First off, study this follower circuit:
http://www.audiodesignguide.com/my/Cool_Follower1.GIF
The outputs in the original design were three pair of the small plastic TIP41/42 type, replace these with one pair of Semelab BUZ900/905. The one pair will be more than enough for 60W/8R or 100W/4R. The input pair may be any small signal type suitable for 10mA or so, the MPSA06/56 would be fine. Use ±15V for the supply for the input pair and 1200 ohms for the bias resistors (instead of the 390 ohm), use a pair of 100 ohm trim pots (instead of the 0.18 ohm). The BUZ need no emitter/source resistors. The capacitor need only be a 1µF (or larger) film type across the gates of the outputs.
Hook the output of this follower straight to ground, since the sources of the BUZ are the case this is easy, just use no insulator and ground the heatsinks!
To convert to a TransNova output stage all we need is a 5K input resistor to the bases of the input pair and a 100K feedback resistor from the same point to the center tap of the power transformer. We will need a regulated ±15V supply for the input pairs, one supply to drive as many channels as needed. We will need a ±40V supply (30-0-30V 150VA or larger), one per channel is needed.
Set the 100 ohm pots to about 25 ohms, then turn on and adjust for the idling current desired in the output stage. After the temperature and bias stabilize (just turn up the bias until the heatsink is as warm as you want it to be, the actual value is unimportant) then trim these again for DC off-set.
This circuit will only work for lateral MOS like the BUZ or the J50/K135 type. IRF type will blow up.
The amount of power in class A is only limited by the size of the heatsinks, and moves into class AB for loud peaks only.
If BUZ 900D/905D are used with minimum bias (100mA) and suitable heatsinking and supply voltages are used, 250W in class AB is possible out of only one pair of outputs.
HI DJK,
You mentioned Transnova on several topic of this forum, have you tried in practice?
I would like to do something similar.
http://www.diyaudio.com/forums/solid-state/15787-opamp-input-stage-2.html#post185211
http://www.diyaudio.com/forums/solid-state/15415-simple-amplifier-2.html#post181215
http://www.diyaudio.com/forums/solid-state/10053-minimalist-fet-amplifier.html#post115106
thanks
Best Regards
Seb
HI,
I would like sim a floating power supply amplifier (transnova) but it seems that ltspice does not like it..
amplifier in normal mode : bufferOPA.asc
amplifier in transnova mode : transnovaOPA.asc
https://sites.google.com/site/opampbasedamp/TRANSNOVA.zip
I forgot something ?
Thanks
I would like sim a floating power supply amplifier (transnova) but it seems that ltspice does not like it..
amplifier in normal mode : bufferOPA.asc
amplifier in transnova mode : transnovaOPA.asc
https://sites.google.com/site/opampbasedamp/TRANSNOVA.zip
I forgot something ?
Thanks
Last edited:
Hi sajti, thanks To your answer
Do you use LT1007 ? I cant read on the screen shot
Hi,
yes. First time I tried LT1001, but it was not fast enough.
Sajti
Even with the lt1007 I have -13v DC on opamp output and 30v DC on Vout.
I also tried with diodes like you to bias mosfet, nothing changes
I also tried with diodes like you to bias mosfet, nothing changes
I found the problem, the TL model is inverted and as the schema itself is inverted, I confused
Last edited:
The input stage of your circuit is not an OP AMP but OTL as the obsolete ca3080 . I have the spice model for it if you want. With op amp you can use, Holloway ,biphase , circuit.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Solid State
- Problem simulating floating power supply amplifier