Audiolab 8000 A bias issues
ok here is a few facts and a few questions around the audiolab 8000 a
---First to notice that units in production till 1992 are equipped with 220V trafos so many of these units with now days 230-240 voltage in Europe will produce almost 43 +43 volts after the capacitors and these few 3-5 volts more per rail will eventually stress the amp and coloration of burned pcb will be there
---Also noticed that either after year 1992 trafos are different so with 230Ac produce almost 40 Volt ( or they simply realized that they are stressing the wrong circuit with the wrong ways.)
---the audio lab 8000S is an sziklai or CFP if you like that means that the double VBE multiplier lays on the drivers .
---Let us forget for a minute the weird bias procedure where a tech is requested to remove covers install probes re install covers and adjust the bias :eek::eek::eek:
---Real problem is that the ventilation designed works only for the outputs and the rest of the amp is tight sealed where in the driver area temperature will increase then the Vbe multiplier will drop the bias according to the temperature ....probably over compensate and reduce the bias more than needed .
so Solution 1 will be to almost over bias the amplifier and expect the VBE multiplier to finally adjust something more reasonable when the pcb are will reach the ""operating'' temperature
Drawbacks will be
--that you still going to stress the pcb and parts around the amp
--will take quite a lot of time for the amp to stabilize a temperature
--And don't know why but i still think that there is still going to be some thermal run away
or Solution 2 drill fins or long ventilation halls above the driver area
---Bias the amp to the manual stated bias
---Defuse all the concentrated heat from inside the box
---Allow the Vbe multiplier to adjust the bias only when amp is stressed
i dont see any drawback here ....
let me not forget how many in a way arrogant designers produced equipment in England and simply missed that one of their amplifier may find its way to a Mediterranean or any other 10 degrees warmer country .
Lets make that 15 C warmer. When I was in Andalucia a couple of weeks ago, temps were hitting 30-35 C there while here temps were struggling to reach 16-18 C...
I was actually expecting more people to jump in with comments on this ....
I am willing to drill the cover anyway and monitor bias for 24 hours in various playing conditions to see how it will go
I may as well do that before and after drilling
Good thing for my testing is that as we speak in Athens we have max temp. of 34 degrees ( having a blast with a second summer my kids are on the way to spent another day on the beach !!! )
Hi Sakis, how are you at 34 degrees? I know about using northern European amplifiers in warm climates too! :redhot:
If the driver stage is the one being thermally compensated, why is there a problem if it gets particularly hot? The Vbe multiplier or diodes should continue to provide fairly linear bias control according to the driver case temperature, whatever it is within reason - if it is in thermal contact. The temperature sensed does not have to track the output stage temperature at all, AFAIK, so this should not affect the bias control.
If you are just trying to reduce driver temperature by ventilation, it should continue to work but perhaps the response curve of the Vbe compensation voltage may not be in the same linear or flat region any more. I would think, as you suspect, that tweaking the bias level to produce lower bias and driver dissipation will result in undercompensation, perhaps runaway but ideally it should track according to Douglas Self''s figuring and graphs in his chapter on "Thermal compensation and thermal dynamics". It's in each edition and not hard to find, even on the web, in defiance of copyright.
Optimally biased CFP stages run cold at typical domestic power output - cold like the sound quality. Then there are several designs with high bias of 70-100 mA, like old Hitachi designs and, of course, the DIY P3a. The high bias designs really complicate thermal compensation because they vary so much from ambient temperature differences and warm-up yet very little from the power output, until you turn up the power and this is reversed. These are a real hassle, so anyone who builds a technically correct comp. circuit for this deserves a medal, IMO. The air-spaced type, as used in the P3a, seems much easier to apply.
Hello Ian ( always nice to see you around my quests ) !!!
I will post a copy of the ""bias adjustment procedure "" according to the service manual which if i manage to understand the two (??? ) (!!!) pages correctly the approach of the designer is as i said
Over bias the amplifier at the bench while tuning ...close the cover allow the amp to built the temperature inside the case and expect the VBE to take it down to a reasonable level after warming up .
Imagine what will happen if the procedure is to take place with ambient temp of 22 degrees and add another 15 or more the amplifiers operating temperature inside the case
we are talking something like 65-70 degrees on the driver heatsink ...that is way too much
I am setting up as we speak a measuring set up for both temperature and bias so in a few days i will have proper graphs of how bias and temperature rising in various operating modes idle - casual listening - average power
70 degrees? Wow! Just to be clear, is the Vbe multiplier attached to the driver(s), not with an air gap. Is that correct?
The procedure you describe sounds like there is no thermal connection there, so it has to be sealed and adjusted carefully to make the setup conditions like the operating conditions. That's awkward, like it sounds and crazy too.
I would still go with your plan to allow air flow or, if possible, increase heatsinking to reduce mean ambient temperatue around the drivers. Perhaps the only problem then is the possibility of hole blockage with dust or trying to use the amp in a bad position etc.
I look forward to your test results anyway. Good luck
yeap that is correct
In this audiolab the VBE is double and also perfectly attached one of them to each driver ...that will make over compensation far more easy
In the only manual I can find (207 Serial nos 87 89) it states that the voltage across the emitter resistor should be 22mV for a quiescent current of 50mA. The circuit has these resistors as 0.22 ohm. So anyone setting it up with 22mV gets a bias current of 100mA.
Is this an error in the manual? I wonder how many of these amps are running hot for this reason.
|All times are GMT. The time now is 07:34 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio