Thanks for the links.
Now that you mention it, I also got some 300 2N3773 , very robust.
So what they told me about Military manufacturing was not a salesman fable 😉
Small World indeed.
I don´t blindly trust labels, "brands", etc. but *measure* everything, and was impressed by performance.
Same way I am impressed by TIP142/147, be it ON-Fairchild or ST made; specs don´t look "that" great but have been incredibly reliable workhorses.
It´s not me, Industry leaders such as Fender - Ampeg - H&K - Crate - Marshal use them massively, there must be a reason for that.
This is an 85W RMS Fender Deluxe amplifier, notice the single TIP14* pair, old style TO3P case.
Very basic heatsinking: they are bolted to the thin (1mm) Aluminum chassis through a heat spreader aluminum bar.
BUT also notice bias diodes (CR10-11)are inserted in holes drilled in said aluminum bar, excellent tracking
Usha used to supply to others outside India for mil grade supplies, now it is defunct...those you had used.
BEL is there, their focus is on their military supplies, I do not know how their supply to the civil market is in terms of ease of availability, or price.
3055HV may be High Voltage, I do not know about 3055MA.
Keltron and CDIL are there too.
You now use which transistors in your amps?
BEL is there, their focus is on their military supplies, I do not know how their supply to the civil market is in terms of ease of availability, or price.
3055HV may be High Voltage, I do not know about 3055MA.
Keltron and CDIL are there too.
You now use which transistors in your amps?
Last edited:
Really VERY basic. In fact you can't transfer the heat sufficiently across of such L-bar to a thin aluminium sheet. Allready the L-bar is too thin to transfer the heat because of its Rth.
Good heat sinking is very important, especially for BJT. They usually need junction temperatures lower than 150 °C. In contrast most MOSFETs work at channel temperatures up to 175 °C. I usually design my heat sinks in a way that the temperature of the device chip always stays 50 K below the device limit.
Exactly - we have since learned that the output transistors are not what was wrong with the original amp. All of the successful applications of the TIP142’s outlined here use the TO-218, which has a much better chance of success than the TO-220 version. If you’re using an insulator of any sort, and a practical size heat sink the most heat you can even pull out of the package is 20 watts. To do better you need forced air and/or live heat sinks. They do both in a typical gaming computer power supply. The bigger package is a huge improvement in case-sink Rth, and makes all the difference in the world in those Fender amps.
Yes it is ironic how common it is with Musical Instrument amp combos in the
85 to 100 watt range.
Which only use a basic aluminum bar for a small heat spreader, then attached to a aluminum chassis. Which actually works rather well for heat dissipation.
MI amps tend to suffer a lot of abuse.
Anyways as far as the question.
Using a pair of Darlington BJT transistors in a T0-220 case
You would have to de rate according to the datasheet.
Within reason basically with 2 devices per rail
or a total of 4 output devices as proposed
Assume you would safely get only 65 to 75 watts with 4R loads
depending upon your heatsink design/size
Unless as mentioned you would need forced air cooling to push that further.
Or add more output devices.
6 to even 8 devises total would be better for 100/120 watts
Assuming small T0-220 packages.
Far as 220 packages Ive had good luck with
BDW42/47
They tend to behave better in real life than TIP
likewise ironically tend to model better in simulation.
If I remember Ft of TIP42/47 is around 3 Mhz and will dip down to 1 Mhz under load
BDW42/47 around 8 Mhz and will maintain around 6 Mhz under heavy load
85 to 100 watt range.
Which only use a basic aluminum bar for a small heat spreader, then attached to a aluminum chassis. Which actually works rather well for heat dissipation.
MI amps tend to suffer a lot of abuse.
Anyways as far as the question.
Using a pair of Darlington BJT transistors in a T0-220 case
You would have to de rate according to the datasheet.
Within reason basically with 2 devices per rail
or a total of 4 output devices as proposed
Assume you would safely get only 65 to 75 watts with 4R loads
depending upon your heatsink design/size
Unless as mentioned you would need forced air cooling to push that further.
Or add more output devices.
6 to even 8 devises total would be better for 100/120 watts
Assuming small T0-220 packages.
Far as 220 packages Ive had good luck with
BDW42/47
They tend to behave better in real life than TIP
likewise ironically tend to model better in simulation.
If I remember Ft of TIP42/47 is around 3 Mhz and will dip down to 1 Mhz under load
BDW42/47 around 8 Mhz and will maintain around 6 Mhz under heavy load
Last edited: