¿ªÔÆÌåÓý

  1. HP-Agilent-Keysight-Equipment
  2. Topics

HP8903B - THD measure on Phono RIAA preamplifier

 

Good morning everyone,
I recently bought an HP8903B audio analyzer, and I'm trying to test a RIAA phono preamp with it. I was able to check the RIAA response curve, but I didn't get decent results as far as the distortion measure is concerned.
 I was thinking of inserting a precise reverse RIAA between the HP8093B output and the phono preamplifier input so that the HP8903B sees it as a line preamp and I can have a correct distortion measurement.
What do you think ?
 Is there any better method?

 Thanks for any advice / suggestion.

-- Cheers
??? Attilio
 

Attilio,
What distortion values did you get? Also, what were the values you expected (and what is the preamp)? Is the level you are using for this testing appropriate for an RIAA preamp??

If indeed too far off, I'd recommend you look into possible sources of noise (such as improper?grounding), as the 8903 will return THD+N. For pure THD readings, I recommend you consider using an audio interface + something like ARTA. You can typically use the "monitor" output of the distortion analyzer to get a conveniently normalized level which would protect your interface and get the best of all worlds.?

Radu.?

On Sun, Feb 20, 2022 at 5:46 AM Attilio <a.alovisetti@...> wrote:
Good morning everyone,
I recently bought an HP8903B audio analyzer, and I'm trying to test a RIAA phono preamp with it. I was able to check the RIAA response curve, but I didn't get decent results as far as the distortion measure is concerned.
 I was thinking of inserting a precise reverse RIAA between the HP8093B output and the phono preamplifier input so that the HP8903B sees it as a line preamp and I can have a correct distortion measurement.
What do you think ?
 Is there any better method?

 Thanks for any advice / suggestion.

-- Cheers
??? Attilio

 

It is true that the freq response curve of RIAA standard deviates from flat, so that THD is not exactly what you think. However, RIAA preamps are typically measured that way. You could build a passive deemphasis network after the preamp to flatten it again and feed into the input of the test set. But you shouldn't be getting horrible THD numbers from a preamp in any case, and usually this indicates noise has crept in. A lot of work goes into microphone and phone cartridge preamps to reduce the noise from the IC's since they have high gain. Did you try filtering using 20KHz LPF for example? Does it have a powerline component to the noise, looking on the scope of the nulled output of the test set?
John Lyles
 

Hi??Attilio,
A reverse RIAA network is what I use.? You can check the accuracy of the EQ curve as you are doing, or if you characterise the reverse RIAA network it becomes more simple.? One thing the network will do for you is reduce the noise floor of the 8903B so you can get better readings.

One thing I had to do is move to a modern audio analyser such as a U8903B or Audio Precision in order to get noise floors low enough.? I was lucky enough to grab an RTX 6001 when they were briefly available, coupled with Multi-instrument (M.I.) software, it is an effective tool.? ARTA also works for quick jobs.? I understand that these are very expensive options, so I hope the 8903B with the reverse RIAA? can give you measurements adequate for your needs or expectations.? I dearly wish I could afford a Keysight U8903B.

-Chris
 

Chris, Attilio,
Indeed, an AP box greatly enhances one's ability to do such measurements. I've used an HP8903B w/ Pete's software for a bit some years back, and it's been good, but I've since?switched to AP. Different ballgame, all this is packaged into the unit/software and you can do orders of magnitude?more things,?and quite a bit more straight forward too.?

Not to discourage you, Attilio, on the contrary, this is really fun and gives you quite a bit of ability to perform measurements?and assess your units or designs. Once you nail this, I recommend looking into one of those instruments. Audio Precision is the leader in this field, and System Ones are not quite so expensive these?days. Even?System Twos, if you have a bit more to put into this.?

Radu.?

On Sun, Feb 20, 2022 at 10:09 AM bhome1959 <bhome@...> wrote:

Hi??Attilio,
A reverse RIAA network is what I use.? You can check the accuracy of the EQ curve as you are doing, or if you characterise the reverse RIAA network it becomes more simple.? One thing the network will do for you is reduce the noise floor of the 8903B so you can get better readings.

One thing I had to do is move to a modern audio analyser such as a U8903B or Audio Precision in order to get noise floors low enough.? I was lucky enough to grab an RTX 6001 when they were briefly available, coupled with Multi-instrument (M.I.) software, it is an effective tool.? ARTA also works for quick jobs.? I understand that these are very expensive options, so I hope the 8903B with the reverse RIAA? can give you measurements adequate for your needs or expectations.? I dearly wish I could afford a Keysight U8903B.

-Chris
 

Hi Chris,
thanks I also think that the reverse RIAA also helps me to reduce the noise and to obtain a more realistic measure of the distortion. The HP8903B in my opinion is still an excellent audio analyzer, certainly the U8903B would be the best but its price is prohibitively expensive for me. With the help of the PC, a GP-IB - USB interface and specific software I can also save and print the sweep graphs from HP8903B.

Hello John Lyles,
I filtered the band that interested me, but I have very high distortion values ??at 20 - 30 Hz and between 12 kHz and 20 kHz, not realistic because viewing the waveform at those frequencies with the oscilloscope distortion (values ??up to 10%).

Hi Radu,
I expected very low distortion values, the line preamp test gave me a result from 20 Hz to 20 kHz below 0.07% of distortion.
I have tried to minimize mains frequency interference (50Hz), but with the tube phono preamp it is not easy (there is a disturbing signal at 50Hz of about 4mV pp at the output). Between the output of the HP8903B and the input of the phono preamp I have inserted a resistive divider to adapt the input level.

-- Cheers
? ? Attilio

 

Thanks Radu,
I understand that the new audio analyzers are the best, but for now I make use of the HP8903B (which I really like) with the help of the PC to view, save and print the graphs.
I also saw the U8903B, very interesting but it costs too much for my finances.

-- Cheers
??? Attilio
 

Hi?Attilio,
Make certain your test signal doesn't exceed 5 mV (as a number) or you may drive the phono preamp beyond it's linear range (a lot lower than clipping).? That will drive distortion numbers up.? Also, a tube stage has much lower transconductance than solid state designs, so you will have much higher distortion figures as you have much lower feedback than a line stage will have, even lower in the lower frequencies.? So do not expect very low distortion figures from (especially) simple phono gain stages.? Tubes also are inherently more noisy than a solid state stage given the same care in design.? I design with tubes, transistors and ICs and you have to match your expectations to the technology you are using.? With tubes, power supply noise is critical as well.

I used an HP 339A with the output to a 3585A, and also a ShibaSuko AD725C.? The new audio analysers are more than a magnitude better, but you do what you can.? You may find that a really good sound card can enhance your capabilities, but they do not tolerate input voltages over 5 Vpk? max.? That's a problem with tube stages and also solid state, so you need to condition and limit the input.? Calibrating this isn't too hard.

-Chris

The 8903B is a fine analyser within it's limitations.? I wish I could afford either of the two current audio analysers, but like you I can't.
 

Hi Chris,
yes, I have reduced the input to the phono preamp so as to have the output signal not clipped in the whole band from 20 Hz to 20 kHz (seen on oscilloscope).
With my friends we have made several phono preamps both tube and solid state. No problem with solid state preamps (fet or low noise opamps), while with tube ones they are fine, but it is very difficult to eliminate the hum at the mains frequency (50 Hz) that is heard especially when the turntable arm is raised from the vinile disc.

-- Cheers
? ? Attilio

1 - 9 of 9