I can't see the reason to pursue the the TL431, in every listening test it never fared as well as the other options. Not used carefully it makes a nice oscillator like Mark points out.
In my opinion the output board should have the drivers, outputs and bias network and the input board the front end regulators. This will make it more versatile.
If you examine the beefcake with the TL431 carefully you will notice a couple of areas of potential problems.
Sometimes the simpler path is the better solution. Don't not go searching for problems that need solutions.
Stabilizing something under steady state conditions does not mean that under dynamic conditions it is going to stay that way. Enough said.
Just my humble opinion.
Jam
In my opinion the output board should have the drivers, outputs and bias network and the input board the front end regulators. This will make it more versatile.
If you examine the beefcake with the TL431 carefully you will notice a couple of areas of potential problems.
Sometimes the simpler path is the better solution. Don't not go searching for problems that need solutions.
Stabilizing something under steady state conditions does not mean that under dynamic conditions it is going to stay that way. Enough said.Just my humble opinion.
Jam
Attachments
Last edited:
Use TO220 devices (like 2sa1837) or the other bloody package I fail to remember like that used for bd139/bd140
Last edited:
Correct.
They require more than the 47uF that Jeff used to make stable.
I like to avoid electrolytic caps as much as possible, just because they are parts that degrade and fail with time, so I would rather use something other than TL431.
Last edited:
Watch out for power supplies too; they often include electrolytic capacitors. Might want to lay out your PSU circuit boards with the goal of super easy removal and replacement of each electrolytic cap. If for example you used caps with screw terminals instead of soldered leads, you could replace them without being required to un-solder the old ones.
It actually oscillated with the 220u that the FirstWatt designs use. 47u corrected the situation (at least at that one frequency).
None of this tells me the TL431 can't be used to good effect. What it does tell me is that it is not a "novice" part, and since there are other solutions available that I do understand I have no need to get into potentially troubled waters.
None of this tells me the TL431 can't be used to good effect. What it does tell me is that it is not a "novice" part, and since there are other solutions available that I do understand I have no need to get into potentially troubled waters.
Obviously, can't avoid them in the power supply, unless you go with battery power.
Yeah I have both snap in and screw terminal caps at my disposal.
Thanks.
Was starting to think you had built one up.
At least you are aware of these possible problems.
Jeff,
Some thoughts........
1) you are using a P-channel device(2SJ109) for the input of the Buzz-Bomb, have you you considered using a N-channel device (2SK389) and inverting the circuit instead. They are slightly easier to obtain and you could easily use 2SK170's if you can't the 389's. I don't think there is an advantage using a P-channel device (higher input capacitance) unless you are using bi-polars(lower noise). If you don't have any 2SK389's I will be glad to send you a few.
2) Your regulators don't use current sources to feed the shunt element, which is a compromise in performance , the current source entails the use of additional parts and heat sinks. A good compromise would be using the original HPA-1 regulator or a good series regulator after a C R C filter.....besides the fact that op-amps give me the creeps.
Overall, outstanding work!
Jam
Some thoughts........
1) you are using a P-channel device(2SJ109) for the input of the Buzz-Bomb, have you you considered using a N-channel device (2SK389) and inverting the circuit instead. They are slightly easier to obtain and you could easily use 2SK170's if you can't the 389's. I don't think there is an advantage using a P-channel device (higher input capacitance) unless you are using bi-polars(lower noise). If you don't have any 2SK389's I will be glad to send you a few.
2) Your regulators don't use current sources to feed the shunt element, which is a compromise in performance , the current source entails the use of additional parts and heat sinks. A good compromise would be using the original HPA-1 regulator or a good series regulator after a C R C filter.....besides the fact that op-amps give me the creeps.
Overall, outstanding work!
Jam
Attachments
Last edited:
......me thinks you put too much importance of the phase of the second harmonic. In my humble there are other parameters that might be....... I feel that the total harmonic structure including the higher orders are of more siginificant value......but it might be interesting to build it both ways.
So are you having Christmas in the States? Wishing you a great Christmas.
. I feel that the total harmonic structure including the higher orders are of more siginificant value......but it might be interesting to build it both ways. So are you having Christmas in the States? Wishing you a great Christmas.
Attachments
Last edited: