The loop at the end?
You can look at several of my amplifier/receiver videos and see the plots of output impedance- the videos are chapter marked which should make it easy to find the output impedance graph.
I am looking forward to going back to your videos now that i have a more clearly understand the test procedures.
My comparison at this time is Bob’s graph in the test procedure. In his example i don’t see anything that resembles that loop back.
Thank you
There is an “upgrade” kit for Hafler DH-220 and DH-500 to replace everything except the power supply and output transistors:
https://audioxpress.com/article/you-can-diy-the-dh-220c-mosfet-power-amplifier-part-1-the-circuit
I rebuilt my 1989 vintage DH-500 and it’s a superb power amp. The DC coupling (via servo) alone made a huge difference in the level and dynamics of bass from a pair of Magnepan MG-IIIa. The “kit” is a set of raw PCBs and a bill of materials at Mouser. It took a few days to get it together, well worth the time and certainly worth the money!
What does that have to do with the Bob Cordell tutorials?
It just has to do with the DH220C design that Rick Savas I did and which is available with PCBs and was a construction project in AudioExpress. It is a full redesign of the Hafler DH220 that is also adaptable to the DH500.
Did you ever look at the David Tillbrook Hitachi MOSFET designs in ETI back in 1981? ETI-477.
They were some of the very first proper amplifier builds I did as a teenager after the Leo Simpson designed Playmaster 40/40 (1976/7).
It was soon after I got into Perreaux and their 2150B. Pretty sure Peter Perreaux was one the absolute first to bring a Hitachi MOSFET application-note based integrated to market and then of course the big power amp lineup all the way to the 5550 into the mid 80s.
I still pickup pro power amps I come across cheap or broken from that era as they often plastered “Mosfet” on the front panels and they are chock full of those lovely Hitachi TO3s. I’ve still got several blank fibreglass unpopulated 477 boards and a few 477 populated boards picked up for a few dollars at a garage sale.
Additional helpful information would be to label each test procedure with what STANDARDS the test complies with. If this also referenced the standard and published it as an attachment at the end of all procedures or a selectable link.
Would it be possible to modify the Load Settings selection? It now takes three clicks to get to the files to load. The menu selection by “JBREED” indicates this feature being implemented.
Hi. I wanted to inform, for those who may not have noticed, that version 3 of his tutorial is available on Bob Cordell’s website
1 Like
Thank you! Comments, suggestions for improvements and mention of errors found are welcome.
2 Likes
Bob’s extensive use of settings files for preparing the QA403 for standardized tests has exposed a curious bug in the QA40x software that I think @matt might want to look into.
It took me a long time to work out what was causing a mains spur (50Hz) to disappear when doing noise settings tests. Also, what was the strong 200Hz all about? And a whole bunch strange frequencies I’d never seen before in the FFT. Much time was spent in my lab looking for the sources of all this noise, including the 16k USB packet noise which appeared and then disappeared.
After a day of thoroughly scratching my head, I tried the 2nd QA403 recently purchased in exactly the same position, same inputs, same cable layout and got exactly the same issues. I moved then to a laptop and the same issues existed.
It comes down to this:
If you create a settings file saved at 192kHz sampling and subsequently load it, either directly or even after doing a ‘new settings’ reset and then loading the settings file, every frequency spur you see on the FFT is 4x what they should be. 50Hz become 200Hz. 250Hz shows as 1000Hz, etc. No wonder I was going crazy.
If you load the settings file again, the problem persists until you manually press the sampling rate buttons and toggle back and forth.
Here’s an example using Bob’s “figure 04 noise.settings file” with a MM (AT91) cartridge load on the inputs in my shielded box. Notice the 200Hz. It made no sense.
Toggle the sample rate button to 48k, back to 192k and then repeat, you get this. Notice 50Hz is where it should be:
Change the sample rate to 96k and save a noise.settings file, load it, run it and you will get this. Notice the mains spur is now moved to 100Hz and everything you see on the FFT is falsely showing a 2x the actual frequency.
The side effect of this bug was also pushing the 16k spike on and off the FFT display making me think I had some other intermittent source of noise.
Until this is fixed in the software, toggle your sampling rate buttons a few times on any loaded high (>48k) sample rate settings files before running any tests and save yourself from this.
Cheers, John
2 Likes
Hi @restorer-john, thanks very much for this. Can you try 1.210 (HERE and see it this works correctly for you?
There was a change in 1.209 where File->New wouldn’t always revert back to 48k in the hardware. Looks like you are 1.209, and so you should see this already. But maybe there was a fix for the fix in 1.210 that isn’t in the release notes.
In 2.210, I can do the following successfully:
- Connect external 1 kHz to the QA403 input
- Do a file->New. Verify freq is correct.
- Set sampling freq to 192k and verify freq is correct.
- Save file as Settings192
- Do a file->New (this reverts to 48k)
- Verify frequency is correct (this confirms the 1.209 bug was fixed)
- Load Settings192
- Verify frequency is correct
The thing that is so insidious about this (as you have found) is that the DAC and ADC are merrily playing a tone that is 4X more than you expect, and yet reporting as you expect. You need an external generator (or main leakage) to figure this out.
1 Like
Hi Jon, thanks for pointing this out. I think there may have been a couple of times where the mains harmonics were not quite where I expected, but I did not drill down as you did. Thanks for all that effort you put in. There will likely be a V4 of the tutorial, and at that point I’ll make any adjustments to the text and/or settings files if that is needed.
Cheers, Bob
1 Like
@matt Success! I can confirm the installation of V1.210 has solved the issue with pre-existing settings files and new saved settings files. Results and measurements are now in-line with expectations.
I think you should consider pulling V1.209 altogether.
V1.209- Here’s the idle gen outputs on loaded settings files for saved sample rates 96 and 192k. Software gen indicates 999.75Hz output:
Frequency counter on the actual output:
Software gen says 1.0004kHz output:
Frequency counter on the actual output:
As far as I can tell, V1.210 solves all these issues. I will keep testing and hang a frequency counter off the outputs as a sanity check. 
Bob’s carefully considered series of settings files for testing are a game changer for new and seasoned users of the QA403.
Hi @restorer-john, yes, you are correct I have deleted the zip associated with 1.209 and added a note to use 1.210 for those that are on pre-releases.
Bob’s carefully considered series of settings files for testing are a game changer for new and seasoned users of the QA403.
Yes, seeing how Bob worked (lots of saved settings files, each for a given task) with his various requests also led to load/save of automated tests. Note that each automated test that you save/load has its own file extension. Shortly you’ll be able to drag an Automated Test measurement from file explorer into the QA40x app and it will cue everything up, ready to run.
4 Likes
Hi @YamahaCA, it’s generally not polite to post private emails between you and someone else. Feel free to summarize if you wish.
2 Likes