Home  Donate  New  Search  Gallery  Reviews  How-To  Books  Links  Workshops  About  Contact

Hitachi HMA-8500 Mk II
100 WPC MOSFET Stereo Amplifier
(1984-1986)
© 2014 KenRockwell.com. All rights reserved.

Intro   Specifications   Measurements   Usage   Recommendations

Please help KenRockwell..com

Hitachi HMA-8500 Mk II

Hitachi HMA-8500 Mk II (rated 100 watts per channel into 8Ω, 35.2 lbs./16 kg, measured 91 watts idle power draw, about $300 used). enlarge. I got this mine at this link directly to them at eBay (see How to Win at eBay).

This free website's biggest source of support is when you use those or any of these links to approved sources when you get anything, regardless of the country in which you live — but I receive nothing for my efforts if you take the chance of buying elsewhere. I get no government hand-outs and run no pledge drives to support my research, so please always use any of these links to approved sources for the best prices, service and selection whenever you get anything. Thanks for helping me help you! Ken.

 

Hitachi HMA-8500 Mk II, rear

Rear, Hitachi HMA-8500 Mk II. enlarge.

 

Hitachi HMA-8500 Mk II, inside

inside the Hitachi HMA-8500 Mk II. enlarge.

 

September 2014    Audio reviews    all reviews

 

Introduction         top

Intro   Specifications   Measurements   Usage   Recommendations

Adorama pays top dollar for your used gear.

Amazon

B&H Photo - Video - Pro Audio

The Hitachi HMA-8500 Mk II is a high quality 100 watts-per-channel MOSFET class AB stereo power amplifier. MOSFETs are solid-state devices that work on principals more similar to vacuum tubes than bipolar transistors — and it sounds great!

I've always loved my Hafler DH-200 that I built from a kit in 1981, and when I learned that it uses Hitachi MOSFETs — the same as used in this amplifier — I figured that along with the awesome big meters, that Hitachi might know a thing or two about making a good amplifier, so I got this one to review.

It's a real dual-mono deign. It even has two power transformers as well as two sets of power supply rectifiers and capacitors. Each transformer has it's own input voltage selector!

This is a 2014 test of a 1980s amplifier.

 

Circuit Design

It's a traditional class AB linear power amplifier in a perforated sheet metal steel case.

It runs hot; it's biased for class A most of the time. Hitachi discovered a source of distortion was varying capacitance of the output devices near turn-off, so by biasing them more towards class A they eliminated this extra source of distortion

The power rails run at ± 58 VDC.

There are two 6800 µF 63 VDC caps and one 1 µF 250v Mylar per rail per channel, total 12 caps and power supply four rails (each channel has independent supplies).

There are two pairs of 2SK134 and 2SJ49 in parallel for each channel.

The meters are fed from the headphone feed. The headphones are fed from the speaker rails (before the speaker relays) through 333-1/3 ohm series resistors.

The protection relay comes before the headphone takeoff. The protection relay waits 5-10 seconds at turn-on, and also trips if there is too mu DC at the outputs.

The meters use a semi-peak half wave detector and drive the needles which have VU ballistics.

Meter bulbs are 14 V, 80 mA, 4mm diameter incandescent lamps with leads. I had to replace mine, and I attempted to match them by DC resistance. Even after doing that, I measured 12.59 V and 12.83 V across the two left bulbs, and 12.78 and 12.96 V across the two right bulbs. As they age, the voltage differential may increase, and as either voltage gets higher, that bulb burns out faster. At 13 V, a 14 V bulb will last 14 times longer than it will at 14 V. Bulb life varies inversely as the 12th power of voltage.

 

History

1977: The HMA-8300 (200 WPC class G) listed at $750.

1978: The HMA-7500, the worlds first MOSFET power amp with 75 WPC listed at $500; the HMA-8300 (200 WPC) went up to $800.

1980: The HMA-7500 MK II (still 75 WPC MOSFET) is listed at $550, and the HMA-6500 listed at $329.95.

1982: The 7500 MK II goes up to $570.

1983: The first HMA-8500 (not Mk II) is 100 WPC MOSFET, 33-1/8 lb. and lists at $700.

1984: This HMA-8500 Mk II debuts with 100 WPC into 8Ω or 4Ω, listed as "Sup. B class," 34-7/8 lb, listing at $650.

 

Specifications         top

Intro   Specifications   Measurements   Usage   Recommendations

 

Output Power

Stereo

100 watts per channel.

20~20,000 Hz at 0.005% THD or

1 kHz at 4Ω or 8Ω at 0.005% THD.

130 WPC at 0.5% IEC.

150 W 8Ω dynamic power

 

Mono (one channel only)

200 watts into 8Ω.

 

Power Bandwidth

5 ~ 100,000 Hz at 50 watts at 0.05% THD.

 

Audio Input

Two stereo pairs of 1/4" RCA jacks.

One is DC coupled, the other AC coupled through an 0.68 µF capacitor.

47 kΩ.

1 V RMS.

 

SMPTE Intermodulation Distortion

0.005% at 50 W.

 

Frequency Response into 8Ω

0.5 ~ 100,000 Hz +0, -1 dB.

 

Crosstalk

> 75 dB at 20 Hz.

> 85 dB at 1 kHz.

> 65 dB at 20 kHz.

 

Signal to Noise Ratio

118 dB A-weighted, referred to 100 watts.

This means < -82.2 dBW or < 219 µV or < -73.2 dBV noise output.

 

Damping Factor

60 into 8Ω, 20~20,000 Hz.

 

Power Input

120, 220 or 240 VAC, 50-60 Hz.

Rated 260 W at 10 WPC, 450W/550 VA at 33-1/3 WPC, 750W at 100 WPC.

Actual power consumption.

 

Quality

Made in Japan.

 

Size

165 x 435 x 320 mm HWD.

6.5 x 17.1 x 12.6 inches HWD.

 

Weight

35.25 pounds (16 kg).

 

Price, USA

1984: $650 list price (equivalent to $1,500 today with inflation).

1985: Still lists at $650.

1986: Still at $650.

1987: gone.

2014: about $300 used.

 

Measurements         top

Intro   Specifications   Measurements   Usage   Recommendations

Gain   Input Levels   Channel Balance   Power Output

Damping Factors   Noise   SNR   DC Offset

Frequency Response   Square Waves   THD    DFD

Headphone Output   Meters

Actual Power Consumption   Mono Operation

 

These measurements were made with a $50,000 Rohde & Schwarz UPL laboratory analyzer. The traces from the Rohde & Schwarz UPL laboratory analyzer are color coded for the Left Channel and for the Right Channel. When they don't lie on top of each other, it's due to channel imbalance. When they do lie on top of each other, the trace turns blue.

Unless otherwise specified, all measurements are RMS, at 1 watt continuous output per channel at 1 kHz with an 8 Ω load, both channels driven with a 120 VAC supply line.

 

Gain     measurements        top

 

Stereo

28.93 dB left, 29.00 dB right.

 

BTL (Bridged mono) Mode

In BTL mode, gain is 34.97 dB.

 

Input Levels     measurements        top

 

Stereo, 8Ω

101 mV input for 1 W output into 8Ω.

318 mV input for 10 W output into 8Ω.

1.01 V for 100 W output at 0.0003% THD into 8Ω.

1.086 V for 116 W output at 0.1% THD into 8Ω.

1.26 V for 153 W 1 mS burst power at 1 burst per second. Meters read 80 W peak with 10 mS bursts at 153 W.

 

Stereo, 4Ω

71.3 mV input for 1 W output into 4Ω.

713.3 mV input for 100 W output into 4Ω.

845 mV input for 135 W output at 0.1% THD into 4Ω.

890 mV for 141 W in 10 mS bursts at a rate of 1 burst per second. Meters read 80 W peak.

 

BTL (bridged mono)

50.4 mV input for 1 W output into 8Ω.

809 mV input for 251 W output into 8Ω at 0.1% THD at 113.2 or 120 VAC.

 

Stereo, unloaded

0.0003% THD at 2.8 V out.

0.00033% THD at 36.8 V out.

0.0004% THD at 37.5 V out, 1.327 V in.

0.0005% THD at 38 V out, 1.35 V in.

0.1% THD at 38.5 V out, 1.357 V in at 118.2 VAC.

1% THD at 1.41 V in (118.4 VAC) 39.6 V out.

10% THD at 44.2 V out, 1.77 V or 118.4 VAC in.

 

Channel Balance     measurements        top

The right channel has 0.0724 dB more gain than the left channel.

 

Power Output     measurements        top

All are continuous RMS measurements at 0.1% THD at 1 kHz into 8Ω at 120 VAC unless otherwise noted:

 

Stereo 8Ω

Hitachi HMA-8500 Mk II power output

HMA-8500 Mk II at clipping into 8Ω at 1 kHz.

116 watts per channel, both channels driven.

 

Burst Power

Burst power at the point of visible clipping on an oscilloscope for 10 cycle (10ms) bursts at 1 kHz is 153 watts RMS.

 

Stereo 4Ω

Into 4 Ω, I measure 135 W at 0.1% THD with 845 mV input.

Into 4 Ω, burst power at the point of visible clipping on an oscilloscope for 10 cycle (10ms) bursts at 1 kHz is 141 watts RMS with a 120.6 VAC power line input.

 

BTL (mono)

251 W continuous into 8Ω with 113 or 120 VAC power supply, one channel.

 

Damping Factors     measurements        top

Stereo
Output Source Impedance
4Ω Damping Factor
8Ω Damping Factor
50 Hz
68.4 mΩ
58.5
117
1 kHz
68.1 mΩ
58.8
117.5
20 kHz
76.6 mΩ
52.2
104

 

Output Noise    measurements        top

Hitachi HMA-8500 Mk II output noise

HMA-8500 Mk II noise spectral analysis.

 

A Weighted
Unweighted
Left
-88.1 dBV A
-82.9 dBV
Right
-88.9 dBV A
-83.0 dBV

 

Signal to Noise Ratio (SNR)     measurements        top

97.53 dB A-weighted SNR (15.9 effective bits) referred to 1 watt into 8Ω (2.83 V or +9.031 dBV).

117.53 dB A-weighted SNR (19.2 effective bits) referred to 100 watts into 8Ω (28.28 V or +29.031 dBV).

118.17 dB A-weighted SNR (19.3 effective bits) referred to 116 watts into 8Ω (30.46 V or +29.657 dBV).

 

Output DC Offset     measurements        top

+5.7 mV left, +5.9 mV right.

 

Frequency Response +0, -0.02 dB     measurements        top

Hitachi HMA-8500 Mk II frequency response

HMA-8500 Mk II frequency response, DC input.

Perfect. Let's expand the vertical scale:

Hitachi HMA-8500 Mk II frequency response

HMA-8500 Mk II frequency response, DC input, expanded scale.

Still perfect except for the slight channel imbalance.

Hitachi HMA-8500 Mk II frequency response

HMA-8500 Mk II frequency response, AC input, expanded scale.

The condenser (AC) inputs are down 0.2 dB at 20 Hz and 0.85 dB at 10 Hz. This is swell.

Let's expand the vertical scale even more:

Hitachi HMA-8500 Mk II frequency response

HMA-8500 Mk II frequency response, DC inputs, greatly expanded scale.

+0.00, -0.02 dB from less than 10 Hz to 20,000 Hz. Perfect.

The HMA-8500 Mk II has two inputs: a DC-coupled input and an AC-coupled ("Condenser") input. Here are their infrasonic frequency responses:

Hitachi HMA-8500 Mk II infrasonic frequency response

HMA-8500 Mk II infrasonic frequency response, 0.1 - 20,000 Hz. Left curve: DC input. Right curve: Capacitor-Coupled input.

The DC input is -3 dB at 0.27 Hz, while the regular capacitor-coupled input is -3 dB at 4.7 Hz. Obviously the servo circuit that nulls eliminates DC also removes some infrasonics.

Here it is with a larger vertical scale to see the infrasonic attenuation of the regular input:

Hitachi HMA-8500 Mk II infrasonic frequency response

HMA-8500 Mk II infrasonic frequency response, 0.1 - 20,000 Hz. Left curve: DC input. Right curve: Capacitor-Coupled input.

