Rsavas -
Since when is there a laptop without a screen and keyboard attached? Having to add a USB powered hu is no big deal, and the QA403 is sure tiny enough to fit on even my messy bench with ease.
Since when is there a laptop without a screen and keyboard attached? Having to add a USB powered hu is no big deal, and the QA403 is sure tiny enough to fit on even my messy bench with ease.
I've used the QA403 with a relatively old PC, and it works fine. Having said that, a $25 USB hub is no big deal. I have not yet tried running it from a laptop. The QA403 actually tells you the USB voltage that it is receiving and the current that it is drawing, and will give you a warning if the voltage is too low. It is highly recommended that a quality USB cable be used, not one of those whose wires are of a very small AWG.
Hi BSST,
I find the concept extremely risky. File compare would probably fail a good percentage.
USB hubs. Bought a self powered one for the bench. Good hub plus decent cables and you shouldn't have any problems.
I find the concept extremely risky. File compare would probably fail a good percentage.
USB hubs. Bought a self powered one for the bench. Good hub plus decent cables and you shouldn't have any problems.
I really like the FFT spectrum display, but the QA403 gives all kinds of useful numerical readouts as well, for example, both THD and THD+N. I believe that few if any traditional THD analyzers, mine included, can produce numerical results for THD by itself. The FFT display is also remarkably good at showing 60 Hz hum and all of its harmonics. Its ability to produce an SNR result is also handy. Ditto a frequency response. It does not provide an electrical output of the distortion residual we all like to look at on an oscilloscope, but it will show a waveform plot of the residual on the PC screen. It also has an oscilloscope function whereby you can see clipping of the sinewave. These are just a few of the features I like.
I'm not saying to throw out your conventional analog test equipment, but this will tend to outperform it and out-function it. Performance-wise, it will run circles around an HP339.
Cheers,
Bob
No, I did not mean to scare you about the USB cable. Just don't use a crappy one. BTW, given the QA403's readout of the voltage it receives, it will largely tell you if one USB cable is better than another.
Cheers,
Bob
Hi Bob,
Given I use an audio analyser, I have to agree with you. But there are times when an old piece of gear with no modern equivalent does some things a lot better. I would say the same for the HP 428B DC clamp-on current meters I use. Tube technology made into the 1980's.
My ancient THD analysers are still on my packed bench because I use them. They are better than any spectrum display for a few things. For setting bias current dynamically, you can't beat a THD meter and scope combination.
Given I use an audio analyser, I have to agree with you. But there are times when an old piece of gear with no modern equivalent does some things a lot better. I would say the same for the HP 428B DC clamp-on current meters I use. Tube technology made into the 1980's.
My ancient THD analysers are still on my packed bench because I use them. They are better than any spectrum display for a few things. For setting bias current dynamically, you can't beat a THD meter and scope combination.
NAND requires an ERASE before WRITE. That's why it's setup in lines.
You read the data out into DDR, then toggle the line, toggle the ERASE line, and then write the block from DDR.
That's why we see Write Amplification with NAND... when your blocks are large and you have to combine the data from several blocks into one. Lots of small writes kills NAND over time.
Yes, it's been a while since I've programmed EEPROM. From my point of view, it all went to NAND in the late 90s.
Crazy, huh.... but there are differences in USB cables... I got rid of all the cheapy ones and started to buy the nicer ones... red for USB-3 (A to C), blue for micro USB (A to micro), gold for USB-C and green ones for the Apple stuff. They look nicer, they make better mechanical connections and by color I can keep them in their own bins and not have to search too much.
Except for OTG... but those I have tagged.
Perhaps I ought to start selling Audiophile Quality USB cables.... Gold Plated.... to go with the Audiophile ethernet switches.... because they have more detailed mid range and extended treble.... you know.. The Absolute Sound will love them.
I did make the exception about hardware people on their bench.
Eons ago I spent like 20 years, off and on, designing ATE. So, I had my bench for development, just like you, to get the ITAs working, but the end product was a test rack for Level One Technicians, E4s in the DoD, and a simple procedure.
But you know, with progress, and so much moving to FPGAs and so on... and the new automated instruments...
I mean, we thought the HP1980B was hot stuff in its day.... press a button and the thing figured it out for you....
Yeah, I'm looking at it... sure wish it were on the Black Friday Sales...
BTW, can you even get an ethernet hub anymore? You can get a managed switch for 50 bucks nowadays...
I wonder if the QA403 will work with a Raspberry... I got some unused ones...
..Looking throug the QA403 forums... I feel bad for you, strictly hardware, guys..... Welcome to programmable devices. ;-) AND, don't pull that plug while it's downloading new firmware...
...welcome to GitLab... ;-)
Last edited:
I wish!
Mind you, I'm looking at new switches, 1 GB, 24 port. The POE types are dear.
In Canada, a managed 24 port 1 GB switch isn't cheap. I think you can still get Ethernet hubs, like 8 port and smaller. 5 port are popular.
Mind you, I'm looking at new switches, 1 GB, 24 port. The POE types are dear.
In Canada, a managed 24 port 1 GB switch isn't cheap. I think you can still get Ethernet hubs, like 8 port and smaller. 5 port are popular.
Switches.... GigE.
5 port unmanaged for less than 20USD.
https://www.amazon.com/gp/aw/d/B07S98YLHM/?_encoding=UTF8&pd_rd_plhdr=t&aaxitk=e06970f67a1f782937d71b85d062018f&hsa_cr_id=0&qid=1731468127&sr=1-1-9e67e56a-6f64-441f-a281-df67fc737124&ref_=sbx_be_s_sparkle_lsi4d_asin_0_img&pd_rd_w=rwExh&content-id=amzn1.sym.8591358d-1345-4efd-9d50-5bd4e69cd942:amzn1.sym.8591358d-1345-4efd-9d50-5bd4e69cd942&pf_rd_p=8591358d-1345-4efd-9d50-5bd4e69cd942&pf_rd_r=AKWBBKE81HW6YA0NDCVQ&pd_rd_wg=Jgmlf&pd_rd_r=b3c80d5f-4d04-4798-9b73-2930912f41cd
Managed for 21 bucks
https://www.amazon.com/TP-LINK-TL-SG105E-5-Port-Gigabit-Version/dp/B00N0OHEMA/ref=sr_1_3?crid=RR0L4EWVMCX2&dib=eyJ2IjoiMSJ9.pysQVRPciWJL9bjScogxDowEd-es96TJ5Zg4_rNuD2NZIKgN2obHtip9higfz7F-dcaUYszt4lnUQGfwva8abjyH2DYX-fG2U99Z_P65zK_N2S_EMYmbBFyNvXWG6juu2hl4Zx1rPnh5snXhjQ8QnSLUaGCod4_e3pzH1c8XDaWxxoPYJ605hSfF7NVRZEGLOCdS1_qqEX6Q7gGdzjw9gAB4ovAx0bPawuks6lr7HeU.idr0mzY_qpznUtS4SV39eU0JC9SXEVCNOKvA7CKSB4I&dib_tag=se&keywords=5+port+managed+ethernet+switch&qid=1731468127&sprefix=5+port+managed+ethernet+switch,aps,179&sr=8-3&th=1
5 port POE for 60 bucks.
https://www.amazon.com/NETGEAR-Gigabit-Ethernet-Switch-GS305EP/dp/B08LR18SC4/ref=sr_1_1_sspa?crid=2CVLZ16GXWYWI&dib=eyJ2IjoiMSJ9.W3picpaF_cFraR3k27zjJsa8USq4hjYsEppCaJAh35-qLOA3W0BH1c-ifvVkKul9ABpkq2pyoNhqWeFAQEilIb0Nb781tL3nxkhoQud1KWjwG2VfAy4c5nRC41kb4e-WJKN2Bhx007UZNZJJiDBv4lMGnJlSCKQu7WemeEXHMmUMIPn73QYQ_Nk-kebWsDIRbtmF537Faopvh1KP6xR3t5IEehQB3CGmonMKF1eDSYY.3621bkH0iX930jvcHoWlO1D3inpm8wQu0XX2oqmdAP4&dib_tag=se&keywords=5+port+managed+ethernet+switch+POE&qid=1731468344&sprefix=5+port+managed+ethernet+switch+poe,aps,211&sr=8-1-spons&ufe=INHOUSE_INSTALLMENTS:US_3M_HL_VENDOR_FUNDING&sp_csd=d2lkZ2V0TmFtZT1zcF9hdGY&th=1
I do have two 24 port and one 16 port managed switches in the LAN, running VLANs and Spanning Tree... but those are to managed my Intranet. At the drops on each bedroom I run 5 or 8 port managed switches that are pretty cheap nowadays.
If Canada is that expensive, maybe we can run a swap... I'll bring you managed switches and you bring me some Canadian....
5 port unmanaged for less than 20USD.
https://www.amazon.com/gp/aw/d/B07S98YLHM/?_encoding=UTF8&pd_rd_plhdr=t&aaxitk=e06970f67a1f782937d71b85d062018f&hsa_cr_id=0&qid=1731468127&sr=1-1-9e67e56a-6f64-441f-a281-df67fc737124&ref_=sbx_be_s_sparkle_lsi4d_asin_0_img&pd_rd_w=rwExh&content-id=amzn1.sym.8591358d-1345-4efd-9d50-5bd4e69cd942:amzn1.sym.8591358d-1345-4efd-9d50-5bd4e69cd942&pf_rd_p=8591358d-1345-4efd-9d50-5bd4e69cd942&pf_rd_r=AKWBBKE81HW6YA0NDCVQ&pd_rd_wg=Jgmlf&pd_rd_r=b3c80d5f-4d04-4798-9b73-2930912f41cd
Managed for 21 bucks
https://www.amazon.com/TP-LINK-TL-SG105E-5-Port-Gigabit-Version/dp/B00N0OHEMA/ref=sr_1_3?crid=RR0L4EWVMCX2&dib=eyJ2IjoiMSJ9.pysQVRPciWJL9bjScogxDowEd-es96TJ5Zg4_rNuD2NZIKgN2obHtip9higfz7F-dcaUYszt4lnUQGfwva8abjyH2DYX-fG2U99Z_P65zK_N2S_EMYmbBFyNvXWG6juu2hl4Zx1rPnh5snXhjQ8QnSLUaGCod4_e3pzH1c8XDaWxxoPYJ605hSfF7NVRZEGLOCdS1_qqEX6Q7gGdzjw9gAB4ovAx0bPawuks6lr7HeU.idr0mzY_qpznUtS4SV39eU0JC9SXEVCNOKvA7CKSB4I&dib_tag=se&keywords=5+port+managed+ethernet+switch&qid=1731468127&sprefix=5+port+managed+ethernet+switch,aps,179&sr=8-3&th=1
5 port POE for 60 bucks.
https://www.amazon.com/NETGEAR-Gigabit-Ethernet-Switch-GS305EP/dp/B08LR18SC4/ref=sr_1_1_sspa?crid=2CVLZ16GXWYWI&dib=eyJ2IjoiMSJ9.W3picpaF_cFraR3k27zjJsa8USq4hjYsEppCaJAh35-qLOA3W0BH1c-ifvVkKul9ABpkq2pyoNhqWeFAQEilIb0Nb781tL3nxkhoQud1KWjwG2VfAy4c5nRC41kb4e-WJKN2Bhx007UZNZJJiDBv4lMGnJlSCKQu7WemeEXHMmUMIPn73QYQ_Nk-kebWsDIRbtmF537Faopvh1KP6xR3t5IEehQB3CGmonMKF1eDSYY.3621bkH0iX930jvcHoWlO1D3inpm8wQu0XX2oqmdAP4&dib_tag=se&keywords=5+port+managed+ethernet+switch+POE&qid=1731468344&sprefix=5+port+managed+ethernet+switch+poe,aps,211&sr=8-1-spons&ufe=INHOUSE_INSTALLMENTS:US_3M_HL_VENDOR_FUNDING&sp_csd=d2lkZ2V0TmFtZT1zcF9hdGY&th=1
I do have two 24 port and one 16 port managed switches in the LAN, running VLANs and Spanning Tree... but those are to managed my Intranet. At the drops on each bedroom I run 5 or 8 port managed switches that are pretty cheap nowadays.
If Canada is that expensive, maybe we can run a swap... I'll bring you managed switches and you bring me some Canadian....
There is an ASIO interface for the QA401. It works as a sound card with that. I don't think one has been hatched for the QA403.
Using a windows PC as the core of a test intrument doesn't really work well long term. I have a Tek DSA8200. Its based on an older Windows. its a real PITA in that there is no upgrade path except to buy a DSA 8300 ($20K used) and you can't change motherboards because the cards are tuned to the motherboard. If you need to look at 10GBPS its what you need. Otherwise huge expensive and quickly obsoleted. (any interest? It is for sale).
I also have 3 ESI Videobridges. These were built on S100 busses and Z80 processors. 40 years later they still work and are close to SOTA for impedance measurements. Its not necessarily bad to use a PC inside BUT you need to get the software right the first time and use serial or GPIB for connectivity to avoid security issues with "networking". (Unless your running centrifuges in Iran, Stuxnet?)
I have managed to get some older hardware to work with Win 10. I have to run 32 bit for the old software to work and some careful tweaking of the system but I have an AP S! with a parallel port Picoscope as a waveform monitor and spectrum analyzer connected to its output. It all works great as a system.
FWIW I also have QA401, RTX6000, Shibasoku AD725D and AM70A, Boonton 1121, Amber 3500 etc. I'm a glutton for punishment. The APs1 is on the bench because its too big and heavy to move. It does have bandwidth to 250 KHz.The QA is a great system with its integrated software whcih cand do most tests with little involvement. The Stand alone intruments are great for "connect, push a few buttons and get some useful info". Not great if you need to plot something quickly.
However even the best tools are useless if you don't now how to use them and an adequte tool can produce execllent result in the hands of an expert.
Using a windows PC as the core of a test intrument doesn't really work well long term. I have a Tek DSA8200. Its based on an older Windows. its a real PITA in that there is no upgrade path except to buy a DSA 8300 ($20K used) and you can't change motherboards because the cards are tuned to the motherboard. If you need to look at 10GBPS its what you need. Otherwise huge expensive and quickly obsoleted. (any interest? It is for sale).
I also have 3 ESI Videobridges. These were built on S100 busses and Z80 processors. 40 years later they still work and are close to SOTA for impedance measurements. Its not necessarily bad to use a PC inside BUT you need to get the software right the first time and use serial or GPIB for connectivity to avoid security issues with "networking". (Unless your running centrifuges in Iran, Stuxnet?)
I have managed to get some older hardware to work with Win 10. I have to run 32 bit for the old software to work and some careful tweaking of the system but I have an AP S! with a parallel port Picoscope as a waveform monitor and spectrum analyzer connected to its output. It all works great as a system.
FWIW I also have QA401, RTX6000, Shibasoku AD725D and AM70A, Boonton 1121, Amber 3500 etc. I'm a glutton for punishment. The APs1 is on the bench because its too big and heavy to move. It does have bandwidth to 250 KHz.The QA is a great system with its integrated software whcih cand do most tests with little involvement. The Stand alone intruments are great for "connect, push a few buttons and get some useful info". Not great if you need to plot something quickly.
However even the best tools are useless if you don't now how to use them and an adequte tool can produce execllent result in the hands of an expert.
I agree with you about traditional test equipment's value. I still have lots of Tek TM500 equipment that I use regularly. Ditto for HP equipment as far back as the 70's and 80s. My favorite AC voltmeter is the HP400EL. It goes to 10 MHz. My favorite test oscillator is the HP653 that goes from 10 Hz to 10 MHz. There are a lot of times when simpler is better. I still have a Tek analog scope in addition to my modern DSOs. I still love my HP3580A analog spectrum analyzer. These were all exceptionally well-engineered pieces of equipment and it is easy to trust their results.
Cheers,
Bob
lol!
Hi Bob, agree completely.
I have a pair of 3580A (cool instrument!) and even some 3581A selective voltmeters, some analogue scopes I keep beside the bench. The 3400A is a wonderful AC voltmeter much like the 400EL except AC only. I have a 1970s HP scope that has a 275 MHz bandwidth, it recently got sick again (more caps I bet).
