Halcro DM 68 Monoblock Amplifiers

[JonMarsh]

This set of DM68's was acquired on Audiogon via a consignment sale at Definitive Audio in Seattle. As they had no shipping containers for the amplifiers (originals discarded in error?) I had to drive up to Seattle to pick them up. They did have the original manual and test documents from Halcro, from 2004.

Here, after unloading to the living room foyer, one can see the condition of the amplifiers physically is quite good.



I will review some of the known information about the technical design of the Halcro Monoblocks first, then present test data from the two amplifiers here.

[Steve Manning]

Very Nice .....

[wkhanna]

:thud:

drop dead gorgeous

[JonMarsh]

Test data

I had time today to take some test data, but finishing technical research and writing something up about the amplifier technology wasn't possible, due to other personal events with the family.

So lets look at a few things- frequency response, THD versus level, THD versus frequency, and a 50Hz FFT and 19+20kHz CCIM FFT.


Frequency response is consistent for the unbalanced and balanced inputs, but doesn't have the LF roll off on the minimal path input. OTOH, the minimal path input is only 450 ohms input impedance, so that's not something most consumer audio gear will be happy with.

Here's 10Hz to 50kHz...




The biggest thing to gripe about is that the frequency response is down about 1/2 dB at 20Hz. Given that the power supply is very well regulated and has no issues with low frequency output, I doubt this will be audible. It's done to provide some system protection from subsonic frequencies- have you ever seen a tone arm and cartridge combo with a high Q arm and cantilever resonance around 6 Hz without good subsonic filters?

Next up is THD versus level at 1 kHz.



Why is this plot essentially a straight downward sloping line? Because that's what a THD+Noise plot looks like versus output level if the residual is all noise and no THD. This is the case right up to about 200W, then distortion starts to rise, and lets say it "clips" at about 250W. (rated at 225W). By the rating method used by many current power amps (max power at 0.1% THD this would be called a 290W amplifier. This was using a scaling for power from an 8 ohm load, but the test load is actually a high power 7 ohm load, so we're delivering a bit more than indicated.

Here's something interesting: Stereophile measured the DM88, and had poorer results with an 8 ohm or any other load, and in point, the distributor said they thought the amplifier was defective.



Now granted, Stereophile has on loan an older, less sensitive analyzer than the APx555 I own. The 8 ohm data doesn't look bad, but it's hard to tell when it "floors"/bottoms out whether it's an analyzer limitation or something with the power amp. And at heavier loads it definitely has rising distortion above 20W. Note that the load in my measurement is actually 7 ohms, not 8, due to the configuration of the paralleled high power resistors used. So I'm working the amp a little harder than you might think.

Obviously, next thing we could do is a sweep of THD versus frequency. I chose to use a level that is representative of fairly loud playback, 50Watts, while not (in principle) using the signal so high that the noise floor looks "unrealistically" low.



This is a lot like the results with the DM10 preamplifier; it seems that one could use that earlier swept result to predict the overall distortion level at almost any audible frequency.

Some amps like to brag about "First Watt" or first 5 watts, as being quite linear. Here, I'd say it's the first 50, or 100, or 200. Take your pick. Of course, you have a right to expect superlatives for what these cost new.

Let's do some FFT's, see if we can get a bit of a handle on whats going on.

Here's a 50W output signal at 50Hz. I was curious to see at this moderate level how much THD versus how much noise, or even power supply noise might be there.




We can see that there is some HD2, HD3, and HD4-

HD2: -135dB relative to signal output
HD3: -145dB relative to signal output
HD4: -155dB relative to signal output

That the HD4 is at all detectable on the FFT is a testament to the APx555; that it is so low is a testament to the DM68.

There's some 120Hz line ripple also visible, but it is also at the -155dB level relative to the signal output. Guess we can give it a pass on this one.


19kHz + 20kHz CCIM is a difficult test for amplifiers; it will usually point out output stage shortcomings that can't be swamped out by negative feedback. There are two distortion artifacts to be concerned about; 1 kHz difference frequency, and side bands at this difference frequency around the test frequency.

This is what the output of the amplifier looks like with the test signal and the frequency beating- two test sines will "beat" and sum to a peak level that is 2x either sine input.




Here's what the FFT spectra on this test looked like, again with about 20VRMS, but 40V peak.





Residual IM harmonics are in the -115 to -135 dB range; note the side bands around the test frequencies are quite low.


This is more typically what one sees, on a recently tested Bel Canto $5K mono block. Note the difference between the two measurements in X-Axis scaling, db relative versus dbV for the DM68. The Bel Canto scale starts at 0dB relative to tested output (200W Peak at 4 ohms); the DM68 scale is pure dbV, and each individual signal is +20, and the combined peak +26; so the total dynamic range must consider that peak level plus the level below 0 dB.





Also, I can safely say that this is the new Benchmark; the Benchmark AHB2 is a fine amplifier in its size and price class, and in general a strong David to many other "Goliath" amplifiers, but in these case we have a clear winner with the Halcro.

Benchmark AHB2 CCIM IM distortion:

[wkhanna]

hate to sound like a broken record......but.......:thud:

Damn! that is performance i thought not even possible.

what are you using to power the amp....iow, how 'clean' is the source?
.....or does it not even matter due to the amps input power circuit topography?

[JonMarsh]

Damn! that is performance i thought not even possible.

Yeah, that's my reaction, too. I even double checked the load to be sure it was the right impedance and drawing load current, but the way the setup is wired there's no way to get a signal to the AP without the proper load connection, as the amplifier is connected to the load, then the signal connections from the load go to the AP.


The amp is plugged in to a plain old 15A service, (you'll probably laugh) using the power cord that came with the DM10 preamp, as it was a fairly burly standard IEC cable (nothing "aftermarket"), and it was already plugged in to my separate outlet feed. With the built in PFC, capable of running down to 85VAC at rated power, 118VAC local (as measured) obviously shouldn't present any problems. The thing is, with the very high power factor this supply will have, you can get almost double the actual power out of the power line before tripping a breaker, compared with a conventional bridge/capacitor power supply, that only pulls current off at the peak of the AC waveform.

An advantage of this kind of power supply, when done right, is that it keeps the AC line "cleaner" for anything else connected locally, as it pulls power more like a resistor; like an incandescent light bulb, for example.

[Steve Manning]

Pretty slick for stuff built over a decade ago.

Though I know it would not be as good as big brother, I wonder how my MC20 would measure.

[wkhanna]

MC20 Stereophile review: LINK

Review Measurements: LINK

[Steve Manning]

Why thank you sir....

[wkhanna]

you are most welcome.....

[JonMarsh]

JA should have had an AP AUX-0025 switching amp filter on all of those measurements. The AUX-0025 was introduced in 2003, if I recall correctly. Of course, I have one for Class D amp measurements.

[csmielke]

Jon,
They are stunning, like something out of the future, perhaps a basis for the Dark Lord's throne - in black of course!
Chris

[JonMarsh]

Yeah, black might be something pretty special, though it also might be a bit too much Death Star for some folks...

Physically, the thing I like is that the actual footprint is pretty small, no where near something like Boulder 2050 mono blocks, which are 18" wide, 27" deep, and weigh 230 lb each (big copper and iron transformers, admittedly much higher output power (1kW). And a THD versus frequency curve like this doesn't especially impress me... especially at low power, like this measurement was made.



And after an hour of testing, things were barely warm... I'm looking forwards to getting a working build going again, and being able to listen to them (the Ardents are in storage and heading to a local colleague soon...)

[wkhanna]

first things first.....

:dancenana::dancenana:arty::dancenana::dancenana:

huge congratulations on the birth of your first grandchild!
our thoughts & well-wishes are with you all.

regarding the THD versus frequency curve generated by Thomas J. Norton, he states,"Except for a common rise at higher frequencies—to percentage levels that are still insignificant—this is outstanding performance. The waveform of the distortion imposed on a 1kHz sinewave, shown in fig.4. Taken at 200W (!) into 2 ohms, this is dominated by a noisy third-harmonic content. Note that the sensitivity of the measuring equipment had to be set at maximum to take this reading, even at this high power output. Into higher impedances of 4 and 8 ohms, the waveforms (not shown) were barely recognizable, even at a 100W output. This latter result makes it clear that the THD+noise shown in fig.3 is dominated by noise, at least below 1kHz."
Read more at http://www.stereophile.com/content/b...GiYQ1242IFo.99"


Figure 4



he has interpreted this distortion as noise.....what is he missing?

[JonMarsh]

At higher frequencies their isn't as much loop gain available to reduce distortion (loop gain has to be rolled off normally at 6 dB/octave, unless one does a two pole/one zero compensation scheme; it still has to cross the loop gain boundary at 6 dB/octave.

So at lower frequencies, it may only show the intrinsic noise of the system. That isn't the case in that plot, though without knowing the actual signal levels for the "distortion residual" it's hard to know how much actual distortion product is there versus noise. Noise can also be lower frequencies, power supply related to line frequency, or other artifacts, such as leakage from digital circuits. This will show up as something other than just "smooth fuzz" that you might reasonably expect high frequency noise to look like.

[wkhanna]

thank you, Maestro ;x(

[Evil Twin]

A most disturbing standard is being set here... this level of perfection is almost evil in it's disdain for lesser capabilities...