DIY Loudspeakers: HOME  INDEX  UPDATES  RESPONSE  WHAT'S NEW

Amps for TQWT and DTQWT speakers
Copyright 2022 © Troels Gravesen
Frequently up-dated with new experiences.

FAQ on bi-amping

Over the last four decades I've had an endless list of amplifiers, mostly diy stuff based on designs published in magazines and mostly class AB solid state. Linsley Hood's 10 watt class A amp and Nelson Pass' 20 watt class A amps may ring a bell for those old enough. Next some Erno Borbely mos-fet amps were built before "The End" appeared in Danish "High Fidelity" magazine. This amp was developed into the LC Audio range of products and I built most of them. After that I started buying amplifiers and Luxman MQ70 (EL34 valves) was my main amplifier for several years until someone brought in some 20 watt PSE 300B mono blocks and my road to aural nirvana took quite a turn as can be seen from my gear pages. Later I've acquired some solid state amps again, the Jungson JA88D-09, low-feedback, partly class A amps to run my Jenzen speakers. Even my 40 watts SET 6C33 triodes couldn't cope with 88-89 dB sensitivity speakers.
Later again, I settled with the Jungson JA88D-09 for my workshop (only used as a power amp), Hypex Ucd digital amp for bi-amping and what appears to be final, the Yoshino EAR 868 PP.
For my main system the EAR 912 + 2 x bridged EAR-861, sometimes supplied with the Hypex Ucd for driving bass modules - bi-amping. I still have to find something better for the money (can't say the GlowMaster was cheap, but worth every cent). Check gear section: #My_System

With regard to the DTQWT speaker, I wish I had a wider selection of amps to refer to but I've only had the opportunity to hear these speakers with a good handful of amps like my 8 watts 300B, 20 wpc Audio Mirror mono blocks, Gain Clone (used for CLIO measurements), Rotel RB981 solid state driven by my WOT line stage, 8 wpc 2A3 push-pull design (discontinued) from DiyHifiSupply, Hong Kong, 15 wpc T16-OTL from Transcendent, 15 wpc 300B design from TriodeDick/Holland (built by DTQWT customer), 25 wpc amp from Jasmin (Sonata, EL156 valves), Nelson Pass FirstWatt JP2 and most recently my JungSon JA88D.
Amoung these I do not recommend the T16OTL, Rotel and Jasmin. BTW: I also tried a pair of (classic) 30 wpc EF86-ECC83-EL34 PP monoblocks designed by Claus Byrith (Denmark). Despite the use off serious global feed-back the sound was dynamic, smooth and easy - but transparency could have been better.
I've had visitors bringing in some PassLab designs and I have to say all of these worked extremely well with the DTQWTs.

Generally I prefer the valve amps over the solid state with the JungSon as an exception. What I would like to try are a few British amps like Croft, Exposure, Naim, Sugden (e.g. A21).

My default amp(s) for DTQWT is the EAR-861 for mid-tweeter and maybe Hypex for bass, i.e. bi-amping. With Hypex' introduction of the FUSION  plate amps I would use the FA251 these days for running the bass driver(s). The FA251 has the advantage of being able to take the signal from the speaker cables (high level input) and makes an easy set-up. Furthermore you may omit the low-pass filter for the bass driver and use the DSP of the Hypex. I haven't tried this an cannot give any solid advice on points of crossover and slopes, but try 100 Hz, 1st order and hear what happens.

Click image to go to page on bi-amping.

Due to room placement I have minimum room-gain and I rush to say that my DTQWTs in a smaller room and placed closer to the corners may sound better in the bass region from other amps with a more tight bass performance.
The listening room plays an integral part of any speaker set-up and will ultimately point back to the amp used, in particular when dealing with low-wattage valve amps with generally poorer bass control.
I recently has a quest for the DTQWT mkIII being suitable for a 2.3 watt SET amps. That is 2-point-3 watts. I can't tell, but I wouldn't recommend it. Maybe the TQWT if you don't play very loud.

The TQWT and DTQWT speakers were designed to offer a speaker alternative to all those investing in low-wattage SET amps, but somewhat to my surprise most customers to these kits use solid state gear. The majority of solid state amps are 50-100 wpc push-pull class B amps and with all respect they may offer a decent compromise with regard to price and performance. To some DTQWT customers there may be more to extract from your source material by trying a variety of low-wattage valve/solid state amps where quality has priority over quantity. It's easy - and cheap - these days to make a 100-200 wpc solid state amp, but the general concept of push-pull class AB designs with global feedback has its shortcomings when it comes to sound quality. To my ears! I rush to say I haven't heard them all so I'm sure there are exceptions.
I've had visitors listening to my DTQWTs driven from my 300B and Audio Mirror amps and unanimously they said: "Well, I guess 300B is not for me..." Why? Because bass is more taught from the GlowMaster. This is my interpretation of their response despite the midrange from the 300B may be a little better, but this is not what we generally favour. Kick-butt is more impressive and it takes an awful lot of time to evaluate the sound of the oboe, the flute, the piano, etc. from two different amps. We need the amps and the speakers for days for careful listening. Never judge an amp from 2 minutes listening like many people do.

As mentioned by Swedish builder Ilpo, the DTQWT can also play seriously loud with low distortion, thus an option for potent surround systems, something I hadn't anticipated when I made the speakers. Here some decent solid state gear may beat all the SET amps by a comfortable margin. Despite high efficiency we need quite a few watts to cope with 100-105 dB peak program material. Pure math comes short here. It doesn't help calculating the watts needed to make 105 dB, we need serious headroom to cope with this kind of application.
That said, the DTQWT and DTQWT-12 are not made for 80-100 square meter living space if you fully want to load some 200-300 cubic metre volume. I have an increasing number of mails requesting speakers for this size rooms - and I suggest taking a look my The Loudspeaker-1-2-3 constructions as well as my Faital-3WC-10-12-15 speakers.

My default recommendation for an amp for the TQWT/DTQWT is a PP valve design like the Cayin A50T, A55T, A88T, etc. These amps usually delivering 20-25 wpc in triode mode and 35-45 watts in pentode mode.
Also the Line Magnetic LM216IA is highly recommended (see November post).
There are numerous variants of these KT88/EL34 PP designs, like the Prima Luna and World Audio designs. Use google to find more - and not least countless diy designs on KT88/EL34/6650.
If you buy a commercial design, be prepared to spend some money on good coupling caps, i.e. super caps! This is of vital importance for making the best of tube amps.

Last, but not least: I have twice seen 8008 main driver's cone ripped apart from using 400 watt solid state amp - or lending the system to children's Saturday Night disco party. This from DTQWT mkII and DTQWT-12. Now, the DTQWT-12 was not made to play louder, only deeper. The main front driver still carries the full low-end signal and 95 dB sensitive speakers should simply not be fed 400 solid state watts. The voice coil was fully intact but the cone was just simply ripped apart. The other reason for this was that the speakers were supposed to energise more than 100 square metres rooms. This is equivalent to a smaller cinema space - and this is a completely different story. Before building speakers we need to analyse what kind of sound level we're aiming at, what is the room size - and volume, and what kind of amplifier are we going to use.

This  VAIC brought in by a visitor did very well with the DTQWTs.

Line Magnetic LM216IA, highly recommended.

My favourite.

I've had the question on bi-amping the DTQWT and there are a number of considerations to address as listed below. Next I have the question on bi-amping plus use of an electronic crossover to substitute the bass drivers' low-pass filter.

First of all I hardly see the rationale for bi-amping a 95 dB system where there's no high-pass filter to the front drivers and only a simple 1st order filter - with an extremely low DCR series coil - to the bass drivers. The system as such requires little power compared to a modern low-efficiency speaker, but should you want to run a 2-4 watt SET on the front drivers, it may certainly benefit from not having to run the bass drivers as well, but this is not without problems.

We need to reiterate the situation as described in the DTQWT file: http://www.troelsgravesen.dk/JA8008_DTQWT.htm

"The DTQWT construction has an 8" front driver and two 10" rear mounted drivers, all connected with positive polarity. They're placed opposite to one another and the rear drivers are fed through a huge coil, thus only adding weight to the lower octaves. All drivers load a center horn and you may ask whether the large bass drivers won't have an impact on the front driver from sharing the same open space and the answer is yes - and this is intentional. Air goes where pressure is low, thus when the rear bass drivers move inwards, they will push air into the large center horn and by doing so they will suck air from the front horn and pull the membrane of the front driver. Obviously the front driver will do the same thing for the bass drivers. We have a situation similar to a push-push configuration of two opposite placed bass drivers in a vented cabinet.
The subjective experience from this is a cleaner midrange, distortion has been lowered at high SPLs and yes, we have more bass. In fact a lot more bass."

Based on experience, the amps running the front and rear drivers have to be the same. Using my Audio Mirror amps for the front drivers plus an e.g. Rotel power amp for the bass drivers didn't perform too well playing at loud levels. I guess that the amps have different phase characteristics in the bass region, thus jeopordize the synergy between front and rear drivers loading the same center horn. Amps being either AC or DC coupled may have an impact on this.

Bi-amping and electronic crossover to the bass: The bass crossover is a low slope 1st order filter and has a significant impact on upper bass/lower mid. Cutting the bass by e.g. 12-24 dB at e.g. 100 Hz will render upper bass/lower mid weakened.
In a situation like this, the front drivers would have to be run directly from the amp, bypassing the electronic crossover. Next the electronic crossover will have to mimic the transfer function of the passive crossover and this may take some experimentation. And again, the phase behavior of the different amps will still have to be taken into consideration.

My current solution for bi-amping is this, which I use this for all larger speakers that simply can't be run from even 40 watt PP tube amps unless you never play very loud.  http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects_files/bi-amping.png.

All in all not that easy - and again: I hardly see the rationale given the system's sensitivity.