Fully Discrete 400 / 700 Driver Circuit (Full Comp Only)

[laatsch55]

Sounds good Nick!!

 

[THD+N]

Frequency Response







Power Output @ 20kHz



FFT @ 20kHz Right



FFT @ 20kHz Left



FFT @ 1kHz Right



FFT @ 1kHz Left



Power Output @ 1kHz



Right Signal to Noise @ 1 Watt



Left Signal to Noise @ 1 Watt



Phase Response @ 1 Watt Stereo, Blue line is phase response



Right Channel Separation @ 1 Watt



Left Channel Separation @ 1 Watt



Signal to Noise @ Rated Output (350 Watts)

 

[oldphaser]

Nick,

It sure is nice to see all those plots.
I will have some questions interpreting a couple of the plots a little later.

I didn't know you had an Audio Precision System. I thought you had an old HP distortion analyzer like a 333A or 334A.

Looking at your plots is spurring me on to buy one as well.
Using my old HP 339A is slow and cumbersome.
What sweep rates are you using?
It also takes me a while to take discrete measurements... adjusting my signal generator to get a specific output level and recording data. All the while heating up Zobel networks, output transistors, fuses, etc.

I have been measuring channel separation at rated output and -3dB for the last few days. I see a difference when I change output levels. I am not so sure that simply my adding 25.442 dB (accounting for 350 W at 8 ohms minus 1W) to a measurement taken with reference to 1W will give me what I am looking for.

I have also found that reversing the line cord can change the signal-to-noise and channel separation measurements. Have you ever tried doing that too? I also plug directing into the AC outlet and not into my Variac when performing these measurements.

I would welcome from anybody out there who has any WOPL's and an Audio Precision system (or any system that has the capability) to generate the same plots.

Ed

 

[THD+N]

BTW, all noise measurements were unweighted. I have an A weighted filter in the AP, but for some reason, it is not being switched into the signal path. I'll have to find out what is going on there.

I have some old HP's and 2 ST 1700B's. I need to let some of that stuff go. If anybody needs one let me know.

Regarding "sweep rate", for a normal test (not involving high power output), I usually use the pre-selected sweep rate, which is "Auto". For test that involve rated power or more than 100 watts, I set the sweep rate and steps to something else to get the test over quickly.

I've noticed lately that you are devoting a lot of time to channel separation. Any particular reason?

When testing, I plug the DUT into a heavy duty power strip on my bench. I've read previously about your experience with reversing the line cord, but have not done so myself. I will have to explore.

I'll post more graphs as testing continues.

 

[laatsch55]

What Model AP is this??

 

[Northwinds]

That's freaking sick you can just hook up to your computer and do this crap and see it right on the screen clearly

 

[Skratch]

Its 2018

 

[oldphaser]

What Model AP is this??

Lee,

I typed in "Audio Precision APWin" (as it appears on one Nick's plots) on Google and got this:

https://www.ap.com/download/apwin-control-software

APWIN Control software. For all System One and System Two models.

Here is a link to version compatability:
https://www.ap.com/technical-library/ap-software-windows-version-compatibility

Here is a link to APWin User's Manual for System Two Cascade:
http://www.waynejones.net/wayne/documents/APWINforSystemTwoCascadeUserManual20_abridged.pdf


Ed

 

[oldphaser]

I've noticed lately that you are devoting a lot of time to channel separation. Any particular reason?

Nick,

Yes. It is as a result of performing a multitude of tests on Phase Linear and other modified Phase Linear amps.

The following appears at http://arrakis-systems.com/pdfs/installtesting.pdf
e) TYPES OF CROSSTALK There are two basic types of crosstalk: capacitive, ground coupled.
f) CAPACITIVE CROSSTALK- This crosstalk is created by capacitive coupling between circuits. This can be radiated by components or caused by power supply coupling. This type of crosstalk increases at 20dB per 10 times increase in frequency. As an example, if crosstalk was -70dB at 1kHz, it would be -50dB at 10kHz.
g) GROUND COUPLED CROSSTALK- This crosstalk is created by signal currents in the audio ground. In this way, one circuit can induce a signal into another circuit through the common ground. This type of crosstalk is independent of frequency and has a strong low frequency component.
h) CAUTIONS
1) Because crosstalk is by definition an interaction between one circuit and another, it can not be easily distinguished from crosstalk created by your test setup.
2) Crosstalk is level dependent because it is masked by the noise floor of your system. Crosstalk tests are usually conducted a few dB below maximum output of your audio product.
3) Most crosstalk in audio products is capacitive and therefore increases with frequency. A crosstalk measurement that is independent of frequency is an indicator that the crosstalk may be induced by your test setup.

Yamaha Sound Reinforcement Handbook section 8.8.1 states:
What Causes Crosstalk?
“Crosstalk can be caused by inductive or capacitive coupling between separate circuits, and can be aggravated by poor circuit design, by poor physical component layout within equipment, and by improper cabling techniques. When audio flows through wires, it creates changing magnetic fields, which cut across nearby conductors, inducing voltages in them. The induced voltage is crosstalk. If two transformers are near one another, and their magnetic fields interact, the crosstalk between the two can be very high. When wires or circuit components carry signal voltages, and are near other wires or signal components, the two can behave as plates of a capacitor (albeit with very low capacitance due to small surface area and large distance). This causes the voltage to capacitively couple between circuits.â€

Ed

 

[Northwinds]

Wouldn't more shielding be the answer to better crosstalk figures? I thought the coating on wires was shielding? Just have to eliminate the exposed ends then? Hide the solder joints under a shielding material? Then the actually physical connections and proximity to other connections causes crosstalk?

 

[WOPL Sniffer]

Nick, what size shoe do you wear? Where was your car manufactured? How many LED bulbs are in your house? What's your favorite color? What's your dogs name? How many miles per gallon does your electric golf cat get? Is the red one green? Can you fly without wings? Why do they call it a sandwich when it has no sand in it? Where is the G-spot? and can you run the 440 in 3 seconds???

If the answers are all yes, did you take into account that your abacus has old batteries??


Huh?

 

[THD+N]

Nick, what size shoe do you wear? Where was your car manufactured? How many LED bulbs are in your house? What's your favorite color? What's your dogs name? How many miles per gallon does your electric golf cat get? Is the red one green? Can you fly without wings? Why do they call it a sandwich when it has no sand in it? Where is the G-spot? and can you run the 440 in 3 seconds???

If the answers are all yes, did you take into account that your abacus has old batteries??


Huh?

Only 3 questions can be answered yes or no.

 

[THD+N]

What Model AP is this??

It is a System One with dual domain option. I have a PCMCIA interface connected to my old laptop for the AP interface. Paid about 5k for it back in 2004.

 

[laatsch55]

Ah....That was before the ATS -1DD then?? I like how the interface looks on the computer screen. Too bad mine won't do that...

 

[oldphaser]

Ah....That was before the ATS -1DD then?? I like how the interface looks on the computer screen. Too bad mine won't do that...

Lee,

I just found this. Haven't had a chance to read all of it yet. Not sure it is applicable for yours. If not, I will keep trying.

https://www.ap.com/technical-library/using-ap-basic-to-control-gpib-devices/

https://www.ap.com/analyzers-accessories/ats-1/ats-1-downloads

https://powermadindustries.wordpres...nd-running-with-an-audio-precision-system-one

https://usermanual.wiki/Document/AUDIOPRECISIONSystemOneInstallationGettingStarted.2940084479.pdf

Ebay

https://www.ebay.com/itm/Audio-Prec...993511?hash=item2840cee827:g:IwQAAOSwCbtZyVtE

or this one that a fellow tried to sell twice in recent weeks but no buyers:
https://www.ebay.com/itm/Audio-Prec...D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557



http://www.diyaudio.com/forums/equi...io-precision-system-sys1-usb-interface-7.html

Ed

 

[laatsch55]

Nope, won't help mine. Mine graphs and plots Ed, just in black and white with it's own printer, and then only one test at a time..

 

[oldphaser]

Nope, won't help mine. Mine graphs and plots Ed, just in black and white with it's own printer, and then only one test at a time..

Lee,

I added a hole bunch of new links to my previous post over the last hour and your response.

Here is another one.

ftp://ftp.dcaudiovisuel.com/dca/AudioPrecision/ATS1A-DD_GPIB_Prog_Ref_Manual_rev2.pdf

Please let me know if any of it helps.

Ed

 

[oldphaser]

Lee,

The following was taken from an ATS-1 data sheet/brochure: https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/Audio-Precision-ATS-1A_ATS-1DD_Datasheet.pdf

IEEE-488.2 GPIB Port: An IEEE-488.2 GPIB interface port is included on the ATS-1. The commands closely model the front panel interface to make software development more productive. ATS1 provides full query back of all manual settings and on-line help to speed up code development. Measurement sweep data is stored in the instrument for quick batch transfers without holding up GPIB bus traffic.

Settled GPIB Readings: Reliable measurements are assured by algorithms inside the instrument which automatically compensate for varying device settling speeds. Settling can be disabled for measurements of jitter or other instantaneous values.

GPIB Software Drivers Audio Precision supports the ATS-1 with a National Instruments LabView and LabWindows CVI driver for C and Basic programmers. The LabWindows driver runs with National Instruments GPIB interface cards for personal computers. The driver speeds development of test routines by eliminating the need to learn the ATS-1 programming mnemonics.

HP8903B GPIB Emulation Mode
(see attachment for image)
ATS-1 also emulates the HP8903B audio analyzer HP command set for a simple replacement of slow and outdated equipment in existing systems. A front panel button selects between the Audio Precision 488.2 (GPIB) command mode and the HPIB command mode.

Ed

 

[oldphaser]

Nick,

Would it be possible for you to either replace the original B+ and B- power supply capacitors with something say twice its value (or add additional capacitors of the same value in parallel) and running the signal-noise ratio tests again?

I am curious to see what if any improvements are actually made. Or are the improvement in the numbers so low as to be hardly discernible.

Additionally, maybe there are some other tests w/ the original Phase Linear B+ and B- power supply capacitor values and with twice the original value that can be made such as a "dynamic headroom test" (IHF-A-202 3.2) or any others that you can think of that would effect the performance of the amplifier at the output terminals.

Ed

 

[THD+N]

Ed,
My 700 already has upgraded supply caps.... Mallory 20000uF/100VDC (125VDC surge). Also, I am using 15024's and 15025's for drivers and outputs.

Attached is the current working circuit that I tested in the graphs posted.