Hello all,
This may be of some help, not sure, I always put one of these at the mains input it's an RFI/EMI filter
I'm going to do a screencast hence the warning
One of Steve Wagner's JLH test videos does show squiggly bits on a square wave, caused by the use of an SMPS, can't recall which one though
This is the one I'm using I did not put in R10 and C6 on Steve's drawing, he didn't either to begin with, a further test with both revealed no difference.
Cheers and thanks - I'm still a very Happy Listener Ooooo I can scarcely believe what I'm hearing now, just got in from my little room.
This may be of some help, not sure, I always put one of these at the mains input it's an RFI/EMI filter
I'm going to do a screencast hence the warning
One of Steve Wagner's JLH test videos does show squiggly bits on a square wave, caused by the use of an SMPS, can't recall which one though
This is the one I'm using I did not put in R10 and C6 on Steve's drawing, he didn't either to begin with, a further test with both revealed no difference.
Cheers and thanks - I'm still a very Happy Listener Ooooo I can scarcely believe what I'm hearing now, just got in from my little room.
Last edited:
It can't do any harm, but I don't think there's any gain by using it,
Keep all ac primary and secondary far away from amp signal, twist them, make the short.
And don't use snubbers on rectifier if you don't have the quasimodo, the can produce bad ringing if incorrect values.
As for smps, a pi filter after DC output might be an idea, meaning a low pass filter that removes switching noise.
It's a simple filter and calculators are online, but you need a scope to see which noise to filter for.
Keep all ac primary and secondary far away from amp signal, twist them, make the short.
And don't use snubbers on rectifier if you don't have the quasimodo, the can produce bad ringing if incorrect values.
As for smps, a pi filter after DC output might be an idea, meaning a low pass filter that removes switching noise.
It's a simple filter and calculators are online, but you need a scope to see which noise to filter for.
Pi and multi-stage LC filters do nothing to the reduce common mode noise which SMPS produce. There's a thread in the power suply forum that is looking into this.
Crack out some schottkeys instead. So much easier to avoid a problem than remove it.
(As I've said for years, one day I'll make a SS choke input for the JLH, just to see what it takes. Much less switching noise there)
Hello all,
More advice taken, I removed the 100 nF caps from the power supply and reconnected the caps into one 40,000 uF line.
Now the hum which was not very much has diminished to the almost inaudible.
Thanks again to all of you.
Cheers -J
More advice taken, I removed the 100 nF caps from the power supply and reconnected the caps into one 40,000 uF line.
Now the hum which was not very much has diminished to the almost inaudible.
Thanks again to all of you.
Cheers -J
Just Google "shielded wire". I think LG refers to this (i.e. a braided wire sheath over insulated wire cable) when he says "lace". It really should be used on all wired runs of signal connections that are prone to noise radiated by AC conductor leads of any kind, power or audio circuits. Without redesigning the circuits, your alternatives are to just endure any environmental or locally generated noise, rebuild the amp. with greater separation between power and signal level conductors or substitute for the cable with ground-connected, sheet metal plates between noise sensitive sections.
Not all shielded wire cable is suitable. Small sizes may have high capacitance/meter which cumulatively affects bandwidth - particularly high impedance circuits. You can measure and compare the capacitance if your instruments are capable of some accuracy in the range of 5-50 pF, which should cover most needs.
Claiming that shielded cable is bad without qualification though, is rather naive when the signal has already passed through enormous lengths of it in multiple circuits of various impedances. The problem with applying shielded cable is in misunderstanding what is meant to do, the type and the conditions for best use, since its not a single product of only one specification, nor is it a universal fix for noise ingress.
Not all shielded wire cable is suitable. Small sizes may have high capacitance/meter which cumulatively affects bandwidth - particularly high impedance circuits. You can measure and compare the capacitance if your instruments are capable of some accuracy in the range of 5-50 pF, which should cover most needs.
Claiming that shielded cable is bad without qualification though, is rather naive when the signal has already passed through enormous lengths of it in multiple circuits of various impedances. The problem with applying shielded cable is in misunderstanding what is meant to do, the type and the conditions for best use, since its not a single product of only one specification, nor is it a universal fix for noise ingress.
Last edited:
I just bought a stereo kit with T- 30 outputs from Hong Kong. I have a filiment transformer 12.6v 10-amp. Need a cap supply board. It looks to be 3 weeks away. Any recommedations? Speaker protection?
Hello CC,
You can see my build here: https://jlh1969classa.blogspot.com/p/my-build.html
No noise at switch on just a bit of speaker wobble and nothing at all at switch off.
And a video here
Cheers - Jim
You can see my build here: https://jlh1969classa.blogspot.com/p/my-build.html
No noise at switch on just a bit of speaker wobble and nothing at all at switch off.
And a video here
Cheers - Jim
It's been mentioned many times in this everlasting, recycling thread, that speaker protection with almost all power amplifiers is primarily about blocking DC voltage from being present at the output terminals. DC melts bass speaker voice coils unless the speaker crossover also contains a huge series cap. which doubly eliminates any DC risk. That fault is usually caused by a failure (i.e. short circuit) of one or both output transistors but it could have started with overheating rather than overdrive or because of other failures, parts sourcing and construction errors or just sloppy assembly work.
Thankfully, since the output is polarised at half the single supply voltage, we already have that thumping great output capacitor on board and caps don't pass DC, right? - JLH '69 Protection solved....no nasty relay contacts to cause trouble either 👌
Thankfully, since the output is polarised at half the single supply voltage, we already have that thumping great output capacitor on board and caps don't pass DC, right? - JLH '69 Protection solved....no nasty relay contacts to cause trouble either 👌
Last edited:
And: Even electrolytic capacitors sound very different. Thus, a tonal tuning would be, is, simply feasible.
I didn't realize these are a different animal. So used to the usual push/pull amps. Cap coupled speakers.. Any caps you prefer?
I've used 3,900 or 4,700 uF caps for some years now. The extra capacitance doesn't make a lot of difference to low bass performance - almost none at all, if your speakers are bookshelf size which many of us will be using because of cost, space restrictions for a second system or maybe yours will be a desktop application etc. Anyways, I have some sensitive, floorstanding speakers that only need a few watts to let the neighbours hear what I like. The convincing bass from just a 10W amplifier never sounded so good to me, at least.
Thx for your description. I wondered if it was important? I'm gonna use low voltage because of my 4 ohm speakers. This will be a home brew.
Load impedance and output capacitance requirements are related. The lower the actual speaker impedance, the more capacitance needed for a given bass response. Check JLH's suggestions covering 3-15 Ohm speakers in the parts tables provided in the original magazine article and referenced many times here.