The diagram appears in application notes on page 4. Shows the Ra + Rb and Ra + Rc combinations.
Page 3 shows Typical Application Circuit that omits the Rb or Rc [ NOT commonly found included on manufactured boards or design circuit schematics] The subject 330uF is on both page 3 & 4.
Perhaps Rb and Rc are usually omitted due to 2 channel use more common.... and to do with + and - detection referenced to OV 'ground' when a single channel input is used ?
I can't understand why the different option presented and most apt use of the 2 options.
'Lost in translation" ?
Page 3 shows Typical Application Circuit that omits the Rb or Rc [ NOT commonly found included on manufactured boards or design circuit schematics] The subject 330uF is on both page 3 & 4.
Perhaps Rb and Rc are usually omitted due to 2 channel use more common.... and to do with + and - detection referenced to OV 'ground' when a single channel input is used ?
I can't understand why the different option presented and most apt use of the 2 options.
'Lost in translation" ?
Last edited:
This diagram shows R5 and R11 [R11 being an unequal value to R5 makes it a Rc on the NEC diagram]
Note Cap is down to 220uF and polarised electro. Seems inappropriate to use polarised there.
Another diagram uses same resistors [referred to as Ra, Rb, Rc, not to be confused with designation on NEC diagram] Cap down to 22uF.
Note Cap is down to 220uF and polarised electro. Seems inappropriate to use polarised there.
Another diagram uses same resistors [referred to as Ra, Rb, Rc, not to be confused with designation on NEC diagram] Cap down to 22uF.
Last edited:
And another on this site [post#14] with both resistors in circuit and 2x inverse series polarised caps.
the easy question to answer is the polarity or not of the filter capacitor.
The input from the speaker is AC and during fault condition can have prolonged +ve or prolonged -ve coming in through the resistors to audio ground.
This fault condition will charge the capacitor to full fault voltage which could be as high as the supply rail.
On this basis the capacitor needs to be Bi-Polar of a rated voltage that exceeds the maximum supply rail voltage.
I use 3u3F MKT for this duty and adjust the resistor values to get the filtering where I want it.
The input from the speaker is AC and during fault condition can have prolonged +ve or prolonged -ve coming in through the resistors to audio ground.
This fault condition will charge the capacitor to full fault voltage which could be as high as the supply rail.
On this basis the capacitor needs to be Bi-Polar of a rated voltage that exceeds the maximum supply rail voltage.
I use 3u3F MKT for this duty and adjust the resistor values to get the filtering where I want it.
Thats my thoughts on the nature of signal received by the Cap. Reading fault diagnosis threads associated with 'name brand' Amps this cap has failed [ down to < 1uF] over time resulting in repeated low level trip. Seems it is needed to smooth signal transients even at low output levels. Has me wondering about taking value down to less than 22uF.
You suggest MKT [non-polar] Andrew, whats the idea using 2x inverse series polarised caps as shown in post #43 ? Wishful thinking?
You suggest MKT [non-polar] Andrew, whats the idea using 2x inverse series polarised caps as shown in post #43 ? Wishful thinking?
470uF+470uF + 56k gives an F-3dB @ ~ 0.012Hz
Are we supposed to copy that nonsense?
How long will it take for the 2Vdc fault to charge that capacitor and trigger the detection?
What if the polar capacitors leak a bit? Will the detection ever trigger?
Who draws these contraptions?
Who reads them without thinking about the consequences?
Are we supposed to copy that nonsense?
How long will it take for the 2Vdc fault to charge that capacitor and trigger the detection?
What if the polar capacitors leak a bit? Will the detection ever trigger?
Who draws these contraptions?
Who reads them without thinking about the consequences?
All good questions... I'm trying to make sense of.
Here is another...
Has anyone simulated or actual loaded fault tested the eBay offerings that have a standard electro 100uF / 50V there to see if they have any DC protection merit?
Might be ok for a while on an Amp that doesn't load above 50V DC under clip or short circuited devices...... but the DC load break capacity of relay is still dubious.
And, whats the reasoning in the NEC application notes that advise similar configurations?
Surely they couldn't have made such a basic error and haven't corrected it for so long, and still another manufacturer has a direct copy of the NEC data sheet.
Here is another...
Has anyone simulated or actual loaded fault tested the eBay offerings that have a standard electro 100uF / 50V there to see if they have any DC protection merit?
Might be ok for a while on an Amp that doesn't load above 50V DC under clip or short circuited devices...... but the DC load break capacity of relay is still dubious.
And, whats the reasoning in the NEC application notes that advise similar configurations?
Surely they couldn't have made such a basic error and haven't corrected it for so long, and still another manufacturer has a direct copy of the NEC data sheet.
Metallised Polypropylene suitable?
So, back to this project set aside due to other demands on my time.
OK, my thoughts and so, question about polarised cap use in this location.
I just wonder why I can't find any similar questioning of that device in the circuit. The voltage ratings in various sections had me go over all components to get it right the first time... anyhow.
I recall reading the cap in that location is to smooth transients [speaker signal] to prevent nuisance tripping. The value increased if such presents.
From that the 3u3F could be low given data sheet [albeit dubious with polar electro used] recommendation.
From that I figured on the lowest seen has been around 22uF up to 330uF.... at at least 150V. AC / DC.
Rather expensive those MKT that large in an array.
Then I came to realise, I've used higher voltage AC rated caps in the trade.... motor run [or start] and power factor correction caps.
1) Motor Start caps commonly used here due to supply at 240VAC are rated around 450V. Higher voltage provides longer service life. They generally have a discharge resistor soldered across the quick connect terminals of a relatively high ohm value easily removed. In the chip data sheet a similar resistor appears anyway.
The higher the rated motor [kW shaft output] the higher the cap value for starting torque]. When used for refrigeration compressor starting I've seen these at up to around 100uF [[these compressors start under almost locked rotor condition], but commonly around 30uf for smaller compressors like a domestic unit.
I think these have a paper type dielectric... sometimes metal can, sometimes plastic case with a pitch type seal. They make a nasty stink / mess when they fail, spraying pitch.
2) Motor Run. Same type voltage rating and sized according to torque required. They are permanently connected in the motor circuit [not switched in under start condition] Modern types are of Metallised Polypropylene construction [5% tolerance typical]. I suspect high ohm discharge resistor in the potted off plastic case.these are used in motors like ceiling fans for speed control and also refrigeration compressors.
3) Power Factor correction. Used in transformer ballast fluro lamp fittings. Denoted as being HPF type [not cheap LPF type]. . There is a requirement here to install caps in all commercial and industrial lamp fittings due to the quantity of fittings used that can have a poor effect on installation power factor.
In a twin 36W fitting you find a 8uF 250VAC rated cap. 7uF in a 65W single and 14uf in a twin 65W lamp fitting.
These have been used for years and the early paper/oil in aluminium can contained PCB.... prompting wholesale retrofitting to modern Metal/Poly type in all public buildings years ago..
So, I have a quantity of modern metal/poly non sealed type caps that can be easily opened to find a bare foil cap with discharge resistor ready for removal.
I thought non potted off might be a bit 'ordinary', but these are approved for use in mains powered lamp fittings and I can't recall finding any over the years in a failed /burned condition... so seal seems not that important.
I'll post an image this weekend.
Question is... have these type caps been used in an audio application rather than their intended application? If not why not?
Seems a cheaper alternative for higher voltage crossovers to me... and they vary in values from 1uF up to an array exceeding 100uF.
So, back to this project set aside due to other demands on my time.
OK, my thoughts and so, question about polarised cap use in this location.
I just wonder why I can't find any similar questioning of that device in the circuit. The voltage ratings in various sections had me go over all components to get it right the first time... anyhow.
I recall reading the cap in that location is to smooth transients [speaker signal] to prevent nuisance tripping. The value increased if such presents.
From that the 3u3F could be low given data sheet [albeit dubious with polar electro used] recommendation.
From that I figured on the lowest seen has been around 22uF up to 330uF.... at at least 150V. AC / DC.
Rather expensive those MKT that large in an array.
Then I came to realise, I've used higher voltage AC rated caps in the trade.... motor run [or start] and power factor correction caps.
1) Motor Start caps commonly used here due to supply at 240VAC are rated around 450V. Higher voltage provides longer service life. They generally have a discharge resistor soldered across the quick connect terminals of a relatively high ohm value easily removed. In the chip data sheet a similar resistor appears anyway.
The higher the rated motor [kW shaft output] the higher the cap value for starting torque]. When used for refrigeration compressor starting I've seen these at up to around 100uF [[these compressors start under almost locked rotor condition], but commonly around 30uf for smaller compressors like a domestic unit.
I think these have a paper type dielectric... sometimes metal can, sometimes plastic case with a pitch type seal. They make a nasty stink / mess when they fail, spraying pitch.
2) Motor Run. Same type voltage rating and sized according to torque required. They are permanently connected in the motor circuit [not switched in under start condition] Modern types are of Metallised Polypropylene construction [5% tolerance typical]. I suspect high ohm discharge resistor in the potted off plastic case.these are used in motors like ceiling fans for speed control and also refrigeration compressors.
3) Power Factor correction. Used in transformer ballast fluro lamp fittings. Denoted as being HPF type [not cheap LPF type]. . There is a requirement here to install caps in all commercial and industrial lamp fittings due to the quantity of fittings used that can have a poor effect on installation power factor.
In a twin 36W fitting you find a 8uF 250VAC rated cap. 7uF in a 65W single and 14uf in a twin 65W lamp fitting.
These have been used for years and the early paper/oil in aluminium can contained PCB.... prompting wholesale retrofitting to modern Metal/Poly type in all public buildings years ago..
So, I have a quantity of modern metal/poly non sealed type caps that can be easily opened to find a bare foil cap with discharge resistor ready for removal.
I thought non potted off might be a bit 'ordinary', but these are approved for use in mains powered lamp fittings and I can't recall finding any over the years in a failed /burned condition... so seal seems not that important.
I'll post an image this weekend.
Question is... have these type caps been used in an audio application rather than their intended application? If not why not?
Seems a cheaper alternative for higher voltage crossovers to me... and they vary in values from 1uF up to an array exceeding 100uF.
Last edited:
Motor run capacitors have been recommended on this Forum for years as an inexpensive audio capacitor.
If this is not appropriate here then perhaps the mods can delete. I originally posted this on the "Speaker Protection Kit for Under $5" thread, but got no replies.
I have used XY protection boards in a couple of Mosfet amplifiers with no problems. I had one in a TPA3116 amp and found that at a reasonable volume level (not mega loud) there was a noise which is hard to describe. It was like there was poor connection in the signal line, a sort of electrical crackling /spitting interference. If I backed off the volume slightly it disappeared. I put this down to the 3116 perhaps getting too much signal or it was not happy in some way.
I have just put together another Class D amp (TK2050) from a board I have had lying around for a few years and added the XY protection board. I had tested the amp without the protection board and it was fine, it played to silly levels without distortion. After I added the protection board it did the same as with the 3116 board. Anyone have any ideas what is going on?
I have used XY protection boards in a couple of Mosfet amplifiers with no problems. I had one in a TPA3116 amp and found that at a reasonable volume level (not mega loud) there was a noise which is hard to describe. It was like there was poor connection in the signal line, a sort of electrical crackling /spitting interference. If I backed off the volume slightly it disappeared. I put this down to the 3116 perhaps getting too much signal or it was not happy in some way.
I have just put together another Class D amp (TK2050) from a board I have had lying around for a few years and added the XY protection board. I had tested the amp without the protection board and it was fine, it played to silly levels without distortion. After I added the protection board it did the same as with the 3116 board. Anyone have any ideas what is going on?
The DC detection circuit may be sending a pulse to the trigger circuit and causing the relay to chatter.
Cam you monitor the output of the detection circuit?
Cam you monitor the output of the detection circuit?
Sadly I have no scope to monitor the output. I suspected that it might be some DC creeping through and that the LED might have flickered if it was intending to go into protection, but it stays off.
You can likely measure the detection circuit with a voltmeter. Assuming it's a conventional detection circuit, measure both AC and DC voltage at the base of the detection transistor(s). If you are reading AC the filter isn't operating correctly, likely poor quality capacitor. If you are measuring DC the amp may have issues, or it might need a couple bleeder resistors added.
Thanks robydream for sharing. Do you by chance have any updated version and would this work with a high powered class-A amplifier as well? What have you used for the temperature switch? Cheers
Last edited: