Marantz PM-90 Review (Vintage Amplifier)

[tomchr]

Wowowow. It looks like a gold plated circuit PCB

Yellow anyway... ENIG gold would be very, very thin and would look pretty on the board. It's standard practice these days to gold plate the boards, though the boards are usually only plated on the pads - which is where the gold matters. The purpose is that the gold provides a planar and clean (inert) surface to solder to. The gold actually dissolves in the solder joint. So the gold doesn't participate electrically in the connections between the components on the board.

ENIG gold is applied by first electroplating a layer of nickel on the copper of the PCB. The board is then dunked in a soup that displaces the surface nickel and replaces it with gold. The resulting gold layer is 0.1-0.2 µm thick. Silver can be deposited in the same way, but that wouldn't have much of a bling factor as it tarnishes quickly.

Another gold plating options commonly offered by PCB manufacturers is to apply relatively thick (0.1-1.3 µm) gold to the copper. This is commonly done to the fingers on PC expansion cards for example.

PCB traces are typically 35 µm thick ("1 oz copper"), though 70 µm is sometimes touted. So even if thick gold is used, it won't have much impact on the resistance of the PCB traces.

Tom

 

[tomchr]

Here's some details on the PM-90:
https://audio-heritage.jp/MARANTZ/amp/pm-90.html

Wow. That milled internal frame is very sexy.

Tom

 

[Doodski]

Wow. That milled internal frame is very sexy.

Tom

It's zinc if memory serves me correct.

 

[restorer-john]

Wow. That milled internal frame is very sexy.

You should see the same die-cast chassis on the PM-95. The entire chassis and internal plates are all copper plated- both sides.

 

[Doodski]

You should see the same die-cast chassis on the PM-95. The entire chassis and internal plates are all copper plated- both sides.View attachment 154934

Is that a zinc sub-chassis too copper plated?

 

[restorer-john]

Is that a zinc sub-chassis too copper plated?

Everything is. Even the painted steel covers are copper plated under the black paint...

I should do a thread on the PM-95 with lots of pics to make people drool.

 

[Doodski]

Everything is. Even the painted steel covers are copper plated under the black paint...

I should do a thread on the PM-95 with lots of pics to make people drool.

I've only worked on 1 Accuphase and it was built very well but nothing like this newer stuff. It must have been a pretty old model. The solder had bubbles in it, had a flat sheen, was very hard and req'd high heat to melt.

EDIT: OOPs wrong thread. Watching a movie and doing ASR stuff...lol

 

[dualazmak]

Although, talking about Accuphase, I (we) became a little bit out of the scope of this thread, let me just touch on one of their business/corporate strategies.

As I already wrote in my post here, Accuphase is intentionally restricting their export sales less than 30 % of their total business revenue in order to keep existing as a pure Japanese "independent and sustainable" company;
http://www.accuphase.com/company_profile/a2_management_policies_2.pdf

Just for your reference, the price of Accuphase E-800 in Japan is JYen 1,078,000 inclusive of our 10 % sales/consumption tax.

 

[amirm]

But this is a Class A/AB amp...

I'm not suggesting that you have different testing protocols for Class A/AB/G/H vs Class D. But my curiosity drives me to want to scratch the surface a bit deeper on this amp.

Tom

This amp has already gone back to its owner. It came in a massive, massive box so I was happy to send it off.

 

[restorer-john]

This amp has already gone back to its owner. It came in a massive, massive box so I was happy to send it off.

I just got out my PM-95 and my back sure is regretting it! Might do a bit of investigation into its low level mains spuriae as @GXAlan was curious. Content for another thread however.

 

[SmackDaddies]

Great looking unit. I hide my Pioneer unit, but that would look great displayed

 

[GXAlan]

I spent some time with the amp today and here are some subjective thoughts. I'm putting it in writing so that one day when QuantAsylums are shipping again, I can try to see if any measurements can correlate with my experience. So as not to offend, all subjective comments are hidden with the spoiler tag.

Spoiler

- Bass seems light. I wasn’t listening loud as only 60W was being pulled from the outlet, and maybe it’s the cognitive bias of the champagne color. But usually newer = better, not worse. I wasn’t listening at high volumes and was using the JBL 4319 which isn’t a bass monster and has a flatter response than a classic Harman downsloping response.

- Soundstage feels very wide. I get great localization with a nice phantom center image, but it also feels like it is stretched out in a pleasant way. That means the center isn’t 100% precise but it isn’t so wide that you cannot position/visualize the solo singer. I almost would describe it as the comparison between box speakers and planar magnetic speakers in terms of the larger-than-life sound but to a much smaller degree. There may be some phase phenomenon since I felt that there was a moment with regular music that there was a sound almost surround-sound like. It makes me wonder if there is some sort of frequency dependent phase or frequency dependent crosstalk. This was the unexpected finding since I would have expected the less impressive crosstalk measurement to result in a narrow soundstage. All of the science I know of is that we cannot hear phase. So it’s not crosstalk or phase but something else.

Looking back at the measurements, one thing that struck me was how well matched the channels are for SNR balance. They match to one decimal point at 5W and 2 decimal points at full power. That’s better than an AHB2! Even with the poorer performing Class A mode, it's still within 0.3 dB matching. Perhaps this contributes to soundstage? I cannot hear small differences in dB either.

But soundstage, which was not something I would imagine being biased toward (not a dual mono design, not great crosstalk) ended up being the standout feature.

——
The Japanese manual provides some added insights too.

1) Peak power is rated at 310W into 4 ohms, which matches Amir’s CEA-2006 measurements. This is especially impressive because it means the 420W into 2 ohms is probably accurate! They did over-rate the power at 130W and 180W into 4 and 8 ohms as opposed to Amir's 125/175W.

2) The advertised THD of 0.0015% that you see on the various websites is actually only for 10W, 8 ohms. Otherwise it’s just 0.008% “SMPTE” standards and 0.03% when looking at 10Hz-50kHz. If you look at Amir’s 45kHz bandwidth measurements, 0.03% = -70 dB which matches nicely.

3) Amir was really impressed by the RIAA equalization. They advertised +/-0.2 dB.

4) SNR is 111 dB A-weighted.
That’s not as good as Amir’s but maybe I don’t think Amir adds A-weighting which will give better numbers.

5) Amir measured -3dB at 80kHz as opposed to 100 kHz.

I think what's most impressive is
a) Manufacturers used to be very honest with advertised performance
b) Manufacturers used to be very detailed with advertised performance such as distortion at different test conditions.
c) The equipment is 40 years old and literally looks and feels brand new.

 

[Mrgoogle]

This is a review and detailed measurements of the Marantz PM-90 Vintage stereo integrated amplifier. It was purchased by a member from japan and drop shipped to me for testing. It came out in 1991 with a price of 160,000 Yen (about $1,500 in today's exchange rate). I looked online and I can only fine one for US $950 but it is more beat up than this unit.

I must say, this is one nice looking amplifier with a look that would fit in even today:

View attachment 154193

The feel of the controls is excellent other than volume control which I found a bit loose. A flip down door gives you additional control including switching between Class AB and Class A mode of operation! Sadly it requires the unit to be powered down to change so no opportunity for instant switching and comparison.

I was pleasantly surprised to see modern, beefy speaker terminals as well:

View attachment 154194

Even more interesting is the inclusion of XLR balanced inputs for CD.

This amplifier is back breakingly heavy due to use of two massive transformers. It is at the limit of what I could lift and carry up two flight of stairs for testing. Sitting here, my back aches just thinking about the reverse trip to pack the darn thing! Speaking of packing, it was immaculately double boxed by Hifi-Do with zero impact to it despite the long journey. Apparently it even came with one year warranty although you have to ship it back to Japan for that.

Speaking of Japan, the unit is made for that market and hence 100 volt rating as opposed to 120 volts we have here in US. If it were mine, I would go ahead and use it at 120 but being that of someone else's and hard to replace, I powered it by both my Lab AC generator and a Variac set to 100 volts. This may have caused some measurement issues as I noted below.

Marantz PM-90 Amplifier Measurements
There is a lot here to test due to dual mode of operation and dual type of input. Let's start with Class A operation with RCA and then XLR:

View attachment 154195

View attachment 154196

Unfortunately our measurements are polluted by high mains noise and harmonics. I tried every which way but could not reduce that impact much. Whether this is due design of the amplifier or longer AC mains connection due to use of lab AC gen or Variac, I can't tell. As is SINAD doesn't appear to be better than modern AVRs with Marantz on their front plate:

View attachment 154197

However, if we look at the FFT, we see only two harmonics and then it disappears into noise. Here is the performance of Marantz SR7015 as a reference:

Noise from power supply is lower there which aids it in getting similar SINAD.

Switching to Class AB mode has no penalty:
View attachment 154198

If anything distortion harmonics go a bit lower although that was not reliable.

To see if there is a difference at all between Class AB and Class A, I turned on the residual mode of Audio Precision in the scope display. This shows us the difference between input and output waveforms. If there is zero crossing error due to class AB operation, we should see more difference around that part of the waveform but we don't:

View attachment 154200

What is there is mostly power supply noise. I changed frequencies, power level down to 1 watt and still saw nothing that would correlate with zero crossing distortion which class A is supposed to solve. Feedback and class AB operation seem to have cured that ailment leaving little room for Class A to do better.

Here is the comparison between Class AB and Class A again with respect to signal to noise ratio:

View attachment 154201

View attachment 154202

Class AB pulls ahead in full power mode because it can create more power. At baseline 5 watts they are similar likely limited by power supply noise.

Crosstalk showed a tiny improvement in favor of class A but it is not material:
View attachment 154203

Bandwidth is good but not exceptional:

View attachment 154204

Notice how amplifier gain changes as you switch modes so for any listening tests, you may have to compensate for that.

Class AB of course has tons more power:

View attachment 154205

View attachment 154206

Overall shape of the graph is average. Wanting to see the impact of the pre-amp vs power amp, I separated the two and fed the unit with my analyzer direct using RCA:

View attachment 154207

As expected, my analyzer is cleaner but impact is on distortion, not noise which is a bit surprising. As is one channel being worse than the other. SINAD reaches almost 100 dB now near full power, good channel, which is very nice.

There is a lot of headroom when using class AB vs Class A:

View attachment 154208

View attachment 154209

Testing amplifier linearity vs frequency we get:
View attachment 154212

View attachment 154213

Class A power limits more at 20 Hz than Class AB. Both show very predictable and smooth responses which is great.

Multitone tests shows slightly lower distortion at higher frequencies in favor of Class A:

View attachment 154210

Effectiveness of feedback lowers at higher frequencies and likely this is the reason class A does better there.

Marantz PM-90 Phono Input Measurements
It was hard to figure out a fair way to measure the performance of its phono stage to compare to stand-alone devices. At first I used tape output but that didn't produce great response. Best was from pre-amp tap to power amp:
View attachment 154215

Sadly the mains noise is there and dwarfs SINAD. There is no sign of harmonic distortion so that is very good. Same situation exists for MC mode:

View attachment 154216

RIAA equalization is probably the best I have measured:
View attachment 154217

That story continues into headroom measurements:

View attachment 154220

You get almost 12 volts of output before it starts to clip! So pops and clicks should be much less noticeable.

Conclusions
The PM-90 does NOT show its age. It produces above average performance compared to today's amplifiers. It does so while running very cool and comfortable. Not once did it get warm to touch or go into protection. Push it hard and it just pulls back on the output but keeps going. Clearly this is a very robust implementation. Can it compete with state of the art amplification today in the form of Benchmark THX and Hypex/Purifi Class D amps? No. There is real innovation there to push down distortion and noise. But we also have a lot of terrible amplifiers today which the PM-90 easily outperforms and does it with so much style.

There is nothing to apologize for and a lot to praise for Marantz PM-90. I hope the company goes back to these roots rather than horrible looks and performance of much of their AV line.

On class A mode, I see no reason to recommend it and many reasons to avoid it. If you run out of power there, you will certainly create audible distortion. Whereas any benefit it might have is literally lost in the noise.

I am happy to recommend the Marantz PM-90.

-----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

any DC offset ?! At maximum volume with input grounded , how much DC at output ?!

 

[musicforcities]

The Totenham Court Road didn't sell vacs as well as audio I promise you and the Edgeware Road was often component shops at one time (Help me please Serge and Keith )

By the mid 1990s to early 2000s when I lived in London, much of Totenham court Road had become a bit seedy and dodgy electronic stores has displaced most of the hifi shops.

