Everything was perfect. Spent an entire day on just Antenna. Cleaned or replaced anything questionable. Terminal, connector or whatever. Took lids off amps and cleaned fans and what ever needed checked. Hard to beat back in the day.
-
You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
Simple protection for 2SC2879's
- Thread starter Shockwave
- Start date
All I can say, is based on what I've learned with these older protection mods and the fuse size required to burn open, prior to the transistor. The 20 amp fuse will pop on a single 2879 before the transistor, if you maintain a reasonable carrier. Average current changes very little on AM when the 4:1 ratio is used. You would think a pair of transistors in push pull would at least be able to handle the next size up fuse at 25 amps. As soon as customers started adding just 5 amps to the size fuse used on a 1 pill, they could easily over drive the amp, pop a 2879 and never burn any of the 25 amp fuses protecting each bank of 2. The same fuse that protects one, protects two, without the headroom to move up one size in fuse.
The only transistor I used protection on back then was the 2879 however, many different types of amps received the same modification. It didn't matter if it was class C or biased AB as in the Messengers, the same 20 amp fuse worked on Toshiba 2979's. The only difference is you got less watts biased than you did in class C, for the same current draw. While I cannot claim to remember the actual current each amp took, to me the important thing was finding what size fuse, could save the transistor.
That means it is likely that a 20 amp fuse may only pop at a 30 amp load. However, what I learned time and time again from customers that kept trying a 25 amp fuse, was adding just 20% to the fuse value that worked on one 2879, would cause one in a pair, to fail before the fuse could. If the current was really being split between two transistors, the largest 30 amp fuse that would fit in the socket, should have been able to protect a pair of 2879's. Push-Pull gives you double the dissipation capabilities, not double the current handling. Doubling the current ratings requires placing push-pull transistors, in parallel which we often see on large AF amplifiers.
The only transistor I used protection on back then was the 2879 however, many different types of amps received the same modification. It didn't matter if it was class C or biased AB as in the Messengers, the same 20 amp fuse worked on Toshiba 2979's. The only difference is you got less watts biased than you did in class C, for the same current draw. While I cannot claim to remember the actual current each amp took, to me the important thing was finding what size fuse, could save the transistor.
That means it is likely that a 20 amp fuse may only pop at a 30 amp load. However, what I learned time and time again from customers that kept trying a 25 amp fuse, was adding just 20% to the fuse value that worked on one 2879, would cause one in a pair, to fail before the fuse could. If the current was really being split between two transistors, the largest 30 amp fuse that would fit in the socket, should have been able to protect a pair of 2879's. Push-Pull gives you double the dissipation capabilities, not double the current handling. Doubling the current ratings requires placing push-pull transistors, in parallel which we often see on large AF amplifiers.
Current is been split between the two transistors in push pull rf amps,
The amps you see on your meter is a measurement of how many electrons flow per second, or coulombs per second,
1 coulomb/sec = 1 amp,
1 amp = 1 coulomb/sec
1 amp = 6.24 quintillion electrons per second
To make the numbers simple lets imagine we are talking about an amplifier operating at 1hz not 27mhz
makes no difference the ratio of time on vs time off stays the same.
a single ended class C rf amp the transistor typically conducts for less than 180 degrees on one half cycle of the waveform lets say it draws 10a & turns off during the opposite half cycle,
it conducts for less than half of 1 second,
in a pushpull pair one transistor conducts less than 180 degrees on one half cycle @10amp & turns off for the opposite half cycle,
its partner conducts for less than 180 degrees on the opposite half cycle @10amp,
they conduct for almost 1 second in total or twice as long as the single ended amplifier each cycle,
If this amp is running at 1 cycle per second & amps = electrons flowing per second,
the single ended amp draws 10a once per second = 10amps
the pp amp draws 10amp twice per second = 20amps
all 3 transistors draw the same current each but the pushpull draws double the current on you amp meter
your meter is indicating electrons or coulombs per second ( AMPERES ) derived from voltage measured across an accurate resistive shunt using ohms law.
Commercial 2 transistor class C pushpull amps typically use 2 x the fuse they use in a single ended amp to get a little less than double the power for twice the current draw,
4 transistors amps use 4x the fuse rating of the single ended amp & draw about double the current of a 2 transistor amp to get a little less than double the 2 transistor amplifiers output,
unless its a RM 4 transistor cb amp which only use 3x the fuse & melt fuse holders & pop fuses for fun,
the same 4 transistor circuit in their HLA amps use 4x the fuse rating & don't melt fuse holders.
The amps you see on your meter is a measurement of how many electrons flow per second, or coulombs per second,
1 coulomb/sec = 1 amp,
1 amp = 1 coulomb/sec
1 amp = 6.24 quintillion electrons per second
To make the numbers simple lets imagine we are talking about an amplifier operating at 1hz not 27mhz
makes no difference the ratio of time on vs time off stays the same.
a single ended class C rf amp the transistor typically conducts for less than 180 degrees on one half cycle of the waveform lets say it draws 10a & turns off during the opposite half cycle,
it conducts for less than half of 1 second,
in a pushpull pair one transistor conducts less than 180 degrees on one half cycle @10amp & turns off for the opposite half cycle,
its partner conducts for less than 180 degrees on the opposite half cycle @10amp,
they conduct for almost 1 second in total or twice as long as the single ended amplifier each cycle,
If this amp is running at 1 cycle per second & amps = electrons flowing per second,
the single ended amp draws 10a once per second = 10amps
the pp amp draws 10amp twice per second = 20amps
all 3 transistors draw the same current each but the pushpull draws double the current on you amp meter
your meter is indicating electrons or coulombs per second ( AMPERES ) derived from voltage measured across an accurate resistive shunt using ohms law.
Commercial 2 transistor class C pushpull amps typically use 2 x the fuse they use in a single ended amp to get a little less than double the power for twice the current draw,
4 transistors amps use 4x the fuse rating of the single ended amp & draw about double the current of a 2 transistor amp to get a little less than double the 2 transistor amplifiers output,
unless its a RM 4 transistor cb amp which only use 3x the fuse & melt fuse holders & pop fuses for fun,
the same 4 transistor circuit in their HLA amps use 4x the fuse rating & don't melt fuse holders.
Interestingly enough, I had this same debate with Luke (BBI) about 5 years ago. He thought I was crazy when I told him a single 2879 in a tuned circuit would produce similar output power and input current, as a pair in push pull. About 6 months after sharing some tuned, single ended circuits with him, I got a message back confirming that he had found the same results after testing.
Keep in mind, the protection modification described here has been in use for 30 years now, on many different types of 2879 amplifiers. Because it is simple and no secret anymore, I suspect someone out here will want to save their Toshiba's and install it. I promise if you try and add just 20% to the fuse size that works on a "1 pill", you will defeat all protection offered by the modification and one 2879 can pop before the 25 amp fuse.
I also trust that you are quite capable of measuring current and found a 24 amp load with a pair of 12 amp transistors. We may be surprised to learn that one 1446, in a well matched single ended circuit, might actually support over 20 amps of collector current too.
Last edited:
I don't dispute the fuse will help protect the single 2879 & it needs to be less than 25a for safety,
with a well tuned single ended amp you can set load impedance optimal for whatever vcc & drive level you intend using, just like a tube amp,
with push pull on a transformer you are stuck with fixed integer ratios & would need an LC network after the transformer to get load impedance anywhere near optimal,
even then you have transformer losses & poor compensation cap choice to contend with,
push pull amps draw twice the current through the fuse while each transistor pulls 1/2 the current,
if that was not true how does a factory untweaked 2x1446 amp pull 20a + reliably on FM,
mine pulled 24a from o beta 1446's with 10w drive using 12amp transistors in sloppy class AB,
a 20a fuse will not hold that current on FM
that is not possible with a single 1446 with ANY tank circuit in any bias class, not even negative bias deep class C
The Davemade style 2 & 4 transistor amps and all the copy/paste clones using 1000pf semco's & big transformers when used on stock volts have the worst efficiency of any amplifier I ever tested & not just poor its disgraceful efficiency
beating a 2x2879 copy/paste amp on stock volts with a single ended optimized amp is possible because they make hardly any power for the crazy current they draw,
the only way to make them work with decent efficiency is run them on the vcc they were designed for.
