I have a NAD C445 tuner. I have recently one of the voltage regulators as it was broken. It is not up and running.
There are four existing voltage regulators which are getting very warm and causing teh tuner to consume 40W:
Regulators IC94 IC91 and IC92 and getting to 75C. In particular IC94 is getting hot without a heatsink. There is a relay that is getting to 45C. The tuner otherwise works fine.
What could be wrong?
Do I just need to replace all the regulators?
Many thanks, George
There are four existing voltage regulators which are getting very warm and causing teh tuner to consume 40W:
- IC91 L7805A with sizable heatsink for FM/AM Amp
- IC92 L7805A with sizable heatsink for OpAmp
- IC93 L7912 for DAB
- IC94 L7812 for vacuum flourescent display
Regulators IC94 IC91 and IC92 and getting to 75C. In particular IC94 is getting hot without a heatsink. There is a relay that is getting to 45C. The tuner otherwise works fine.
What could be wrong?
Do I just need to replace all the regulators?
Many thanks, George
Last edited:
The regulators are not likely to be the problem, if the regulated voltages are ok.
Is your measured AC line close to the rated AC line voltage of the tuner?
Power Requirement: 230-240V AC, 50Hz
Is your measured AC line close to the rated AC line voltage of the tuner?
Power Requirement: 230-240V AC, 50Hz
The service manual does not give any spec for the PSU. Though, the windings are reading 15-0-15 and 9-0-9. So, I guess this is all fine.
I have just put my watt meter and found that the unit is consuming 12W, which is exactly as the spec. Not the 40W I previously mentioned. Sorry.
During my previous repair I have accidentally skipped the relay when connecting the PSU. As a result the standby power was 12W and exactly the same as the on power.
Sounds like these L7812 voltage regulators will just get very hot. Unless replacing them may stop them from getting so hot?
I have just put my watt meter and found that the unit is consuming 12W, which is exactly as the spec. Not the 40W I previously mentioned. Sorry.
During my previous repair I have accidentally skipped the relay when connecting the PSU. As a result the standby power was 12W and exactly the same as the on power.
Sounds like these L7812 voltage regulators will just get very hot. Unless replacing them may stop them from getting so hot?
No, the same equation will still hold for new replacement regulators.
P ( Watts ) = V ( DC Volts ) x I ( DC Amperes ) .
P ( Watts ) = V ( DC Volts ) x I ( DC Amperes ) .
In my experience, heating caused by inefficient design is a common problem. Often there are linear regulators that overheat just because using a torroid which has optimal voltages for the power stage and not suitable for the drivers and control circuits. Or there is a fancy chassis with not enough ventilation. In your case, if the regulator input voltage is too high, you can insert a series resistor, as shown in the 78xx spec, or even a pre-regulator to spread the heat. Also try running with cover off.
Hey @rayma, removing the cover might show if ventilation contributes to the problem.
I did some measurements on a NAD C320 and it had excessive voltages applied to the regulators and the mainboard was blocking the bottom air intake.
I did some measurements on a NAD C320 and it had excessive voltages applied to the regulators and the mainboard was blocking the bottom air intake.
For device cooling to be effective, you have to have the right package and the right heat sink,
as well as a place for the air to flow. For this tuner, prevention is the best cure.
as well as a place for the air to flow. For this tuner, prevention is the best cure.
I see on the schematic in post #1 that some of the rectifying diodes are 1N4003 (makes sense) and some 1N4148 (seems way too small). Is that an error on the schematic or did they really use 1N4148s?
Regarding the voltage, 9 V and 15 V transformer secondary voltages don't seem excessive to me. When I take into account mains tolerances, ripple, regulator drop-out voltage, diode voltage drop and the effect of the spiky current waveforms, I usually end up using an 8 V or 9 V transformer for 5 V supplies and a 15 V transformer for 12 V supplies with 78xx or 79xx regulators.
Regarding the voltage, 9 V and 15 V transformer secondary voltages don't seem excessive to me. When I take into account mains tolerances, ripple, regulator drop-out voltage, diode voltage drop and the effect of the spiky current waveforms, I usually end up using an 8 V or 9 V transformer for 5 V supplies and a 15 V transformer for 12 V supplies with 78xx or 79xx regulators.
Modern 78xx/79xx have less than 1.5V dropout voltage so today 9V AC for 5V regulated is excessive just like 15V AC for 12V regulated is. Both can easily be lower voltages resulting in lower temperatures. Those 1N4148 used as rectifier diodes is an odd choice as they are quite wimpy with regards to current. I would replace those for nice 1A Schottky diodes and use lower AC voltages. Depending on the brand of 78xx/79xx 7V AC for 5V regulated is no problem and 12V AC for 12V regulated also works fine in low current demand situations. I am sure when you measure the voltages of the now used transformer voltages are way higher. Like 2 x 9V and 2 x 18V, this won't make matters any better.
In fact it is the cause of the higher temperatures and the only thing achieved is burning power to heat. Besides a waste of energy also a cause for deterioration/defects of which you got proof as one of them died already.
In fact it is the cause of the higher temperatures and the only thing achieved is burning power to heat. Besides a waste of energy also a cause for deterioration/defects of which you got proof as one of them died already.
Last edited:
https://www.ti.com/lit/gpn/lm7800
The LM7805 output voltage is guaranteed to be in spec when the input voltage is at least 7.5 V, the LM7812 output voltage when its input voltage is at least 14.5 V.
That's actually worse than I remember from decades old datasheets, I thought it was 7.3 V for a 5 V regulator. Maybe that was for a different brand.
Dropout voltage 2 V typical, but as that's a typical value, you can't count on it.
The LM7805 output voltage is guaranteed to be in spec when the input voltage is at least 7.5 V, the LM7812 output voltage when its input voltage is at least 14.5 V.
That's actually worse than I remember from decades old datasheets, I thought it was 7.3 V for a 5 V regulator. Maybe that was for a different brand.
Dropout voltage 2 V typical, but as that's a typical value, you can't count on it.
I do this on a monthly basis. It works with action/verify method and in 9 out of 10 cases also with still too high margins so no issues with mains voltage variation etc. Just like high semiconductor temperatures simply never improve lifetime of parts/devices. There is also nothing to be found against it in reality otherwise it would not make sense to change stuff and it is already overheating so there IS an issue to solve. NAD and its sub brands have many longevity hampering features like this all implemented by design* by clever engineers. It can be fixed.
*Like IC95 (7805) is connected to the output of IC94 (7812)....Yup 7V and extra dissipated power for IC94. Could have been connected to the 9V AC with own rectifier and cap resulting in IC94 being less hot. Saved 2 diodes and a cap!
As they have become hot and one of them already being defective replacing the 50 cent 78xx/79xx for known good ones with lower noise and lower dropout voltage is a no brainer. When possible larger heatsinks (if you have those) are a nice addition but not the solution.
GCK303: it may help to mention brand name and type of the relay. High sensitivity coil types exist often by the same brand, also resulting in lower temperatures.
*Like IC95 (7805) is connected to the output of IC94 (7812)....Yup 7V and extra dissipated power for IC94. Could have been connected to the 9V AC with own rectifier and cap resulting in IC94 being less hot. Saved 2 diodes and a cap!
As they have become hot and one of them already being defective replacing the 50 cent 78xx/79xx for known good ones with lower noise and lower dropout voltage is a no brainer. When possible larger heatsinks (if you have those) are a nice addition but not the solution.
GCK303: it may help to mention brand name and type of the relay. High sensitivity coil types exist often by the same brand, also resulting in lower temperatures.
Last edited: