"I think there are two issues here. The ultra low frequency noise from the wiring, and the switc hing noise from the diodes.
Why do the caps not smooth the noise out?"
Because the caps are causing it.
Why do the caps not smooth the noise out?"
Because the caps are causing it.
Well,
This brings up another issue, what is the wiring to and between the caps?
How many ufd are the caps?
It is possible that the wire is not heavy enough, or the traces, whatever you are using.
You should test the rails themselves, as mentioned.
I'm still completely lost on what you need a thermistor for if you use a slow start in the primary?? Lost on what a relay is doing in the secondary at all. Lost on what effect a barrier strip will have, removed or not. IF there was a problem with radiation from wiring or a barrier strip, then shielding would be indicated, but I doubt that this is where the problem comes from.
The MUR diodes are soft recovery, iirc...
...and davada, I am unclear on why the caps are causing this noise, I use far more capacitance than he does in my Symphony No.1 amplifiers and I don't get that noise. However I do use copper buss bar and 10ga. wire. (mine is a total of ~500,000ufd. @ ~55vdc)
BigE, it's current, not voltage. Charging up big caps take big current on each cycle, so think big spikes of current, not voltage. So Vdrop on ur wiring = switching noise on ur rails.
The hard ground I refer to, and I should have been clear is the chassis ground vs. the signal ground.
Otoh, you might want to also check for ground loops by lifting the chassis safety ground at the outlet ("3 prong ground lift adapter). Probably nothing there to be discovered, but maybe.
Somewhere in there something you have done is amiss. A picture might really show it (or not).
I would have suggested an old school Tektronix over a new fangled digital scope, unless it was a high-end new fangled digital scope. Fwiw. I do not like the aliasing in the display on the ones I have seen thus far.
_-_-
This brings up another issue, what is the wiring to and between the caps?
How many ufd are the caps?
It is possible that the wire is not heavy enough, or the traces, whatever you are using.
You should test the rails themselves, as mentioned.
I'm still completely lost on what you need a thermistor for if you use a slow start in the primary?? Lost on what a relay is doing in the secondary at all. Lost on what effect a barrier strip will have, removed or not. IF there was a problem with radiation from wiring or a barrier strip, then shielding would be indicated, but I doubt that this is where the problem comes from.
The MUR diodes are soft recovery, iirc...
...and davada, I am unclear on why the caps are causing this noise, I use far more capacitance than he does in my Symphony No.1 amplifiers and I don't get that noise. However I do use copper buss bar and 10ga. wire. (mine is a total of ~500,000ufd. @ ~55vdc)
BigE, it's current, not voltage. Charging up big caps take big current on each cycle, so think big spikes of current, not voltage. So Vdrop on ur wiring = switching noise on ur rails.
The hard ground I refer to, and I should have been clear is the chassis ground vs. the signal ground.
Otoh, you might want to also check for ground loops by lifting the chassis safety ground at the outlet ("3 prong ground lift adapter). Probably nothing there to be discovered, but maybe.
Somewhere in there something you have done is amiss. A picture might really show it (or not).
I would have suggested an old school Tektronix over a new fangled digital scope, unless it was a high-end new fangled digital scope. Fwiw. I do not like the aliasing in the display on the ones I have seen thus far.
_-_-
if you remove the filter caps all the switching noise goes away..
The switching noise is proportional to the value of capacitance.
Consider the instantaneous current at peak sine and the fact that a capacitor is essentially a short circuit to the rectifiers. The rectifiers have some resistance and a voltage drop. So the switching noise increases with instantaneous current. The caps at 80000uF are far too slow to absorb RF.
The caps are mounted on Tea Bag's PS board. The traces on the board are more like ground planes and have minimal distance between them.... ie no loop area. I believe that Tea Bag used 4oz copper.
The caps are 22,000 uF each arranged as CRC so that each C is 44,000 uF
Each channel has 8 of them.
bear said:
A slow/soft start operates very quickly. The cap bank takes much longer to come up to voltage, and will draw extreme amounts of current for a longer duration. A thermistor and a bypass relay is a common implementation for a slow *charge* circuit. A soft start does nothing for the cap bank inrush current.
Hence, the thermistor, bypass relay and blocking relay to ensure that the thermistor is cool enough to work properly.
bear said:
OK. The chassis is connected to the IEC safety earth with a 2-3" piece of wire.
bear said:
Sorry, that is something I will not do.
bear said:
I've misplaced my camera, I think it is about 2hrs away.
I have posted some earlier shots on the F5T build thread. Shots at 3715/3716
http://www.diyaudio.com/forums/pass-labs/207103-f5-turbo-builders-thread-372.html
Assuming EMI is found not to be an issue...... ground-loop is highly suspected... other than major screw up somewhere.
I float everything from ac power ground. I use an isolation transformer.
THx-RNMarsh
I float everything from ac power ground. I use an isolation transformer.
THx-RNMarsh
Last edited:
My 2 centimes here -- I would remove everything relating to the 50/60Hz issue from this thread as it really doesn't relate to the QA400 itself, but an idiosyncratic issue for the user.
It's just putting tons of chaff into the thread into what is, otherwise, a discussion of an otherwise very fine instrument.
It's just putting tons of chaff into the thread into what is, otherwise, a discussion of an otherwise very fine instrument.
You are correct, it does not belong here, other than to identify a use for the QA400 that is not common -- ie it is more than just a distortion analyzer.
perhaps a mod can clean it up and stuff the unrelated posts into another thread?
perhaps a mod can clean it up and stuff the unrelated posts into another thread?
Yes, ah 'spose, at this point the discussion can move to the related F5 thread. Big E could post the link...
_-_-
_-_-
My guess is that in order to get the probes to match precisely (be able to null down to the lower limit of the measurement capability) they can't be random cables or lengths.
Only a guess.
Only a guess.
Actually I think the QA190 probe issue is safety. Making sure you don't have exposed 120V is a real issue. All the similar products are made the same and cost 5X what the QA400 costs. Its a real bargain.
At the moment I have no use for one but I may later.
Since BigE has a Canadian flag I thought maybe QA had changed there policy.
Using the QA400 Tips
For anyone who's new to using the QA400
I've posted some tips for it in my blog.
See, Using the QA400 Tits.
Sorry I can't link to it.
For anyone who's new to using the QA400
I've posted some tips for it in my blog.
See, Using the QA400 Tits.
Sorry I can't link to it.
