that will already damage the PCB, it's indeed crap!
They were only 'better' cos less leakage and smaller size. Today, Lo-Leakage Al. electrolytics rival Tant leakage and size too.
They were also thought to be OK with 0V across them which we now know to be liquid BS.
I have some correspondence with Guru Scott Wurcer on the less well known sonic 'features' of various capacitors for LN use. Most (all?) polystyrenes are microphonic. Tants and the Golden Pinnae Micas introduce noise. MicBuilders members have confirmed the wooshy noise with Tants. used in the signal path at capacitor mike levels.
Bear in mind that some 'PSU decoupling caps' in certain circuits are actually in the signal path.
If you like these 'features' or aren't bothered about LN ... and don't mind replacing Tants regularly, by all means use them.
Last edited:
“The question is: should the J112 be replaced by another reference? “
As @jp8 said: “J113, 2N3819 and BF245 ...” can all work.
I recently replaced J112 for J111 in a build and also had great success with BF245 in KM84 & U87 builds in the past. Adjust Drain voltage (10V-10.7V to 11V depending)
to adjust Bias for least distortion.
That’s what R3** is for.
WATCH OUT WITH PIN OUT COMPATIBILITY THOUGH!!!
PS: “This WA circuit is indeed crap (for 700€ a pair = a shame)” 700€ ?!?
Real shame, yes!
Merry Christmas you all !
️
M
As @jp8 said: “J113, 2N3819 and BF245 ...” can all work.
I recently replaced J112 for J111 in a build and also had great success with BF245 in KM84 & U87 builds in the past. Adjust Drain voltage (10V-10.7V to 11V depending)
to adjust Bias for least distortion.
That’s what R3** is for.
WATCH OUT WITH PIN OUT COMPATIBILITY THOUGH!!!
PS: “This WA circuit is indeed crap (for 700€ a pair = a shame)” 700€ ?!?
Real shame, yes!
Merry Christmas you all !
️M
Cqwet Dbdfte
Well-known member
- Joined
- Oct 2, 2023
- Messages
- 662
That Zener is a noise source, a sharp knee zener (low Z) has more noise than the hi Z. Inadequate capacitive bypassing would feed that noise right into the signal. Check with a scope.Hi
I disassembled my broken WA84. As I already said it suffers from noise & cracking. The mechanical aspect of this microphone is crap. The "tube" that receives the capsule connection peak is just soldered between the 2 PCBs and it is painful to unsolder / resolder. The PCB itself is not quality either and a kind of "mud" recovers the components and melts with the solder when you try to re-solder the components.
In short, I redid all the soldering (re-melt them) and removed the 1uF tantalum in series to measure it. It is good. I then changed the jfet (J112 installed despite the fact that it should be a BF245 according to some web review) for an MPF102. No improvement ...
One thing to note is that by dint of unsolder / resolder the PCBs they start to get damaged...
So I decided, next time working on the microphone again, to change all the components (except the OPT). On this occasion, and against all odds, I will replace the tantalums with chemical but quality ones (Panasonic FC or Elna Silmic II). According to the KM84 schematic, all 5 capacitors (except polystyren & Wima MKP 100nF) are 4,7uF and 50v is ok for coupling as for filtering. In another hand, Banzai (https://www.banzaimusic.com/Tantalum/) does sell tantalums (can we rely on their quality ?)
The question is: should the J112 be replaced by another reference? and if so, which one do you recommend? On the original schem. it is said 2N3891 or J113... or similar (?). I got a stock (~30) J113...
Regards
Otherwise crackling noise may be from a bad solderjoint, failing dielectrics, failing capsule, surface tracking on PCB, and EMI/RFI. Check with cellphone next to mike and turn on / off "airplane mode" , which will send a few bursts of RF to sync up with local tower.