To see crossover notches, nothing beats a high quality notch filter with autotune. So basically a THD meter.
Hi Bob, agree completely.
I have a pair of 3580A (cool instrument!) and even some 3581A selective voltmeters, some analogue scopes I keep beside the bench. The 3400A is a wonderful AC voltmeter much like the 400EL except AC only. I have a 1970s HP scope that has a 275 MHz bandwidth, it recently got sick again (more caps I bet).
To see crossover notches, nothing beats a high quality notch filter with autotune. So basically a THD meter.
Most of us have seen how the measured THD on a THD analyzer changes with time, especially at high power levels. This happens because the temperature of the power transistors (junction, case, heatsink) changes with time as a result of time-dependent changes in power dissipation in the output stage. As a result, the optimum class AB bias current is not maintained (think Oliver criteria) over temperature swings. This phenomena can even make it difficult to set the quiescent bias for its optimum value. Another way to evaluate this phenomenon is to operate the amplifier at high power, maybe 1/3 rated power for some period and then cutting power to zero and monitoring the quiescent bias change with time as the output transistors cool down. Bias temperature compensation with transistors or diodes mounted on the heat sink or on a power transistor case are not nearly fast enough to track such junction temperature changes.
The QA403 measures distortion in signal bursts that collect enough data for an FFT. These bursts can be made to be quite short. You can set the QA403 to measure maximum burst power or short-term THD with a single short test burst with a very low duty cycle. One can just hit CTRL+space to get a single burst measurement. The duration of the burst is controlled by the chosen sample rate and FFT size. Bursts of less than 100 ms are possible. This greatly reduces the time-dependent temperature shifts in the amplifier during the measurement. Although it will not fix the dynamic bias shift problems in the amplifier, it will allow you to set the bias more accurately for minimum distortion at a realistic power levels for crossover distortion, maybe 10 watts, for example. You can get a distortion reading before too much shift has occurred.
This is also handy for testing a power amplifier safely at high power levels. In some cases, you can see near-instantaneous maximum output power before the power supply has had much time to sag or even before some protection circuits have time to intervene. For these purposes, a relatively small FFT with a high sample rate of 192 kHz can provide good results. In some ways, this is analogous to analog tone-burst testing of amplifiers. It can also enable you to test an amplifier's high-power capabilities with a smaller-wattage load resistor. Finally, and perhaps most importantly, it can help you avoid blowing up your amplifier with really punishing high-power tests, such as with a 2-ohm or even 1-ohm load and/or at high frequencies like 20 kHz. It can also help in investigating the behavior of your protection circuits.
Cheers,
Bob
The QA403 measures distortion in signal bursts that collect enough data for an FFT. These bursts can be made to be quite short. You can set the QA403 to measure maximum burst power or short-term THD with a single short test burst with a very low duty cycle. One can just hit CTRL+space to get a single burst measurement. The duration of the burst is controlled by the chosen sample rate and FFT size. Bursts of less than 100 ms are possible. This greatly reduces the time-dependent temperature shifts in the amplifier during the measurement. Although it will not fix the dynamic bias shift problems in the amplifier, it will allow you to set the bias more accurately for minimum distortion at a realistic power levels for crossover distortion, maybe 10 watts, for example. You can get a distortion reading before too much shift has occurred.
This is also handy for testing a power amplifier safely at high power levels. In some cases, you can see near-instantaneous maximum output power before the power supply has had much time to sag or even before some protection circuits have time to intervene. For these purposes, a relatively small FFT with a high sample rate of 192 kHz can provide good results. In some ways, this is analogous to analog tone-burst testing of amplifiers. It can also enable you to test an amplifier's high-power capabilities with a smaller-wattage load resistor. Finally, and perhaps most importantly, it can help you avoid blowing up your amplifier with really punishing high-power tests, such as with a 2-ohm or even 1-ohm load and/or at high frequencies like 20 kHz. It can also help in investigating the behavior of your protection circuits.
