Hi,
I would like to build this circuit from EE magazine and run it from a single 50v supply the current op amp TLO72 will only work up to 40v max. Any suggestions or would it be better to go with a discrete op amp such as the ones on the ESP site
I would like to build this circuit from EE magazine and run it from a single 50v supply the current op amp TLO72 will only work up to 40v max. Any suggestions or would it be better to go with a discrete op amp such as the ones on the ESP site
Thanks all. Looks like OPA551 would be a good choice. Would splitting R6/R11 into two 1K8 resistors in series and then bootstrapping them with a 100uf cap to the output hold any advantage and any thoughts on improving the circuit ?
Opamps can have their supplies bootstrapped from the amp output, avoiding the need for higher voltage opamp - so long as the opamp rails straddle both the amps input and output signals at all times. Thus you might have a +/-25V output swing and a +/- 1V input swing, which are at most 24V apart. bootstrap +/-15V supplies with 50% of the output's ac signal and your supplies could swing from -2.5/+27.5V to -27.5/+2.5V
There are various instances knocking around this site and elsewhere.
https://www.analog.com/en/analog-dialogue/raqs/raq-issue-168.html
There are various instances knocking around this site and elsewhere.
https://www.analog.com/en/analog-dialogue/raqs/raq-issue-168.html
Thanks for the replies. Printek I like the use of complementary pair for the 30 watt 2SA1943/2SA5200 are a lot easier to get and are better than the shown Darlington transistors. Thanks Mark will read up on that
Darlington are used for the current gain, which is probably 10 times more than the Toshiba.
So you may have (will have) difficulty in providing the required base current for large current swings.
An intermediate stage will be needed for that current (I.e. TL072 is only good into a 2k load and above if I recall right.)
I'd suggest a BJT buffer between Op Amp and Output stage, with those large BJTs.
That might make whatever feedback loop you had in mind, redundant, or difficult to implement the way you hope.
Speaking from very basic experience, making something very much like that with TL072/LF351, driving a diode biased push pull stage using some To220 case bjt (I forget which), and it made some....3W with heavy THD past 2W.
Strictly Class B, and I could have upped quiescent output stage current, but it was a weekend project, and I never bothered.
It was a weekend when I learned of the TL072 habit of phase inversion, when driven close to the rails. I'm sure some of the more modern suggestions have probably solved that issue (not sure if the LF351 does the same, I think not, which is why I swapped for that opamp)
So you may have (will have) difficulty in providing the required base current for large current swings.
An intermediate stage will be needed for that current (I.e. TL072 is only good into a 2k load and above if I recall right.)
I'd suggest a BJT buffer between Op Amp and Output stage, with those large BJTs.
That might make whatever feedback loop you had in mind, redundant, or difficult to implement the way you hope.
Speaking from very basic experience, making something very much like that with TL072/LF351, driving a diode biased push pull stage using some To220 case bjt (I forget which), and it made some....3W with heavy THD past 2W.
Strictly Class B, and I could have upped quiescent output stage current, but it was a weekend project, and I never bothered.
It was a weekend when I learned of the TL072 habit of phase inversion, when driven close to the rails. I'm sure some of the more modern suggestions have probably solved that issue (not sure if the LF351 does the same, I think not, which is why I swapped for that opamp)
Last edited:
Now if they still made these. LM12CL 80V max power supply, 60V max into 8 ohms, 80W.
https://media.digikey.com/pdf/Data Sheets/Texas Instruments PDFs/K04RLM12_LM12CL.pdf
https://media.digikey.com/pdf/Data Sheets/Texas Instruments PDFs/K04RLM12_LM12CL.pdf
I think of going the route of building a discrete op amp like something on Rod Elliots ( project 17 ) site ESP as this easily run at a higher voltage and current it's for a pair of pa speakers so it doesn't necessarily have to be the best quality. I agree LM12 would be perfect but they stopped making them a long time ago