Hitachi specifies -12 dB at 1 Hz for the AC-coupled input, which is exactly what I measure.

Let's look at the ultrasonic response to 110 kHz:

Hitachi HMA-8500 Mk II ultrasonic frequency response

HMA-8500 Mk II ultrasonic frequency response, 1,000 - 110,000 Hz.

Golly, flat to beyond 110 kHz. Let's expand the scale:

Hitachi HMA-8500 Mk II ultrasonic frequency response

HMA-8500 Mk II ultrasonic frequency response, 1,000 - 110,000 Hz.

Down only 0.4 dB at 100 kHz. Good enough for my ears, and exactly as Hitachi specifies.

This is driving an 8Ω resistor, which doesn't count in the real world. Let's see what it does driving a real loudspeaker:

Hitachi HMA-8500 Mk II frequency response driving a B&W 805 loudspeaker

HMA-8500 Mk II driving a pair of B&W Matrix 805 loudspeakers. 0 dB is the level when driving an 8Ω resistor.

Marvelous, there's less than ±0.05 dB variation from flat, the rest is all overall level variations from channel imbalance.

 

Square Waves     measurements        top

Sony SCD-X777ES

50 Hz square wave, AC input. (HP 209A, TEK TDS3052.)

 

Sony SCD-X777ES

50 Hz square wave, DC input. (HP 209A, TEK TDS3052.)

 

Sony SCD-X777ES

1 kHz square wave, DC input. (HP 209A, TEK TDS3052.)

 

Sony SCD-X777ES

10 kHz square wave, DC input. (HP 209A, TEK TDS3052.)

These are perfect. The square wave is supposed to do that when AC coupled.

 

THD: 0.0002% (-114 dB)     measurements        top

The Hitachi HMA-8500 MK II has the lowest distortion of any power amplifier I've ever measured. Only the Sony TA-N77ES comes close.

All these THD measurements are measuring pure distortion components only (THD only), not noise and distortion (THD+N). THD+N numbers would be worse than THD alone shown here.

 

THD versus output power

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II THD versus output power.

 

Harmonic Distortion at 10 milliwatts

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II THD at 10 mW.

At only 10 milliwatts it has less THD than most amplifiers at any level. Note the expanded scale at the bottom at 0.0002%.

The increased THD at low frequencies is actually the analyzer reading hum at the output.

 

Harmonic Distortion at 1 watt

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II THD at 1 W.

This is nothing, Note the expanded scale at the bottom to 0.0001%, far less than the residual of most analyzers. The apparent increase at the low frequencies is again actually hum — and it's still completely negligible.

 

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II THD at 1 W into 4 Ω.

Distortion is a little higher into 4Ω, but not by much.

 

Harmonic Distortion at 100 watts

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II THD at 100 W.

Note the expanded scale at the bottom at 0.0002%. This is the lowest I've measured as of August 2014.

 

Harmonic Distortion Components at 1 watt

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II harmonic distortion content at 1 W.

This is nothing; less than most analyzer's limits.

 

Harmonic Distortion Components at 100 watts

Hitachi HMA-8500 Mk II THD

HMA-8500 Mk II harmonic distortion content at 100 W.

This is still nothing, and not how the only distortion components are second and third harmonics.

 

19+20 kHz Difference-Frequency Distortion (DFD)

per DIN IEC 268-3 or 118:

HMA-8500 Mk II DFD

DFD at 1 W RMS total output.

This is superb. Only the Sony TA-N77ES measures better here.

HMA-8500 Mk II DFD

DFD at 10 W RMS total output.

 

HMA-8500 Mk II DFD

DFD at 50 W RMS total output.

Oddly the right channel is worse than the left channel at high power at high frequencies, but don't worry, since not only is this still superb performance, if you ever actually used this much output at 19 kHz, you'll fry your tweeters instantly.

 

Headphone Output     measurements        top

The headphone output is simply fed from the amplifier's outputs via a resistor.

More precisely, each headphone channel is fed from the output bus via a relay and then three 1kΩ 1/2 watt resistors in parallel (333 Ω effective). (The speakers are fed from this bus via other relays.)

The headphone output is active regardless of the settings of the speaker switches, but is muted as the amp turns on.

Even though Hitachi didn't bother to provide a dedicated headphone output amplifier, they did include a few relays so that if you run the HMA-8500 Mk II in BTL (bridged mono) mode the headphone output will revert to proper mono, with both channels in-phase.

The HMA-8500 Mk II is so quiet that there is no noise audible with sensitive 58Ω headphones like the Senal SMH-1000, and of course packs enough power to melt just about any headphones if you try.

It should put out 25 V into 600 Ω (over 1 watt or +30 dBm).

It should put out 3.5 V into 32 Ω (380 mW or +25 dBm).

 

Meters     measurements        top

Hitachi HMA-8500 Mk II Meter Scale

Hitachi HMA-8500 Mk II Meter.

The meters, the most outstanding thing about this amplifier, are big and legible.

The needles move behind the clear plastic scales. The scales and needles are lit by four 14 V, 80 mA, 4mm diameter incandescent lamps with leads, held in by greenish silicon boots, which give the meters a light green glow. Each side uses two bulbs in series connected across 26VDC.

The meters have VU ballistics and are fed from a half-wave quasi-peak detector. They read about half way between average and true peak power as read on the Technics SH-9020. Funny thing with the VU ballistics is that even if the Hitachi reads the peak well, its needles get there about a third of a second behind the SH-9020.

The meters read the voltage level of the output bus, as fed from the headphone relay.

The meters are more guides than precision instruments. They read only the positive half of the waveform, and they have plenty of overshoot.

Like the meters used by the Krell in Forbidden Planet, they barely move at all when played softly or at reasonable levels. This is why these big meters were so short lived: once people realized that they rarely used any power at all, they realized they either never needed big amplifiers, or that by just turning it down a few clicks even a small amplifier is all they would ever need.

 

Meter Accuracy

Actual output RMS
Meters Read
0
0
10 mW
10 mW
100 mW
95 mW
1 W
0.92 W / -21 dB
10 W
9 W
100 W
97 W
153 W for 10 mS
80 W
120 W
115 W
150 W
150 W
200 W
190 W

 

Frequency Response

From DC input, read at meters:

-10 dB at 1.0 Hz

-3 dB at 4.2 Hz

-1 dB at 8.5 Hz

Flat 20 - 32,000 Hz

-1 dB at 90,000 Hz

 

Burst Response

Measured at 100 W at 60 bpm or 1 burst per second. The overshoots were greatest at 30 BPM (1 burst every 2 seconds).

Burst duration
Burst Duty Cycle
Metered Percent Power
Metered Decibels
1,000 mS
100%
100%
0.0 dB
500 mS
50%
145%
+1.7 dB
200 mS
20%
170%
+2.1 dB
100 mS
10%
150%
+1.7 dB
50 mS
5%
120%
+1 dB
20 mS
2%
80%
-1 dB
10 mS
1%
50%
-3 dB
5 mS
0.5%
20%
-7 dB
2 mS
0.2%
2%
-17 dB
1 mS
0.1%
0.1%
-30 dB

Thus it's a roughly peak-reading half-wave meter and it has VU-style display ballistics. It's similar to the Sony TA-N77ES, except that the Sony has much quicker peak-hold capture, and much poorer low-frequency response.

This roughly 10 mS peak response is about the same as our ear's perception of clipping distortion, in other words, shorter periods of clipping usually aren't noticed (even if your tweeters will feel it).

With 10 mS bursts at 145W, the meters read 80W.

 

Measured Power Consumption     measurements        top

The HMA-8500 Mk II runs relatively hot. Keep it well ventilated.

Hitachi says leave at least 25 cm (10") above it for ventilation.

Even well ventilated, the sticker warns not to touch the top cover, and it does get hot enough that I wouldn't want to keep my hand on it too long. This is normal.

Output per channel, both driven
Power Consumption
Efficiency
Power off
zero.
n/a
Idle, cold
86 W
0%
Idle
91 W at 119.8 VAC
0%
10 mW
92 W at 120.2 VAC
0.022%
100 mW
95 W at 120.2 VAC
0.21%
1 W
111 W at 119.6 VAC
1.8%
10 W
179 W at 119.1 VAC
11.1%
100 W
407 W at 119 VAC
49.1%
116 W
419 W (4.34 A) at 120 VAC
55.3%
135 W (4Ω)
586 W at 120 VAC
46%

 

BTL (mono) Operation    measurements        top

BTL means bridged transformerless, or mono operation.

