• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

VSSA Lateral MosFet Amplifier

Just another nice thing to report, with cap multiplier and in balance mode, the GND can be disconnected from the main power supply and the output stays the same (offset less than 2mV) also the sound does not change at all. It seems that the CM virtually creates the necessary GND successfully.

Even started from 0V, power off completely to full power, there is no issue at all. So I can go with mono-polar power supply (should be around 70V).

Another nice feature of VSSA :D
 
Even started from 0V, power off completely to full power, there is no issue at all. So I can go with mono-polar power supply (should be around 70V).

Another nice feature of VSSA :D

WoW it works stable without GND, complete H-bridge. Very good forecast also for the First One since it can source-sink 16 A, ideally when used in full ballanced configuration. :up:

Congratulationa on courage experimental work you did, providing us very useful informations, thanks Radu. :cheers:
 
Well, I am still in a test setup with only one channel, and it seems that with the Cap multiplier (that is an active circuit) it works very well in that way. However, I will use a power bleeder/divider for each channel in the final setup, the way it was also pointed out by BYRTT a few pages ago. I have at hand some matched 1kOHM/25W so I will use those.

And will use separate supply for each channel, to minimize any surprise :)

Cheers!
 
Just another nice thing to report, with cap multiplier and in balance mode, the GND can be disconnected from the main power supply and the output stays the same (offset less than 2mV) also the sound does not change at all. It seems that the CM virtually creates the necessary GND successfully.

Even started from 0V, power off completely to full power, there is no issue at all. So I can go with mono-polar power supply (should be around 70V).

Another nice feature of VSSA :D
Very interesting CM can do the divider job, at least under test setup.
Well, I am still in a test setup with only one channel, and it seems that with the Cap multiplier (that is an active circuit) it works very well in that way. However, I will use a power bleeder/divider for each channel in the final setup, the way it was also pointed out by BYRTT a few pages ago. I have at hand some matched 1kOHM/25W so I will use those.

And will use separate supply for each channel, to minimize any surprise :)

Cheers!
Am i wrong dividers should read 10kOHM/25W.

I attach simulation-plot of gained groupdelay in lows by using 6,8mF for feedback caps instead of 2,2mF, also X-config/virtuel-point gain can be seen compared to standard VSSA module. Wish you good setup/soldering and thanks for reports.
 

Attachments

  • feedbackCaps.PNG
    feedbackCaps.PNG
    67.8 KB · Views: 470
Just another nice thing to report, with cap multiplier and in balance mode, the GND can be disconnected from the main power supply and the output stays the same (offset less than 2mV) also the sound does not change at all. It seems that the CM virtually creates the necessary GND successfully...
I thought of doing something like that when I read about your transformers with the four secondary windings (something I can't easily duplicate without a custom transformer). The two RC circuits in the CM filter may be acting like a divider which gives you the virtual ground reference in the middle. I don't know if that would work under all conditions though, and I am not sure how to test to be 100% certain.

(Virtual ground diagram in Post #2822 from BYRTT)

edit:

BTW, my earlier question in Post #2831,
has the VSSA module been tested into 4-ohm load, up to 45V??
 
Last edited:
I haven't test it in 4 OHMS at all... I thought I did read something like that but now I am not sure at all, and I quite doubt about that. In the doc that accompanies my last VSSA set arrived just today I can read that usable voltage range is up to +/-45V that usable load is 4 or 8 OHMs and that maximum output current is 8A (which is quite logical).
Now if we want to see what 8A is into 4 OHMS load (considering it pure resistive) the resulting voltage is 32V. With addition of voltage drop on the MOSFET of 5-6V we have a maximum supply of +/-37V-38V.
Hoping that my math is correct, I would not go over that voltage limit for 4OHMS load if using regulated voltage. If using unregulated supplies, maybe a +/-35 should be a reasonable limit.
I will defer my opinion to anything Andrej will comment, but that's what I see as the limits. It would be a pity to destroy such jewels by torturing above the docummented ratings, especially as dual output pair is not easy to find.

@Pete:
My 500VA trafo is with 4 windings 28Vac each, but I will use in this case a 650VA trafo with 2 windings of 50Vac each. Still on the limit a bit, but better; and I have no intentions to use my amplifier in a PA environment :) Due to the way BOSE 901 loudspeaker works (with an active EQ in the chain) I need more reserve on the bass peaks. Otherwise I am quite happy even with simple VSSA setup (modded with shunts instead of R-D groups) on most music material.

I hope everything will go well in all conditions, considering separate supplies for each channel. I will report that anyway; if anyone has suggestions related to especially testing I will be glad to read them.
Cheers!
 
.....BTW, my earlier question in Post #2831,
has the VSSA module been tested into 4-ohm load, up to 45V??

Hi this is guess, lets say under Andrej's many tests, he found sweetspot soundwise for the calibration belonging to exact +/-45V and therefore recommend this voltage. If this is true members with other voltage rails maybe can experiment better soundwise results by wasting some time in other calibration specs for CSS/VAS/ALF-bias.
Also this is guess, many doesn't have more than 8A RMS current and then 4R setup survive, or else input voltage throttle can be used as limit (In post #811 2R is allowed if output RMS current doesn't exeed 8A).
Here diagram for ALF +/-45V unlimited current into 4R load, power very short moment ~506watts .........:fight:....:yell:....:h_ache:....:redhot:.....:flame:.....:dead:
 
Last edited:
I often read comments about 4 Ohms speakers in this thread. As a 8 Ohm boomer will usually present a 6 Ohms minimal impedance, so 4 ohms -> 3, I don't see the interest to use 4 Ohms impedance speakers. It can be explained for car use + amp without SMPS, but i don't see any other advantage.
Please, those of you who use 4 Ohms speakers,can-you explain-me why and what are their models ?
Pure curiosity :)
 
Today i've powered up my VSSA modules from LazyCat for the first time.
Adjustments went ok, also a bit smoke came out (i dont know from where) but all voltages seems to be ok and nothing looks damaged.

I should be worried if the VAS current is not rock stable ? Cold measure 113-155mV and tried to rise the temp a little bit with hair dryer current went to 130mV.

DC output remained all the time under +/-5mV.
 
Last edited:
Hi guys :wave2:

ALF08NP has according to datasheet continous drain current Id=8 A, meaning constant DC current capability.

Converting this into AC results in 11,313 Ap peak drain current, multiplying this with 4 Ohm load, results in 45,25 Vp.

So at +/-45 V PSU there is no chance to break ALF08NP on 4 Ohm load with just AC drain current.

Also at Id=8 A, Vds is almost 10 V, meaning 35 Vp max load output voltage.

In any case at +/-45 V there's no risk at all for ALF08NP, also tested in practice - VSSA driving 4 Ohm speakers in many systems.

The main and most important are SOAR conditions at that voltage PSU, so one has to provide 0,5°K heatsink and very good thermal connection.

Regards, Lazy Cat :cheers:
 
Last edited:
Today i've powered up my VSSA modules from LazyCat for the first time.
Adjustments went ok, also a bit smoke came out but all voltages seems to be ok.

I should be worried if the VAS current is not rock stable ? Cold measure 113-155mV and tried to rise the temp a little bit with hair dryer current went to 130mV.

DC output remained all the time under +/-5mV.

Smoke :confused:

VAS bias normally goes from 12 mA (120 mV reading between TP1-TP2, TP3-TP4) to 15 mA (150 mV reading between TP1-TP2, TP3-TP4), cold start to warmed amp.

VAS thermal bias variation has no influence on output bias current.
 
Esperado_
8 is to 7 as 4 is to 3.5. Nit picking of course. Motors, and speakers as they are the same principle as a motor, need current to function well. If speakers are made with higher impedances, they need higher voltages to deliver more current in order to work. Over the years the standard speaker has dropped from 16 to 8 and now to 4 ohms so they can recieve more current. More efficient?
I may not be using the proper terms but I think you can get my meaning.
 
I should be worried if the VAS current is not rock stable ? Cold measure 113-155mV and tried to rise the temp a little bit with hair dryer current went to 130mV...
You might have to explain what you mean by 113-155.

Btw, the input should always be shorted during calibration or this kind of test, to make sure there is no phantom signal, or interference, getting into the circuit. If it was not shorted, you should probably repeat the test.

If the test voltage is fluctuating, or going up and down on the meter, without touching anything, like touching the trim pots for example, that would be wrong.

The VAS current (and therefore this test voltage) will increase from cold to hot, because the VAS transistors are BJTs, and they have a positive temperature coefficient. Because they are on the main heatsink in this case, and because the heatsink is being heated by the idle power dissipation of the ALFET (about 14~15 watts), the current can go from 10 to about 14 mA, which translates to 100 mV to 140 mV, cold to hot.

Increasing from cold to hot is not the same as being unstable.

The reading should stabilize after a few minutes, and after that, it should not change as long as the heatsink temperature stays the same.
 
Last edited:
Maybe "unstable" is not the right word as he current goes from 120 to 140 with temperature not without reason.
I will repeat measurements with input shorted.

I'm really frustrated because i don't know where that smoke came from. There are no signs of burning but i know what i saw and smell.

Thanks Andrej and Pete !
 
Maybe "unstable" is not the right word as he current goes from 120 to 140 with temperature not without reason.
I will repeat measurements with input shorted.

I'm really frustrated because i don't know where that smoke came from. There are no signs of burning but i know what i saw and smell.

Thanks Andrej and Pete !
Make sure that you have the 22-ohm series resistors on the rails (or the "bulb tester")

Also, I would look at the 10-ohm resistors on the board, those in series with the rails, make sure the capacitors are installed in the correct orientation, that the circuit board is installed flat and making contact with the heatsink, signs of burned paint on any component, etc. Sometimes the basic stuff can be the easiest to miss, because you are looking for something more complicated.