Not in cases where beta droop will be the dominant flaw.
Base currents at peak output in greatest need of a match.
Base currents balance in quiescent state are completely
manageable by the combo of global loop and offset CCS.
If quiescent HFE doesn't match, bothers almost nothing.
But the maximum base current here is fixed by Q8, and
this loop can't to jack to fix it if the maximum currents
required by output bases do not match.
If one 2n3055 peak flattens earlier than other, match
at any lesser current won't give you that information.
This example is slightly exaggerated, but not by much.
Attachments
Last edited:
JLH probably abused beta droop of old style power devices
to control thermal runaway. Look at OnSemi datasheet for
2n3055, and observe all temperature curves meet at 4.5A.
Also note how the higher temp droops at the faster rate...
This won't protect newer better devices that aren't driven
into droop. Gain just keeps going up with temp...
----
Yeah, I got a method of thermal tracking I can share.
http://www.diyaudio.com/forums/solid-state/197707-shunt-regulated-jlh.html
Best was probably the version presented at post #22.
I don't want to waste my time tracking thermals and hope
the curves somehow cancel. Instead I will make darn sure
both output currents are forced equal and opposite, and the
sum is held to a reference that shrinks with temperature.
You don't want to mess with matching pairs, adjusting
bias and offset, and worrying about droop vs runaway?
I believe all can be done with one less transistor than
JLH-2005, and works in simulation. The prototype still
a pile of parts in my kitchen, so I can't say for sure yet.
One I get the bugs (if any) worked out of SRJLH-2012,
I don't mind anyone wants to make a PCB. I have no
experience in layout. All my peers at work prefer I stay
busy in the test lab, and don't bother our layout guys
with such questions.
Last edited:
I am very curious to know how hot that heat sink gets. If I would guess it would be around 90 degrees centigrade or even higher after a few minutes. If the heat sink can be touched, you are not running near the bias current that you should be.
It's depend on which heatsink to use. When mounted on a small heatsink, it can boiling water after 20 minutes. With a large heatsink the temp won't be over 70C even after a few hours.
New boards fab out
New version PCB is fab out!
Modifications:
1. connected collectors of Q8 and Q3 (which was missing on the first version)
2. modified footprint of potentiometers. Multi turn pots can also be mounted
3. fixed the bug of small hole for Q8.
4. added a 220uf cap on feedback path. This is an optional component.If you don't need it just short 2 pins of the cap
5. footprint of Q1 and Q2 are modified to which TO-3PL transistors like 2SC3280 can be assembly on board directly.
Free boards keep on offering:
FREE JLH -2005 Class-A Amplifier Board - Boards | Kits | Components | Modules | Tools
Also thinking about add a tube into a JLH design. Any suggestions?
New version PCB is fab out!
Modifications:
1. connected collectors of Q8 and Q3 (which was missing on the first version)
2. modified footprint of potentiometers. Multi turn pots can also be mounted
3. fixed the bug of small hole for Q8.
4. added a 220uf cap on feedback path. This is an optional component.If you don't need it just short 2 pins of the cap
5. footprint of Q1 and Q2 are modified to which TO-3PL transistors like 2SC3280 can be assembly on board directly.
Free boards keep on offering:
FREE JLH -2005 Class-A Amplifier Board - Boards | Kits | Components | Modules | Tools
Also thinking about add a tube into a JLH design. Any suggestions?
Attachments
Finally built a chassis (sort of) for my JLH.
Still needs some covering up before it's allowed in the house as a permanent feature.
There's still a tiny amount of hum, which I can live with, and sometimes a weird buzzing when a phono input is half plugged in - I suspect to do with a lack of ground connection from the source somehow. Anyway, not an issue really.
Also quite reassured by the fact it weighs 9.3kg.
Still needs some covering up before it's allowed in the house as a permanent feature.
There's still a tiny amount of hum, which I can live with, and sometimes a weird buzzing when a phono input is half plugged in - I suspect to do with a lack of ground connection from the source somehow. Anyway, not an issue really.
Also quite reassured by the fact it weighs 9.3kg.
Attachments
I do not know the technical reason it may not be a good practice to subject power lines to the magnetic field of a transformer. But I am pretty sure that it would not do any harm to lay out power lines away from the transformer if possible.
Sorry for my worry too much noise plus I have no proper electronic education et al.
Hi Chris,
You're right, it might be possible there's induced hum, but I think the hum (not that there is much at all - only when there's no signal and you're really close to the speaker) can probably be reduced with further capcitance and possibly some filtering (basically CRC filtering). The weird noise I can only guess is a grounding issue of some sort.
Cheers,
Phil
You're right, it might be possible there's induced hum, but I think the hum (not that there is much at all - only when there's no signal and you're really close to the speaker) can probably be reduced with further capcitance and possibly some filtering (basically CRC filtering). The weird noise I can only guess is a grounding issue of some sort.
Cheers,
Phil
