Become a Site Supporter and Never see Ads again!

Author Topic: NEW SNAKE OIL: R-09HR DECK MODIFICATION  (Read 20101 times)

0 Members and 1 Guest are viewing this topic.

Offline guysonic

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1366
  • WISDOM FOR ALL TIMES
    • Sonic Studios DSM Stereo-Surround Microphone Systems
NEW SNAKE OIL: R-09HR DECK MODIFICATION
« on: September 06, 2008, 08:41:37 AM »
INTRODUCTION:

Taperssection members are no stranger to modified gear offerings.  While many well known businesses and individuals are involved in modification services, here at Sonic Studios I tend to do things a bit differently than most in this business. 

Sonic Studios may well be the very first to offer deck modification with the arrival of Sony's TCD-D7 DAT deck in 91-92' where the stock deck's mic powering feature was workable, but somewhat marginal for my DSM mic power requirements.  Later TCD-D100/PCM-M1 Sony models plainly would not power DSM mics without very audible poor performance, and occasional gross clipping type distortion from having way-too-low available mic powering current, and therefore, these particular Sony DAT models absolutely required MOD-3 internal deck modification, or any number of external mic power solutions.

Started mod services with a simple (MOD-1) D7 internal deck in-house modification, but potential warranty issues prompted me to contact the then HiFi Sales/Service (the Oade Brothers) to sell my customer new & MOD-1 service on their owned decks.  HiFi SALES was a certified Sony sales/service outfit, so I was hoping to avoid potential warranty issues by offloading mod service to them, and did in fact instruct them schematic/PCB and clear suggestions on soldering technique/equipment on how to initially do the MOD-1 upgrade (the very first deck mod for them) on up to a hundred 7/D8 model decks. 

However, with more experience, a more involved MOD-2 version upgrade seemed necessary, and Oade Brothers (now known as Oade) declined dong more technically complex board work,  and also were finding their own resources taxed to provide the simple MOD-1 modification customers timely services as was too often requested.  Sony warranty issues were then resolved to everyone's satisfaction and I took on all deck modification service for my customers, and also did a few MOD-2's for OADE.

So from about 93-94' onward, Sonic Studios did all MOD-2/3 mic powering Sony deck upgrades on D7/8/100 and M1 DAT models so that DSM mics could be directly deck mic-input powered with no mic performance issues. MOD-2/3 upgrades did have a secondary effect of reducing input noise by 1-2 dB, but this improvement was mostly not audible or advertised.

COMMON MODIFICATION

The usual tact for those doing modified gear service, is to claim custom work done makes for a better performing deck.  Therefore, customers must take the moder's word solely on faith, and subsequent good word referrals of satisfied customers.

The most common modification is to reduce an external preamp or deck's first stage microphone preamp gain by 12-to-20dB so high output mics recording loud POP/ROCK venues do not clip distort the gear's mic preamplifier.  Alternate workaround to deck mod is to apply attenuation to the output of the microphone itself, feeding the mic pre with a lower signal. 

However, some preamplifiers/deck's have a rather noisy mic input and so this tack does not reduce noise, and will often make preamp noise more noticeable with having the pre's gain turned up to compensate for the mic attenuation. 

ALSO, if the pre or deck is also powering the mic, this precludes using an external in-series  attenuation network that disconnects the mics from the powering pre or deck input. 

So in this case, a modification lowering the mic pre 1st stage gain is preferable.  And this modification involves changing the value of just two resistors, lowering the pre's gain by 12-to-20 dB, AND this action automatically lowers noise by 12-20 dB; exactly the same amount of the pre's gain reduction!

While those doing such a modification rarely if ever state before/after specifications or test data (the moder's snake oil factor), customer's faith is most always vindicated by an after-modification experience of no more preamp clipping, and very audible lower preamp noise. 

So the input clipping/noise reduction modification, often done solely based on customer's faith accepting the moder's word, and previous customer referrals, is most often successful in performing real technical magic on the gear, and in this case, involves the changing two internal resistor values with another!   

Yes, it's still considered by me 'snake oil' due to the non-disclosure environment of the moder.  However, it does work to both generate customer appreciation, and an income source for the moder who's providing an on-demand technical service with clear results-proven customer value.

MORE SNAKE OIL MODS

Besides the anti-input-clipping mod that simply decreases an input stage's gain/noise with 2 resistor swaps, there's a second type of mod with maximum snake oil content

This is where the modification claims lower distortion/noise improving performance exclusively separate from gain reduction by replacing preamplifier IC s and other components refining the audio quality of the modified gear.  The problem with these type mods is without before/after specifications/graphs commonly derived from simple bench testing, this type of modification is the most suspect to have 'high snake oil' content. 

Being an experienced systems engineer/PCB board designer with over 40 years working discreet and IC components on very low level analog signal circuits, I know that simply changing an IC and/or a few components may or may not improve performance, and may even hurt performance due to the many variables that determine analog output qualities.

In summary, and in other words, modifications beyond gain reduction change not documenting before/after performance in clear specifications/measurement graphs are all suspect even if customers rave about experiencing improved gear quality.  We all know how subjective personal audition based reports of quality can be, and I personally feel this is no way to verify the truth behind potentially high content snake oil type gear mods.

EDIROL R-09HR MOD-4

Fast forward to the present state of the art compact flash deck, specifically the Edirol/Roland R-09HR model I technically reviewed a few months ago.  See review at: www.sonicstudios.com/r09hrrev.htm

The R-09 and newest R-09HR is NOT a candidate for needing gain reduction modifications as it's mic input is quite good at handling very high level mic inputs.  However, the R-09 does have mic (power on or off) input noise issues, and though much improved, the R-09HR has very audible noise with mic power on to make this mode almost unusable for anything but louder POP/ROCK venue recording when powering higher output mics.

As discussed in several threads, and my site's mic powering page, DSM mic powering MODs are available for current flash deck models Marantz PMD-620, Edirol's R-09, and now announcing MOD-4 for R-09HR that differs from all previous MODs in advertising both DSM mic powering AND NOISE REDUCTION at least when mic power is turned on, and chance of MOD-4 version 2 reducing non-audible, but quality affecting high frequency noise. 

While mic powering upgrade usually does not include preamplifier noise reduction features, I personally like the R-09HR a bit more that previous compact flash models, and especially motivated to improve the noise/quality performance because of its excellent build quality, discrete very low noise dual mic preamp IC, easy 2 AA cell powering, and uncommon 24bit/88.2K  .wav recording feature.

Also different from snake oil type mods, MOD-4 IS FULLY DOCUMENTED  not depending on 'faith' or highly subjective customer's audition based proclamations of worth.

While the DSM mic powering part of MOD-4 is self evident by first trying a stock deck mic power on feature finding the DSM  does not work in the least, to having MOD-4 modified R-09HR fully operating any DSM mic without any performance issues.  Furthermore, the very audible excessive noise heard with mic power turned on is NOW GONE!   

To document MOD-4 (v1) noise reduction performance see the below before/after graphs.

Since MOD-4 does NOT change noise performance over stock deck w/mic power off, MOD-4 power off plot is removed for simpler viewing of the below graph. 

The graph above shows that MOD-4 accomplishes reducing the mic power on input noise to be NOW VIRTUALLY IDENTICAL TO THE STOCK DECK WITH MIC POWER OFF. MOD-4 version 2 may reduce or even eliminate the higher frequency noise, and may also further lower the low bandwidth noise (that's still 3-4dB higher below 20 cycles than mic power off mode) regardless of mic power mode as there's a few areas in the circuitry I'm waiting for ordered parts that may make further noise reduction improvements, so stay tuned for version 2 documented results.

The R-09HR review posted on my site has a 15 second .wav file comparing previous R-09 mic input noise followed with 5 seconds of newer STOCK R-09HR MIC PIP=OFF, then 5 seconds of R-09HR PIP=ON as depicted in the noise graph above, but amplified by 35 dB so very audible.

If interested, anyone can download this file using it to audibly hear MOD-4 results as it will sound EXACTLY the same as the MIDDLE 15 second portion of the STOCK R-09HR recorded with mic power off.  This comparison file is at: www.sonicstudios.com/-35dB_R-09vsHR.wav

THE REALITY OF DOING MODIFICATION SERVICE

The physical aspects of PCB surface mount circuit modification is physically and technically challenging.  While many of us find viewing deck circuit diagrams and enlarged images of PCB component placement challenging to our experience, and made more practical/possible with the availability of very detailed service manuals, the fact remains that removing/replacing the every shrinking size of currently used components is quite challenging.  Not only is skill most needed, special soldering/optical tools are most required to do this type of work successfully.

While many doing gear modifications perform this as an in-house service same as I've done since the early 90's, businesses like Oade now outsource this work to a specialized 'deck repair' company that handles the actual design/physical realities of modification work.  The physical challenges and difficulty in working with current surface mounted component size may be better appreciated with viewing the component size comparison image below.

 

As shown above, chip resistors, and (not shown) same sized but slightly thicker chip type capacitors, are shrinking to be the size of a pencil dot!   In the 90's, most of the Sony DAT decks used the 805 sized chip components, and in later years also included a mix to 603 size on 5-6 layer PCB boards. 

While there are many 603 sized resistor/capacitor chips inside the R-09HR, MOD-4 mostly deals with the smallest 402 size (actually EIA 0402) resistors/chip capacitors on the analog (center) board inside this model.  So for doing effective modification work, not only is uncommon understanding of component/circuit/PCB physics needed, but also a significant investment in specialized solder, soldering tools, and optical gear essential for proven quality results. 

In summary, this is not the kind of work well suited for the inexperienced and poorly equipped, unless those claiming to do this kind of service have solely 'faith' customers requiring no documented reality to moder's claims of mod-service value, and who then spread the good news of their subjective satisfaction with having paid the price of admission for (maybe having) a 'high content snake oil' mod label on their gear. 

Please make no mistake, I have no argument with those doing/accepting 'high snake oil content' modification sold at face value, just suggest this should not be confused with really knowing the truth about untested/undocumented gear's actual value in terms of audio quality/performance.
« Last Edit: September 07, 2008, 06:04:20 AM by guysonic »
"mics? I no got no mics!  Besides, I no have to show you no stink'n mics!" stxxlth taper's disclaimer

DSM HRTF STEREO-SURROUND RECORDING SYSTEMS WEBSITE: http://www.sonicstudios.com

Offline digifish_music

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1014
    • digifish music
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #1 on: September 07, 2008, 12:30:33 AM »
Thanks for the detailed explanation, interesting you can reduce the low frequency noise when the power is on. What was causing this noise btw?

Also you say...

Quote
MOD-4 version 2 may reduce or even eliminate the higher frequency noise, and may also further lower the low bandwidth noise regardless of mic power mode, so stay tuned for version 2 documented results.

...why 'may reduce/eliminate'? Do your own plots show a reduction or not?

BTW: In my experience with the low-frequency noise, it's not audible as it is lost in the natural rumble present in all recording locations except the best designed studios.

digifish
« Last Edit: September 07, 2008, 12:32:46 AM by digifish_music »
- What's this knob do?

Offline guysonic

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1366
  • WISDOM FOR ALL TIMES
    • Sonic Studios DSM Stereo-Surround Microphone Systems
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #2 on: September 07, 2008, 05:59:04 AM »
Thanks for the detailed explanation, interesting you can reduce the low frequency noise when the power is on. What was causing this noise btw?

Also you say...

Quote
MOD-4 version 2 may reduce or even eliminate the higher frequency noise, and may also further lower the low bandwidth noise regardless of mic power mode, so stay tuned for version 2 documented results.

...why 'may reduce/eliminate'? Do your own plots show a reduction or not?

BTW: In my experience with the low-frequency noise, it's not audible as it is lost in the natural rumble present in all recording locations except the best designed studios.

digifish
REPLY 1: A 5 MFD capacitor in the mic bias circuit needing enlarging.  MOD-4 increases this to a 350 MFD value.

REPLY 2: As now edited into the original post above:

"The graph above shows that MOD-4 accomplishes reducing the mic power on input noise to be NOW VIRTUALLY IDENTICAL TO THE STOCK DECK WITH MIC POWER OFF. MOD-4 version 2 may reduce or even eliminate the higher frequency noise, and may also further lower the low bandwidth noise (that's still 3-4dB higher below 20 cycles than mic power off mode) regardless of mic power mode as there's a few areas in the circuitry I'm waiting for ordered parts that may make further noise reduction improvements, so stay tuned for version 2 documented results.

The R-09HR review posted on my site has a 15 second .wav file comparing previous R-09 mic input noise followed with 5 seconds of newer STOCK R-09HR MIC PIP=OFF, then 5 seconds of R-09HR PIP=ON as depicted in the noise graph above, but amplified by 35 dB so very audible.

If interested, anyone can download this file using it to audibly hear MOD-4 results as it will sound EXACTLY the same as the MIDDLE 15 second portion of the STOCK R-09HR recorded with mic power off.  This comparison file is at: www.sonicstudios.com/-35dB_R-09vsHR.wav"

In addition, many of us do understand types of recording situations tend to mask mic/preamp noise.  It may help to remember many owners recording pure acoustic subjects finding the original R-09 mic input mostly unusable due to very audible preamp noise throughout 'low level' sound recordings.
As is audible in the noise sample, the STOCK R-09HR's mic preamp is very much quieter to not be an issue for most doing low level acoustic recordings.  However, during the 'silent spaces' common to some studio/nature ambient situations there's better chance of hearing the stock deck's 20+ dB increased low bandwidth noise over the MOD-4 deck version.

As ordered MOD-4 version 2 parts arrive next week, hoping there's chance of reducing that inaudible, but annoying 'looking' high frequency buzz noise that in my opinion is challenging the quality of at least acoustic voice/music instrument recordings with audible coloration.
"mics? I no got no mics!  Besides, I no have to show you no stink'n mics!" stxxlth taper's disclaimer

DSM HRTF STEREO-SURROUND RECORDING SYSTEMS WEBSITE: http://www.sonicstudios.com

Offline illconditioned

  • Trade Count: (5)
  • Needs to get out more...
  • *****
  • Posts: 2552
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #3 on: September 07, 2008, 03:24:54 PM »
Dear Guy,

Do you know the setup of Plug-in-power for the Edirol R09HR?  That is, the voltage and source resistor?  I believe the original R09 is 2.5v with a 2.2k resistor.  Is the HR different?

By the way, I just ordered a Sony PCM-D50.  That has 2.8v through a 4.7k resistor.  I think you provided similar information to me before about this.

Thanks,
  Richard
Please DO NOT mail me with tech questions.  I will try to answer in the forums when I get a chance.  Thanks.

Sample recordings at: http://www.soundmann.com.

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (12)
  • Needs to get out more...
  • *****
  • Posts: 10982
  • Gender: Male
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #4 on: September 07, 2008, 09:02:50 PM »
Guy,

I'd like to express my appreciation for your in-depth analysis of the stock recorders and of your modifications - backed by actual posted plots reflecting your measurements.  I've referred to your reviews many times as the only available measured comparison between various recorders.  Some of your posts can at times come across as self-promotional and perhaps a bit hard on other builders and some may be difficult to interpret, but all that is all overshadowed in my mind by the outstanding contribution of your posted reviews and in-depth measurements not available anywhere else that I am aware of.  Data speaks louder than words. So again, thanks for the good work and invaluable contributions!

One thing that I notice from looking at the R-09 and R-09HR graphs you've posted here and on your site, is that the R-09HR's line-input noise floor when recording in 88.2 kHz mode is higher than when recoding in 44.1 kHz mode.  From looking at your measurements, it appears about the same as the older R-09 in 44.1 mode.  Granted that subjectively, it's possible that the extended bandwidth of the higher sampling rates available in the R-09HR may be perceptible in some form or another, possibly even after SRC to 44.1, it seems there is a trade off to be made with noise performance.

Using the R-09 line-in almost exclusively, I've been watching with interest to try and determine if the the newer R-09HR's line-in performance is significantly improved.  As far as I can determine, the answer is perhaps, though probably not significantly. For that reason, I've concluded that upgrading to the newer version if using the line-in and external preamplification exclusively is not nearly as compelling as for those using the mic-input where the improved performance is significant.

Interested in your comments.
« Last Edit: September 07, 2008, 10:27:06 PM by Gutbucket »
volition > vibrations > voltages > numeric values | numeric values > voltages > vibrations > virtual teleportation time-machine experience

"Narrow or widely spaced microphone configurations are preferred. It is well-known experience that pure coincidence microphone concepts are not able to produce a satisfying natural spatial impression, due to the lack of adequate interchannel temporal relations (time-of-arrival, phase, correlation)" -Günther Theile
"The mix of the Double M/S signals with a large A/B configuration of omnis results in the spacious sound that is often desired. This option also provides decorrelated low-frequency signals." -Helmut Wittek

Offline digifish_music

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1014
    • digifish music
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #5 on: September 07, 2008, 09:39:40 PM »
Guy,

Using the R-09 line-in almost exclusively.  I've been watching with interest to try and determine if the the newer R-09HR's line-in performance is significantly improved.  As far as I can determine, the answer is perhaps, though probably not significantly. For that reason, I've concluded that upgrading to the newer version if using the line-in and external preamplification exclusively is not nearly as compelling as for those using the mic-input where the improved performance is significant.

Interested in your comments.

I used the R09 now R09HR almost exclusively for line-in and agree with your conclusions. My excitement about the HR has to do with mic-in, larger screen and build quality.

digifish
- What's this knob do?

Offline digifish_music

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1014
    • digifish music
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #6 on: September 07, 2008, 09:49:08 PM »

REPLY 2: As now edited into the original post above:

The R-09HR review posted on my site has a 15 second .wav file comparing previous R-09 mic input noise followed with 5 seconds of newer STOCK R-09HR MIC PIP=OFF, then 5 seconds of R-09HR PIP=ON as depicted in the noise graph above, but amplified by 35 dB so very audible.

If interested, anyone can download this file using it to audibly hear MOD-4 results as it will sound EXACTLY the same as the MIDDLE 15 second portion of the STOCK R-09HR recorded with mic power off.  This comparison file is at: www.sonicstudios.com/-35dB_R-09vsHR.wav"

In addition, many of us do understand types of recording situations tend to mask mic/preamp noise.  It may help to remember many owners recording pure acoustic subjects finding the original R-09 mic input mostly unusable due to very audible preamp noise throughout 'low level' sound recordings.
As is audible in the noise sample, the STOCK R-09HR's mic preamp is very much quieter to not be an issue for most doing low level acoustic recordings.  However, during the 'silent spaces' common to some studio/nature ambient situations there's better chance of hearing the stock deck's 20+ dB increased low bandwidth noise over the MOD-4 deck version.

As ordered MOD-4 version 2 parts arrive next week, hoping there's chance of reducing that inaudible, but annoying 'looking' high frequency buzz noise that in my opinion is challenging the quality of at least acoustic voice/music instrument recordings with audible coloration.


Thanks for your response....

Looking at your file in a spectral editor it appears that there may be more digi-noise (note the horizontal stripes starting at ~8 kHz the top of the plot) in your modified HR (middle section below).



I also see a wide high-frequency noise-band at the top of both HR sections (starts at ~ 16 kHz) that I have not seen before in HR recordings, any comments on that?

It will be interesting to see how you get on reducing this, given the Japanese engineers claimed it was not possible within their budget constraints. Your mods don't look particularly expensive :)

digifish
« Last Edit: September 07, 2008, 09:57:04 PM by digifish_music »
- What's this knob do?

Offline rastasean

  • in paradise
  • Trade Count: (23)
  • Needs to get out more...
  • *****
  • Posts: 3665
  • Gender: Male
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #7 on: September 08, 2008, 01:19:00 PM »
for all your hard work, you get a Grateful Dead Ticket Sales The Other Ones
+t
Advice is a form of nostalgia, dispensing it is a way of fishing the past from the disposal, wiping it off, painting over the ugly parts and recycling it for more than it’s worth.

Offline guysonic

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1366
  • WISDOM FOR ALL TIMES
    • Sonic Studios DSM Stereo-Surround Microphone Systems
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #8 on: September 14, 2008, 05:25:26 PM »
NEW SNAKE OIL PART TWO: EDIROL R-09HR MOD-4 (Version 2)

MOD-4v2 (Version 2) attempts reducing higher bandwidth MIC PREAMP (SPIKE) NOISE unaffected by MOD-4v1.   While this HF spike noise is well below –102dBu, and by itself totally inaudible, there’s chance this type of noise will mix with sensitive recordings producing some audible coloration. So I felt reducing this type of noise, even at such a low level, was worth an attempt. 

Usual first steps in improving performance/reducing analog audio noise is making certain all IC’s have sufficient power supply bypass capacitors in place.  Looking at the mic preamplifier and associated analog processing circuitry, it’s apparent virtually all (center board) IC’s have very tiny (physical size) 0402 ceramic chip 0.1 MFD value power supply bypass capacitors.  0.1 MFD seems a minimum, but likely adequate bypass capacitance value.  So first step (a shotgun approach) was MOD-4v2 increased bypass value with 0402 size 4.7 MFD chips for about a dozen critical preamp IC/transistor components. 

Unfortunately, and somewhat expected, THIS HAD NO EFFECT WHATSOEVER in reducing spike noise, and encouraged me to drill deeper into the analog circuit functions for finding at-cause components.

Noticed while doing the noise plots previously posted spike noise actually decreased ~2-3dB with MIC PIP=ON verses OFF mode.  See graph below.

Possibly the additional PIP loading from the 2.2K/4.7K bias resistors clued me onto the MIC input circuitry (before the preamp IC) as possible area where spike noise (from various bias voltage supplies) was entering the audio path.   As the schematic was studied in greater depth, it also became quite apparent the analog (center/top PCB boards) had a MULTITUDE of unused sections/components all labeled NOT IN USE (NIU), with actual pads/traces still on the PCB supporting NIU functions. 

Surprisingly, this makes both top/center PCBs ‘PROTO’ PRODUCTION BOARDS leaving HUNDREDS of engineering refinement/feature options open in the final production stages.  See schematic samples below.

As mentioned, +VA and preamp BIAS supplies were next to see if spike noise was associated with their functions.  These analog voltage/bias supplies are IC switch/amp buffered sources using moderately large input/output noise reduction filtering capacitors, AND are being used for ALL the analog sections for power supply (+VA) and preamp input (false 0 volts ref. ground) BIAS functions, so it seemed reasonable increasing cap value size in both the +VA/preamp bias supply had good chance to reduce spike noise. 

HOWEVER, this tact also FAILED making NO DIFFERENCE WHATSOEVER!   

See below marked up bias supply schematic, AND this also partially showing AC/DC digital-to-analog GROUND PLANE connection scheme (next topic) used by Edirol engineers.

As some of you versed in the arts may already realize, choice of Digital/Analog System PCB ground plane connection is one of the most critical system decisions with most certain effect for causing/reducing analog signal noise. 

This is also one of the MOST DIFFICULT for engineers as ground plane connection points are most always a function of good/bad PCB designer layout decisions.  In best case, PCB design is greatly influenced by knowledgeable circuit/system engineer’s specific instruction on critical signal/ground path priorities for the designer to closely follow.   

MY BACKSTORY: Personally, I’ve done both engineering, and multi-layer analog PCB design of most critical extremely low signal level instrumentation.  One, an analog/digital mass measuring system, I engineered the analog section only, with the all-on-one digital/analog PCB design done by an outside company BUT, I also documented all critical system path circuit trace priorities, and the PCB designer had me ‘looking over his shoulder’ every step to insure no variations crept into the board design.   Yes, the designer complained a bit stating never had anyone specified so extensively, and monitored design work so closely. However, it also was the first time in our experience a PCB of this complexity and critical low noise performance WORKED THE FIRST TIME as expected needing NO CHANGES!  Soon after completion, the mass-measurement instrument was flown by NASA to map the gravity (variations) of the planet.

Even with Edirol engineering/PCB designers on the ‘same page,’ the physical reality of a manufactured and component-populated proto production board is ALWAYS (with very rare exception) somewhat different than hoped.  Proto-PCB REFINEMENT of ground-plane connections requires adjustments for lowest system noise and doing this adjustment step should be anticipated by experienced engineers/designers.

THIS IS EXACTLY the tact used by Edirol wisely implementing a multitude of digital-to-analog ground plane connections ALL OVER THE BOARD, AND then loading them INITIALLY with 0.1 MFD chip capacitors, THEN LATER deciding on JUST WHERE to physically place the DC connection point for best effect.  I imagine they simply went around the board shorting out the loaded .1 MFD chip caps spanning the digital/analog ground plane with a jumper to see what worked best, then replaced the chip cap with a zero ohm chip resistor (R210) as shown in the schematic.  (SEE CIRCUIT DIAGRAM ABOVE)

Personally having exhausted all other options in reducing the spike noise, I temporarily removed R210 chip resistor and installed an alternate grounding system using 20GA 450 strand wire having many times the copper of the digital/analog PCB board to see what happens to the noise.  (SEE PCB IMAGE BELOW)

AS OFTEN HAPPENS, the alternate grounding scheme MADE THINGS WORSE.  See noise graph below.


MYSTERIOUS SPIKE NOISE SOLVED?

I am disappointed in NOT solving spike noise, and still wondered what’s causing this noise clearly showing in test recordings using mic input?  The most logical, and by extensive process of elimination as it turns out, is the IC5 CODEC internal digital/analog system having a low level ‘bleed’ factor.  YES, this CODEC is a refined design with separate EXTERNAL digital/analog power/ground paths almost eliminating internal cross-system noise.   

However, internally I think there’s still digital-to-analog system (maybe capacitance/radio type) path coupling showing up as very low level noise not removable short of a CODEC redesign of some sort.  Luckily it’s so low level to NOT be at all audible.  I also think the LINE input is more noise immune working a lower gain factor than the MIC input so internal noise is not so amplified, and maybe the input trace path/impedance of the LINE signal path is different/lower so spike noise is not showing in the tests.

SUMMARY:

FIRST, like to acknowledge the engineering skill shown in this Edirol product.  It’s at least evident to me they wisely planned for refinement/feature decisions in the design of the two (top/center) boards.  Evidence of a NIU switching supply (top board) leads me to think they seriously considered having a mic 48 volt phantom power feature, but later deciding this not so practical in the final product.

The low noise performance of the R-09HR is testament to wise engineering practices made most clear to me by closer circuit study, and then trying to improve on a nearly perfect implementation.  Other than my MOD-4 (v1) sole improvement of allowing direct DSM mic powering (see PCB layout below), and (not shown) greatly reducing the mic PIP low frequency noise, I don’t feel there’s any further practical performance improvement possible. 
 
This includes changing the mic preamp dual amp IC1 (see below).


TO MOD OR NOT MOD

Question of audible benefits changing out IC1 for something else is certainly debatable. AND I HAVE TO ADMIT thinking this would be a great new (high snake oil) requested MOD-5 service most likely soon be offered by those in the mod business, and admittedly highly desired by customers looking to own a modified 09HR deck.

Unfortunately, at least for my choices of doing business, is tempered with experience with good reason to doubt the stock IC amplifier is holding back audio quality.  But rather the entire mic preamp circuit design surrounding IC1 is, in my experience, most likely at cause for any and all audible shortcomings. 

Upgrading a circuit like this, in my opinion, is one with internal pre circuit having multiple signal coupling capacitors and other circuit mechanisms raising many performance quality questions not easily addressed with simple parts swap.  So solely changing out IC1 will do nothing real or audible to improve the stock implementation but admittedly satisfying to mod customers if only for having a (high snake oil) ‘modified brand-name’ labeled deck.   

As already mentioned by others, seems most sure-performance upgrade is an external preamplifier totally eliminating all mic input issues.  External LINE feed rests these concerns solely on personal choice of external preamp features/audio quality.

It seems a good time to note the uncompromising quality of internal mic preamp component choices I found inside the 09HR.  Even though I personally would’ve chosen a different preamp circuit design altogether, if at all practical, the Edirol engineer’s choice of signal coupling capacitors used inside showed great concern for preserving audio quality by using as many surface mount FILM TYPE capacitors possible to carry the mic input/preamp output signal.  And where film caps where not practical because of size, then carefully chosen tantalum types seemed evident.  In other words, optimum audio quality was a definite engineered intention regardless of component cost with this model, and this leaves little or no room for someone else to make improvements over the stock product.   

With a few other companies producing low costing flash decks with questionable audio/build quality and with proven issues, I find it refreshing Edirol engineers have learned from past shortcomings (remembering the R-09) in having now produced a truly great and affordable flash audio recording product. 

It’s also now evident to me this model is a bit over-built (not a bad thing) because of the extensive use of highest quality components, and with many additional parts allowing performance refinement tuning of the ‘over-loaded’ proto-production board.  All this adds up to a model with lowered profit margins, but one that hopefully should make for increased sales (quantity=profitability factor) by appreciating recordists.

CONCLUSION

Most should agree the R-09HR is an excellent performer, has excellent build quality, and most easy to operate/power.  All this at a great selling price as-is.  And for at least DSM mic customers, the MOD-4 upgrade does succeed to allow direct PIP powering with reduced low frequency noise.   

For recordists with a self-powered mics not relying on R-09HR’s too low (for most mics) 2.9 volt PIP, the stock deck makes for great 2-3 piece recording system of exceptional small size/audio quality. 

So I recommend everybody buy one showing Edirol they did the right thing with their extra effort/refinement build costs.  This ‘cost-is-no-object-tact’ definitely lowered the profit margin, and this is ONLY made up by increased model sales quantity. 

If Edirol sells sufficient quantity of R-09HR decks making the model reasonably profitable, this should further the motivation in producing MORE lower profit, but much higher quality/refined audio devices eventually making this type of engineered production quite profitable resulting in larger sales quantity from knowledgeable customers recognizing/appreciating a company’s best efforts producing higher quality/performing products regardless of time/cost factors.

While this is my personal tact in producing very low quantity custom audio products, Edirol is making mass produced products at lowest cost in a highly ‘look-alike’ competitive market. Custom proto-production refinement like found in the R-09HR model seems (these days) uncommon. 

So if you want this tact to succeed, vote YES with your dollars showing the audio mass recording gear market this is the way products should be produced!
"mics? I no got no mics!  Besides, I no have to show you no stink'n mics!" stxxlth taper's disclaimer

DSM HRTF STEREO-SURROUND RECORDING SYSTEMS WEBSITE: http://www.sonicstudios.com

Offline illconditioned

  • Trade Count: (5)
  • Needs to get out more...
  • *****
  • Posts: 2552
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #9 on: September 14, 2008, 07:16:06 PM »
NEW SNAKE OIL PART TWO: EDIROL R-09HR MOD-4 (Version 2)


Wow!  Thanks for all the hard work.  I think I learned something... think hard before attempting gear mods.

By the way, do you mind saying what your mod was to PIP for the DSM mics?  I'm guessing you increased the supply resistor from 2.2k, right?  I expect this would increase the gain by a few dB (by reducing the load on the mic input) of the preamp... as long as there was enough current there to drive your mics.  It seems other units, like the Sony PCM-D50, have higher values, like 4.7k.  Sony Minidisc have 6.8k.

Thanks again...
  Richard
Please DO NOT mail me with tech questions.  I will try to answer in the forums when I get a chance.  Thanks.

Sample recordings at: http://www.soundmann.com.

Offline digifish_music

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1014
    • digifish music
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #10 on: September 14, 2008, 09:11:11 PM »
NEW SNAKE OIL PART TWO: EDIROL R-09HR MOD-4 (Version 2)
...

Nice! +T

...and is consistent with what R&D Japan told me, they tried and failed, but were not overly concerned given the low level and con/prosumer nature of the recorder.

digifish
« Last Edit: September 14, 2008, 09:13:14 PM by digifish_music »
- What's this knob do?

Offline guysonic

  • Trade Count: (0)
  • Taperssection All-Star
  • ****
  • Posts: 1366
  • WISDOM FOR ALL TIMES
    • Sonic Studios DSM Stereo-Surround Microphone Systems
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #11 on: September 14, 2008, 10:32:01 PM »
NEW SNAKE OIL PART TWO: EDIROL R-09HR MOD-4 (Version 2)


Wow!  Thanks for all the hard work.  I think I learned something... think hard before attempting gear mods.

By the way, do you mind saying what your mod was to PIP for the DSM mics?  I'm guessing you increased the supply resistor from 2.2k, right?  I expect this would increase the gain by a few dB (by reducing the load on the mic input) of the preamp... as long as there was enough current there to drive your mics.  It seems other units, like the Sony PCM-D50, have higher values, like 4.7k.  Sony Minidisc have 6.8k.

Thanks again...
  Richard


You're correct in assuming a bias limiting resistor change from the stock 2.2K value to identical to D50's 4.7K value. 

Experience powering DSM mics directly off some Sony MD inputs shows a 6.8K is way to high a value for at least my brand of mics causing soft performance and occasional gross distortion from being 'current starved.' 

As mentioned in the original post, MOD-4's low frequency noise reduction was the result of increasing the mic bias supply's pre-filter capacitor value from 47MFD to 350MFD.
"mics? I no got no mics!  Besides, I no have to show you no stink'n mics!" stxxlth taper's disclaimer

DSM HRTF STEREO-SURROUND RECORDING SYSTEMS WEBSITE: http://www.sonicstudios.com

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (12)
  • Needs to get out more...
  • *****
  • Posts: 10982
  • Gender: Male
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #12 on: September 15, 2008, 07:45:39 PM »
Fantastic contribution to the forum, Guy.  Big thanks for sharing your highly informed and thorough analysis. +T

I just encouraged a friend to vote with his dollars for this candidate. If I didn't already have two R-09s I'd be on the short list for this one too.
volition > vibrations > voltages > numeric values | numeric values > voltages > vibrations > virtual teleportation time-machine experience

"Narrow or widely spaced microphone configurations are preferred. It is well-known experience that pure coincidence microphone concepts are not able to produce a satisfying natural spatial impression, due to the lack of adequate interchannel temporal relations (time-of-arrival, phase, correlation)" -Günther Theile
"The mix of the Double M/S signals with a large A/B configuration of omnis results in the spacious sound that is often desired. This option also provides decorrelated low-frequency signals." -Helmut Wittek

Offline flintstone

  • Trade Count: (2)
  • Taperssection Member
  • ***
  • Posts: 765
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #13 on: September 17, 2008, 11:44:16 AM »
Yes, we should vote with our dollars for the best small recorder designs.
The Sony PCM-D50 should be considered in this category. 

The PCM-D50 costs about $450, the R-09HR about $330 from
authorized dealers. D50 features not found in the R-09HR:
-- a preamp that doesn't need modifying to correct flaws
-- low cut filter selectable for 75 or 150 hz
-- 5 second pre-roll buffer
-- very good quality built-in mics
-- optical digital input as well as mic and line in
-- a battery that's good for 10+ hours of recording
-- 4 GB of flash memory built in
-- a particularly effective, yet unobtrusive, limiter
-- an effective noise reduction circuit that improves 16 bit recordings
-- best build quality under $1000

Your cash, your choice.

Flintstone

Offline rastasean

  • in paradise
  • Trade Count: (23)
  • Needs to get out more...
  • *****
  • Posts: 3665
  • Gender: Male
Re: NEW SNAKE OIL: R-09HR DECK MODIFICATION
« Reply #14 on: September 17, 2008, 11:54:40 AM »


Your cash, your choice.

Flintstone

now you just complicate the hell out of things! I really don't know what one I like more now.
Advice is a form of nostalgia, dispensing it is a way of fishing the past from the disposal, wiping it off, painting over the ugly parts and recycling it for more than it’s worth.

 

RSS | Mobile
Page created in 0.36 seconds with 41 queries.
© 2002-2017 Taperssection.com
Powered by SMF
Website Design by Foxtrot Media, Inc., a Baltimore Website Company