Cheers,
Bob
Mr. Cordell -
Can you take a look at your QA403 and tell me which flavor of USB connector it uses?
Can you take a look at your QA403 and tell me which flavor of USB connector it uses?
ASIO interface
I found my answers after a little investigation
https://quantasylum.com/blogs/news/asio-arrives?_pos=1&_sid=bef4942f1&_ss=r
https://www.lewitt-audio.com/blog/what-is-asio
Hey Bob, how about trying to install Etienne Dechamps QA401 ASIO driver, looks like it's been updated to run on the QA403
https://github.com/dechamps/ASIO401
Someone should buy Etienne Dechamps a pint for his hard work and abilities
More investigation
https://quantasylum.com/products/qa403-audio-analyzer
USB type B
Good info, thx. Now I wonder who wrote the ASIO driver for QA401. If it was done for the QA401, why not the QA403? I do not know enough, who to ask? TonyEE?
I found my answers after a little investigation
https://quantasylum.com/blogs/news/asio-arrives?_pos=1&_sid=bef4942f1&_ss=r
https://www.lewitt-audio.com/blog/what-is-asio
Hey Bob, how about trying to install Etienne Dechamps QA401 ASIO driver, looks like it's been updated to run on the QA403
https://github.com/dechamps/ASIO401
Someone should buy Etienne Dechamps a pint for his hard work and abilities
More investigation
https://quantasylum.com/products/qa403-audio-analyzer
USB type B
Last edited:
My old PCI audio cards, in Windows PC, ran an ASIO driver, it came with the installation software. I do recall the Windows built in driver had serious real time issues, thus ASIO. But again, it was installed within the audio device manufacturer's "driver".
I went to Firewire and USB audio a while back. Running under Windows (XT, Win7, etc..) the older M-Audio and Nuforce devices did require a software driver install but the newer ones, and the Linux based platforms don't need it. I think Apple, based on Unix, doesn't need it either, but I'm not much into Apple... so ask someone else.
IMHO, the best platform to run Audio nowadays is Android, Raspbian and Linux. I've used CentOS and Ubuntu with great success at work. At home I use a number of Android tablets, phones and a Chromebook. They all drive audio devices over USB OTG very well.
I'm running the RME with its own software under Win 10. I should try Android but in this case it's an interesting AD/DA beast so its software runs on Windows... Add 600 bucks for a Dell i5 MFF machine - I run an i7.
I did ask earlier if the QA401 and QA403 run on a Raspberry, or any kind of Linux. I didn't see it in their forums so I asked, no one has replied yet. With a Linux based platform, like a tiny Raspberry 5, you could run audio devices over USB without worrying about latencies. But, in the case of the QA401/403 that's a lot more than just an audio device. But, if they support the USB Audio interface, then they should not require additional drivers to play audio.
EDIT: there are some people in these forums that know a lot more about Raspberry as an audio device... @birdbox ... did you present the Raspberry hat last time?
I went to Firewire and USB audio a while back. Running under Windows (XT, Win7, etc..) the older M-Audio and Nuforce devices did require a software driver install but the newer ones, and the Linux based platforms don't need it. I think Apple, based on Unix, doesn't need it either, but I'm not much into Apple... so ask someone else.
IMHO, the best platform to run Audio nowadays is Android, Raspbian and Linux. I've used CentOS and Ubuntu with great success at work. At home I use a number of Android tablets, phones and a Chromebook. They all drive audio devices over USB OTG very well.
I'm running the RME with its own software under Win 10. I should try Android but in this case it's an interesting AD/DA beast so its software runs on Windows... Add 600 bucks for a Dell i5 MFF machine - I run an i7.
I did ask earlier if the QA401 and QA403 run on a Raspberry, or any kind of Linux. I didn't see it in their forums so I asked, no one has replied yet. With a Linux based platform, like a tiny Raspberry 5, you could run audio devices over USB without worrying about latencies. But, in the case of the QA401/403 that's a lot more than just an audio device. But, if they support the USB Audio interface, then they should not require additional drivers to play audio.
EDIT: there are some people in these forums that know a lot more about Raspberry as an audio device... @birdbox ... did you present the Raspberry hat last time?
Last edited: