Thanks for the dremel tip, but I dont need it Im thinking big🙂
Yeah the UGS is a nice project, but its quite an expensive amp to build, and quite alot of work.
Im not sure how much to bias it, maybe two amps, that would be around 300 watts of heat with 80V rails. I have 4 sinks that would make a big chassis that could handle this sort of heat, but Im not sure yet.
that depends on whether you plan to adopt optimum ClassAB bias and also on the value of emitter resistor used.
At low enough voltage the Re has a much lower, minimum limit and could vary between 0r05 and 0r47.
As rail voltage is raised the minimum Re rises to maintain thermal stability of the output devices (Cordell's rule).
Take two values of Re and a 12pr output stage.
Re . . . Vre . . . Ib . . . total I . . . I max ClassA
0r1 . . 22mV . 220mA . 2.64A . . . 5.2Apk
0r33 . . 25mV . 76mA . . 0.9A . . . 1.8Apk
The 0r1Re might be used on a +-50V and would dissipate ~270W for 108W of ClassA into 8r0. Similar ClassA to a Krell KSA100
The 0r33Re might be used on a +-80V and would dissipate ~150W for 13W of ClassA into 8r0.
The bias and ClassA output current is totally dependent on the Re adopted.
Last edited:
Thanks Andrew, I was wondering about that. Just one question, where did you get the Vre voltage from?
AndrewT,
Thanks for that bit of information. I am beginning to learn that there is an exponential amount of tweeking involved when you change anything from the basic format.
Luke, As to whether all mica's are silver -- I do not know -- just that the ones I checked are. Now I always thought they were all silver until you posed that particular point. Any mica caps I use in the future will be researched for there content in the datasheets.
I see the caps you ordered are Wima. Do you like the way they sound? I have been trying to use Rifa/Evox films but they are hard to source in the U.S.
Has anyone ever read a review of a Leach amp. I wonder how they stack up against Audiophile graded units -- those with golden ears. They have a great reputation here at least.
Tad
Thanks for that bit of information. I am beginning to learn that there is an exponential amount of tweeking involved when you change anything from the basic format.
Luke, As to whether all mica's are silver -- I do not know -- just that the ones I checked are. Now I always thought they were all silver until you posed that particular point. Any mica caps I use in the future will be researched for there content in the datasheets.
I see the caps you ordered are Wima. Do you like the way they sound? I have been trying to use Rifa/Evox films but they are hard to source in the U.S.
Has anyone ever read a review of a Leach amp. I wonder how they stack up against Audiophile graded units -- those with golden ears. They have a great reputation here at least.
Tad
by convention, class A operation is when current flows in a device for the whole 360* cycle of input sgnal......
no, there were mica caps in the early days, up to about late 60's they are still being used.
exactly.
That's why I showed -4.99mA for a 5mA bias.
At the lowest current the device is still working as a current source/sink.
It is not off and it is not handing over to another device.
It alone contributes to the output signal for the whole 360degrees of waveform.
from this Forum.
You need to read the right threads.
Vre (where re includes the internal resistance of the transistor) should be 26mV @ ~Tj=25degC. This optimum voltage falls slightly with increasing temperature.
At high Re values the internal resistance is relatively small and so the proportion of the 26mV appearing across the external resistor is quite high. I chose 25mV for the example.
As the external Re drops the internal r becomes a higher proportion of the total emitter resistance.
I adopted 22mV for the voltage across the external Re.
This might be quite a bit lower.
PMA has produced amplifier figures that showed the optimum bias was of the order of 180 to 190mA for Re=0r1.
What is often forgotten about is why the 26mV changes with temperature. Is it necessary to change the Vre to take account of variations in Tj requiring a change in that 26mv? I think so.
For a completely different reason I like to see a fall in bias current as the amplifier heats up from a stable quiescent temperature. This is for thermal stability.
However, the bias voltage also moves in the correct direction if that 26mV does indeed drop as Tj rises above 25degC.
Any thoughts?
AndrewT,
All of these measurements are conducting in a quiescent state without music being played. Where does all of this stability go once the speaker coils begin to heat up and add there own signature to the recipe?
I would think with all of the external variables involved the quiescent bias would need to be extremely conservative to sustain reliablilty.
The numbers you presented earlier on Re resistor values are depressing. For a high voltage amp to remain stable --class A-- huge amounts of available output are wasted in the emitters. I had thought the Vbe was responsible for the majority of output thermal management -- hence the Thermal Trac device.
Tad
All of these measurements are conducting in a quiescent state without music being played. Where does all of this stability go once the speaker coils begin to heat up and add there own signature to the recipe?
I would think with all of the external variables involved the quiescent bias would need to be extremely conservative to sustain reliablilty.
The numbers you presented earlier on Re resistor values are depressing. For a high voltage amp to remain stable --class A-- huge amounts of available output are wasted in the emitters. I had thought the Vbe was responsible for the majority of output thermal management -- hence the Thermal Trac device.
Tad
those emitter resistors acts as local feedback if you will, any increase in collector current translate to higher voltage across these resistor, thus lowering the Vbe of the device, and preventing thermal run-away from happening....
there is a comprehensive discussion about it in chapter 4 of the national semiconductors' Audio/Radio Handbook, published in 1980...i still have my copy of that book.
such is the "nature of the beast" classA amps run efficiencies in the 25% to 50% regions, more closer to 25%actually.....
that is the reason it does not gain wide use....not like classAB..
Last edited:
Leach clone pcb
Hi Tad,
Just received my boards in Toronto this morning, Great job.Credos to you and Jens.
Hi Tad,
Just received my boards in Toronto this morning, Great job.Credos to you and Jens.
PCBs arrived yesterday, Big thanks to everybody involved !
Can I use NJW0281/0302 as output devices instead of the LMJ3281/1302 as suggested by the BOM ? Do I need to lower the Vcc ?
Can I use NJW0281/0302 as output devices instead of the LMJ3281/1302 as suggested by the BOM ? Do I need to lower the Vcc ?
Hi,
have you read Cordell? It might help explain to you what I think you are misunderstanding.
Re=0r33 is a fairly common value. Most would not consider it as wasting potential output power. Re=0r1 is decidedly towards the low end and must be considered only in conjunction with the amps various stabilities.
A 6pair output stage with Re=0.33 wastes the same power as a single pair with Re=0r055. Have you ever seen a reliable high power amplifier with Re<0r1?
Andrew T and Tony, thanks for your explanations.
Loboone, I got my boards yesterday,where are you sourcing parts. What rails and semis to you plan to use. You from Hamilton or just living there?
Loboone, I got my boards yesterday,where are you sourcing parts. What rails and semis to you plan to use. You from Hamilton or just living there?
AndrewT.
I got involved in some of Cordell's thread on Thermal tracs and on the biasing subject. As usual I can only pick up pits and pieces. I continue to read and try to absorb what I can. When I started the whole DIY thing it was because of my inability to repair some old Mcintosh amps I had acquired. Being totally inexperienced in electronic circuitry made it a hard road. Thanks to you and many others I have acquired some very useful skills in this area.
Most of my projects continue to function very well and I will keep working on new ones. Success is indeed rewarding even at my extended age.
To All: We need to get together and come up with some suitable -- affordable -- output resistors. Many that are affordable are not non-inductive or is that a big deal. Any suggestions?
Tad
I got involved in some of Cordell's thread on Thermal tracs and on the biasing subject. As usual I can only pick up pits and pieces. I continue to read and try to absorb what I can. When I started the whole DIY thing it was because of my inability to repair some old Mcintosh amps I had acquired. Being totally inexperienced in electronic circuitry made it a hard road. Thanks to you and many others I have acquired some very useful skills in this area.
Most of my projects continue to function very well and I will keep working on new ones. Success is indeed rewarding even at my extended age.
To All: We need to get together and come up with some suitable -- affordable -- output resistors. Many that are affordable are not non-inductive or is that a big deal. Any suggestions?
Tad
- Status
- Not open for further replies.