Exactly, so consequently you must still be insisting the transformer was wrongly specified. Your argument though, is against recommended design guidelines for 3 terminal regulators and would apply to all commercial linear power supplies. I put it to you again, that reducing the raw DC supply abitrarily until the regulator device is cool enough for your liking, is the wrong approach to 3 terminal regulators or indeed any semiconductor application. Can you even be certain there will be no dropouts by significantly reducing the raw DC input to any or all of the 3 supplies?
I suspect you haven't considered the poor regulation of most small transformers and the relatively heavy ripple that needs to be accommodated here. If you read the service notes (p16), you'll see problems were already anticipated with low or unstable mains supplies, when the primary winding turns were increased from PCB issue 3 on.
I thought that but couldn't find any current products, apart from a couple of power amplifiers on their website that were identifiably solid state. Perhaps I assume too much in thinking that standard regulators would not be used in their audio circuits.
O/T....
I'd be fascinated to learn what effect the temperature of a 3 terminal reg. has on sound quality - or is that more a justification for utilizing the full rating rather than just forcing a high die temp? Temperatures that are considered hot or cool are relative to our personal comfort so how is there a relationship between BJTs or ICs running within their ratings at some particular temperature range? It may be true that certain designs or regs do seem to confer better sound quality when hot but why can't a sub-zero temp. reg. equally sound better in another application?
Otherwise, I'd be fairly certain it's only a convenient association of heat with good audio which may prove to be coincidence rather than universally true.
Associating heat with tubes and class A is now a strong psychological drawcard in audio. People seem to fall for it as a desirable quality like the association of noisy mufflers with engine power or even UHF stripline PCB layout with high-end preamps!
I don't believe that semiconductors NEED to run hot and have stated so a few times on this Forum.
I firmly believe that performance AND reliability is improved by using adequate heatsinking for the duty to be performed.
A 3pin regulator changes performance when the output current is set to different levels.
Running hot, to me implies, that the higher levels of current capability are being used and/or higher levels of voltage drop have been imposed. That usually means that some form of heat removal beyond the bare chip is required.
One could run the 3pin reg with a little bit of Z shaped aluminium plate (guessing at 30C/W) or use a 1C/W heatsink. The electrical loading is the same. the operating temperature will be very different. I used a 15C/W heatsink that just fits the space.
I thought about using the surrounding aluminium "box" but since it protects against Mains interference and Mains faults I decided not to.
It is not electrically connected to the PCB but could modified to become "earthed" to the Chassis.
I firmly believe that performance AND reliability is improved by using adequate heatsinking for the duty to be performed.
A 3pin regulator changes performance when the output current is set to different levels.
Running hot, to me implies, that the higher levels of current capability are being used and/or higher levels of voltage drop have been imposed. That usually means that some form of heat removal beyond the bare chip is required.
One could run the 3pin reg with a little bit of Z shaped aluminium plate (guessing at 30C/W) or use a 1C/W heatsink. The electrical loading is the same. the operating temperature will be very different. I used a 15C/W heatsink that just fits the space.
I thought about using the surrounding aluminium "box" but since it protects against Mains interference and Mains faults I decided not to.
It is not electrically connected to the PCB but could modified to become "earthed" to the Chassis.
In a desing of this type electrical desing & packaging is sometime done by different people or team .... and there is economic and architectural issue... this may explain why the regulator run so hot .
Of course a heatsink or a better air flow will help.
According to post 10 mains voltage must be betwen 216 and 253 Volts ... about 15% variation.
Reducing mains voltage will reduce heating of regulator and the unit should work properly as long as voltage is betwen limits of 216 & 253 Volts.
I do not suggest to reduce voltage lower than that ... just work betwen these limits.... closer to the lower one.
Of course a heatsink or a better air flow will help.
According to post 10 mains voltage must be betwen 216 and 253 Volts ... about 15% variation.
Reducing mains voltage will reduce heating of regulator and the unit should work properly as long as voltage is betwen limits of 216 & 253 Volts.
I do not suggest to reduce voltage lower than that ... just work betwen these limits.... closer to the lower one.
Since the supply is 240V and 216V is nominally only 10% lower, you would be wasting an isolated external transformer to make such a small change. The range you quote though, is actually the (existing in 1983) mains supply standard of +5%, -10%. That simply says it will work correctly on the UK power grid but that is not the tuner's operating specification.
The real problem is that if the mains 240V supply then drops below its nominal voltage (and it will at some point) the reduced AC will fall below 216V. Your desperation to win an argument is showing but to help your cause, you could read up and analyse the implications in your suggestions and do a little math first.
Eventually, you will have to accept that the product is going to run hot because it was designed with minimal (no) heatsinking and if you want it to be cool but still operate reliably, you will have to increase that heatsinking substantially because reducing the voltage supply to a linear regulator will prove futile, however much you think it should be otherwise.
Let's be realistic. The tuner is working fine. No changes are required if there is no damage and AndrewT can simply close the PSU box, refit the case and play the tuner much as it has been for the past 30 years. You can't touch or see the reg, so you don't know it is hot but it works very reliably. Why fiddle with it?
The real problem is that if the mains 240V supply then drops below its nominal voltage (and it will at some point) the reduced AC will fall below 216V. Your desperation to win an argument is showing but to help your cause, you could read up and analyse the implications in your suggestions and do a little math first.
Eventually, you will have to accept that the product is going to run hot because it was designed with minimal (no) heatsinking and if you want it to be cool but still operate reliably, you will have to increase that heatsinking substantially because reducing the voltage supply to a linear regulator will prove futile, however much you think it should be otherwise.
Let's be realistic. The tuner is working fine. No changes are required if there is no damage and AndrewT can simply close the PSU box, refit the case and play the tuner much as it has been for the past 30 years. You can't touch or see the reg, so you don't know it is hot but it works very reliably. Why fiddle with it?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.