I am searching for a 47uF 25v audio grade electrolytic capacitor. Mouser have 100v rated devices at that 47uF value (Nichicon).
The cap is DC decoupling on the output of a 5534 op amp in an audio preamplifier circuit.
What downsides, if any, are there to using a 100v rated cap? I'm vaguely aware that voltage ratings should be kept as reasonably close to a circuit's specified rating as possible, but this is 4 times higher.
Any comments welcome!
The cap is DC decoupling on the output of a 5534 op amp in an audio preamplifier circuit.
What downsides, if any, are there to using a 100v rated cap? I'm vaguely aware that voltage ratings should be kept as reasonably close to a circuit's specified rating as possible, but this is 4 times higher.
Any comments welcome!
Mouser Australia seems to have them:
https://au.mouser.com/c/passive-com...citors-radial-leaded/?q=UKL1E470MED&instock=y
Hugo
https://au.mouser.com/c/passive-com...citors-radial-leaded/?q=UKL1E470MED&instock=y
Hugo
You probably mean that they will be used as coupling caps?
They should be able to withstand full rail voltages at power on preferably with a margin. Way higher voltage ratings have no detrimental effects except bizarre physical sizes, possible antenna working of too long lead wires and possibly PCB tracks coming loose. Often combined with amateur melt glue and extension wire situations. This is category "I just do something".
If you really need 47 µF then electrolytic caps are unavoidable. In normal cases smaller values can be used (use the formula instead of assumption model and always calculate worst case). Then Wima MKS 10 µF 50V in 5 mm pitch is to be preferred and they also do not need replacement in time. Small, way better properties, true bipolar, no replacement tick a few boxes.
Often electrolytic caps in oversized values are used because of cost reasons, uniform values in the device ("one size fits all") as that is cheaper and to calculate value loss because of aging in the process.
They should be able to withstand full rail voltages at power on preferably with a margin. Way higher voltage ratings have no detrimental effects except bizarre physical sizes, possible antenna working of too long lead wires and possibly PCB tracks coming loose. Often combined with amateur melt glue and extension wire situations. This is category "I just do something".
If you really need 47 µF then electrolytic caps are unavoidable. In normal cases smaller values can be used (use the formula instead of assumption model and always calculate worst case). Then Wima MKS 10 µF 50V in 5 mm pitch is to be preferred and they also do not need replacement in time. Small, way better properties, true bipolar, no replacement tick a few boxes.
Often electrolytic caps in oversized values are used because of cost reasons, uniform values in the device ("one size fits all") as that is cheaper and to calculate value loss because of aging in the process.
Last edited:
Hi, Hugo.
Thanks for sharing the link, however, the UKL series aren't listed by Nichicon as audio. The UFW and UKW series are suitable. In both cases, only 100v rated caps are manufactured by Nichicon in those audio series devices.
Mouser has 63v rated caps by Chemi-con, I may need to employ those.
@jean-paul Thanks for the suggestions, most appreciated. I took delivery of the 100v caps some months back, and they're not too big (8mm diameter). I don't know enough about design to determine why the author chose electros over polyester, bipolar, etc.
Here's the schematic and cap in question.
I'll go back to Douglas Self's audio design book to see if I can flesh out more info.
Cheers,
Tim
Here's the schematic and cap in question.
I'll go back to Douglas Self's audio design book to see if I can flesh out more info.
Cheers,
Tim
10 µF 50V film caps are perfectly possible. Certainly when R62 and R63 are changed to 47 kOhm. These are now chosen uncomfortably low (their function is reference to GND) and then they are the load with the useful load in parallel. As you can see uniformity in value is done. The decoupling caps are also 47 µF.
If you know the input impedance of the following stage you can calculate everything to be 100% accurate. Possibly even 4.7 µF is then enough. C30 and C31 can also be superior film caps in 5 mm. Power supply decoupling does not need film caps.
* Please look up the values of the resistors in the RC filters of supply lines. 180 Ohm is quite unusual and IMHO too high.
If you know the input impedance of the following stage you can calculate everything to be 100% accurate. Possibly even 4.7 µF is then enough. C30 and C31 can also be superior film caps in 5 mm. Power supply decoupling does not need film caps.
* Please look up the values of the resistors in the RC filters of supply lines. 180 Ohm is quite unusual and IMHO too high.
Last edited:
Any 47 µF 25 V electrolytic, made by a renowned manufacturer, will do. No need for »audio grade«, and, regarding R64, R65, no need even for low ESR.
Best regards!
Best regards!
I'd go with Nichicon UES here. Cheap, small, vanishingly low THD:
https://www.mouser.es/c/passive-com...njxO2yNQowEX5-rineYURaK8O8YfQQo9N4yTCy7UXzB0u
https://www.mouser.es/c/passive-com...njxO2yNQowEX5-rineYURaK8O8YfQQo9N4yTCy7UXzB0u
Replace NE5534 with OPA1655, remove C34/35 compensation caps, instead of coupling capacitors C30/31/36/37 put jumpers, remove R62/63 and instead of R64/65 (1k) put 22-47ohms. Check the square signal with an oscilloscope, if necessary add parallel to R58/59 compensation caps (10pF or more). It will be much better. Those modern OPAs have a small DC offset, no coupling caps are needed.
Last edited:
Very good suggestions and audio wise indeed better but no GND reference means heavy plops and DC coupling might not be welcome at the following stage (and the previous stage may have DC offset too). We have only seen a part of the device, a radical approach without knowing the total may not be beneficial to the total.
R62/R63. I would do the same as you did but only if I would know the complete schematic and things are well adapted to each other.
It will only lose the soft seductive sound of NE5532/34 in combination with coupled electrolytes. 😊
You mean its a coupling capacitor? It is expected to pass the full audio spectrum when loaded with 1k, so 47µF is plenty, and it'll handle a few mA at worst - easy. Audio grade is a meaningless concept in my opinion - different roles require different specs from a component - coupling caps in a low power signal path don't have to be exceptional at all, just be working and quiet.