• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.


A new TSSA amp is on the stairs.

We wanted to preserve the 2 stage design and on the same time lower the output impedance and distortion without changing the basic circuit design.

We have succeed in doing it with an lift in airiness in the upper region. From midrange and up. Better control in the Bass.

The basic circuit known in the TSSA V1.6 is exactly the same. But by carefully redesign the current sources we have been able to raise the openloop gain by 7.5 times.

All idle currents tracks really fine. Within a few seconds from power up, idle currents are set. Meaning true "power on and play".

There is no chip solution on the AMP only PURE DISCRETE solution.

We have ended up with an amp in the same formfactor as the TSSA V1.6 so it is upgradeable. All holes are identical.

It can be run at 400mA as well as up to 1.5Amp idle with prober cooling,

First measurements show:

  • Distortion level into 25W into 8R is 0.06%
  • Bandwidth is 1.4MHz.

The new board is named TSSA V1.7 and will be available within 2 weeks.

To be safe Monday the 10. of february.

The modules will be delivered assembled and tested. The only thing needed to be fitted is the mosfets.

There is also an TSSA V1.7 HV (High voltage) on its way, but we need to hear the sonics performance.

- Sonny
Last edited:
Some preview on the Data from the TSSA V1.7

Bandwidth is 1.5MHz
Slewrate is 44V/us Not fast, but 2stage design make it not possible to exceed it with lateral mosfet.
Zout is 50mOhm.

THD measurements below is taken at bias current of only 250mA. So it is real class AB as well.

Noise about 5KHz is Notebook interface that plays with us.

THD into 24W 8R is 0.034%
THD into 1W 8R is 0.014%

Harmonics falls off nicely.


  • thd 1KHz 0.25A bias 1W 8R.pdf
    38.5 KB · Views: 158
  • thd 1KHz 0.25A bias 24Watt 8R.pdf
    38.9 KB · Views: 143
Hello vgeorge and buzzforb.

In last minute i have added an onboard LC filter. After i have talked with Vgeorge i have decided to ad it. The reason for this is that vgeorge asked me if i have had problem using SMPS500R from connexelectronics.

I have to state right away that there is nothing wrong with the SMPS500R but the scope result showed an ringing of 5MHz every ~4usec..
Regardless of what amp we talk about PSRR is vanishing low at 5MHz.
Only solution is an EMI filter -> LC filter the inductance should have resonance as high as possible (>6MHz) capacitance should also have an resonance above 10MHz.

The FETS i use is ALF16N/P16 or EZW20N/P20.

They are 250Watt package, but i would not use them for more than 80 - 100W into 8R.
The HV version will be able to run from +/-45VDC. The main difference on this standard and the HV version is that we have to change the cascode from JFET too Bipolar due to the higher voltage. It is an whole new PCB so i cannot say anything about sound quality before we have build the first set.

It was only planed to use one set mosfets.

George and i talked about an high quality amp for Horn systems. A note to this. I talked with Maz from Odense, about an NANO AMP. actually smaller than the TSSA but it will make more sense to keep the form factor.

So this so called NANO amp will be same size as the TSSA but it will be based on the circuit from the Mirand A2. But only with one set output transistor. It will be 20 - 80Watt into 8R and can be run in CLASS A and AB depending on taste.
Not only horn system. But it will be as quiet or more as the TSSA. Distortion is also better. Actually i think you would be very pleased with the NANO AMP.

Last week we had a listening session. Michael G. and I. My wife was also present. The amps was the TSSA V1.6 with V1.7 changes and the MIRAND A2.

Michael and I was sitting in the sweet spot. My wife was sitting sideways in another sofa. She was knitting. Now what amazed me was the she suddenly said that the MIRAND A2 sounded better! She is not trained in listening. But from a not optimum position she had no problem in hear the difference.

Bussforb : We will find a solution.
I must admit, the Nano sounds interesting. Have you ever gone to the trouble of doing a comparison of through gain devices vs the SMD you currently use? The reason I ask about the output stage enlargement is my affection for the FE/driver stage you haev created. While my system will be semi HE, at 93dB, it is not a horn system and 50-100W amp is appealing. Another thing I have found is that i tend to like amps that are either bridged or ones that use higher rails. Seems to add a dynamic capabilty that pairs well with dynamic speakers. Balanced./bridged designs are a way to get this voltage headroom and still maintain higher bias, although it can also be achieved with SE designs and higher rails if you are willing to parallel outputs and create a FE that can drive the capacitance.
Sorry if I have cluttered the thread. Very much looking forward to the new designs. Have the fets here waiting:)

Hi, i have added the schematic.

The Main difference on the TSSA V1.6 and TSSA V1.7 is the more "advanced" current generators which makes it possible to raise the gain by 15 times.

It depends on heavy use of multipacks transistors. So no matching is required. ;)

The board uses BCM847/857 (Matched pairs) and THAT340S.

Secondly 1 trimmer is removed. The use of a JFET current source is hugely depended on the supply voltage . ... In other words .. worse than standard CCS with two BJT'S!!!

So trimming should be easier.

Added an ferrite bead (10Amp DC, [email protected]) together with an 10uF 50V X7R to form an effective lowpass filter for noise from supply ... Main contributor is ringing from diodes.

Resistor added to bypass capacitors C8 and C9 to prevent ringing due to parasitic inductances.

Gatestoppers added to the mosfet, but it does not reduce performance regarding bandwidth and distortion.

The schematic show a 47pF millercap. This gives a bandwidth of ~2.7MHz.

So all in all a serious upgrade. And the people that have listened to until now is very happy with the result.

The updates are not cheap but effective.

Best regards

Sonny Andersen


  • TSSA-v17-SCH.pdf
    75.3 KB · Views: 729

TSSA V1.7 will be priced at :

1 - 3 modules 115€/module.
4 - 10 modules 100€/module.

Old customers will receive an discount to the prices above. Please contact me at [email protected]

I was asked if they could run bridged. It is possible and will with +/-32V supply produce in the order of 120 - 150Watt into 8R. But they will need a balanced source.

In an bridged stereo set you will need 4 modules.

We will post pictures of such an setup.

- Sonny
Last edited:
I was not happy with the HV version I had designed. So this amp will be changed to be an dual TSSA V1.7 module running balanced with an optional discrete inverter onboard to make it possible to feed it from an single ended source.

The new TSSA V1.7 BL (HV is renamed to BL) will cost €200/module.

But the system voltage should be +/-32VDC. So it is possible to reuse powersupply from the TSSA V1.6.

- Sonny

PS: Below is the PCB design of the TSSA V1.7


  • tssav17assembly.jpg
    242.3 KB · Views: 734