 

[Dougey_Jones]

This product is similar in look and build to my vintage Marantz SR-96 AVR, which was Marantz first THX certified product, and AFAIK the first AVR to carry the THX cert. I still use it sometimes purely as an amplifier with the Pre/Amp jumpers removed and with an SMSL SU-9 plugged directly into the Amp inputs and will say that the sound is very impressive.

 

[deprogrammed]

Wow! Japan has two power system networks! I’m shocked!

Japan’s bifurcated power system is a holdover from the 19th century, when early power ventures were small in scale and highly localized. In Tokyo, entrepreneurs who were already providing electric lighting in a limited area, using direct current, decided to expand their business by importing high-voltage alternating-current generators from Germany. The German equipment, purchased from the company that became AEG, worked on a frequency of 50 Hz. Meanwhile, the local power providers in Osaka brought in 60 Hz generators from the United States, supplied by the predecessor of General Electric Company (cut and paste from web)

 

[ousi]

Actually, the one I really wanna see is the last Sony ES integrated - the elusive TA-FA777ES. It's supposed to be the last good Sony amplification, no not today's junk. Maybe the TA-A1ES was an exception but it's discontinued already.

 

[GXAlan]

Actually, the one I really wanna see is the last Sony ES integrated - the elusive TA-FA777ES. It's supposed to be the last good Sony amplification, no not today's junk. Maybe the TA-A1ES was an exception but it's discontinued already.

That seems rarer than the TA-E1 and TA-N1! There are two sets available in eBay.

I am really curious about today’s Denon PMA SX1 Limited which is only available in Japan. That it two steps up from the PMA-A110 that we have here.

The old POA-S1 monoblock is interesting. I remember seeing one in the USA at a “The Good Guys” and in my internet sleuthing, I have only seen a US voltage model sold once.

 

[ousi]

That seems rarer than the TA-E1 and TA-N1! There are two sets available in eBay.

I am really curious about today’s Denon PMA SX1 Limited which is only available in Japan. That it two steps up from the PMA-A110 that we have here.

The old POA-S1 monoblock is interesting. I remember seeing one in the USA at a “The Good Guys” and in my internet sleuthing, I have only seen a US voltage model sold once.

Yea, those were legendary stuff. The TA-E1 and TA-N1 they list on eBay is ridiculously priced though. The TA-E90 and TA-N90 might work but it's not top of the line. If one can get the TA-NR1/TA-NR10 would be awesome (the TA-NR1 did come to US).

I've heard good things about the SX1 Limited. I did audition the SX1 (not limited) in Tokyo back in 2019 and it sounds pretty wonderful and neutral. Thought for a long time if I should just get it home with me to US, but the voltage conversion is always the wildcard.

Japanese companies kept their best in JDM unfortunately...

 

[dualazmak]

Hello friends,

Sony TA-A1ES has been a really long-selling quasi class-A integrated amp in audio scene, and it is still manufactured and widely available at least in Japan. It has been occupying a unique position in the market for long period after its launch in 2013. I hope your web browser would properly translate this comprehensive page on TA-A1ES and also this spec page into English.

In December last year, I became much interested in SONY's description of;
"The TA-A1ES uses a current feedback type power amplifier to achieve a sound with excellent focus of the sound image without phase rotation up to the ultra-high range. "
(This looks somewhat similar to Accuphase's description for A-36 as shared here.)
and,
"In addition, by applying sufficient bias, the heat generation of the transistor becomes almost constant, there is no sudden temperature change at the semiconductor chip level, phase rotation in the ultra-high range is suppressed, you can enjoy a sound with good focus, and the emitter resistance is deleted. "

In January this year, therefore, I intensively evaluated TA-A1ES in my multichannel multi-driver multi-way multi-amplifier project, as shared in my posts #307 through #309 on the thread. Then I finally decided to use TA-A1ES to directly and dedicatedly drive Be-tweeters and metal horn super tweeters in my multichannel multi-way system as sharted in my post #311 there.

In my latest system configuration, TA-A1ES is dedicatedly driving Be-tweeters, and Yamaha A-S301 is dedicatedly driving metal horn super tweeters Fostex T925A.

I found and confirmed the quasi class-A operation of TA-A1ES is really nice in driving high Fq efficient SP drivers in multi-driver multi-way system, even though we should be careful about its physical layout since the top cover would be rather warm in quasi class-A configuration.