Akai GX-620 Reel to Reel Tape Deck Measurements

[pseudoid]

Imagine the crating and freight bill to send the machine to Seattle! It'd be like mailing a pallet of used paving bricks. )

It'd be cheaper if @amirm just provided room&board for anyone that is willing to bring any worthy equipment to be tested to his airbnb...
[want for a missing emoji]

 

[pma]

Fixed by selecting the correct units (dBr). The software provides dBFS display as that’s needed by another feature.

THD and THD+N result must be independent of selected scale, should it be dBFS, dBr, dBV - THD and THD+N are relative numbers and must not change with selected Y scale.

REW has many highly sophisticated functions, and they become obvious just in a "less-usual" case of measurements, like these old tape recorders. In case of distortion setting, it is the AES17-2015 standard notch:

Let me quote from REW manual:

If the Use AES17-2015 standard notch option is selected the fundamental power for THD and THD+N calculations will be the power within a one
octave span around the fundamental frequency. If that option is not selected it will be the power in the main lobe of the fundamental. When the
fundamental level approaches the noise floor using the standard notch will produce a much higher figure than the fundamental main lobe, in those
cases it is better not to use the standard notch setting.
If the Highlight fundamental option is selected the portion of the response which contributes to the fundamental power for THD and THD+N
calculations will be highlighted. The response outside the highlighted region is used for the noise and noise+distortion calculations.

These options are very useful if the fundamental is surrounded by wow/flutter or jitter, like in case of tape recorders.

The power for THD and THD+N calculation is now derived from power over the octave band around the fundamental frequency. Without this option, the THD+N calculation "sees" high noise just next to the fundamental frequency line and gets crazy, as one can see in post #1 and shows too bad SNR and THD+N (TD+N).

Without AES17-2015 notch, REW shows this:

 

[pkane]

REW has many highly sophisticated functions, and they become obvious just in a "less-usual" case of measurements, like these old tape recorders. In case of distortion setting, it is the AES17-2015 standard notch:

View attachment 240133

Let me quote from REW manual:



These options are very useful if the fundamental is surrounded by wow/flutter or jitter, like in case of tape recorders.

View attachment 240135

The power for THD and THD+N calculation is now derived from power over the octave band around the fundamental frequency. Without this option, the THD+N calculation "sees" high noise just next to the fundamental frequency line and gets crazy, as one can see in post #1 and shows too bad SNR and THD+N (TD+N).

Without AES17-2015 notch, REW shows this:

View attachment 240136

No argument, Pavel. REW is an awesome piece of software and John is an amazing developer. Multitone isn't designed to compete with REW, although it does have some similar features, as they were requested by users. AES17 notch is one of these.

 

[Draki]

Analog Open Reel Tape Machines | BALLFINGER | The Ultimate Analog Sound

Open Reel Tape Machines Following the revival of vinyl, tape is now making a comeback in the world of music. As the highest quality medium…

www.ballfinger.de

 

[audio2920]

It does look cool!! Very nice.

In the film world we used to mitigate for calibration errors by shooting pink noise at the head of each roll (on the machine that was about to record to it) such that when a recording was put on a different replay machine, it could be brought back to reasonably flat, even if the recorder's line up wandered or variance in tape stock introduced issues.

As others have said, with analogue you really need a reference tape, unless all you want to do is record and playback on the same machine. For the latter it obviously doesn't massively matter if the EQ is wrong, so long as the errors in record EQ are counteracted on repro.

I'm not familiar with the GX-620, presumably it has some EQ adjustment available in the guts somewhere...??

 

[Stephen H]

No argument, Pavel. REW is an awesome piece of software and John is an amazing developer. Multitone isn't designed to compete with REW, although it does have some similar features, as they were requested by users. AES17 notch is one of these.

View attachment 240138

As powerful as REW is, I prefer using your software.

For clarity, using the dBr option should give accurate TD/TD+N results, correct?

Also, I take it the notch setting is not necessary when measuring devices such as DACs etc.?

Please forgive my ignorance, this is not my area of expertise.

It does look cool!! Very nice.

In the film world we used to mitigate for calibration errors by shooting pink noise at the head of each roll (on the machine that was about to record to it) such that when a recording was put on a different replay machine, it could be brought back to reasonably flat, even if the recorder's line up wandered or variance in tape stock introduced issues.

As others have said, with analogue you really need a reference tape, unless all you want to do is record and playback on the same machine. For the latter it obviously doesn't massively matter if the EQ is wrong, so long as the errors in record EQ are counteracted on repro.

I'm not familiar with the GX-620, presumably it has some EQ adjustment available in the guts somewhere...??

Thanks! At some point in the future I will have a go at calibrating the Deck properly. It will give me something to do. I can purchase an MRL tape from eBay.

For now, my plan is to pay the electronics some more attention. The pre-amp board has had a complete re-cap. It only makes sense for the PSU to undergo the same treatment before anything goes bang. The deck is not used a lot, so I'm really not overly fussed about the uneven frequency response. Obviously it would be nice for it to be flatter, but it's hardly a priority right now. You can adjust the bias etc. via some trim pots on the pre-amp board.

 

[pkane]

As powerful as REW is, I prefer using your software.

For clarity, using the dBr option should give accurate TD/TD+N results, correct?

Also, I take it the notch setting is not necessary when measuring devices such as DACs etc.?

Please forgive my ignorance, this is not my area of expertise.

TD+N is always relative, so dBr setting isn't required and will not change anything. I've released a new version of Multitone (1.0.54) that changes THD calculation to make it always relative to the fundamental, rather than based on the current units. This way you don't need to worry about unit selection when looking at the THD value. dBr is still a better choice for THD+N/THD charts as that scales the fundamental to the 0dB line, making it easier to read the relative amplitudes directly from the chart. Harmonic amplitudes are still shown in the units selected.

The AES17 notch is needed if there is a large amount of noise/jitter/distortion around the fundamental, as was the case with your measurement. Multitone doesn't use AES17 notch by default because it's really not suitable for multitone measurements, but it is fine to use for a single tone. I believe it's the default notch for Audio Precision measurements of THD and THD+N.

 

[Stephen H]

TD+N is always relative to the fundamental, so dBr setting isn't required and will not change anything. I've released a new version of Multitone (1.0.54) that changes THD calculation to make it always relative to the fundamental, rather than based on the current units. This way you don't need to worry about unit selection when looking at THD value. dBr is still a better choice for THD+N/THD charts as that scales the fundamental to the 0dB line, making it easier to read the relative amplitudes directly from the chart. Harmonic amplitudes are still shown in the units selected.

The AES17 notch is needed if there is a large amount of noise/jitter/distortion around the fundamental, as was the case with your measurement. Multitone doesn't use AES17 by default because it's really not suitable for multitone measurements, but it is fine to use for a single tone. I believe it's the default notch for Audio Precision measurements of THD and THD+N.

Thank you for clarifying. All makes sense. I will update to the latest version.

 

[Deleted member 53587]

after or during all this testing, did you enjoy the music ?

 

[Ralph_Cramden]

Many of the great masterpieces of jazz were recorded on these (RVG's) primitive tape decks. Would they be any better if recorded today?

 

[eliash]

As a fellow with a nice Pioneer RT-909 a-sittin' in the basement also waiting for mechanical rehab (i.e., by me)... umm, yeah, no. Sorry.

More on topic: definitely thinking (i.e., agreeing) that tweaking the bias will help our OP, but his rolloff appeared smooth, and lowering bias tends to produce a hump in the high treble, doesn't it?

True, but it extends the upper freg. limit a bit. After all it depends on your taste, what personal optimum you choose.
Btw., this is similar to tweaking phono carts, there are even more variables than the capacitive load, like tip shape, cantilever design and material...i.e. commerially available styli for a certain cart.

Basically I remember adjusting the bias in direct A-B comparison when recording and monitoring with a Technics RS-1506. It had this 3-step bias selector, which was sometimes not sufficient without finer adjustments inside the machine.
It was always a trade-off between between treble brightness and distortion perception...When using 7.5ips it was crucial to find the right bias. At 15ips, the situation was already relaxed.
When measuring and adjusting the overall frequency response at 7.5ips I tried not to exceed a +/-3dB spec below 16KHz (referenced to low frequencies like 400/1000Hz). At 15ips +/-1.5dB were possible...and no tonal degradation listening to my beloved Beyer DT-880s HP (1st gen from the 80´s!), even recording from CDs...

 

[bennybbbx]

here you can get service manual for free. maybe record and playback head do not fit or bias setting is bad for this tape. here you see how to correct. you need download this pdf . online view did not work https://www.audioservicemanuals.com/a/akai/akai-gx/82313-akai-gx-620-service-manual

 

[John B]

Thanks for doing this. Very cool! My heart flutters whenever I see analog gear on here.

 

[mhardy6647]

My heart flutters whenever I see analog gear on here.

all I can say with regards to your flutter pun is...

wow.

 

[Entropy]

Thanks for the review. No matter how terrible it measures, I still find Reel to reel to be quite charming.

 

[Kef ice 24]

Did you try to demagnetice th heads??

 

[Stephen H]

Did you try to demagnetice th heads??

Indeed. The heads were demagnetised recently before I did the test.

 

[IVX]

the idea: I think the spectrum of demagnetized tape(no bias) on the calibrated R2R(with special test tape), may be used as a reference to adjust the frequency response of uncalibrated R2Rs. Maybe we need to use the same type of reference tape but anyway, don't need to buy a special testing tape is a big deal.

 

[restorer-john]

the idea: I think the spectrum of demagnetized tape(no bias) on the calibrated R2R(with special test tape), may be used as a reference to adjust the frequency response of uncalibrated R2Rs. Maybe we need to use the same type of reference tape but anyway, don't need to buy a special testing tape is a big deal.

What are you talking about??

 

[IVX]

l think if the FFT of Apex 456 tape's noise played with two different R2R is identical, then these 2 R2Rs have the same frequency response. If so, hence you can show me your FFT result with Apex 456 digitized in Australia, let's believe that your R2R has a perfect frequency response. Mine has bad freq.resp and I'm in China, and around me no reliable source for a nice test-tape, but I also have original 456. Is that still funky or can I EQ my R2R with your noise as the reference?