In BTL mode, gain is 34.97 dB.

50.4 mV gives 1 W output into 8Ω.

809 mV gives 251 W output into 8Ω at 0.1% THD.

I didn't test burst power in BTL.

HMA-8500 Mk II frequency response, mono mode

HMA-8500 Mk II frequency response driving 8Ω in MONO mode (DC and AC-coupled inputs shown).

Frequency response is unchanged.

THD runs around 0.001% at 1 W.

In BTL mode, both meters read the same, and set themselves to the 4Ω setting, meaning each reads properly for 8Ω when driven by the BTL amplifier.

The dual meter scales are confusing. Ignore the BTL scale except in BTL (Mono) mode. You read 8Ω and 4Ω speakers off the same lower scale; the meter changes its calibration internally as you set the 4Ω/8Ω switch.

 

Usage         top

Intro   Specifications   Measurements   Usage   Recommendations   More

The amp's protection relays don't click on for 5-10 seconds after turn-on. While the protection relay is cut-off for any reason, the meter range LED blinks.

The speaker outputs look like real 3/4" banana jacks, but they aren't! They are just plastic knobs trying to look like bananas. They are at slightly more than 3/4," and even after you muscle a standard dual plug into them, there';s no sound because they have no metal in them! The red and black caps are just handles to screw-in wire. Pooh.

This amp gets hot; let it sit by itself and don't stack other components on top of it.

BTL (mono) mode uses only the LEFT input and the two red B speaker terminals.

Protection trips the speaker relays if the amp detects either that it's safe-operating area has been exceeded, or if there's too much DC on the speaker outputs. Because of this, the amp trips the relays at 100 W into 8 Ω at sine-wave frequencies of less than 2.5 Hz.

DC detect sums channels, may not work in BTL

 

Recommendations         top

Intro   Specifications   Measurements   Usage   Recommendations

The HMA-8500 Mk II measures extraordinarily well and sounds great.

What the measurements tell us when looking at 4Ω performance is that, as many Japanese amplifiers, this is a lab queen. That means it's bred to test well, and less so to drive real loudspeakers. Oddly the little Hafler DH-120 can belt out more peak power into 4Ω than this much larger amplifier which uses twice as many of the same output devices!

It doesn't put out much more power at 4Ω than 8Ω, and its 4Ω output power doesn't vary much with power supply input voltage.

This HMA-8500 Mk II sounds wonderful, even with 4Ω speakers like the B&W 802, but for maximum power, an American amplifier will deliver much more into 4Ω.

This amplifier works great into 4Ω, but it's much happier at 8Ω.

I love this thing, even with my 4Ω speakers. It looks great with its big meters; I just realize that it really only puts out about 140W burst power into 4Ω or 8Ω.

Tell yourself it's a Class A amplifier, which most of the time it is. That's why it runs so warm.

 

Help me help you         top

I support my growing family through this website, as crazy as it might seem.

The biggest help is when you use any of these links when you get anything, regardless of the country in which you live. It costs you nothing, and is this site's, and thus my family's, biggest source of support. These places have the best prices and service, which is why I've used them since before this website existed. I recommend them all personally.

If you find this page as helpful as a book you might have had to buy or a workshop you may have had to take, feel free to help me continue helping everyone.

If you've gotten your gear through one of my links or helped otherwise, you're family. It's great people like you who allow me to keep adding to this site full-time. Thanks!

If you haven't helped yet, please do, and consider helping me with a gift of $5.00.

As this page is copyrighted and formally registered, it is unlawful to make copies, especially in the form of printouts for personal use. If you wish to make a printout for personal use, you are granted one-time permission only if you PayPal me $5.00 per printout or part thereof. Thank you!

 

Thanks for reading!

 

 

Mr. & Mrs. Ken Rockwell, Ryan and Katie.

 

Home  Donate  New  Search  Gallery  Reviews  How-To  Books  Links  Workshops  About  Contact

August 2014