Amp Lab: Demeter HPB1-800D

This Article Was Originally Published On: July 1st, 2014 #Issue 14.

The front panel and chassis are musician-approved matte black, with black input jacks and black knobs. White labeling provides contrast against the black panel, making the labeling easy to read, despite the numerous controls and features available from the front panel. The front panel also sports three silver buttons (EQ bypass, effects loop bypass and mute), which imparts a cool aesthetic. The only splash of color is provided by the red power button. Overall, the external construction of the front panel appears well put together, and well thought out.

Likewise, the back panel features the normal suspects, namely, a voltage selector, IEC connector, SpeakonTM connectors, various jacks for a tuner output, effects loop and various outputs. Overall, the external construction of the back panel also appears well put together, and well thought out, except for one control – the fan switch. The HPB1-800D has a cooling fan that is manually set on or off using a switch on the back panel. The switch for the fan appears to be labeled with a stick-on label (such as from a label maker), and is labeled in all capital letters, in contrast to the neatly screened white lettering in all lower case for the rest of the front and back panel controls. Moreover, the switch position – located off the center line and between the voltage selector and IEC connector – make the inclusion of the fan feel more like an afterthought, or a revision, rather than an integral part of the design.

With reference to Fig. 1, taking a peek under the hood, we see that the device includes three main circuit boards, including a preamp board (seen to the bottom), a power amp board (seen towards the middle/top) and an output board (seen to the top). The preamp board has a few nice surprises up its sleeve. First, note that the preamp board is transformer-isolated from the power amp board by a Jensen transformer (see the white and silver transformer in the lower left hand corner just above the input jacks). This is an interesting idea. Demeter claims that this transformer improves noise performance, regardless of whether the musician uses the preamp only, power amp only or entire amp. The transformer also imparts a bit of color to the signal. Cool! Also, note that the preamp includes a tube circuit built around a 12AX7 preamp tube (an Electro-Harmonix 12AX7 was in our test unit). As an added bonus, another Jensen transformer provides transformer-balanced DI output.

There were a few things that bugged me about the internal construction of this amp. As can be seen in the preamp board, there are a number of contact pads that do not include components soldered in place. This makes me suspect that the circuit board is either reused from another design, or is a generic board that serves multiple products. I think this can be a great efficiency, and can be perfectly well expected. However, with reference to Fig. 2, it appears that some engineering change occurred to the design that the circuit board could not accommodate. It looks like some of the contact pins on the mute switch are serving as the connection points to some capacitors, resistors, and at least one diode. I get that this is a normal occurrence in the evolution of designs. I would even look past this in a budget product. But this is a boutique, high-end design that carries a boutique, high-end price tag. As such, I would have liked to seen this circuitry brought out onto a daughter card or otherwise cleaned up. I personally like things looking as neat and tidy on the inside as they do on the outside.

FIG.2 Internal Wiring

The Tone Stack

Referring to Fig. 3, with the parametric EQ bypassed and all of the tone controls of the preamp section set to “noon,” the HPB1-800D exhibits a mild midrange dip at around 300Hz, and a nice high-end boost relative to the low end, as seen by the black trace. Overall, we measured +/- 6.2dB, 20Hz-20kHz in the “all tone controls at noon” setting. The red trace is our “optimally flat” setting. We were able to achieve +/- 1.6dB 20Hz- 20kHz, and +/- 0.9dB, 80Hz-80kHz. With reference to Fig. 4, a photo illustration demonstrates the position where we recorded our optimally flat setting.

FIG. 3 Optimally Flat vs All Controls at Noon

FIG. 4 Optimally Flat

The preamp tone stack includes controls for Treble, Middle, Bass, and Presence. The HBP1-800D also includes a 2-band fully parametric EQ section that the user can use or bypass. I really like the flexibility afforded by the design of the tone stack surrounding the preamplifier, in combination with the two parametric bands. I also like the way that the tone stack in the preamp provides a range of control that compliments, rather than duplicates, that of the parametric bands.

Referring to Figs. 5-10, sweeps are illustrated for each of the tone controls. The black trace is the “noon” setting, the blue trace is the control fully counterclockwise, and the red trace is the control fully clockwise. With reference to Fig. 5, the Bass sweep demonstrates a low shelving filter with ample low end control and no relevant high band interaction. With reference to Fig. 6, the Mid control exhibits the traditional signs of a passive tone stack. Note that at fully counterclockwise, the greatest mid dip is achieved at about 600Hz. As the Mid control is turned up, a mid boost hump is never really achieved. This is a great idea, because the HPB1-800D provides two fully parametric bands (discussed below). As such, this design provides two different ways to treat the ever- important midrange frequencies, namely the Mid control and the mid band of the parametric section.

FIG. 5 Bass Sweep

FIG. 6 Mid Sweep

With reference to Fig. 7, the Treble sweep shows that there is quite a bit of opportunity for boosting the higher frequencies on this amp. Note, however, in contradiction to the Bass sweep of Fig. 5, the Treble control does exhibit low band interaction. The Treble does not appear to change the shape of the low end, but there is a change in gain. With reference to Fig. 8, the Presence control provides a nice compliment to the Mid control, adjusting frequencies over 600Hz, but with minimal band interaction with lower frequencies.

FIG. 7 Treble Sweep

FIG. 8 Presence Sweep

With reference to Fig. 9, should you find yourself in need of low-end adjustment, you should be able to achieve your goal with the low band parametric. The low band parametric operates from about 30Hz to about 300Hz. With reference to Fig. 10, the mid band of the parametric section extends from just over 300Hz to a bit over 3kHz.

FIG. 9 Low Band Parametric Sweep

FIG. 10 Mid Band Parametric Sweep

Output Power

The HBP-1-800D provides a generous amount of power on tap. Due to tube preamp section, this amp leans more towards the “warm” side, rather than the squeaky clean side. After extensive testing, I concluded that running the tests to about 2.5 % THD+N seemed about right. At this level, there were no signs of hard clipping on the peaks, and the signal appeared upon visual inspection to preserve the shape of the input signal.

At 8 ohms, I measured 547 watts; 2.5% THD+N filtered 20Hz-20kHz; 219 mVrms 1kHz sin Input. However, burst power hit an impressive 614 watts; 2.5% THD+N filtered 20Hz-20kHz; 1kHz CEA 2006.

At 4 ohms, I measured 786 watts; 2.5% THD+N filtered 20Hz-20kHz; 213 mVrms 1kHz sin Input. Burst power hit 998 watts; 2.5% THD+N filtered 20Hz-20kHz; 1kHz CEA 2006. Allowing the distortion level to creep up to 3.5% THD+N (still reasonable, given that this distortion was generated by the tube in the preamp, and was thus not exhibiting significant clipping), I measured about 815 watts, with burst close to 1,000 watts, at 4 ohms.

One issue that occurred during output power testing was that the HBP1-800D has thermal management that shuts the output section off until the circuitry cools down. The onset of thermal can happen rapidly if you hit the amp hard enough. The amp gracefully recovers every time, but as Fig.11 shows, the amp can remain offline for just over 4.5 seconds.

FIG. 11 Thermal Management

As a bass player, I would not be concerned about the thermal management of this amp. The nature of the bass guitar is that the spikes in the bass signal are of relatively short duration, compared to a test analyzer that is capable of cranking out large, continuously running sine waves. Moreover, there can be about a 20dB difference in a typical bass between the peaks and steady state of the sustained notes. As such, even driving the amp hard, it is unlikely that the thermal management will shut down, except maybe if the player is using a lot of compression. In our testing, we deliberately pushed the HBP1- 800D into shut down several times in a row to monitor its performance and to detect any issues, and in each instance, the amp gracefully recovered.

Referring to Fig. 12, the gain structure of the amplifier is extremely linear over a wide range of inputs. In Fig. 12, the onset of clipping occurs where the trace begins dropping off its horizontal path. Referring to Fig. 13, the linearity of the amplifier is further demonstrated. The angled trace shows that the output power of the amp tracks extremely consistently with the input level.

FIG. 12 Gain

FIG. 13 Level

Referring to Fig. 14, the THD+N levels follow a predictable and stable path as a function of output power. Note that the HBP1-800D is capable of THD+N well under 1% for small output levels. However, the bulk of the output power comes at a THD+N level over 1%. For this amp design, that is perfectly fine, because that little bit of distortion is character brought to you courtesy of the preamp tube.

FIG. 14 THD+N Ratio vs Measured Level

To put you all at ease, refer to Fig. 15. This figure demonstrates a scope trace of the output of the HBP1-800D at 1% THD+N. At 1% distortion, this signal shows no signs of hard clipping. Fig. 16 shows a scope trace of the output of the HBP1-800D at 2.5 % THD+N and again, the signal looks clean. Referring to Fig. 17, at 5.7% THD+N, there are now clear signs that the preamp tube is compressing. Note that the lower lobes appear narrower than the upper lobes. Referring to Fig. 18, at 16.4% THD+N, we see nice signs of asymmetric distortion. However, thanks to the design of the tube preamp, even high levels of distortion result in graceful tube compression that should sound fat and present.

FIG. 15 Scope Monitor 1% THD+N

FIG. 16 Scope Monitor 2.5% THD+N

FIG. 17 Scope Monitor 5.7% THD+N

FIG. 18 Scope Monitor 16.4% THD+N

Conclusion

This amp is likely to fit the requirements for many a musician. There is a flexible EQ section, a parametric section that can be bypassed or tuned in when needed, and a host of convenience features, such as an effects loop (controllable from the front panel), a dedicated tuner output, a DI output and line level outputs. I really like the use of Jensen transformers. I am sure other companies make fine transformers, but for decades, I have known that I can trust Jensens to bring pro-level performance. I really like the tube preamp design and the inclusion of a presence control. I also really like the use of high quality op-amps in the analog section of the amp. My only real downside is the attention to those final little details. The fan switch and labeling, and the patchwork re-wiring inside the amp make this feel “fixed,” compared to organic and complete.

Bonus

While we were digging around with the HBP1-800D, we decided to sneak in a bonus look at the Minnie 800D. If you are looking for a power amp, this may just fit your need. As a power amp, the Minnie 800D is one of the most insanely flexible amps that I have come across with regard to mating to a preamp or other input source. I threw signals a the Minnie 800D from a few mVrms to over 10 Vrms, and with a slight tweak of the single gain control, I was able to pull the input signal in clean as a whistle (well under 1% THD+N). Kudos, here.

Without the tube preamp of the HBP1-800D, I held the Minnie 800D to a stricter standard for output power. At 8 ohms, I measured 530 watts; 1% THD+N filtered 20Hz- 20kHz; 930 mVrms 1kHz sin Input. Burst power hit 610 watts; 1% THD+N filtered 20Hz-20kHz; 1kHz CEA 2006. At 4 ohms, I measured 775 watts; 1% THD+N filtered 20Hz-20kHz; 800 mVrms 1kHz sin Input. Burst power hit 970 watts; 1% THD+N filtered 20Hz-20kHz; 1kHz CEA 2006. Allowing the distortion level to creep up to 2% THD+N, I measured about 810 watts, with burst well over 1,000 watts.

In this regard, the gentle compression of the tube preamp of the HBP1-800D helps eek out just a tad more power. Then again, the difference is for all intents and purposes negligible.

As Fig. 19 illustrates, the Minnie 800D also features thermal protection. However, the time to recovery varies from under 1 second to just over 2 seconds. Also, note that the time that the signal remains active can vary from under 0.5 seconds to tripping. Again, I want to stress that this test is conducted under extremely stressful bench conditions to see if the amp has thermal shutdown and whether the amp properly recovers. In each instance, not only did the amp recover, it was able to supply burst power in excess of its steady state maximum output power with no artifacts.

FIG. 19 Minnie 800D Scope Thermal Protection

Referring to Figs. 20-22, the Minnie 800D exhibits stable gain, but note that the drop off is a harder edge, compared to the HBP1-800D. This is due to the gentle compression of the preamp tube in the HBP1-800D. Fig. 21 demonstrates that the Minnie 800D is extremely linear in matching input level changes to output level changes. Fig. 22 shows that the Minnie 800D is extremely clean, with distortion levels well under 1% THD+N until reaching the limits of steady state maximum output power.

FIG. 20 Minnie 800D Gain

FIG. 21 Minnie 800 D Level

FIG. 22 Minnie 800D THD+N Ratio vs Measured Level

1-5 (unacceptable to impeccable)

On-bench

SONIC PROFILE:

Low: Nice and meaty
Mids: Nicely present without being overbearing
Highs: Sweet and clean