Rework it with proper turns ratios & caps for stock volts you will see efficiency goes up like a Saturn V & your single ended amp is getting its lips cut off.
i just noticed you added a bit above,
I don't believe a 1446 will handle 24a FM with any tank circuit,
but i have seen a Dave style 2x2879 pulling enough current to suck the life out of an astron rs70 on FM & not because of rfi,
i have also seen the same style amp using the same components & ferrite pulling 50a FM been used in a mobile, tested on my bird 43 & bird load, good quality 500a shunt + fluke 87,
will a single 2879 withstand 50+ amps continuous like a pp pair will ? i have never seen that, i would be amazed,
with a well tuned single ended amp you can set load impedance optimal for whatever vcc & drive level you intend using, just like a tube amp,
with push pull on a transformer you are stuck with fixed integer ratios & would need an LC network after the transformer to get load impedance anywhere near optimal,
even then you have transformer losses & poor compensation cap choice to contend with,
push pull amps draw twice the current through the fuse while each transistor pulls 1/2 the current,
if that was not true how does a factory untweaked 2x1446 amp pull 20a + reliably on FM,
mine pulled 24a from o beta 1446's with 10w drive using 12amp transistors in sloppy class AB,
a 20a fuse will not hold that current on FM
that is not possible with a single 1446 with ANY tank circuit in any bias class, not even negative bias deep class C
The Davemade style 2 & 4 transistor amps and all the copy/paste clones using 1000pf semco's & big transformers when used on stock volts have the worst efficiency of any amplifier I ever tested & not just poor its disgraceful efficiency
beating a 2x2879 copy/paste amp on stock volts with a single ended optimized amp is possible because they make hardly any power for the crazy current they draw,
the only way to make them work with decent efficiency is run them on the vcc they were designed for.
Rework it with proper turns ratios & caps for stock volts you will see efficiency goes up like a Saturn V & your single ended amp is getting its lips cut off.
i just noticed you added a bit above,
I don't believe a 1446 will handle 24a FM with any tank circuit,
but i have seen a Dave style 2x2879 pulling enough current to suck the life out of an astron rs70 on FM & not because of rfi,
i have also seen the same style amp using the same components & ferrite pulling 50a FM been used in a mobile, tested on my bird 43 & bird load, good quality 500a shunt + fluke 87,
will a single 2879 withstand 50+ amps continuous like a pp pair will ? i have never seen that, i would be amazed,
Last edited:
Thank you Bob. You managed to beat a stubborn Italian into agreement with your sound technical description of the flaws found in most of our 2879 amplifiers. Thirty years I ago I did not have the knowledge to perfect matching issues in broadband amplifiers and just couldn't imagine that everyone on this band, failed to achieve a decent match. Fact is an 8 pill was just a pair of four pills with and extra combiner and splitter. The only time they seemed to change the turns ratios was with a 2 pill.
Later when building my own, I found out that when you do that, it results in very inefficient input and output impedances that are far too low. I already admitted that I'm easily seeing twice the output from a pair of the high voltage MOSFET's when properly matched, than on the single ended circuit. Now I'll admit I never put nearly as much effort into matching a 2879 amplifier and assumed the dozen or so models this had been installed and tested in, could not all be wrong from the start.
What you've taken the time to break down in no uncertain terms, is in exact alignment with what I'm seeing today, having gone much deeper into design. It would be silly to still think it does not apply to the common 2879 amplifiers. They clearly had design flaws that additional matching would reveal. So, depending on who built your amplifier, your optimal fuse size, will vary...
PS: Thanks for getting trough to me, before getting mad at me.
Later when building my own, I found out that when you do that, it results in very inefficient input and output impedances that are far too low. I already admitted that I'm easily seeing twice the output from a pair of the high voltage MOSFET's when properly matched, than on the single ended circuit. Now I'll admit I never put nearly as much effort into matching a 2879 amplifier and assumed the dozen or so models this had been installed and tested in, could not all be wrong from the start.
What you've taken the time to break down in no uncertain terms, is in exact alignment with what I'm seeing today, having gone much deeper into design. It would be silly to still think it does not apply to the common 2879 amplifiers. They clearly had design flaws that additional matching would reveal. So, depending on who built your amplifier, your optimal fuse size, will vary...
PS: Thanks for getting trough to me, before getting mad at me.
- No one is chatting at the moment.