BigE try to get the QA400 as far away from the amp as possible. I think this will make a difference. The field intensity from the amp's power transformer decays with the square of the distance.
The idea is to get things working this way before using the differential probe. It makes the trouble shooting easier. One less thing.
Last edited:
You might not win with the shielding. Shielding only works on voltage field not on magnetic fields. The QA400 is a very sensitive instrument and is strongly effected by mag fields. There is nothing but a thing aluminum case housing the QA. Maybe try some magnetic shielding tricks. A large steel garbage can is an inexpensive but very effective solution. Put the QA in the can not the amp.
BigE,
Cover my vacation time and flight to your location and I'm there! I think you could use some hands on help
You could be right.
OK, right now, the QA400 actually touches the amp chassis. The cables I have coming are 8 footers and double shielded This ought to help quite a bit.
The QA190 will need BNC couplers to get more distance from the QA400. The QA190 has hardwired BNC cables attached, and they are short.
On a loopback test, The noise on the charts is present anywhere in the basement. There are very many GU-10 pot light fixtures in the ceiling. Perhaps a correlation? There is also a water meter transmitter in the same room. A few feet from my bench, but about 30 from the amp.
On a loopback test, The noise on the charts is present anywhere in the basement. There are very many GU-10 pot light fixtures in the ceiling. Perhaps a correlation? There is also a water meter transmitter in the same room. A few feet from my bench, but about 30 from the amp.
OK, now your higher harmonics are at more, some much more, than -95dB, and your basic analysis noise floor is at -120dB, so you're getting a pretty good measurement, especially in view of the location of the QA400 and computer. Get the QA400 away from the amp -- David obviously was reading my mail -- or my mind -- again.
A 2nd H at about -70dB is high at 1W. I guess I should look at some info about this amp. RE running the bias up -- you probably want just enough to hold most of your actual music signals in the class A region -- I'm having a hard time seeing a 75W/ch Class A amp biased up to full-on class A not turning the heat sinks red and burning the house down...
I'm going to change my mind again and recommend not using the QA190 diff probe, even though it would really help, because I doubt that it will have distortion low enough to be useful. A better solution would be to cobble up a diff amp stage out of an OPA134 or an LME49710 with unity gain and powered by a couple of 9V batteries, to drive the input of the QA400. That's what I would do.
A 2nd H at about -70dB is high at 1W. I guess I should look at some info about this amp. RE running the bias up -- you probably want just enough to hold most of your actual music signals in the class A region -- I'm having a hard time seeing a 75W/ch Class A amp biased up to full-on class A not turning the heat sinks red and burning the house down...
I'm going to change my mind again and recommend not using the QA190 diff probe, even though it would really help, because I doubt that it will have distortion low enough to be useful. A better solution would be to cobble up a diff amp stage out of an OPA134 or an LME49710 with unity gain and powered by a couple of 9V batteries, to drive the input of the QA400. That's what I would do.
I doubt it. It was in a way his idea.
Richard did this with my oscillator when taking measurement on it.
Worked quite well.
The amp is 10" tall and 20" deep. Two heatsinks per channel.
10.080" - HeatsinkUSA, LLC Store
My buddy dazed2 made a 150 watt version with 4 heatsinks per channel that runs stone cold. It is a four box affair, with the transformers separate from the amps, and using all the audiophile goodies. I made mine with house wiring to prove a point.
I am going to try to lower H2 by adjusting a pot in the amp... it is made precisely for the purpose of giving you the type of distortion you prefer.
Hopefully, the longer BNC's come tomorrow, and I can find the darn camera.
10.080" - HeatsinkUSA, LLC Store
My buddy dazed2 made a 150 watt version with 4 heatsinks per channel that runs stone cold. It is a four box affair, with the transformers separate from the amps, and using all the audiophile goodies. I made mine with house wiring to prove a point.
I am going to try to lower H2 by adjusting a pot in the amp... it is made precisely for the purpose of giving you the type of distortion you prefer.
Hopefully, the longer BNC's come tomorrow, and I can find the darn camera.
Anything special about the can? Can I use a cookie tin? should it have a lid? Does it need to be earthed? can I use a plastic tub wrapped in metal tape?
Oh and here is a link to the amp details. Mine is a V3:
http://www.firstwatt.com/pdf/art_f5_turbo.pdf
http://www.firstwatt.com/pdf/art_f5_turbo.pdf
The seam needs to overlap by a couple of inches. A garbage can is ideal. Comes with a lid and has more overlap than necessary. It doesn't need to be grounded but grounding won't hurt. The metal should be ferrous in nature. Iron. steel. The field will rap around the can with out penetrating it. This is what you want to achieve.
No, plastic wrapped with foil won't work. It need to be a fairly heavy gauge. No cookie tin.
OK on the V3 amp. I consider the cascode of the input JFETs critical, and the use of C3 and C4 as noted by Pass would be a prime requirement for me, too. Then tune P3 for lowest distortion -- seems like Mies Van der Rohe was essentially correct -- less is more.
Other than that, well, the longer cables for the QA400, the small garbage can or other heavy steel can with an overlapping lid because you are measuring in a tough environment noise-wise, and you should be good to go.
BTW, looking at Pass' AP 1kHz THD curve, it starts to flatten out below 2W output, indicating that the distortion is starting to run into the noise, and that gives an idea about where the noise floor is, and you're essentially seeing it at your -120dB...
Other than that, well, the longer cables for the QA400, the small garbage can or other heavy steel can with an overlapping lid because you are measuring in a tough environment noise-wise, and you should be good to go.
BTW, looking at Pass' AP 1kHz THD curve, it starts to flatten out below 2W output, indicating that the distortion is starting to run into the noise, and that gives an idea about where the noise floor is, and you're essentially seeing it at your -120dB...
Thanks! Perhaps doubling the amp input to 0.4 V or -8 dB will give better results?
This will give 4 watts output, and a voltage to the analyzer of 1.2V, or 0.76 dB. Still within range of the QA400.
And, I need to get a small steel can....
davida, would mu metal work at all if just pasted to the enclosure or will the EMI enter via the cable penetrations?
This will give 4 watts output, and a voltage to the analyzer of 1.2V, or 0.76 dB. Still within range of the QA400.
And, I need to get a small steel can....
davida, would mu metal work at all if just pasted to the enclosure or will the EMI enter via the cable penetrations?
I don't. I haven't tried mu metal. Use what's available first for a can and see if it helps before spending.
Gee, don't change the amp before you can make reasonable tests!!
Have you looked at the supply rails with a scope? The amp is more or less under load all the time, so that will show something... then you could look at the supply rails with the QA, but I hesitate to suggest that, at least until you feel more at home with it... blowing it up won't do any good.
Have you looked at the supply rails with a scope? The amp is more or less under load all the time, so that will show something... then you could look at the supply rails with the QA, but I hesitate to suggest that, at least until you feel more at home with it... blowing it up won't do any good.
No scope.... had to return it to TEquipment due to it's failure to start up. Should be back next week.
Oh no, I am not going to change it, I have to determine how first. Building the platform will take some time. May buy a welder of some sort.
Whoa!
http://www.instructables.com/id/How-to-weld-aluminum/
Oh no, I am not going to change it, I have to determine how first. Building the platform will take some time. May buy a welder of some sort.
Whoa!
http://www.instructables.com/id/How-to-weld-aluminum/