FF85WK and RS225-8 Passive WAW / FAST

[BYRTT]

DATS sweep 225-8 sealed looks fine.

From post 1 comparing schematic to hardware stumbled over what white wire on cap do but maybe it was just from some pre-stage exercise.

 

[xrk971]

Yes white wire was a dangling vestigial remnant from an earlier xo attempt. Not connected to anything.

So we are not any closer to finding the source of the Bermuda Triangle's strange effects in my room.

There is only amp, crossover, or mic as the problem left.

 

[BYRTT]

Yes white wire was a dangling vestigial remnant from an earlier xo attempt. Not connected to anything.

So we are not any closer to finding the source of the Bermuda Triangle's strange effects in my room.

There is only amp, crossover, or mic as the problem left.

Thanks clearing guess then you take things one by one as time allow : ) further can think of is your REW up to date and how is its settings, in same regards are you on good old Win7 or had it updated to Win10, and last what about OB effects as Grasso789 also point to.

 

[xrk971]

REW v5.1
Using Win7 still - don't think OS has anything to do with it could it? Maybe a bad sound card?
I will use Dagger next to eliminate OB.

I can't imagine it's the amp. This is perhaps one of the best amps I have ever used. I just tried using a 53v dual rail SMPS with my custom CRCLC filter last night and the power supply rail at 51v with no music playing has 1mV ripple. The speakers have never been so dead silent before.

It just shows that low cost SMPS with the right filtering can be used for very nice HiFi class AB power supply - the ripple is less than what a toroidal trafo can do. Continuous current is not as high though. I don't think anyone would believe you can get 20ppm ripple with a $20 SMPS unless I took a photo of the Fluke DMM AC volts reading.

 

[BYRTT]

No don't think OS has anything to do with it when you on Win7, it was more the problems for USB ports and audio bwaslo told us about Win8/8.1 and guess Win10 in engine room is Win8/8.1

Your amp looks great be proud, also a pleasure kind of feel it makes your day better : )

Not that i think it is so but take care not to go to far in values for that CRCLC filter just to have good spec readouts on a DMM. The more R in in each filter really clean ripple but also slow the chain and it is important speed is up to application and think also that all cascaded filters shall have same relative speed and energy. At this webpage Duncan's Amp Pages there is a fun tool under software called PSU Designer II, for fun try model a cascade of 10-15 RC filters and see how clean stuff goes into nano volt area but also how many seconds from start of power up before power is ready downstream at load.

 

[jacozz]

X.
Is that a current source amp?
I don't know much about these things, but I remember A Nelson Pass test with current vs voltage amps: Lookie here:
http://www.firstwatt.com/pdf/art_cs_amps.pdf

 

[xrk971]

X.
Is that a current source amp?
I don't know much about these things, but I remember A Nelson Pass test with current vs voltage amps: Lookie here:
http://www.firstwatt.com/pdf/art_cs_amps.pdf

It's the VHEX+ designed by members Vzaichenko and Jwilhelm. Ffrom what I am told, it is a "non-standard current drive approach to voltage amplification". More info here:
vzaudio

Schematic here:
http://www.diyaudio.com/forums/atta...lifier-simulated-tina-vhex_v1.1_schematic.jpg

PCB's are $27/pair - BOM ended up about $45. A very nice amp for the price.

Byrtt:

Thanks for the kind words. It's one of my better amps - fit and finish wise I had a lot of practice on soldering leading up to it

Not that i think it is so but take care not to go to far in values for that CRCLC filter just to have good spec readouts on a DMM. The more R in in each filter really clean ripple but also slow the chain and it is important speed is up to application and think also that all cascaded filters shall have same relative speed and energy. At this webpage Duncan's Amp Pages there is a fun tool under software called PSU Designer II, for fun try model a cascade of 10-15 RC filters and see how clean stuff goes into nano volt area but also how many seconds from start of power up before power is ready downstream at load.

I did simulate this and the propagation delay was not bad at all as the limiter is the 3.4ohm resistor between the 1st and second cap banks, third is connected via low R inductor.

 

[xrk971]

The Bermuda Triangle Mystery Solved

Ok, after some detective work following the footsteps of Mr Holmes, I used method of elimination to find what was causing my terrible frequency response and step response measurements that seemed to be messed up with a huge +10dB bass hump followed by a mid bass dip.

I switched amps from my new VHEX+ MOSFET class AB back to my basic TPA3116D2 that I have used for the past 3 years. Did not help.

I switched back to an old crossover with the Dagger that I knew worked before. Did not help same old problem.

I switched microphones from UMIK-1 back to UMM-6, did not help.

So after having eliminated everything but REW software (newer version) and the computer itself (new laptop since last one died), it must remain with these two. I decided to use the audio out from my USB DAC (a Behringer UCA202) rather than the built in headphone outs. No matter how unlikely the answer is, it must be correct when all other options are eliminated.

It was like the Gods of Olympus all arrived at once - Shazammm! Everything works now and it works like gangbusters. So check out the results now.

Here is the measured crossover plot (with VHEX+ amp and passive XO and UMM-6 mic, UCA202 audio out - via headphone jack):

I am still wondering why the FF85WKeN has such a HF drop off at zero degrees incidence. Checking into that now - but the rest looks wonderful.

Here is the measured phase (just as a 1st order transient perfect XO should look like - flat):

Here is the measured distortion (now excellent with no bass hump, will need to redo again as there is some ambient junk coming from the washer/dryer running upstairs above the lab):

And the real clincher is the impulse and step response - a darn near perfect right triangle:

Here is a square wave recorded at 2kHz:

Here is a square wave at about the XO frequency of 700Hz - a challenge because the two drivers have to be balanced and in phase:

All is right in the lab again and the veil of the Bermuda Triangle has been lifted. There is a terrible hidden bass boost EQ lurking somewhere on my PC that I cannot find when using headphone outs.

 

[jacozz]

Good work X!
I don't know what to say about the decision by the pc-manufacturer to make a fixed non removable EQ?

 

[wesayso]

X,

Try and go to: Control Panel > Sound > "Speakers" or "Headphone" or "Line out" etc, Properties > Enhancements = (in the scroll down menu) Loudness Equalization or Bass Boost

Look if you can find this on the Enhancement tab of your headphone out:

It might be that something is selected there by default...

 

[xrk971]

I'll give it a look when I get back. Could it be something as simple but lame as that?

 

[zman01]

X,

What sound chip do you have on the new notebook?

 

[xrk971]

X,

Try and go to: Control Panel > Sound > "Speakers" or "Headphone" or "Line out" etc, Properties > Enhancements = (in the scroll down menu) Loudness Equalization or Bass Boost

Look if you can find this on the Enhancement tab of your headphone out:

It might be that something is selected there by default...

I checked and nothing was applied. I even checked the box "disable all enhancements" and it still had the problem.

 

[xrk971]

X,

What sound chip do you have on the new notebook?

It's an HP 2760p. The built in sound card looks like an Intel with an IDT driver. But looking at the Control Panel, I see a thing called SRS.

And sure enough, in the icon bar at the bottom SRS sound enhancement was running with this EQ as default:

What a piece of work that was. I turned it all off and lets see if that helps. Another great piece of unwanted bloatware.

Just tested it with SRS disabled and it works fine from the laptop headphone output.

 

[xrk971]

Measurements with UMIK-1

I repeated measurements with UMIK-1, calibrated by Cross Spectrum Labs. Calibrated absolute SPL with handheld sound meter. The UMIK-1 brings up the HF a bit and lowers the mids a bit for an even flatter response.

Here is XO plot (light blue is no felt absorber on shelf created by setback):

Phase and SPL note that variation is about +/- 1.5dB a very good result for a simple XO with no notch filters:

HD (HVAC not running so cleaner than earlier one);

IR/SR (triangle appears straighter):

All taken with UCA202, VHEX+ amp (now with 300VA +/-44vdc power supply), 0.5m.

 

[zman01]

It's an HP 2760p. The built in sound card looks like an Intel with an IDT driver. But looking at the Control Panel, I see a thing called SRS.

And sure enough, in the icon bar at the bottom SRS sound enhancement was running with this EQ as default:

I had suspected as such... annoying indeed.

Good to know that you've got it under control now.

Btw, you are running this with the same passive XO you shared in post #1 of this thread?

 

[xrk971]

I had suspected as such... annoying indeed.

Good to know that you've got it under control now.

Btw, you are running this with the same passive XO you shared in post #1 of this thread?

The schematic diagram topology is same but some values have changed. I will update the diagram soon but basically the main high pass cap is now 60uF and the low pass coil is about 4.12mH. The resistors and 2.2uF on the Zobel remain unchanged. I was trying to get back to an earlier XO design when debugging that bass bump but the values eneded up working pretty well given the flat
response.

I am going to investigate the use of a very cost effective cap and coil combo. There are $0.20 MLCC caps that are rated 10uF ea at 50v. I will try using qnty 6 x and add a single MKT 2.2uF bypass and I think Coilcraft has some nice high current ferrite core chokes for the 4mH value that are reasonably priced.

 

[BYRTT]

Ok, after some detective work following the footsteps of Mr Holmes.....All is right in the lab again and the veil of the Bermuda Triangle has been lifted.....

Good news

 

[xrk971]

Revised XO schematic diagram

Here is the revised crossover used for latest measurements and for blind test.

 

[xrk971]

FF85WKeN vs Pristine ABX listening test thread

http://www.diyaudio.com/forums/full...est-enabled-ff85wk-round-6-a.html#post4737993