Simple Symetrical Amplifier

[Lazy Cat]

SSA Power Amp - drain output & CCS

Improved and probably the last version

 

[Joachim Gerhard]

I have only 10N20R and 10P20R. They have halve the current rating then the laterals you use.
Should the PSU voltage go down with these or can i parallel 2 ?

 

[Lazy Cat]

SSA Power Amp - drain output & CCS & parallel mosfets

Output peak voltage headroom is always welcomed, so stay on +/-50 V rails PSU and parallel the outputs. Same gender type of mosfets should be matched within +/-5 % to successfully split output power dissipation evenly among them.

 

[Lazy Cat]

SSA Power Amp - source output & CCS

SSA standard

 

[Lazy Cat]

New A-75 Accuphase power amplifier uses SSA circuit in its voltage amplification stage MCS+, more here

 

[UltimateX86]

Hi LC,

Can you help me improve this ?

it's based on your vssa

An externally hosted image should be here but it was not working when we last tested it.

QuasiSSA.asc

R5 / R55 are not necessarily a good solution...

thanks

 

[UltimateX86]

in fact 20v supply on VP VN is enough to have 28v output

 

[Lazy Cat]

Simple and yet excellent DC coupled balanced circlotron. Ran the sim, looks very good. For the start implementing some familiar design improvements to get better THD at higher output levels, like enhanced VAS, driver stage, more parallel outputs, would do the job and make it stellar. Need some time to see what can be done.

 

[UltimateX86]

With 20V we could use TO-92 for vas like BC560C

A driver stage would be good

 

[Lazy Cat]

Yes, no problem with BC560C or enhanced VAS, driver stage is also crucial to get low THD driving laterals but then it wouldn't be simple symmetrical amplifier anymore.
Already found my best VAS, tried them all. Will be inside new First One modules.

 

[de Ocampo]

It is good to see updates of the SSA, thank you LC.
Hope to see an SSA using Hexfet output.

 

[UltimateX86]

I created a multithreaded application to run incremental ltspice batch simulations with as much parameter as desired and taking measurements to keep only the adapted ranges

It seems that the most suitable current for the VAS is 2.7mA with BC560C

409591 simulations and only 255 combination kept, 1h with my laptop (4thread)but only a few minutes with my dual xeon (32thread)

https://sites.google.com/site/solidcirclotron/BRUTE_FORCE.ods

    "listeParam": [
        {
            "name": "PARAM_RA",
            "value": 100,
            "min": 100,
            "max": 1000,
            "inc": 10
        },
        {
            "name": "PARAM_RB",
            "value": 1000,
            "min": 1000,
            "max": 10000,
            "inc": 2
        }
    ],
    "listeMeas": [
        {
            "code": "Ie(Q1)",
            "min": 0.001,
            "max": 0.016
        },
        {
            "code": "Is(M3)",
            "min": 0.2,
            "max": 0.5
        }
    ]

 

[N101N]

R5 / R55 are not necessarily a good solution...

They should go to the output source terminals.

 

[UltimateX86]

yes, with the differential feedback it works

thanks

 

[banat]

R5/R55 can be bootstrapped also .

 

[UltimateX86]

hi banat it's nice to see you again, you already talked about bootstrap but I had not managed to work

 

[banat]

Hi X , it is my pleasure !

I think in that configuration VAS bootstrap will work just fine , just split both R5/R55 and find optimum values for new four colectors resistors and two bootstrapp capacitors .

Best Regards !

 

[N101N]

It works regardless of feedback. I personally dislike trickeries like bootstrapping.
You could consider referring the emitter resistors of the input stage to ground.

 

[padamiecki]

Hi, this circlo swings!
do you have optional phase splitter with the gain to power it?

 

[UltimateX86]

Hi pada

The UGS could be a good preamp for a circlotron

UGS adventures